Commit ce7339e80a11e0a8f466e779212ee328a7ddc9e5 | zig

author	Alex Rønne Petersen <alex@alexrp.com> 2025-07-16 04:52:10+02:00
committer	Alex Rønne Petersen <alex@alexrp.com> 2025-08-30 06:36:40+02:00
log	ce7339e80a11e0a8f466e779212ee328a7ddc9e5
tree	911ae7cce5f776dc6276a84d4e640092cf40d7f3
parent	b7a8c045efab89c0fab0d3d1c678442c2b373066
signature	Signed by SSH key SHA256:7B/LJ7bpR1eX8aCXSr4mtd5M45VMPKcx9zY8e95b5QM

zig cc: update intrinsic headers to LLVM 21

45 files changed, 31764 insertions(+), 27580 deletions(-)

lib/include/__clang_spirv_builtins.h created+217

...	...	@@ -0,0 +1,217 @@
	1	/*===---- spirv_builtin_vars.h - SPIR-V built-in ---------------------------===
	2	*
	3	* Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
	4	* See https://llvm.org/LICENSE.txt for license information.
	5	* SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
	6	*
	7	*===-----------------------------------------------------------------------===
	8	*/
	9
	10	#ifndef __SPIRV_BUILTIN_VARS_H
	11	#define __SPIRV_BUILTIN_VARS_H
	12
	13	#if __cplusplus >= 201103L
	14	#define __SPIRV_NOEXCEPT noexcept
	15	#else
	16	#define __SPIRV_NOEXCEPT
	17	#endif
	18
	19	#pragma push_macro("__size_t")
	20	#pragma push_macro("__uint32_t")
	21	#pragma push_macro("__uint64_t")
	22	#define __size_t __SIZE_TYPE__
	23	#define __uint32_t __UINT32_TYPE__
	24
	25	#define __SPIRV_overloadable __attribute__((overloadable))
	26	#define __SPIRV_convergent __attribute__((convergent))
	27	#define __SPIRV_inline __attribute__((always_inline))
	28
	29	#define __global __attribute__((opencl_global))
	30	#define __local __attribute__((opencl_local))
	31	#define __private __attribute__((opencl_private))
	32	#define __constant __attribute__((opencl_constant))
	33	#ifdef __SYCL_DEVICE_ONLY__
	34	#define __generic
	35	#else
	36	#define __generic __attribute__((opencl_generic))
	37	#endif
	38
	39	// Check if SPIR-V builtins are supported.
	40	// As the translator doesn't use the LLVM intrinsics (which would be emitted if
	41	// we use the SPIR-V builtins) we can't rely on the SPIRV32/SPIRV64 etc macros
	42	// to establish if we can use the builtin alias. We disable builtin altogether
	43	// if we do not intent to use the backend. So instead of use target macros, rely
	44	// on a __has_builtin test.
	45	#if (__has_builtin(__builtin_spirv_num_workgroups))
	46	#define __SPIRV_BUILTIN_ALIAS(builtin) \
	47	__attribute__((clang_builtin_alias(builtin)))
	48	#else
	49	#define __SPIRV_BUILTIN_ALIAS(builtin)
	50	#endif
	51
	52	// Builtin IDs and sizes
	53
	54	extern __SPIRV_BUILTIN_ALIAS(__builtin_spirv_num_workgroups) __size_t
	55	__spirv_NumWorkgroups(int);
	56	extern __SPIRV_BUILTIN_ALIAS(__builtin_spirv_workgroup_size) __size_t
	57	__spirv_WorkgroupSize(int);
	58	extern __SPIRV_BUILTIN_ALIAS(__builtin_spirv_workgroup_id) __size_t
	59	__spirv_WorkgroupId(int);
	60	extern __SPIRV_BUILTIN_ALIAS(__builtin_spirv_local_invocation_id) __size_t
	61	__spirv_LocalInvocationId(int);
	62	extern __SPIRV_BUILTIN_ALIAS(__builtin_spirv_global_invocation_id) __size_t
	63	__spirv_GlobalInvocationId(int);
	64
	65	extern __SPIRV_BUILTIN_ALIAS(__builtin_spirv_global_size) __size_t
	66	__spirv_GlobalSize(int);
	67	extern __SPIRV_BUILTIN_ALIAS(__builtin_spirv_global_offset) __size_t
	68	__spirv_GlobalOffset(int);
	69	extern __SPIRV_BUILTIN_ALIAS(__builtin_spirv_subgroup_size) __uint32_t
	70	__spirv_SubgroupSize();
	71	extern __SPIRV_BUILTIN_ALIAS(__builtin_spirv_subgroup_max_size) __uint32_t
	72	__spirv_SubgroupMaxSize();
	73	extern __SPIRV_BUILTIN_ALIAS(__builtin_spirv_num_subgroups) __uint32_t
	74	__spirv_NumSubgroups();
	75	extern __SPIRV_BUILTIN_ALIAS(__builtin_spirv_subgroup_id) __uint32_t
	76	__spirv_SubgroupId();
	77	extern __SPIRV_BUILTIN_ALIAS(__builtin_spirv_subgroup_local_invocation_id)
	78	__uint32_t __spirv_SubgroupLocalInvocationId();
	79
	80	// OpGenericCastToPtrExplicit
	81
	82	extern __SPIRV_overloadable
	83	__SPIRV_BUILTIN_ALIAS(__builtin_spirv_generic_cast_to_ptr_explicit)
	84	__global void __spirv_GenericCastToPtrExplicit_ToGlobal(__generic void ,
	85	int) __SPIRV_NOEXCEPT;
	86	extern __SPIRV_overloadable
	87	__SPIRV_BUILTIN_ALIAS(__builtin_spirv_generic_cast_to_ptr_explicit)
	88	__global const void *
	89	__spirv_GenericCastToPtrExplicit_ToGlobal(__generic const void *,
	90	int) __SPIRV_NOEXCEPT;
	91	extern __SPIRV_overloadable
	92	__SPIRV_BUILTIN_ALIAS(__builtin_spirv_generic_cast_to_ptr_explicit)
	93	__global volatile void *
	94	__spirv_GenericCastToPtrExplicit_ToGlobal(__generic volatile void *,
	95	int) __SPIRV_NOEXCEPT;
	96	extern __SPIRV_overloadable
	97	__SPIRV_BUILTIN_ALIAS(__builtin_spirv_generic_cast_to_ptr_explicit)
	98	__global const volatile void *
	99	__spirv_GenericCastToPtrExplicit_ToGlobal(__generic const volatile void *,
	100	int) __SPIRV_NOEXCEPT;
	101	extern __SPIRV_overloadable
	102	__SPIRV_BUILTIN_ALIAS(__builtin_spirv_generic_cast_to_ptr_explicit)
	103	__local void __spirv_GenericCastToPtrExplicit_ToLocal(__generic void ,
	104	int) __SPIRV_NOEXCEPT;
	105	extern __SPIRV_overloadable
	106	__SPIRV_BUILTIN_ALIAS(__builtin_spirv_generic_cast_to_ptr_explicit)
	107	__local const void *
	108	__spirv_GenericCastToPtrExplicit_ToLocal(__generic const void *,
	109	int) __SPIRV_NOEXCEPT;
	110	extern __SPIRV_overloadable
	111	__SPIRV_BUILTIN_ALIAS(__builtin_spirv_generic_cast_to_ptr_explicit)
	112	__local volatile void *
	113	__spirv_GenericCastToPtrExplicit_ToLocal(__generic volatile void *,
	114	int) __SPIRV_NOEXCEPT;
	115	extern __SPIRV_overloadable
	116	__SPIRV_BUILTIN_ALIAS(__builtin_spirv_generic_cast_to_ptr_explicit)
	117	__local const volatile void *
	118	__spirv_GenericCastToPtrExplicit_ToLocal(__generic const volatile void *,
	119	int) __SPIRV_NOEXCEPT;
	120	extern __SPIRV_overloadable
	121	__SPIRV_BUILTIN_ALIAS(__builtin_spirv_generic_cast_to_ptr_explicit)
	122	__private void *
	123	__spirv_GenericCastToPtrExplicit_ToPrivate(__generic void *,
	124	int) __SPIRV_NOEXCEPT;
	125	extern __SPIRV_overloadable
	126	__SPIRV_BUILTIN_ALIAS(__builtin_spirv_generic_cast_to_ptr_explicit)
	127	__private const void *
	128	__spirv_GenericCastToPtrExplicit_ToPrivate(__generic const void *,
	129	int) __SPIRV_NOEXCEPT;
	130	extern __SPIRV_overloadable
	131	__SPIRV_BUILTIN_ALIAS(__builtin_spirv_generic_cast_to_ptr_explicit)
	132	__private volatile void *
	133	__spirv_GenericCastToPtrExplicit_ToPrivate(__generic volatile void *,
	134	int) __SPIRV_NOEXCEPT;
	135	extern __SPIRV_overloadable
	136	__SPIRV_BUILTIN_ALIAS(__builtin_spirv_generic_cast_to_ptr_explicit)
	137	__private const volatile void *
	138	__spirv_GenericCastToPtrExplicit_ToPrivate(__generic const volatile void *,
	139	int) __SPIRV_NOEXCEPT;
	140
	141	// OpGenericCastToPtr
	142
	143	static __SPIRV_overloadable __SPIRV_inline __global void *
	144	__spirv_GenericCastToPtr_ToGlobal(__generic void *p, int) __SPIRV_NOEXCEPT {
	145	return (__global void *)p;
	146	}
	147	static __SPIRV_overloadable __SPIRV_inline __global const void *
	148	__spirv_GenericCastToPtr_ToGlobal(__generic const void *p,
	149	int) __SPIRV_NOEXCEPT {
	150	return (__global const void *)p;
	151	}
	152	static __SPIRV_overloadable __SPIRV_inline __global volatile void *
	153	__spirv_GenericCastToPtr_ToGlobal(__generic volatile void *p,
	154	int) __SPIRV_NOEXCEPT {
	155	return (__global volatile void *)p;
	156	}
	157	static __SPIRV_overloadable __SPIRV_inline __global const volatile void *
	158	__spirv_GenericCastToPtr_ToGlobal(__generic const volatile void *p,
	159	int) __SPIRV_NOEXCEPT {
	160	return (__global const volatile void *)p;
	161	}
	162	static __SPIRV_overloadable __SPIRV_inline __local void *
	163	__spirv_GenericCastToPtr_ToLocal(__generic void *p, int) __SPIRV_NOEXCEPT {
	164	return (__local void *)p;
	165	}
	166	static __SPIRV_overloadable __SPIRV_inline __local const void *
	167	__spirv_GenericCastToPtr_ToLocal(__generic const void *p,
	168	int) __SPIRV_NOEXCEPT {
	169	return (__local const void *)p;
	170	}
	171	static __SPIRV_overloadable __SPIRV_inline __local volatile void *
	172	__spirv_GenericCastToPtr_ToLocal(__generic volatile void *p,
	173	int) __SPIRV_NOEXCEPT {
	174	return (__local volatile void *)p;
	175	}
	176	static __SPIRV_overloadable __SPIRV_inline __local const volatile void *
	177	__spirv_GenericCastToPtr_ToLocal(__generic const volatile void *p,
	178	int) __SPIRV_NOEXCEPT {
	179	return (__local const volatile void *)p;
	180	}
	181	static __SPIRV_overloadable __SPIRV_inline __private void *
	182	__spirv_GenericCastToPtr_ToPrivate(__generic void *p, int) __SPIRV_NOEXCEPT {
	183	return (__private void *)p;
	184	}
	185	static __SPIRV_overloadable __SPIRV_inline __private const void *
	186	__spirv_GenericCastToPtr_ToPrivate(__generic const void *p,
	187	int) __SPIRV_NOEXCEPT {
	188	return (__private const void *)p;
	189	}
	190	static __SPIRV_overloadable __SPIRV_inline __private volatile void *
	191	__spirv_GenericCastToPtr_ToPrivate(__generic volatile void *p,
	192	int) __SPIRV_NOEXCEPT {
	193	return (__private volatile void *)p;
	194	}
	195	static __SPIRV_overloadable __SPIRV_inline __private const volatile void *
	196	__spirv_GenericCastToPtr_ToPrivate(__generic const volatile void *p,
	197	int) __SPIRV_NOEXCEPT {
	198	return (__private const volatile void *)p;
	199	}
	200
	201	#pragma pop_macro("__size_t")
	202	#pragma pop_macro("__uint32_t")
	203	#pragma pop_macro("__uint64_t")
	204
	205	#undef __SPIRV_overloadable
	206	#undef __SPIRV_convergent
	207	#undef __SPIRV_inline
	208
	209	#undef __global
	210	#undef __local
	211	#undef __constant
	212	#undef __generic
	213
	214	#undef __SPIRV_BUILTIN_ALIAS
	215	#undef __SPIRV_NOEXCEPT
	216
	217	#endif /* __SPIRV_BUILTIN_VARS_H */

lib/include/__stdarg_va_arg.h+2-2

...	...	@@ -10,8 +10,8 @@
10	10	#ifndef va_arg
11	11
12	12	#if defined(__STDC_VERSION__) && __STDC_VERSION__ >= 202311L
13		/* C23 does not require the second parameter for va_start. */
14		#define va_start(ap, ...) __builtin_va_start(ap, 0)
	13	/* C23 uses a special builtin. */
	14	#define va_start(...) __builtin_c23_va_start(__VA_ARGS__)
15	15	#else
16	16	/* Versions before C23 do require the second parameter. */
17	17	#define va_start(ap, param) __builtin_va_start(ap, param)

lib/include/altivec.h+82-79

...	...	@@ -17525,70 +17525,73 @@ vec_bperm(vector unsigned long long __a, vector unsigned char __b) {
17525	17525
17526	17526	/* vec_reve */
17527	17527
17528		static inline __ATTRS_o_ai vector bool char vec_reve(vector bool char __a) {
	17528	static __inline__ __ATTRS_o_ai vector bool char vec_reve(vector bool char __a) {
17529	17529	return __builtin_shufflevector(__a, __a, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6,
17530	17530	5, 4, 3, 2, 1, 0);
17531	17531	}
17532	17532
17533		static inline __ATTRS_o_ai vector signed char vec_reve(vector signed char __a) {
	17533	static __inline__ __ATTRS_o_ai vector signed char
	17534	vec_reve(vector signed char __a) {
17534	17535	return __builtin_shufflevector(__a, __a, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6,
17535	17536	5, 4, 3, 2, 1, 0);
17536	17537	}
17537	17538
17538		static inline __ATTRS_o_ai vector unsigned char
	17539	static __inline__ __ATTRS_o_ai vector unsigned char
17539	17540	vec_reve(vector unsigned char __a) {
17540	17541	return __builtin_shufflevector(__a, __a, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6,
17541	17542	5, 4, 3, 2, 1, 0);
17542	17543	}
17543	17544
17544		static inline __ATTRS_o_ai vector bool int vec_reve(vector bool int __a) {
	17545	static __inline__ __ATTRS_o_ai vector bool int vec_reve(vector bool int __a) {
17545	17546	return __builtin_shufflevector(__a, __a, 3, 2, 1, 0);
17546	17547	}
17547	17548
17548		static inline __ATTRS_o_ai vector signed int vec_reve(vector signed int __a) {
	17549	static __inline__ __ATTRS_o_ai vector signed int
	17550	vec_reve(vector signed int __a) {
17549	17551	return __builtin_shufflevector(__a, __a, 3, 2, 1, 0);
17550	17552	}
17551	17553
17552		static inline __ATTRS_o_ai vector unsigned int
	17554	static __inline__ __ATTRS_o_ai vector unsigned int
17553	17555	vec_reve(vector unsigned int __a) {
17554	17556	return __builtin_shufflevector(__a, __a, 3, 2, 1, 0);
17555	17557	}
17556	17558
17557		static inline __ATTRS_o_ai vector bool short vec_reve(vector bool short __a) {
	17559	static __inline__ __ATTRS_o_ai vector bool short
	17560	vec_reve(vector bool short __a) {
17558	17561	return __builtin_shufflevector(__a, __a, 7, 6, 5, 4, 3, 2, 1, 0);
17559	17562	}
17560	17563
17561		static inline __ATTRS_o_ai vector signed short
	17564	static __inline__ __ATTRS_o_ai vector signed short
17562	17565	vec_reve(vector signed short __a) {
17563	17566	return __builtin_shufflevector(__a, __a, 7, 6, 5, 4, 3, 2, 1, 0);
17564	17567	}
17565	17568
17566		static inline __ATTRS_o_ai vector unsigned short
	17569	static __inline__ __ATTRS_o_ai vector unsigned short
17567	17570	vec_reve(vector unsigned short __a) {
17568	17571	return __builtin_shufflevector(__a, __a, 7, 6, 5, 4, 3, 2, 1, 0);
17569	17572	}
17570	17573
17571		static inline __ATTRS_o_ai vector float vec_reve(vector float __a) {
	17574	static __inline__ __ATTRS_o_ai vector float vec_reve(vector float __a) {
17572	17575	return __builtin_shufflevector(__a, __a, 3, 2, 1, 0);
17573	17576	}
17574	17577
17575	17578	#ifdef __VSX__
17576		static inline __ATTRS_o_ai vector bool long long
	17579	static __inline__ __ATTRS_o_ai vector bool long long
17577	17580	vec_reve(vector bool long long __a) {
17578	17581	return __builtin_shufflevector(__a, __a, 1, 0);
17579	17582	}
17580	17583
17581		static inline __ATTRS_o_ai vector signed long long
	17584	static __inline__ __ATTRS_o_ai vector signed long long
17582	17585	vec_reve(vector signed long long __a) {
17583	17586	return __builtin_shufflevector(__a, __a, 1, 0);
17584	17587	}
17585	17588
17586		static inline __ATTRS_o_ai vector unsigned long long
	17589	static __inline__ __ATTRS_o_ai vector unsigned long long
17587	17590	vec_reve(vector unsigned long long __a) {
17588	17591	return __builtin_shufflevector(__a, __a, 1, 0);
17589	17592	}
17590	17593
17591		static inline __ATTRS_o_ai vector double vec_reve(vector double __a) {
	17594	static __inline__ __ATTRS_o_ai vector double vec_reve(vector double __a) {
17592	17595	return __builtin_shufflevector(__a, __a, 1, 0);
17593	17596	}
17594	17597	#endif
...	...	@@ -17721,42 +17724,42 @@ typedef vector signed int unaligned_vec_sint __attribute__((aligned(1)));
17721	17724	typedef vector unsigned int unaligned_vec_uint __attribute__((aligned(1)));
17722	17725	typedef vector float unaligned_vec_float __attribute__((aligned(1)));
17723	17726
17724		static inline __ATTRS_o_ai vector signed char vec_xl(ptrdiff_t __offset,
17725		const signed char *__ptr) {
	17727	static __inline__ __ATTRS_o_ai vector signed char
	17728	vec_xl(ptrdiff_t __offset, const signed char *__ptr) {
17726	17729	return (unaligned_vec_schar )(__ptr + __offset);
17727	17730	}
17728	17731
17729		static inline __ATTRS_o_ai vector unsigned char
	17732	static __inline__ __ATTRS_o_ai vector unsigned char
17730	17733	vec_xl(ptrdiff_t __offset, const unsigned char *__ptr) {
17731	17734	return (unaligned_vec_uchar)(__ptr + __offset);
17732	17735	}
17733	17736
17734		static inline __ATTRS_o_ai vector signed short
	17737	static __inline__ __ATTRS_o_ai vector signed short
17735	17738	vec_xl(ptrdiff_t __offset, const signed short *__ptr) {
17736	17739	signed char __addr = (signed char )__ptr + __offset;
17737	17740	return (unaligned_vec_sshort )__addr;
17738	17741	}
17739	17742
17740		static inline __ATTRS_o_ai vector unsigned short
	17743	static __inline__ __ATTRS_o_ai vector unsigned short
17741	17744	vec_xl(ptrdiff_t __offset, const unsigned short *__ptr) {
17742	17745	signed char __addr = (signed char )__ptr + __offset;
17743	17746	return (unaligned_vec_ushort )__addr;
17744	17747	}
17745	17748
17746		static inline __ATTRS_o_ai vector signed int vec_xl(ptrdiff_t __offset,
17747		const signed int *__ptr) {
	17749	static __inline__ __ATTRS_o_ai vector signed int
	17750	vec_xl(ptrdiff_t __offset, const signed int *__ptr) {
17748	17751	signed char __addr = (signed char )__ptr + __offset;
17749	17752	return (unaligned_vec_sint )__addr;
17750	17753	}
17751	17754
17752		static inline __ATTRS_o_ai vector unsigned int
	17755	static __inline__ __ATTRS_o_ai vector unsigned int
17753	17756	vec_xl(ptrdiff_t __offset, const unsigned int *__ptr) {
17754	17757	signed char __addr = (signed char )__ptr + __offset;
17755	17758	return (unaligned_vec_uint )__addr;
17756	17759	}
17757	17760
17758		static inline __ATTRS_o_ai vector float vec_xl(ptrdiff_t __offset,
17759		const float *__ptr) {
	17761	static __inline__ __ATTRS_o_ai vector float vec_xl(ptrdiff_t __offset,
	17762	const float *__ptr) {
17760	17763	signed char __addr = (signed char )__ptr + __offset;
17761	17764	return (unaligned_vec_float )__addr;
17762	17765	}
...	...	@@ -17766,20 +17769,20 @@ typedef vector signed long long unaligned_vec_sll __attribute__((aligned(1)));
17766	17769	typedef vector unsigned long long unaligned_vec_ull __attribute__((aligned(1)));
17767	17770	typedef vector double unaligned_vec_double __attribute__((aligned(1)));
17768	17771
17769		static inline __ATTRS_o_ai vector signed long long
	17772	static __inline__ __ATTRS_o_ai vector signed long long
17770	17773	vec_xl(ptrdiff_t __offset, const signed long long *__ptr) {
17771	17774	signed char __addr = (signed char )__ptr + __offset;
17772	17775	return (unaligned_vec_sll )__addr;
17773	17776	}
17774	17777
17775		static inline __ATTRS_o_ai vector unsigned long long
	17778	static __inline__ __ATTRS_o_ai vector unsigned long long
17776	17779	vec_xl(ptrdiff_t __offset, const unsigned long long *__ptr) {
17777	17780	signed char __addr = (signed char )__ptr + __offset;
17778	17781	return (unaligned_vec_ull )__addr;
17779	17782	}
17780	17783
17781		static inline __ATTRS_o_ai vector double vec_xl(ptrdiff_t __offset,
17782		const double *__ptr) {
	17784	static __inline__ __ATTRS_o_ai vector double vec_xl(ptrdiff_t __offset,
	17785	const double *__ptr) {
17783	17786	signed char __addr = (signed char )__ptr + __offset;
17784	17787	return (unaligned_vec_double )__addr;
17785	17788	}
...	...	@@ -17790,13 +17793,13 @@ static inline __ATTRS_o_ai vector double vec_xl(ptrdiff_t __offset,
17790	17793	typedef vector signed __int128 unaligned_vec_si128 __attribute__((aligned(1)));
17791	17794	typedef vector unsigned __int128 unaligned_vec_ui128
17792	17795	__attribute__((aligned(1)));
17793		static inline __ATTRS_o_ai vector signed __int128
	17796	static __inline__ __ATTRS_o_ai vector signed __int128
17794	17797	vec_xl(ptrdiff_t __offset, const signed __int128 *__ptr) {
17795	17798	signed char __addr = (signed char )__ptr + __offset;
17796	17799	return (unaligned_vec_si128 )__addr;
17797	17800	}
17798	17801
17799		static inline __ATTRS_o_ai vector unsigned __int128
	17802	static __inline__ __ATTRS_o_ai vector unsigned __int128
17800	17803	vec_xl(ptrdiff_t __offset, const unsigned __int128 *__ptr) {
17801	17804	signed char __addr = (signed char )__ptr + __offset;
17802	17805	return (unaligned_vec_ui128 )__addr;
...	...	@@ -17991,64 +17994,64 @@ vec_load_splats(unsigned long long __offset, const float *__ptr) {
17991	17994
17992	17995	#define vec_xstd2 vec_xst
17993	17996	#define vec_xstw4 vec_xst
17994		static inline __ATTRS_o_ai void
	17997	static __inline__ __ATTRS_o_ai void
17995	17998	vec_xst(vector signed char __vec, ptrdiff_t __offset, signed char *__ptr) {
17996	17999	(unaligned_vec_schar )(__ptr + __offset) = __vec;
17997	18000	}
17998	18001
17999		static inline __ATTRS_o_ai void
	18002	static __inline__ __ATTRS_o_ai void
18000	18003	vec_xst(vector unsigned char __vec, ptrdiff_t __offset, unsigned char *__ptr) {
18001	18004	(unaligned_vec_uchar )(__ptr + __offset) = __vec;
18002	18005	}
18003	18006
18004		static inline __ATTRS_o_ai void
	18007	static __inline__ __ATTRS_o_ai void
18005	18008	vec_xst(vector signed short __vec, ptrdiff_t __offset, signed short *__ptr) {
18006	18009	signed char __addr = (signed char )__ptr + __offset;
18007	18010	(unaligned_vec_sshort )__addr = __vec;
18008	18011	}
18009	18012
18010		static inline __ATTRS_o_ai void vec_xst(vector unsigned short __vec,
18011		ptrdiff_t __offset,
18012		unsigned short *__ptr) {
	18013	static __inline__ __ATTRS_o_ai void vec_xst(vector unsigned short __vec,
	18014	ptrdiff_t __offset,
	18015	unsigned short *__ptr) {
18013	18016	signed char __addr = (signed char )__ptr + __offset;
18014	18017	(unaligned_vec_ushort )__addr = __vec;
18015	18018	}
18016	18019
18017		static inline __ATTRS_o_ai void vec_xst(vector signed int __vec,
18018		ptrdiff_t __offset, signed int *__ptr) {
	18020	static __inline__ __ATTRS_o_ai void
	18021	vec_xst(vector signed int __vec, ptrdiff_t __offset, signed int *__ptr) {
18019	18022	signed char __addr = (signed char )__ptr + __offset;
18020	18023	(unaligned_vec_sint )__addr = __vec;
18021	18024	}
18022	18025
18023		static inline __ATTRS_o_ai void
	18026	static __inline__ __ATTRS_o_ai void
18024	18027	vec_xst(vector unsigned int __vec, ptrdiff_t __offset, unsigned int *__ptr) {
18025	18028	signed char __addr = (signed char )__ptr + __offset;
18026	18029	(unaligned_vec_uint )__addr = __vec;
18027	18030	}
18028	18031
18029		static inline __ATTRS_o_ai void vec_xst(vector float __vec, ptrdiff_t __offset,
18030		float *__ptr) {
	18032	static __inline__ __ATTRS_o_ai void vec_xst(vector float __vec,
	18033	ptrdiff_t __offset, float *__ptr) {
18031	18034	signed char __addr = (signed char )__ptr + __offset;
18032	18035	(unaligned_vec_float )__addr = __vec;
18033	18036	}
18034	18037
18035	18038	#ifdef __VSX__
18036		static inline __ATTRS_o_ai void vec_xst(vector signed long long __vec,
18037		ptrdiff_t __offset,
18038		signed long long *__ptr) {
	18039	static __inline__ __ATTRS_o_ai void vec_xst(vector signed long long __vec,
	18040	ptrdiff_t __offset,
	18041	signed long long *__ptr) {
18039	18042	signed char __addr = (signed char )__ptr + __offset;
18040	18043	(unaligned_vec_sll )__addr = __vec;
18041	18044	}
18042	18045
18043		static inline __ATTRS_o_ai void vec_xst(vector unsigned long long __vec,
18044		ptrdiff_t __offset,
18045		unsigned long long *__ptr) {
	18046	static __inline__ __ATTRS_o_ai void vec_xst(vector unsigned long long __vec,
	18047	ptrdiff_t __offset,
	18048	unsigned long long *__ptr) {
18046	18049	signed char __addr = (signed char )__ptr + __offset;
18047	18050	(unaligned_vec_ull )__addr = __vec;
18048	18051	}
18049	18052
18050		static inline __ATTRS_o_ai void vec_xst(vector double __vec, ptrdiff_t __offset,
18051		double *__ptr) {
	18053	static __inline__ __ATTRS_o_ai void vec_xst(vector double __vec,
	18054	ptrdiff_t __offset, double *__ptr) {
18052	18055	signed char __addr = (signed char )__ptr + __offset;
18053	18056	(unaligned_vec_double )__addr = __vec;
18054	18057	}
...	...	@@ -18056,16 +18059,16 @@ static inline __ATTRS_o_ai void vec_xst(vector double __vec, ptrdiff_t __offset,
18056	18059
18057	18060	#if defined(__POWER8_VECTOR__) && defined(__powerpc64__) && \
18058	18061	defined(__SIZEOF_INT128__)
18059		static inline __ATTRS_o_ai void vec_xst(vector signed __int128 __vec,
18060		ptrdiff_t __offset,
18061		signed __int128 *__ptr) {
	18062	static __inline__ __ATTRS_o_ai void vec_xst(vector signed __int128 __vec,
	18063	ptrdiff_t __offset,
	18064	signed __int128 *__ptr) {
18062	18065	signed char __addr = (signed char )__ptr + __offset;
18063	18066	(unaligned_vec_si128 )__addr = __vec;
18064	18067	}
18065	18068
18066		static inline __ATTRS_o_ai void vec_xst(vector unsigned __int128 __vec,
18067		ptrdiff_t __offset,
18068		unsigned __int128 *__ptr) {
	18069	static __inline__ __ATTRS_o_ai void vec_xst(vector unsigned __int128 __vec,
	18070	ptrdiff_t __offset,
	18071	unsigned __int128 *__ptr) {
18069	18072	signed char __addr = (signed char )__ptr + __offset;
18070	18073	(unaligned_vec_ui128 )__addr = __vec;
18071	18074	}
...	...	@@ -18075,51 +18078,51 @@ static inline __ATTRS_o_ai void vec_xst(vector unsigned __int128 __vec,
18075	18078
18076	18079	#if defined(__POWER10_VECTOR__) && defined(__VSX__) && \
18077	18080	defined(__SIZEOF_INT128__)
18078		static inline __ATTRS_o_ai void vec_xst_trunc(vector signed __int128 __vec,
18079		ptrdiff_t __offset,
18080		signed char *__ptr) {
	18081	static __inline__ __ATTRS_o_ai void vec_xst_trunc(vector signed __int128 __vec,
	18082	ptrdiff_t __offset,
	18083	signed char *__ptr) {
18081	18084	*(__ptr + __offset) = (signed char)__vec[0];
18082	18085	}
18083	18086
18084		static inline __ATTRS_o_ai void vec_xst_trunc(vector unsigned __int128 __vec,
18085		ptrdiff_t __offset,
18086		unsigned char *__ptr) {
	18087	static __inline__ __ATTRS_o_ai void
	18088	vec_xst_trunc(vector unsigned __int128 __vec, ptrdiff_t __offset,
	18089	unsigned char *__ptr) {
18087	18090	*(__ptr + __offset) = (unsigned char)__vec[0];
18088	18091	}
18089	18092
18090		static inline __ATTRS_o_ai void vec_xst_trunc(vector signed __int128 __vec,
18091		ptrdiff_t __offset,
18092		signed short *__ptr) {
	18093	static __inline__ __ATTRS_o_ai void vec_xst_trunc(vector signed __int128 __vec,
	18094	ptrdiff_t __offset,
	18095	signed short *__ptr) {
18093	18096	*(__ptr + __offset) = (signed short)__vec[0];
18094	18097	}
18095	18098
18096		static inline __ATTRS_o_ai void vec_xst_trunc(vector unsigned __int128 __vec,
18097		ptrdiff_t __offset,
18098		unsigned short *__ptr) {
	18099	static __inline__ __ATTRS_o_ai void
	18100	vec_xst_trunc(vector unsigned __int128 __vec, ptrdiff_t __offset,
	18101	unsigned short *__ptr) {
18099	18102	*(__ptr + __offset) = (unsigned short)__vec[0];
18100	18103	}
18101	18104
18102		static inline __ATTRS_o_ai void vec_xst_trunc(vector signed __int128 __vec,
18103		ptrdiff_t __offset,
18104		signed int *__ptr) {
	18105	static __inline__ __ATTRS_o_ai void vec_xst_trunc(vector signed __int128 __vec,
	18106	ptrdiff_t __offset,
	18107	signed int *__ptr) {
18105	18108	*(__ptr + __offset) = (signed int)__vec[0];
18106	18109	}
18107	18110
18108		static inline __ATTRS_o_ai void vec_xst_trunc(vector unsigned __int128 __vec,
18109		ptrdiff_t __offset,
18110		unsigned int *__ptr) {
	18111	static __inline__ __ATTRS_o_ai void
	18112	vec_xst_trunc(vector unsigned __int128 __vec, ptrdiff_t __offset,
	18113	unsigned int *__ptr) {
18111	18114	*(__ptr + __offset) = (unsigned int)__vec[0];
18112	18115	}
18113	18116
18114		static inline __ATTRS_o_ai void vec_xst_trunc(vector signed __int128 __vec,
18115		ptrdiff_t __offset,
18116		signed long long *__ptr) {
	18117	static __inline__ __ATTRS_o_ai void vec_xst_trunc(vector signed __int128 __vec,
	18118	ptrdiff_t __offset,
	18119	signed long long *__ptr) {
18117	18120	*(__ptr + __offset) = (signed long long)__vec[0];
18118	18121	}
18119	18122
18120		static inline __ATTRS_o_ai void vec_xst_trunc(vector unsigned __int128 __vec,
18121		ptrdiff_t __offset,
18122		unsigned long long *__ptr) {
	18123	static __inline__ __ATTRS_o_ai void
	18124	vec_xst_trunc(vector unsigned __int128 __vec, ptrdiff_t __offset,
	18125	unsigned long long *__ptr) {
18123	18126	*(__ptr + __offset) = (unsigned long long)__vec[0];
18124	18127	}
18125	18128	#endif

lib/include/amxavx512intrin.h+1-1

...	...	@@ -228,7 +228,7 @@
228	228	/// dst.byte[i] := a.row[row_index].byte[row_chunk+i]
229	229	/// ENDFOR
230	230	/// \endcode
231		#define _tile_movrow(a, b) __builtin_ia32_tilemovrow(a, b)
	231	#define _tile_movrow(a, b) ((__m512i)__builtin_ia32_tilemovrow(a, b))
232	232
233	233	/// This is internal intrinsic. C/C++ user should avoid calling it directly.
234	234

lib/include/amxcomplexintrin.h+4-6

...	...	@@ -135,9 +135,8 @@ _tile_cmmrlfp16ps_internal(unsigned short m, unsigned short n, unsigned short k,
135	135	/// The 1st source tile. Max size is 1024 Bytes.
136	136	/// \param src1
137	137	/// The 2nd source tile. Max size is 1024 Bytes.
138		__DEFAULT_FN_ATTRS_COMPLEX
139		static void __tile_cmmimfp16ps(__tile1024i *dst, __tile1024i src0,
140		__tile1024i src1) {
	138	static __inline__ void __DEFAULT_FN_ATTRS_COMPLEX
	139	__tile_cmmimfp16ps(__tile1024i *dst, __tile1024i src0, __tile1024i src1) {
141	140	dst->tile = _tile_cmmimfp16ps_internal(src0.row, src1.col, src0.col,
142	141	dst->tile, src0.tile, src1.tile);
143	142	}
...	...	@@ -158,9 +157,8 @@ static void __tile_cmmimfp16ps(__tile1024i *dst, __tile1024i src0,
158	157	/// The 1st source tile. Max size is 1024 Bytes.
159	158	/// \param src1
160	159	/// The 2nd source tile. Max size is 1024 Bytes.
161		__DEFAULT_FN_ATTRS_COMPLEX
162		static void __tile_cmmrlfp16ps(__tile1024i *dst, __tile1024i src0,
163		__tile1024i src1) {
	160	static __inline__ void __DEFAULT_FN_ATTRS_COMPLEX
	161	__tile_cmmrlfp16ps(__tile1024i *dst, __tile1024i src0, __tile1024i src1) {
164	162	dst->tile = _tile_cmmrlfp16ps_internal(src0.row, src1.col, src0.col,
165	163	dst->tile, src0.tile, src1.tile);
166	164	}

lib/include/amxmovrstransposeintrin.h+1-1

...	...	@@ -197,4 +197,4 @@ static void __tile_2rpntlvwz1rst1(__tile1024i dst0, __tile1024i dst1,
197	197
198	198	#undef __DEFAULT_FN_ATTRS
199	199	#endif /* __x86_64__ */
200		#endif /* __AMX_MOVRS_TRANSPOSEINTRIN_H */
		\ No newline at end of file
	200	#endif /* __AMX_MOVRS_TRANSPOSEINTRIN_H */

lib/include/amxtf32transposeintrin.h+1-1

...	...	@@ -8,7 +8,7 @@
8	8	*/
9	9	#ifndef __IMMINTRIN_H
10	10	#error \
11		"Never use <amxtf32tranposeintrin.h> directly; include <immintrin.h> instead."
	11	"Never use <amxtf32transposeintrin.h> directly; include <immintrin.h> instead."
12	12	#endif // __IMMINTRIN_H
13	13
14	14	#ifndef __AMX_TF32TRANSPOSEINTRIN_H

lib/include/andes_vector.h created+16

...	...	@@ -0,0 +1,16 @@
	1	//===----- andes_vector.h - Andes Vector definitions ----------------------===//
	2	//
	3	// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
	4	// See https://llvm.org/LICENSE.txt for license information.
	5	// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
	6	//
	7	//===----------------------------------------------------------------------===//
	8
	9	#ifndef _ANDES_VECTOR_H_
	10	#define _ANDES_VECTOR_H_
	11
	12	#include "riscv_vector.h"
	13
	14	#pragma clang riscv intrinsic andes_vector
	15
	16	#endif //_ANDES_VECTOR_H_

lib/include/arm_acle.h+11-41

...	...	@@ -29,47 +29,16 @@ extern "C" {
29	29
30	30	/* 7 SYNCHRONIZATION, BARRIER AND HINT INTRINSICS */
31	31	/* 7.3 Memory barriers */
32		#if !__has_builtin(__dmb)
33		#define __dmb(i) __builtin_arm_dmb(i)
34		#endif
35		#if !__has_builtin(__dsb)
36		#define __dsb(i) __builtin_arm_dsb(i)
37		#endif
38		#if !__has_builtin(__isb)
39		#define __isb(i) __builtin_arm_isb(i)
40		#endif
	32	void __dmb(unsigned int);
	33	void __dsb(unsigned int);
	34	void __isb(unsigned int);
41	35
42	36	/* 7.4 Hints */
43
44		#if !__has_builtin(__wfi)
45		static __inline__ void __attribute__((__always_inline__, __nodebug__)) __wfi(void) {
46		__builtin_arm_wfi();
47		}
48		#endif
49
50		#if !__has_builtin(__wfe)
51		static __inline__ void __attribute__((__always_inline__, __nodebug__)) __wfe(void) {
52		__builtin_arm_wfe();
53		}
54		#endif
55
56		#if !__has_builtin(__sev)
57		static __inline__ void __attribute__((__always_inline__, __nodebug__)) __sev(void) {
58		__builtin_arm_sev();
59		}
60		#endif
61
62		#if !__has_builtin(__sevl)
63		static __inline__ void __attribute__((__always_inline__, __nodebug__)) __sevl(void) {
64		__builtin_arm_sevl();
65		}
66		#endif
67
68		#if !__has_builtin(__yield)
69		static __inline__ void __attribute__((__always_inline__, __nodebug__)) __yield(void) {
70		__builtin_arm_yield();
71		}
72		#endif
	37	void __wfi(void);
	38	void __wfe(void);
	39	void __sev(void);
	40	void __sevl(void);
	41	void __yield(void);
73	42
74	43	#if defined(__ARM_32BIT_STATE) && __ARM_32BIT_STATE
75	44	#define __dbg(t) __builtin_arm_dbg(t)
...	...	@@ -872,8 +841,9 @@ __gcspopm() {
872	841	return __builtin_arm_gcspopm(0);
873	842	}
874	843
875		static __inline__ const void * __attribute__((__always_inline__, __nodebug__, target("gcs")))
876		__gcsss(const void *__stack) {
	844	static __inline__ void *__attribute__((__always_inline__, __nodebug__,
	845	target("gcs")))
	846	__gcsss(void *__stack) {
877	847	return __builtin_arm_gcsss(__stack);
878	848	}
879	849	#endif

lib/include/arm_fp16.h+89-89

...	...	@@ -34,408 +34,408 @@ typedef __fp16 float16_t;
34	34	float16_t __ret; \
35	35	float16_t __s0 = __p0; \
36	36	float16_t __s1 = __p1; \
37		__ret = (float16_t) __builtin_neon_vabdh_f16(__s0, __s1); \
	37	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vabdh_f16(__s0, __s1)); \
38	38	__ret; \
39	39	})
40	40	#define vabsh_f16(__p0) __extension__ ({ \
41	41	float16_t __ret; \
42	42	float16_t __s0 = __p0; \
43		__ret = (float16_t) __builtin_neon_vabsh_f16(__s0); \
	43	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vabsh_f16(__s0)); \
44	44	__ret; \
45	45	})
46	46	#define vaddh_f16(__p0, __p1) __extension__ ({ \
47	47	float16_t __ret; \
48	48	float16_t __s0 = __p0; \
49	49	float16_t __s1 = __p1; \
50		__ret = (float16_t) __builtin_neon_vaddh_f16(__s0, __s1); \
	50	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vaddh_f16(__s0, __s1)); \
51	51	__ret; \
52	52	})
53	53	#define vcageh_f16(__p0, __p1) __extension__ ({ \
54	54	uint16_t __ret; \
55	55	float16_t __s0 = __p0; \
56	56	float16_t __s1 = __p1; \
57		__ret = (uint16_t) __builtin_neon_vcageh_f16(__s0, __s1); \
	57	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vcageh_f16(__s0, __s1)); \
58	58	__ret; \
59	59	})
60	60	#define vcagth_f16(__p0, __p1) __extension__ ({ \
61	61	uint16_t __ret; \
62	62	float16_t __s0 = __p0; \
63	63	float16_t __s1 = __p1; \
64		__ret = (uint16_t) __builtin_neon_vcagth_f16(__s0, __s1); \
	64	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vcagth_f16(__s0, __s1)); \
65	65	__ret; \
66	66	})
67	67	#define vcaleh_f16(__p0, __p1) __extension__ ({ \
68	68	uint16_t __ret; \
69	69	float16_t __s0 = __p0; \
70	70	float16_t __s1 = __p1; \
71		__ret = (uint16_t) __builtin_neon_vcaleh_f16(__s0, __s1); \
	71	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vcaleh_f16(__s0, __s1)); \
72	72	__ret; \
73	73	})
74	74	#define vcalth_f16(__p0, __p1) __extension__ ({ \
75	75	uint16_t __ret; \
76	76	float16_t __s0 = __p0; \
77	77	float16_t __s1 = __p1; \
78		__ret = (uint16_t) __builtin_neon_vcalth_f16(__s0, __s1); \
	78	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vcalth_f16(__s0, __s1)); \
79	79	__ret; \
80	80	})
81	81	#define vceqh_f16(__p0, __p1) __extension__ ({ \
82	82	uint16_t __ret; \
83	83	float16_t __s0 = __p0; \
84	84	float16_t __s1 = __p1; \
85		__ret = (uint16_t) __builtin_neon_vceqh_f16(__s0, __s1); \
	85	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vceqh_f16(__s0, __s1)); \
86	86	__ret; \
87	87	})
88	88	#define vceqzh_f16(__p0) __extension__ ({ \
89	89	uint16_t __ret; \
90	90	float16_t __s0 = __p0; \
91		__ret = (uint16_t) __builtin_neon_vceqzh_f16(__s0); \
	91	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vceqzh_f16(__s0)); \
92	92	__ret; \
93	93	})
94	94	#define vcgeh_f16(__p0, __p1) __extension__ ({ \
95	95	uint16_t __ret; \
96	96	float16_t __s0 = __p0; \
97	97	float16_t __s1 = __p1; \
98		__ret = (uint16_t) __builtin_neon_vcgeh_f16(__s0, __s1); \
	98	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vcgeh_f16(__s0, __s1)); \
99	99	__ret; \
100	100	})
101	101	#define vcgezh_f16(__p0) __extension__ ({ \
102	102	uint16_t __ret; \
103	103	float16_t __s0 = __p0; \
104		__ret = (uint16_t) __builtin_neon_vcgezh_f16(__s0); \
	104	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vcgezh_f16(__s0)); \
105	105	__ret; \
106	106	})
107	107	#define vcgth_f16(__p0, __p1) __extension__ ({ \
108	108	uint16_t __ret; \
109	109	float16_t __s0 = __p0; \
110	110	float16_t __s1 = __p1; \
111		__ret = (uint16_t) __builtin_neon_vcgth_f16(__s0, __s1); \
	111	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vcgth_f16(__s0, __s1)); \
112	112	__ret; \
113	113	})
114	114	#define vcgtzh_f16(__p0) __extension__ ({ \
115	115	uint16_t __ret; \
116	116	float16_t __s0 = __p0; \
117		__ret = (uint16_t) __builtin_neon_vcgtzh_f16(__s0); \
	117	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vcgtzh_f16(__s0)); \
118	118	__ret; \
119	119	})
120	120	#define vcleh_f16(__p0, __p1) __extension__ ({ \
121	121	uint16_t __ret; \
122	122	float16_t __s0 = __p0; \
123	123	float16_t __s1 = __p1; \
124		__ret = (uint16_t) __builtin_neon_vcleh_f16(__s0, __s1); \
	124	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vcleh_f16(__s0, __s1)); \
125	125	__ret; \
126	126	})
127	127	#define vclezh_f16(__p0) __extension__ ({ \
128	128	uint16_t __ret; \
129	129	float16_t __s0 = __p0; \
130		__ret = (uint16_t) __builtin_neon_vclezh_f16(__s0); \
	130	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vclezh_f16(__s0)); \
131	131	__ret; \
132	132	})
133	133	#define vclth_f16(__p0, __p1) __extension__ ({ \
134	134	uint16_t __ret; \
135	135	float16_t __s0 = __p0; \
136	136	float16_t __s1 = __p1; \
137		__ret = (uint16_t) __builtin_neon_vclth_f16(__s0, __s1); \
	137	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vclth_f16(__s0, __s1)); \
138	138	__ret; \
139	139	})
140	140	#define vcltzh_f16(__p0) __extension__ ({ \
141	141	uint16_t __ret; \
142	142	float16_t __s0 = __p0; \
143		__ret = (uint16_t) __builtin_neon_vcltzh_f16(__s0); \
	143	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vcltzh_f16(__s0)); \
144	144	__ret; \
145	145	})
146	146	#define vcvth_n_s16_f16(__p0, __p1) __extension__ ({ \
147	147	int16_t __ret; \
148	148	float16_t __s0 = __p0; \
149		__ret = (int16_t) __builtin_neon_vcvth_n_s16_f16(__s0, __p1); \
	149	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vcvth_n_s16_f16(__s0, __p1)); \
150	150	__ret; \
151	151	})
152	152	#define vcvth_n_s32_f16(__p0, __p1) __extension__ ({ \
153	153	int32_t __ret; \
154	154	float16_t __s0 = __p0; \
155		__ret = (int32_t) __builtin_neon_vcvth_n_s32_f16(__s0, __p1); \
	155	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vcvth_n_s32_f16(__s0, __p1)); \
156	156	__ret; \
157	157	})
158	158	#define vcvth_n_s64_f16(__p0, __p1) __extension__ ({ \
159	159	int64_t __ret; \
160	160	float16_t __s0 = __p0; \
161		__ret = (int64_t) __builtin_neon_vcvth_n_s64_f16(__s0, __p1); \
	161	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vcvth_n_s64_f16(__s0, __p1)); \
162	162	__ret; \
163	163	})
164	164	#define vcvth_n_u16_f16(__p0, __p1) __extension__ ({ \
165	165	uint16_t __ret; \
166	166	float16_t __s0 = __p0; \
167		__ret = (uint16_t) __builtin_neon_vcvth_n_u16_f16(__s0, __p1); \
	167	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vcvth_n_u16_f16(__s0, __p1)); \
168	168	__ret; \
169	169	})
170	170	#define vcvth_n_u32_f16(__p0, __p1) __extension__ ({ \
171	171	uint32_t __ret; \
172	172	float16_t __s0 = __p0; \
173		__ret = (uint32_t) __builtin_neon_vcvth_n_u32_f16(__s0, __p1); \
	173	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vcvth_n_u32_f16(__s0, __p1)); \
174	174	__ret; \
175	175	})
176	176	#define vcvth_n_u64_f16(__p0, __p1) __extension__ ({ \
177	177	uint64_t __ret; \
178	178	float16_t __s0 = __p0; \
179		__ret = (uint64_t) __builtin_neon_vcvth_n_u64_f16(__s0, __p1); \
	179	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcvth_n_u64_f16(__s0, __p1)); \
180	180	__ret; \
181	181	})
182	182	#define vcvth_s16_f16(__p0) __extension__ ({ \
183	183	int16_t __ret; \
184	184	float16_t __s0 = __p0; \
185		__ret = (int16_t) __builtin_neon_vcvth_s16_f16(__s0); \
	185	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vcvth_s16_f16(__s0)); \
186	186	__ret; \
187	187	})
188	188	#define vcvth_s32_f16(__p0) __extension__ ({ \
189	189	int32_t __ret; \
190	190	float16_t __s0 = __p0; \
191		__ret = (int32_t) __builtin_neon_vcvth_s32_f16(__s0); \
	191	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vcvth_s32_f16(__s0)); \
192	192	__ret; \
193	193	})
194	194	#define vcvth_s64_f16(__p0) __extension__ ({ \
195	195	int64_t __ret; \
196	196	float16_t __s0 = __p0; \
197		__ret = (int64_t) __builtin_neon_vcvth_s64_f16(__s0); \
	197	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vcvth_s64_f16(__s0)); \
198	198	__ret; \
199	199	})
200	200	#define vcvth_u16_f16(__p0) __extension__ ({ \
201	201	uint16_t __ret; \
202	202	float16_t __s0 = __p0; \
203		__ret = (uint16_t) __builtin_neon_vcvth_u16_f16(__s0); \
	203	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vcvth_u16_f16(__s0)); \
204	204	__ret; \
205	205	})
206	206	#define vcvth_u32_f16(__p0) __extension__ ({ \
207	207	uint32_t __ret; \
208	208	float16_t __s0 = __p0; \
209		__ret = (uint32_t) __builtin_neon_vcvth_u32_f16(__s0); \
	209	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vcvth_u32_f16(__s0)); \
210	210	__ret; \
211	211	})
212	212	#define vcvth_u64_f16(__p0) __extension__ ({ \
213	213	uint64_t __ret; \
214	214	float16_t __s0 = __p0; \
215		__ret = (uint64_t) __builtin_neon_vcvth_u64_f16(__s0); \
	215	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcvth_u64_f16(__s0)); \
216	216	__ret; \
217	217	})
218	218	#define vcvtah_s16_f16(__p0) __extension__ ({ \
219	219	int16_t __ret; \
220	220	float16_t __s0 = __p0; \
221		__ret = (int16_t) __builtin_neon_vcvtah_s16_f16(__s0); \
	221	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vcvtah_s16_f16(__s0)); \
222	222	__ret; \
223	223	})
224	224	#define vcvtah_s32_f16(__p0) __extension__ ({ \
225	225	int32_t __ret; \
226	226	float16_t __s0 = __p0; \
227		__ret = (int32_t) __builtin_neon_vcvtah_s32_f16(__s0); \
	227	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vcvtah_s32_f16(__s0)); \
228	228	__ret; \
229	229	})
230	230	#define vcvtah_s64_f16(__p0) __extension__ ({ \
231	231	int64_t __ret; \
232	232	float16_t __s0 = __p0; \
233		__ret = (int64_t) __builtin_neon_vcvtah_s64_f16(__s0); \
	233	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vcvtah_s64_f16(__s0)); \
234	234	__ret; \
235	235	})
236	236	#define vcvtah_u16_f16(__p0) __extension__ ({ \
237	237	uint16_t __ret; \
238	238	float16_t __s0 = __p0; \
239		__ret = (uint16_t) __builtin_neon_vcvtah_u16_f16(__s0); \
	239	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vcvtah_u16_f16(__s0)); \
240	240	__ret; \
241	241	})
242	242	#define vcvtah_u32_f16(__p0) __extension__ ({ \
243	243	uint32_t __ret; \
244	244	float16_t __s0 = __p0; \
245		__ret = (uint32_t) __builtin_neon_vcvtah_u32_f16(__s0); \
	245	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vcvtah_u32_f16(__s0)); \
246	246	__ret; \
247	247	})
248	248	#define vcvtah_u64_f16(__p0) __extension__ ({ \
249	249	uint64_t __ret; \
250	250	float16_t __s0 = __p0; \
251		__ret = (uint64_t) __builtin_neon_vcvtah_u64_f16(__s0); \
	251	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcvtah_u64_f16(__s0)); \
252	252	__ret; \
253	253	})
254	254	#define vcvth_f16_u16(__p0) __extension__ ({ \
255	255	float16_t __ret; \
256	256	uint16_t __s0 = __p0; \
257		__ret = (float16_t) __builtin_neon_vcvth_f16_u16(__s0); \
	257	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vcvth_f16_u16(__s0)); \
258	258	__ret; \
259	259	})
260	260	#define vcvth_f16_s16(__p0) __extension__ ({ \
261	261	float16_t __ret; \
262	262	int16_t __s0 = __p0; \
263		__ret = (float16_t) __builtin_neon_vcvth_f16_s16(__s0); \
	263	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vcvth_f16_s16(__s0)); \
264	264	__ret; \
265	265	})
266	266	#define vcvth_f16_u32(__p0) __extension__ ({ \
267	267	float16_t __ret; \
268	268	uint32_t __s0 = __p0; \
269		__ret = (float16_t) __builtin_neon_vcvth_f16_u32(__s0); \
	269	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vcvth_f16_u32(__s0)); \
270	270	__ret; \
271	271	})
272	272	#define vcvth_f16_s32(__p0) __extension__ ({ \
273	273	float16_t __ret; \
274	274	int32_t __s0 = __p0; \
275		__ret = (float16_t) __builtin_neon_vcvth_f16_s32(__s0); \
	275	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vcvth_f16_s32(__s0)); \
276	276	__ret; \
277	277	})
278	278	#define vcvth_f16_u64(__p0) __extension__ ({ \
279	279	float16_t __ret; \
280	280	uint64_t __s0 = __p0; \
281		__ret = (float16_t) __builtin_neon_vcvth_f16_u64(__s0); \
	281	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vcvth_f16_u64(__s0)); \
282	282	__ret; \
283	283	})
284	284	#define vcvth_f16_s64(__p0) __extension__ ({ \
285	285	float16_t __ret; \
286	286	int64_t __s0 = __p0; \
287		__ret = (float16_t) __builtin_neon_vcvth_f16_s64(__s0); \
	287	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vcvth_f16_s64(__s0)); \
288	288	__ret; \
289	289	})
290	290	#define vcvth_n_f16_u32(__p0, __p1) __extension__ ({ \
291	291	float16_t __ret; \
292	292	uint32_t __s0 = __p0; \
293		__ret = (float16_t) __builtin_neon_vcvth_n_f16_u32(__s0, __p1); \
	293	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vcvth_n_f16_u32(__s0, __p1)); \
294	294	__ret; \
295	295	})
296	296	#define vcvth_n_f16_s32(__p0, __p1) __extension__ ({ \
297	297	float16_t __ret; \
298	298	int32_t __s0 = __p0; \
299		__ret = (float16_t) __builtin_neon_vcvth_n_f16_s32(__s0, __p1); \
	299	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vcvth_n_f16_s32(__s0, __p1)); \
300	300	__ret; \
301	301	})
302	302	#define vcvth_n_f16_u64(__p0, __p1) __extension__ ({ \
303	303	float16_t __ret; \
304	304	uint64_t __s0 = __p0; \
305		__ret = (float16_t) __builtin_neon_vcvth_n_f16_u64(__s0, __p1); \
	305	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vcvth_n_f16_u64(__s0, __p1)); \
306	306	__ret; \
307	307	})
308	308	#define vcvth_n_f16_s64(__p0, __p1) __extension__ ({ \
309	309	float16_t __ret; \
310	310	int64_t __s0 = __p0; \
311		__ret = (float16_t) __builtin_neon_vcvth_n_f16_s64(__s0, __p1); \
	311	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vcvth_n_f16_s64(__s0, __p1)); \
312	312	__ret; \
313	313	})
314	314	#define vcvth_n_f16_u16(__p0, __p1) __extension__ ({ \
315	315	float16_t __ret; \
316	316	uint16_t __s0 = __p0; \
317		__ret = (float16_t) __builtin_neon_vcvth_n_f16_u16(__s0, __p1); \
	317	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vcvth_n_f16_u16(__s0, __p1)); \
318	318	__ret; \
319	319	})
320	320	#define vcvth_n_f16_s16(__p0, __p1) __extension__ ({ \
321	321	float16_t __ret; \
322	322	int16_t __s0 = __p0; \
323		__ret = (float16_t) __builtin_neon_vcvth_n_f16_s16(__s0, __p1); \
	323	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vcvth_n_f16_s16(__s0, __p1)); \
324	324	__ret; \
325	325	})
326	326	#define vcvtmh_s16_f16(__p0) __extension__ ({ \
327	327	int16_t __ret; \
328	328	float16_t __s0 = __p0; \
329		__ret = (int16_t) __builtin_neon_vcvtmh_s16_f16(__s0); \
	329	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vcvtmh_s16_f16(__s0)); \
330	330	__ret; \
331	331	})
332	332	#define vcvtmh_s32_f16(__p0) __extension__ ({ \
333	333	int32_t __ret; \
334	334	float16_t __s0 = __p0; \
335		__ret = (int32_t) __builtin_neon_vcvtmh_s32_f16(__s0); \
	335	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vcvtmh_s32_f16(__s0)); \
336	336	__ret; \
337	337	})
338	338	#define vcvtmh_s64_f16(__p0) __extension__ ({ \
339	339	int64_t __ret; \
340	340	float16_t __s0 = __p0; \
341		__ret = (int64_t) __builtin_neon_vcvtmh_s64_f16(__s0); \
	341	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vcvtmh_s64_f16(__s0)); \
342	342	__ret; \
343	343	})
344	344	#define vcvtmh_u16_f16(__p0) __extension__ ({ \
345	345	uint16_t __ret; \
346	346	float16_t __s0 = __p0; \
347		__ret = (uint16_t) __builtin_neon_vcvtmh_u16_f16(__s0); \
	347	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vcvtmh_u16_f16(__s0)); \
348	348	__ret; \
349	349	})
350	350	#define vcvtmh_u32_f16(__p0) __extension__ ({ \
351	351	uint32_t __ret; \
352	352	float16_t __s0 = __p0; \
353		__ret = (uint32_t) __builtin_neon_vcvtmh_u32_f16(__s0); \
	353	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vcvtmh_u32_f16(__s0)); \
354	354	__ret; \
355	355	})
356	356	#define vcvtmh_u64_f16(__p0) __extension__ ({ \
357	357	uint64_t __ret; \
358	358	float16_t __s0 = __p0; \
359		__ret = (uint64_t) __builtin_neon_vcvtmh_u64_f16(__s0); \
	359	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcvtmh_u64_f16(__s0)); \
360	360	__ret; \
361	361	})
362	362	#define vcvtnh_s16_f16(__p0) __extension__ ({ \
363	363	int16_t __ret; \
364	364	float16_t __s0 = __p0; \
365		__ret = (int16_t) __builtin_neon_vcvtnh_s16_f16(__s0); \
	365	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vcvtnh_s16_f16(__s0)); \
366	366	__ret; \
367	367	})
368	368	#define vcvtnh_s32_f16(__p0) __extension__ ({ \
369	369	int32_t __ret; \
370	370	float16_t __s0 = __p0; \
371		__ret = (int32_t) __builtin_neon_vcvtnh_s32_f16(__s0); \
	371	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vcvtnh_s32_f16(__s0)); \
372	372	__ret; \
373	373	})
374	374	#define vcvtnh_s64_f16(__p0) __extension__ ({ \
375	375	int64_t __ret; \
376	376	float16_t __s0 = __p0; \
377		__ret = (int64_t) __builtin_neon_vcvtnh_s64_f16(__s0); \
	377	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vcvtnh_s64_f16(__s0)); \
378	378	__ret; \
379	379	})
380	380	#define vcvtnh_u16_f16(__p0) __extension__ ({ \
381	381	uint16_t __ret; \
382	382	float16_t __s0 = __p0; \
383		__ret = (uint16_t) __builtin_neon_vcvtnh_u16_f16(__s0); \
	383	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vcvtnh_u16_f16(__s0)); \
384	384	__ret; \
385	385	})
386	386	#define vcvtnh_u32_f16(__p0) __extension__ ({ \
387	387	uint32_t __ret; \
388	388	float16_t __s0 = __p0; \
389		__ret = (uint32_t) __builtin_neon_vcvtnh_u32_f16(__s0); \
	389	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vcvtnh_u32_f16(__s0)); \
390	390	__ret; \
391	391	})
392	392	#define vcvtnh_u64_f16(__p0) __extension__ ({ \
393	393	uint64_t __ret; \
394	394	float16_t __s0 = __p0; \
395		__ret = (uint64_t) __builtin_neon_vcvtnh_u64_f16(__s0); \
	395	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcvtnh_u64_f16(__s0)); \
396	396	__ret; \
397	397	})
398	398	#define vcvtph_s16_f16(__p0) __extension__ ({ \
399	399	int16_t __ret; \
400	400	float16_t __s0 = __p0; \
401		__ret = (int16_t) __builtin_neon_vcvtph_s16_f16(__s0); \
	401	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vcvtph_s16_f16(__s0)); \
402	402	__ret; \
403	403	})
404	404	#define vcvtph_s32_f16(__p0) __extension__ ({ \
405	405	int32_t __ret; \
406	406	float16_t __s0 = __p0; \
407		__ret = (int32_t) __builtin_neon_vcvtph_s32_f16(__s0); \
	407	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vcvtph_s32_f16(__s0)); \
408	408	__ret; \
409	409	})
410	410	#define vcvtph_s64_f16(__p0) __extension__ ({ \
411	411	int64_t __ret; \
412	412	float16_t __s0 = __p0; \
413		__ret = (int64_t) __builtin_neon_vcvtph_s64_f16(__s0); \
	413	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vcvtph_s64_f16(__s0)); \
414	414	__ret; \
415	415	})
416	416	#define vcvtph_u16_f16(__p0) __extension__ ({ \
417	417	uint16_t __ret; \
418	418	float16_t __s0 = __p0; \
419		__ret = (uint16_t) __builtin_neon_vcvtph_u16_f16(__s0); \
	419	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vcvtph_u16_f16(__s0)); \
420	420	__ret; \
421	421	})
422	422	#define vcvtph_u32_f16(__p0) __extension__ ({ \
423	423	uint32_t __ret; \
424	424	float16_t __s0 = __p0; \
425		__ret = (uint32_t) __builtin_neon_vcvtph_u32_f16(__s0); \
	425	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vcvtph_u32_f16(__s0)); \
426	426	__ret; \
427	427	})
428	428	#define vcvtph_u64_f16(__p0) __extension__ ({ \
429	429	uint64_t __ret; \
430	430	float16_t __s0 = __p0; \
431		__ret = (uint64_t) __builtin_neon_vcvtph_u64_f16(__s0); \
	431	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcvtph_u64_f16(__s0)); \
432	432	__ret; \
433	433	})
434	434	#define vdivh_f16(__p0, __p1) __extension__ ({ \
435	435	float16_t __ret; \
436	436	float16_t __s0 = __p0; \
437	437	float16_t __s1 = __p1; \
438		__ret = (float16_t) __builtin_neon_vdivh_f16(__s0, __s1); \
	438	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vdivh_f16(__s0, __s1)); \
439	439	__ret; \
440	440	})
441	441	#define vfmah_f16(__p0, __p1, __p2) __extension__ ({ \
...	...	@@ -443,7 +443,7 @@ typedef __fp16 float16_t;
443	443	float16_t __s0 = __p0; \
444	444	float16_t __s1 = __p1; \
445	445	float16_t __s2 = __p2; \
446		__ret = (float16_t) __builtin_neon_vfmah_f16(__s0, __s1, __s2); \
	446	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vfmah_f16(__s0, __s1, __s2)); \
447	447	__ret; \
448	448	})
449	449	#define vfmsh_f16(__p0, __p1, __p2) __extension__ ({ \
...	...	@@ -451,142 +451,142 @@ typedef __fp16 float16_t;
451	451	float16_t __s0 = __p0; \
452	452	float16_t __s1 = __p1; \
453	453	float16_t __s2 = __p2; \
454		__ret = (float16_t) __builtin_neon_vfmsh_f16(__s0, __s1, __s2); \
	454	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vfmsh_f16(__s0, __s1, __s2)); \
455	455	__ret; \
456	456	})
457	457	#define vmaxh_f16(__p0, __p1) __extension__ ({ \
458	458	float16_t __ret; \
459	459	float16_t __s0 = __p0; \
460	460	float16_t __s1 = __p1; \
461		__ret = (float16_t) __builtin_neon_vmaxh_f16(__s0, __s1); \
	461	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vmaxh_f16(__s0, __s1)); \
462	462	__ret; \
463	463	})
464	464	#define vmaxnmh_f16(__p0, __p1) __extension__ ({ \
465	465	float16_t __ret; \
466	466	float16_t __s0 = __p0; \
467	467	float16_t __s1 = __p1; \
468		__ret = (float16_t) __builtin_neon_vmaxnmh_f16(__s0, __s1); \
	468	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vmaxnmh_f16(__s0, __s1)); \
469	469	__ret; \
470	470	})
471	471	#define vminh_f16(__p0, __p1) __extension__ ({ \
472	472	float16_t __ret; \
473	473	float16_t __s0 = __p0; \
474	474	float16_t __s1 = __p1; \
475		__ret = (float16_t) __builtin_neon_vminh_f16(__s0, __s1); \
	475	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vminh_f16(__s0, __s1)); \
476	476	__ret; \
477	477	})
478	478	#define vminnmh_f16(__p0, __p1) __extension__ ({ \
479	479	float16_t __ret; \
480	480	float16_t __s0 = __p0; \
481	481	float16_t __s1 = __p1; \
482		__ret = (float16_t) __builtin_neon_vminnmh_f16(__s0, __s1); \
	482	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vminnmh_f16(__s0, __s1)); \
483	483	__ret; \
484	484	})
485	485	#define vmulh_f16(__p0, __p1) __extension__ ({ \
486	486	float16_t __ret; \
487	487	float16_t __s0 = __p0; \
488	488	float16_t __s1 = __p1; \
489		__ret = (float16_t) __builtin_neon_vmulh_f16(__s0, __s1); \
	489	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vmulh_f16(__s0, __s1)); \
490	490	__ret; \
491	491	})
492	492	#define vmulxh_f16(__p0, __p1) __extension__ ({ \
493	493	float16_t __ret; \
494	494	float16_t __s0 = __p0; \
495	495	float16_t __s1 = __p1; \
496		__ret = (float16_t) __builtin_neon_vmulxh_f16(__s0, __s1); \
	496	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vmulxh_f16(__s0, __s1)); \
497	497	__ret; \
498	498	})
499	499	#define vnegh_f16(__p0) __extension__ ({ \
500	500	float16_t __ret; \
501	501	float16_t __s0 = __p0; \
502		__ret = (float16_t) __builtin_neon_vnegh_f16(__s0); \
	502	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vnegh_f16(__s0)); \
503	503	__ret; \
504	504	})
505	505	#define vrecpeh_f16(__p0) __extension__ ({ \
506	506	float16_t __ret; \
507	507	float16_t __s0 = __p0; \
508		__ret = (float16_t) __builtin_neon_vrecpeh_f16(__s0); \
	508	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vrecpeh_f16(__s0)); \
509	509	__ret; \
510	510	})
511	511	#define vrecpsh_f16(__p0, __p1) __extension__ ({ \
512	512	float16_t __ret; \
513	513	float16_t __s0 = __p0; \
514	514	float16_t __s1 = __p1; \
515		__ret = (float16_t) __builtin_neon_vrecpsh_f16(__s0, __s1); \
	515	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vrecpsh_f16(__s0, __s1)); \
516	516	__ret; \
517	517	})
518	518	#define vrecpxh_f16(__p0) __extension__ ({ \
519	519	float16_t __ret; \
520	520	float16_t __s0 = __p0; \
521		__ret = (float16_t) __builtin_neon_vrecpxh_f16(__s0); \
	521	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vrecpxh_f16(__s0)); \
522	522	__ret; \
523	523	})
524	524	#define vrndh_f16(__p0) __extension__ ({ \
525	525	float16_t __ret; \
526	526	float16_t __s0 = __p0; \
527		__ret = (float16_t) __builtin_neon_vrndh_f16(__s0); \
	527	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vrndh_f16(__s0)); \
528	528	__ret; \
529	529	})
530	530	#define vrndah_f16(__p0) __extension__ ({ \
531	531	float16_t __ret; \
532	532	float16_t __s0 = __p0; \
533		__ret = (float16_t) __builtin_neon_vrndah_f16(__s0); \
	533	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vrndah_f16(__s0)); \
534	534	__ret; \
535	535	})
536	536	#define vrndih_f16(__p0) __extension__ ({ \
537	537	float16_t __ret; \
538	538	float16_t __s0 = __p0; \
539		__ret = (float16_t) __builtin_neon_vrndih_f16(__s0); \
	539	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vrndih_f16(__s0)); \
540	540	__ret; \
541	541	})
542	542	#define vrndmh_f16(__p0) __extension__ ({ \
543	543	float16_t __ret; \
544	544	float16_t __s0 = __p0; \
545		__ret = (float16_t) __builtin_neon_vrndmh_f16(__s0); \
	545	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vrndmh_f16(__s0)); \
546	546	__ret; \
547	547	})
548	548	#define vrndnh_f16(__p0) __extension__ ({ \
549	549	float16_t __ret; \
550	550	float16_t __s0 = __p0; \
551		__ret = (float16_t) __builtin_neon_vrndnh_f16(__s0); \
	551	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vrndnh_f16(__s0)); \
552	552	__ret; \
553	553	})
554	554	#define vrndph_f16(__p0) __extension__ ({ \
555	555	float16_t __ret; \
556	556	float16_t __s0 = __p0; \
557		__ret = (float16_t) __builtin_neon_vrndph_f16(__s0); \
	557	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vrndph_f16(__s0)); \
558	558	__ret; \
559	559	})
560	560	#define vrndxh_f16(__p0) __extension__ ({ \
561	561	float16_t __ret; \
562	562	float16_t __s0 = __p0; \
563		__ret = (float16_t) __builtin_neon_vrndxh_f16(__s0); \
	563	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vrndxh_f16(__s0)); \
564	564	__ret; \
565	565	})
566	566	#define vrsqrteh_f16(__p0) __extension__ ({ \
567	567	float16_t __ret; \
568	568	float16_t __s0 = __p0; \
569		__ret = (float16_t) __builtin_neon_vrsqrteh_f16(__s0); \
	569	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vrsqrteh_f16(__s0)); \
570	570	__ret; \
571	571	})
572	572	#define vrsqrtsh_f16(__p0, __p1) __extension__ ({ \
573	573	float16_t __ret; \
574	574	float16_t __s0 = __p0; \
575	575	float16_t __s1 = __p1; \
576		__ret = (float16_t) __builtin_neon_vrsqrtsh_f16(__s0, __s1); \
	576	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vrsqrtsh_f16(__s0, __s1)); \
577	577	__ret; \
578	578	})
579	579	#define vsqrth_f16(__p0) __extension__ ({ \
580	580	float16_t __ret; \
581	581	float16_t __s0 = __p0; \
582		__ret = (float16_t) __builtin_neon_vsqrth_f16(__s0); \
	582	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vsqrth_f16(__s0)); \
583	583	__ret; \
584	584	})
585	585	#define vsubh_f16(__p0, __p1) __extension__ ({ \
586	586	float16_t __ret; \
587	587	float16_t __s0 = __p0; \
588	588	float16_t __s1 = __p1; \
589		__ret = (float16_t) __builtin_neon_vsubh_f16(__s0, __s1); \
	589	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vsubh_f16(__s0, __s1)); \
590	590	__ret; \
591	591	})
592	592	#endif

lib/include/arm_neon.h+22139-19621

...	...	@@ -24,7 +24,9 @@
24	24	#ifndef __ARM_NEON_H
25	25	#define __ARM_NEON_H
26	26
27		#ifndef __ARM_FP
	27	#if !defined(__arm__) && !defined(__aarch64__) && !defined(__arm64ec__)
	28	#error "<arm_neon.h> is intended only for ARM and AArch64 targets"
	29	#elif !defined(__ARM_FP)
28	30	#error "NEON intrinsics not available with the soft-float ABI. Please use -mfloat-abi=softfp or -mfloat-abi=hard"
29	31	#else
30	32
...	...	@@ -123,26 +125,45 @@ typedef struct poly64x2x4_t {
123	125
124	126	#define __ai static __inline__ __attribute__((__always_inline__, __nodebug__))
125	127
	128	#if !defined(__LITTLE_ENDIAN__)
	129	#if defined(__aarch64__) \|\| defined(__arm64ec__)
	130	#define __lane_reverse_64_32 1,0
	131	#define __lane_reverse_64_16 3,2,1,0
	132	#define __lane_reverse_64_8 7,6,5,4,3,2,1,0
	133	#define __lane_reverse_128_64 1,0
	134	#define __lane_reverse_128_32 3,2,1,0
	135	#define __lane_reverse_128_16 7,6,5,4,3,2,1,0
	136	#define __lane_reverse_128_8 15,14,13,12,11,10,9,8,7,6,5,4,3,2,1,0
	137	#else
	138	#define __lane_reverse_64_32 1,0
	139	#define __lane_reverse_64_16 3,2,1,0
	140	#define __lane_reverse_64_8 7,6,5,4,3,2,1,0
	141	#define __lane_reverse_128_64 0,1
	142	#define __lane_reverse_128_32 1,0,3,2
	143	#define __lane_reverse_128_16 3,2,1,0,7,6,5,4
	144	#define __lane_reverse_128_8 7,6,5,4,3,2,1,0,15,14,13,12,11,10,9,8
	145	#endif
	146	#endif
126	147	#ifdef __LITTLE_ENDIAN__
127	148	#define splatq_lane_bf16(__p0, __p1) __extension__ ({ \
128	149	bfloat16x8_t __ret; \
129	150	bfloat16x4_t __s0 = __p0; \
130		__ret = (bfloat16x8_t) __builtin_neon_splatq_lane_bf16((int8x8_t)__s0, __p1, 11); \
	151	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_splatq_lane_bf16(__builtin_bit_cast(int8x8_t, __s0), __p1, 11)); \
131	152	__ret; \
132	153	})
133	154	#else
134	155	#define splatq_lane_bf16(__p0, __p1) __extension__ ({ \
135	156	bfloat16x8_t __ret; \
136	157	bfloat16x4_t __s0 = __p0; \
137		bfloat16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
138		__ret = (bfloat16x8_t) __builtin_neon_splatq_lane_bf16((int8x8_t)__rev0, __p1, 11); \
139		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	158	bfloat16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	159	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_splatq_lane_bf16(__builtin_bit_cast(int8x8_t, __rev0), __p1, 11)); \
	160	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
140	161	__ret; \
141	162	})
142	163	#define __noswap_splatq_lane_bf16(__p0, __p1) __extension__ ({ \
143	164	bfloat16x8_t __ret; \
144	165	bfloat16x4_t __s0 = __p0; \
145		__ret = (bfloat16x8_t) __builtin_neon_splatq_lane_bf16((int8x8_t)__s0, __p1, 11); \
	166	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_splatq_lane_bf16(__builtin_bit_cast(int8x8_t, __s0), __p1, 11)); \
146	167	__ret; \
147	168	})
148	169	#endif
...	...	@@ -151,22 +172,22 @@ typedef struct poly64x2x4_t {
151	172	#define splat_lane_bf16(__p0, __p1) __extension__ ({ \
152	173	bfloat16x4_t __ret; \
153	174	bfloat16x4_t __s0 = __p0; \
154		__ret = (bfloat16x4_t) __builtin_neon_splat_lane_bf16((int8x8_t)__s0, __p1, 11); \
	175	__ret = __builtin_bit_cast(bfloat16x4_t, __builtin_neon_splat_lane_bf16(__builtin_bit_cast(int8x8_t, __s0), __p1, 11)); \
155	176	__ret; \
156	177	})
157	178	#else
158	179	#define splat_lane_bf16(__p0, __p1) __extension__ ({ \
159	180	bfloat16x4_t __ret; \
160	181	bfloat16x4_t __s0 = __p0; \
161		bfloat16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
162		__ret = (bfloat16x4_t) __builtin_neon_splat_lane_bf16((int8x8_t)__rev0, __p1, 11); \
163		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	182	bfloat16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	183	__ret = __builtin_bit_cast(bfloat16x4_t, __builtin_neon_splat_lane_bf16(__builtin_bit_cast(int8x8_t, __rev0), __p1, 11)); \
	184	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
164	185	__ret; \
165	186	})
166	187	#define __noswap_splat_lane_bf16(__p0, __p1) __extension__ ({ \
167	188	bfloat16x4_t __ret; \
168	189	bfloat16x4_t __s0 = __p0; \
169		__ret = (bfloat16x4_t) __builtin_neon_splat_lane_bf16((int8x8_t)__s0, __p1, 11); \
	190	__ret = __builtin_bit_cast(bfloat16x4_t, __builtin_neon_splat_lane_bf16(__builtin_bit_cast(int8x8_t, __s0), __p1, 11)); \
170	191	__ret; \
171	192	})
172	193	#endif
...	...	@@ -175,22 +196,22 @@ typedef struct poly64x2x4_t {
175	196	#define splatq_laneq_bf16(__p0, __p1) __extension__ ({ \
176	197	bfloat16x8_t __ret; \
177	198	bfloat16x8_t __s0 = __p0; \
178		__ret = (bfloat16x8_t) __builtin_neon_splatq_laneq_bf16((int8x16_t)__s0, __p1, 43); \
	199	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_splatq_laneq_bf16(__builtin_bit_cast(int8x16_t, __s0), __p1, 43)); \
179	200	__ret; \
180	201	})
181	202	#else
182	203	#define splatq_laneq_bf16(__p0, __p1) __extension__ ({ \
183	204	bfloat16x8_t __ret; \
184	205	bfloat16x8_t __s0 = __p0; \
185		bfloat16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
186		__ret = (bfloat16x8_t) __builtin_neon_splatq_laneq_bf16((int8x16_t)__rev0, __p1, 43); \
187		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	206	bfloat16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	207	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_splatq_laneq_bf16(__builtin_bit_cast(int8x16_t, __rev0), __p1, 43)); \
	208	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
188	209	__ret; \
189	210	})
190	211	#define __noswap_splatq_laneq_bf16(__p0, __p1) __extension__ ({ \
191	212	bfloat16x8_t __ret; \
192	213	bfloat16x8_t __s0 = __p0; \
193		__ret = (bfloat16x8_t) __builtin_neon_splatq_laneq_bf16((int8x16_t)__s0, __p1, 43); \
	214	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_splatq_laneq_bf16(__builtin_bit_cast(int8x16_t, __s0), __p1, 43)); \
194	215	__ret; \
195	216	})
196	217	#endif
...	...	@@ -199,22 +220,22 @@ typedef struct poly64x2x4_t {
199	220	#define splat_laneq_bf16(__p0, __p1) __extension__ ({ \
200	221	bfloat16x4_t __ret; \
201	222	bfloat16x8_t __s0 = __p0; \
202		__ret = (bfloat16x4_t) __builtin_neon_splat_laneq_bf16((int8x16_t)__s0, __p1, 43); \
	223	__ret = __builtin_bit_cast(bfloat16x4_t, __builtin_neon_splat_laneq_bf16(__builtin_bit_cast(int8x16_t, __s0), __p1, 43)); \
203	224	__ret; \
204	225	})
205	226	#else
206	227	#define splat_laneq_bf16(__p0, __p1) __extension__ ({ \
207	228	bfloat16x4_t __ret; \
208	229	bfloat16x8_t __s0 = __p0; \
209		bfloat16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
210		__ret = (bfloat16x4_t) __builtin_neon_splat_laneq_bf16((int8x16_t)__rev0, __p1, 43); \
211		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	230	bfloat16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	231	__ret = __builtin_bit_cast(bfloat16x4_t, __builtin_neon_splat_laneq_bf16(__builtin_bit_cast(int8x16_t, __rev0), __p1, 43)); \
	232	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
212	233	__ret; \
213	234	})
214	235	#define __noswap_splat_laneq_bf16(__p0, __p1) __extension__ ({ \
215	236	bfloat16x4_t __ret; \
216	237	bfloat16x8_t __s0 = __p0; \
217		__ret = (bfloat16x4_t) __builtin_neon_splat_laneq_bf16((int8x16_t)__s0, __p1, 43); \
	238	__ret = __builtin_bit_cast(bfloat16x4_t, __builtin_neon_splat_laneq_bf16(__builtin_bit_cast(int8x16_t, __s0), __p1, 43)); \
218	239	__ret; \
219	240	})
220	241	#endif
...	...	@@ -222,22 +243,22 @@ typedef struct poly64x2x4_t {
222	243	#ifdef __LITTLE_ENDIAN__
223	244	__ai __attribute__((target("bf16,neon"))) float32x4_t vbfdotq_f32(float32x4_t __p0, bfloat16x8_t __p1, bfloat16x8_t __p2) {
224	245	float32x4_t __ret;
225		__ret = (float32x4_t) __builtin_neon_vbfdotq_f32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
	246	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vbfdotq_f32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 41));
226	247	return __ret;
227	248	}
228	249	#else
229	250	__ai __attribute__((target("bf16,neon"))) float32x4_t vbfdotq_f32(float32x4_t __p0, bfloat16x8_t __p1, bfloat16x8_t __p2) {
230	251	float32x4_t __ret;
231		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
232		bfloat16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
233		bfloat16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
234		__ret = (float32x4_t) __builtin_neon_vbfdotq_f32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 41);
235		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	252	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	253	bfloat16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	254	bfloat16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
	255	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vbfdotq_f32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 41));
	256	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
236	257	return __ret;
237	258	}
238	259	__ai __attribute__((target("bf16,neon"))) float32x4_t __noswap_vbfdotq_f32(float32x4_t __p0, bfloat16x8_t __p1, bfloat16x8_t __p2) {
239	260	float32x4_t __ret;
240		__ret = (float32x4_t) __builtin_neon_vbfdotq_f32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
	261	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vbfdotq_f32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 41));
241	262	return __ret;
242	263	}
243	264	#endif
...	...	@@ -245,22 +266,22 @@ __ai __attribute__((target("bf16,neon"))) float32x4_t __noswap_vbfdotq_f32(float
245	266	#ifdef __LITTLE_ENDIAN__
246	267	__ai __attribute__((target("bf16,neon"))) float32x2_t vbfdot_f32(float32x2_t __p0, bfloat16x4_t __p1, bfloat16x4_t __p2) {
247	268	float32x2_t __ret;
248		__ret = (float32x2_t) __builtin_neon_vbfdot_f32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
	269	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vbfdot_f32(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 9));
249	270	return __ret;
250	271	}
251	272	#else
252	273	__ai __attribute__((target("bf16,neon"))) float32x2_t vbfdot_f32(float32x2_t __p0, bfloat16x4_t __p1, bfloat16x4_t __p2) {
253	274	float32x2_t __ret;
254		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
255		bfloat16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
256		bfloat16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
257		__ret = (float32x2_t) __builtin_neon_vbfdot_f32((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 9);
258		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	275	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	276	bfloat16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	277	bfloat16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
	278	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vbfdot_f32(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 9));
	279	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
259	280	return __ret;
260	281	}
261	282	__ai __attribute__((target("bf16,neon"))) float32x2_t __noswap_vbfdot_f32(float32x2_t __p0, bfloat16x4_t __p1, bfloat16x4_t __p2) {
262	283	float32x2_t __ret;
263		__ret = (float32x2_t) __builtin_neon_vbfdot_f32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
	284	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vbfdot_f32(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 9));
264	285	return __ret;
265	286	}
266	287	#endif
...	...	@@ -268,22 +289,22 @@ __ai __attribute__((target("bf16,neon"))) float32x2_t __noswap_vbfdot_f32(float3
268	289	#ifdef __LITTLE_ENDIAN__
269	290	__ai __attribute__((target("bf16,neon"))) float32x4_t vbfmlalbq_f32(float32x4_t __p0, bfloat16x8_t __p1, bfloat16x8_t __p2) {
270	291	float32x4_t __ret;
271		__ret = (float32x4_t) __builtin_neon_vbfmlalbq_f32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
	292	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vbfmlalbq_f32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 41));
272	293	return __ret;
273	294	}
274	295	#else
275	296	__ai __attribute__((target("bf16,neon"))) float32x4_t vbfmlalbq_f32(float32x4_t __p0, bfloat16x8_t __p1, bfloat16x8_t __p2) {
276	297	float32x4_t __ret;
277		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
278		bfloat16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
279		bfloat16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
280		__ret = (float32x4_t) __builtin_neon_vbfmlalbq_f32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 41);
281		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	298	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	299	bfloat16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	300	bfloat16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
	301	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vbfmlalbq_f32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 41));
	302	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
282	303	return __ret;
283	304	}
284	305	__ai __attribute__((target("bf16,neon"))) float32x4_t __noswap_vbfmlalbq_f32(float32x4_t __p0, bfloat16x8_t __p1, bfloat16x8_t __p2) {
285	306	float32x4_t __ret;
286		__ret = (float32x4_t) __builtin_neon_vbfmlalbq_f32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
	307	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vbfmlalbq_f32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 41));
287	308	return __ret;
288	309	}
289	310	#endif
...	...	@@ -291,22 +312,22 @@ __ai __attribute__((target("bf16,neon"))) float32x4_t __noswap_vbfmlalbq_f32(flo
291	312	#ifdef __LITTLE_ENDIAN__
292	313	__ai __attribute__((target("bf16,neon"))) float32x4_t vbfmlaltq_f32(float32x4_t __p0, bfloat16x8_t __p1, bfloat16x8_t __p2) {
293	314	float32x4_t __ret;
294		__ret = (float32x4_t) __builtin_neon_vbfmlaltq_f32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
	315	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vbfmlaltq_f32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 41));
295	316	return __ret;
296	317	}
297	318	#else
298	319	__ai __attribute__((target("bf16,neon"))) float32x4_t vbfmlaltq_f32(float32x4_t __p0, bfloat16x8_t __p1, bfloat16x8_t __p2) {
299	320	float32x4_t __ret;
300		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
301		bfloat16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
302		bfloat16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
303		__ret = (float32x4_t) __builtin_neon_vbfmlaltq_f32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 41);
304		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	321	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	322	bfloat16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	323	bfloat16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
	324	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vbfmlaltq_f32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 41));
	325	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
305	326	return __ret;
306	327	}
307	328	__ai __attribute__((target("bf16,neon"))) float32x4_t __noswap_vbfmlaltq_f32(float32x4_t __p0, bfloat16x8_t __p1, bfloat16x8_t __p2) {
308	329	float32x4_t __ret;
309		__ret = (float32x4_t) __builtin_neon_vbfmlaltq_f32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
	330	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vbfmlaltq_f32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 41));
310	331	return __ret;
311	332	}
312	333	#endif
...	...	@@ -314,17 +335,17 @@ __ai __attribute__((target("bf16,neon"))) float32x4_t __noswap_vbfmlaltq_f32(flo
314	335	#ifdef __LITTLE_ENDIAN__
315	336	__ai __attribute__((target("bf16,neon"))) float32x4_t vbfmmlaq_f32(float32x4_t __p0, bfloat16x8_t __p1, bfloat16x8_t __p2) {
316	337	float32x4_t __ret;
317		__ret = (float32x4_t) __builtin_neon_vbfmmlaq_f32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
	338	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vbfmmlaq_f32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 41));
318	339	return __ret;
319	340	}
320	341	#else
321	342	__ai __attribute__((target("bf16,neon"))) float32x4_t vbfmmlaq_f32(float32x4_t __p0, bfloat16x8_t __p1, bfloat16x8_t __p2) {
322	343	float32x4_t __ret;
323		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
324		bfloat16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
325		bfloat16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
326		__ret = (float32x4_t) __builtin_neon_vbfmmlaq_f32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 41);
327		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	344	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	345	bfloat16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	346	bfloat16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
	347	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vbfmmlaq_f32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 41));
	348	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
328	349	return __ret;
329	350	}
330	351	#endif
...	...	@@ -338,10 +359,10 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x8_t vcombine_bf16(bfloat16x4_
338	359	#else
339	360	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vcombine_bf16(bfloat16x4_t __p0, bfloat16x4_t __p1) {
340	361	bfloat16x8_t __ret;
341		bfloat16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
342		bfloat16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	362	bfloat16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	363	bfloat16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
343	364	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 1, 2, 3, 4, 5, 6, 7);
344		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	365	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
345	366	return __ret;
346	367	}
347	368	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t __noswap_vcombine_bf16(bfloat16x4_t __p0, bfloat16x4_t __p1) {
...	...	@@ -354,7 +375,7 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x8_t __noswap_vcombine_bf16(bf
354	375	#define vcreate_bf16(__p0) __extension__ ({ \
355	376	bfloat16x4_t __ret; \
356	377	uint64_t __promote = __p0; \
357		__ret = (bfloat16x4_t)(__promote); \
	378	__ret = __builtin_bit_cast(bfloat16x4_t, __promote); \
358	379	__ret; \
359	380	})
360	381	__ai __attribute__((target("bf16,neon"))) float32_t vcvtah_f32_bf16(bfloat16_t __p0) {
...	...	@@ -364,22 +385,22 @@ __ai __attribute__((target("bf16,neon"))) float32_t vcvtah_f32_bf16(bfloat16_t _
364	385	}
365	386	__ai __attribute__((target("bf16,neon"))) bfloat16_t vcvth_bf16_f32(float32_t __p0) {
366	387	bfloat16_t __ret;
367		__ret = (bfloat16_t) __builtin_neon_vcvth_bf16_f32(__p0);
	388	__ret = __builtin_bit_cast(bfloat16_t, __builtin_neon_vcvth_bf16_f32(__p0));
368	389	return __ret;
369	390	}
370	391	#ifdef __LITTLE_ENDIAN__
371	392	#define vduph_lane_bf16(__p0, __p1) __extension__ ({ \
372	393	bfloat16_t __ret; \
373	394	bfloat16x4_t __s0 = __p0; \
374		__ret = (bfloat16_t) __builtin_neon_vduph_lane_bf16((bfloat16x4_t)__s0, __p1); \
	395	__ret = __builtin_bit_cast(bfloat16_t, __builtin_neon_vduph_lane_bf16(__s0, __p1)); \
375	396	__ret; \
376	397	})
377	398	#else
378	399	#define vduph_lane_bf16(__p0, __p1) __extension__ ({ \
379	400	bfloat16_t __ret; \
380	401	bfloat16x4_t __s0 = __p0; \
381		bfloat16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
382		__ret = (bfloat16_t) __builtin_neon_vduph_lane_bf16((bfloat16x4_t)__rev0, __p1); \
	402	bfloat16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	403	__ret = __builtin_bit_cast(bfloat16_t, __builtin_neon_vduph_lane_bf16(__rev0, __p1)); \
383	404	__ret; \
384	405	})
385	406	#endif
...	...	@@ -395,9 +416,9 @@ __ai __attribute__((target("bf16,neon"))) bfloat16_t vcvth_bf16_f32(float32_t __
395	416	#define vdupq_lane_bf16(__p0_1, __p1_1) __extension__ ({ \
396	417	bfloat16x8_t __ret_1; \
397	418	bfloat16x4_t __s0_1 = __p0_1; \
398		bfloat16x4_t __rev0_1; __rev0_1 = __builtin_shufflevector(__s0_1, __s0_1, 3, 2, 1, 0); \
	419	bfloat16x4_t __rev0_1; __rev0_1 = __builtin_shufflevector(__s0_1, __s0_1, __lane_reverse_64_16); \
399	420	__ret_1 = __noswap_splatq_lane_bf16(__rev0_1, __p1_1); \
400		__ret_1 = __builtin_shufflevector(__ret_1, __ret_1, 7, 6, 5, 4, 3, 2, 1, 0); \
	421	__ret_1 = __builtin_shufflevector(__ret_1, __ret_1, __lane_reverse_128_16); \
401	422	__ret_1; \
402	423	})
403	424	#endif
...	...	@@ -413,9 +434,9 @@ __ai __attribute__((target("bf16,neon"))) bfloat16_t vcvth_bf16_f32(float32_t __
413	434	#define vdup_lane_bf16(__p0_3, __p1_3) __extension__ ({ \
414	435	bfloat16x4_t __ret_3; \
415	436	bfloat16x4_t __s0_3 = __p0_3; \
416		bfloat16x4_t __rev0_3; __rev0_3 = __builtin_shufflevector(__s0_3, __s0_3, 3, 2, 1, 0); \
	437	bfloat16x4_t __rev0_3; __rev0_3 = __builtin_shufflevector(__s0_3, __s0_3, __lane_reverse_64_16); \
417	438	__ret_3 = __noswap_splat_lane_bf16(__rev0_3, __p1_3); \
418		__ret_3 = __builtin_shufflevector(__ret_3, __ret_3, 3, 2, 1, 0); \
	439	__ret_3 = __builtin_shufflevector(__ret_3, __ret_3, __lane_reverse_64_16); \
419	440	__ret_3; \
420	441	})
421	442	#endif
...	...	@@ -424,15 +445,15 @@ __ai __attribute__((target("bf16,neon"))) bfloat16_t vcvth_bf16_f32(float32_t __
424	445	#define vduph_laneq_bf16(__p0, __p1) __extension__ ({ \
425	446	bfloat16_t __ret; \
426	447	bfloat16x8_t __s0 = __p0; \
427		__ret = (bfloat16_t) __builtin_neon_vduph_laneq_bf16((bfloat16x8_t)__s0, __p1); \
	448	__ret = __builtin_bit_cast(bfloat16_t, __builtin_neon_vduph_laneq_bf16(__s0, __p1)); \
428	449	__ret; \
429	450	})
430	451	#else
431	452	#define vduph_laneq_bf16(__p0, __p1) __extension__ ({ \
432	453	bfloat16_t __ret; \
433	454	bfloat16x8_t __s0 = __p0; \
434		bfloat16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
435		__ret = (bfloat16_t) __builtin_neon_vduph_laneq_bf16((bfloat16x8_t)__rev0, __p1); \
	455	bfloat16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	456	__ret = __builtin_bit_cast(bfloat16_t, __builtin_neon_vduph_laneq_bf16(__rev0, __p1)); \
436	457	__ret; \
437	458	})
438	459	#endif
...	...	@@ -448,9 +469,9 @@ __ai __attribute__((target("bf16,neon"))) bfloat16_t vcvth_bf16_f32(float32_t __
448	469	#define vdupq_laneq_bf16(__p0_5, __p1_5) __extension__ ({ \
449	470	bfloat16x8_t __ret_5; \
450	471	bfloat16x8_t __s0_5 = __p0_5; \
451		bfloat16x8_t __rev0_5; __rev0_5 = __builtin_shufflevector(__s0_5, __s0_5, 7, 6, 5, 4, 3, 2, 1, 0); \
	472	bfloat16x8_t __rev0_5; __rev0_5 = __builtin_shufflevector(__s0_5, __s0_5, __lane_reverse_128_16); \
452	473	__ret_5 = __noswap_splatq_laneq_bf16(__rev0_5, __p1_5); \
453		__ret_5 = __builtin_shufflevector(__ret_5, __ret_5, 7, 6, 5, 4, 3, 2, 1, 0); \
	474	__ret_5 = __builtin_shufflevector(__ret_5, __ret_5, __lane_reverse_128_16); \
454	475	__ret_5; \
455	476	})
456	477	#endif
...	...	@@ -466,9 +487,9 @@ __ai __attribute__((target("bf16,neon"))) bfloat16_t vcvth_bf16_f32(float32_t __
466	487	#define vdup_laneq_bf16(__p0_7, __p1_7) __extension__ ({ \
467	488	bfloat16x4_t __ret_7; \
468	489	bfloat16x8_t __s0_7 = __p0_7; \
469		bfloat16x8_t __rev0_7; __rev0_7 = __builtin_shufflevector(__s0_7, __s0_7, 7, 6, 5, 4, 3, 2, 1, 0); \
	490	bfloat16x8_t __rev0_7; __rev0_7 = __builtin_shufflevector(__s0_7, __s0_7, __lane_reverse_128_16); \
470	491	__ret_7 = __noswap_splat_laneq_bf16(__rev0_7, __p1_7); \
471		__ret_7 = __builtin_shufflevector(__ret_7, __ret_7, 3, 2, 1, 0); \
	492	__ret_7 = __builtin_shufflevector(__ret_7, __ret_7, __lane_reverse_64_16); \
472	493	__ret_7; \
473	494	})
474	495	#endif
...	...	@@ -483,7 +504,7 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x8_t vdupq_n_bf16(bfloat16_t _
483	504	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vdupq_n_bf16(bfloat16_t __p0) {
484	505	bfloat16x8_t __ret;
485	506	__ret = (bfloat16x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
486		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	507	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
487	508	return __ret;
488	509	}
489	510	#endif
...	...	@@ -498,7 +519,7 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t vdup_n_bf16(bfloat16_t __
498	519	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vdup_n_bf16(bfloat16_t __p0) {
499	520	bfloat16x4_t __ret;
500	521	__ret = (bfloat16x4_t) {__p0, __p0, __p0, __p0};
501		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	522	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
502	523	return __ret;
503	524	}
504	525	#endif
...	...	@@ -512,9 +533,9 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t vget_high_bf16(bfloat16x8
512	533	#else
513	534	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vget_high_bf16(bfloat16x8_t __p0) {
514	535	bfloat16x4_t __ret;
515		bfloat16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	536	bfloat16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
516	537	__ret = __builtin_shufflevector(__rev0, __rev0, 4, 5, 6, 7);
517		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	538	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
518	539	return __ret;
519	540	}
520	541	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_high_bf16(bfloat16x8_t __p0) {
...	...	@@ -528,21 +549,21 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_high_bf16(b
528	549	#define vgetq_lane_bf16(__p0, __p1) __extension__ ({ \
529	550	bfloat16_t __ret; \
530	551	bfloat16x8_t __s0 = __p0; \
531		__ret = (bfloat16_t) __builtin_neon_vgetq_lane_bf16((bfloat16x8_t)__s0, __p1); \
	552	__ret = __builtin_bit_cast(bfloat16_t, __builtin_neon_vgetq_lane_bf16(__s0, __p1)); \
532	553	__ret; \
533	554	})
534	555	#else
535	556	#define vgetq_lane_bf16(__p0, __p1) __extension__ ({ \
536	557	bfloat16_t __ret; \
537	558	bfloat16x8_t __s0 = __p0; \
538		bfloat16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
539		__ret = (bfloat16_t) __builtin_neon_vgetq_lane_bf16((bfloat16x8_t)__rev0, __p1); \
	559	bfloat16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	560	__ret = __builtin_bit_cast(bfloat16_t, __builtin_neon_vgetq_lane_bf16(__rev0, __p1)); \
540	561	__ret; \
541	562	})
542	563	#define __noswap_vgetq_lane_bf16(__p0, __p1) __extension__ ({ \
543	564	bfloat16_t __ret; \
544	565	bfloat16x8_t __s0 = __p0; \
545		__ret = (bfloat16_t) __builtin_neon_vgetq_lane_bf16((bfloat16x8_t)__s0, __p1); \
	566	__ret = __builtin_bit_cast(bfloat16_t, __builtin_neon_vgetq_lane_bf16(__s0, __p1)); \
546	567	__ret; \
547	568	})
548	569	#endif
...	...	@@ -551,21 +572,21 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_high_bf16(b
551	572	#define vget_lane_bf16(__p0, __p1) __extension__ ({ \
552	573	bfloat16_t __ret; \
553	574	bfloat16x4_t __s0 = __p0; \
554		__ret = (bfloat16_t) __builtin_neon_vget_lane_bf16((bfloat16x4_t)__s0, __p1); \
	575	__ret = __builtin_bit_cast(bfloat16_t, __builtin_neon_vget_lane_bf16(__s0, __p1)); \
555	576	__ret; \
556	577	})
557	578	#else
558	579	#define vget_lane_bf16(__p0, __p1) __extension__ ({ \
559	580	bfloat16_t __ret; \
560	581	bfloat16x4_t __s0 = __p0; \
561		bfloat16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
562		__ret = (bfloat16_t) __builtin_neon_vget_lane_bf16((bfloat16x4_t)__rev0, __p1); \
	582	bfloat16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	583	__ret = __builtin_bit_cast(bfloat16_t, __builtin_neon_vget_lane_bf16(__rev0, __p1)); \
563	584	__ret; \
564	585	})
565	586	#define __noswap_vget_lane_bf16(__p0, __p1) __extension__ ({ \
566	587	bfloat16_t __ret; \
567	588	bfloat16x4_t __s0 = __p0; \
568		__ret = (bfloat16_t) __builtin_neon_vget_lane_bf16((bfloat16x4_t)__s0, __p1); \
	589	__ret = __builtin_bit_cast(bfloat16_t, __builtin_neon_vget_lane_bf16(__s0, __p1)); \
569	590	__ret; \
570	591	})
571	592	#endif
...	...	@@ -579,9 +600,9 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t vget_low_bf16(bfloat16x8_
579	600	#else
580	601	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vget_low_bf16(bfloat16x8_t __p0) {
581	602	bfloat16x4_t __ret;
582		bfloat16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	603	bfloat16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
583	604	__ret = __builtin_shufflevector(__rev0, __rev0, 0, 1, 2, 3);
584		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	605	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
585	606	return __ret;
586	607	}
587	608	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bfloat16x8_t __p0) {
...	...	@@ -594,14 +615,14 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
594	615	#ifdef __LITTLE_ENDIAN__
595	616	#define vld1q_bf16(__p0) __extension__ ({ \
596	617	bfloat16x8_t __ret; \
597		__ret = (bfloat16x8_t) __builtin_neon_vld1q_bf16(__p0, 43); \
	618	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vld1q_bf16(__p0, 43)); \
598	619	__ret; \
599	620	})
600	621	#else
601	622	#define vld1q_bf16(__p0) __extension__ ({ \
602	623	bfloat16x8_t __ret; \
603		__ret = (bfloat16x8_t) __builtin_neon_vld1q_bf16(__p0, 43); \
604		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	624	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vld1q_bf16(__p0, 43)); \
	625	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
605	626	__ret; \
606	627	})
607	628	#endif
...	...	@@ -609,14 +630,14 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
609	630	#ifdef __LITTLE_ENDIAN__
610	631	#define vld1_bf16(__p0) __extension__ ({ \
611	632	bfloat16x4_t __ret; \
612		__ret = (bfloat16x4_t) __builtin_neon_vld1_bf16(__p0, 11); \
	633	__ret = __builtin_bit_cast(bfloat16x4_t, __builtin_neon_vld1_bf16(__p0, 11)); \
613	634	__ret; \
614	635	})
615	636	#else
616	637	#define vld1_bf16(__p0) __extension__ ({ \
617	638	bfloat16x4_t __ret; \
618		__ret = (bfloat16x4_t) __builtin_neon_vld1_bf16(__p0, 11); \
619		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	639	__ret = __builtin_bit_cast(bfloat16x4_t, __builtin_neon_vld1_bf16(__p0, 11)); \
	640	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
620	641	__ret; \
621	642	})
622	643	#endif
...	...	@@ -624,14 +645,14 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
624	645	#ifdef __LITTLE_ENDIAN__
625	646	#define vld1q_dup_bf16(__p0) __extension__ ({ \
626	647	bfloat16x8_t __ret; \
627		__ret = (bfloat16x8_t) __builtin_neon_vld1q_dup_bf16(__p0, 43); \
	648	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vld1q_dup_bf16(__p0, 43)); \
628	649	__ret; \
629	650	})
630	651	#else
631	652	#define vld1q_dup_bf16(__p0) __extension__ ({ \
632	653	bfloat16x8_t __ret; \
633		__ret = (bfloat16x8_t) __builtin_neon_vld1q_dup_bf16(__p0, 43); \
634		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	654	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vld1q_dup_bf16(__p0, 43)); \
	655	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
635	656	__ret; \
636	657	})
637	658	#endif
...	...	@@ -639,14 +660,14 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
639	660	#ifdef __LITTLE_ENDIAN__
640	661	#define vld1_dup_bf16(__p0) __extension__ ({ \
641	662	bfloat16x4_t __ret; \
642		__ret = (bfloat16x4_t) __builtin_neon_vld1_dup_bf16(__p0, 11); \
	663	__ret = __builtin_bit_cast(bfloat16x4_t, __builtin_neon_vld1_dup_bf16(__p0, 11)); \
643	664	__ret; \
644	665	})
645	666	#else
646	667	#define vld1_dup_bf16(__p0) __extension__ ({ \
647	668	bfloat16x4_t __ret; \
648		__ret = (bfloat16x4_t) __builtin_neon_vld1_dup_bf16(__p0, 11); \
649		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	669	__ret = __builtin_bit_cast(bfloat16x4_t, __builtin_neon_vld1_dup_bf16(__p0, 11)); \
	670	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
650	671	__ret; \
651	672	})
652	673	#endif
...	...	@@ -655,16 +676,16 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
655	676	#define vld1q_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
656	677	bfloat16x8_t __ret; \
657	678	bfloat16x8_t __s1 = __p1; \
658		__ret = (bfloat16x8_t) __builtin_neon_vld1q_lane_bf16(__p0, (int8x16_t)__s1, __p2, 43); \
	679	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vld1q_lane_bf16(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 43)); \
659	680	__ret; \
660	681	})
661	682	#else
662	683	#define vld1q_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
663	684	bfloat16x8_t __ret; \
664	685	bfloat16x8_t __s1 = __p1; \
665		bfloat16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
666		__ret = (bfloat16x8_t) __builtin_neon_vld1q_lane_bf16(__p0, (int8x16_t)__rev1, __p2, 43); \
667		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	686	bfloat16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	687	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vld1q_lane_bf16(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 43)); \
	688	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
668	689	__ret; \
669	690	})
670	691	#endif
...	...	@@ -673,16 +694,16 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
673	694	#define vld1_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
674	695	bfloat16x4_t __ret; \
675	696	bfloat16x4_t __s1 = __p1; \
676		__ret = (bfloat16x4_t) __builtin_neon_vld1_lane_bf16(__p0, (int8x8_t)__s1, __p2, 11); \
	697	__ret = __builtin_bit_cast(bfloat16x4_t, __builtin_neon_vld1_lane_bf16(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 11)); \
677	698	__ret; \
678	699	})
679	700	#else
680	701	#define vld1_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
681	702	bfloat16x4_t __ret; \
682	703	bfloat16x4_t __s1 = __p1; \
683		bfloat16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
684		__ret = (bfloat16x4_t) __builtin_neon_vld1_lane_bf16(__p0, (int8x8_t)__rev1, __p2, 11); \
685		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	704	bfloat16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	705	__ret = __builtin_bit_cast(bfloat16x4_t, __builtin_neon_vld1_lane_bf16(__p0, __builtin_bit_cast(int8x8_t, __rev1), __p2, 11)); \
	706	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
686	707	__ret; \
687	708	})
688	709	#endif
...	...	@@ -698,8 +719,8 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
698	719	bfloat16x8x2_t __ret; \
699	720	__builtin_neon_vld1q_bf16_x2(&__ret, __p0, 43); \
700	721	\
701		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
702		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	722	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	723	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
703	724	__ret; \
704	725	})
705	726	#endif
...	...	@@ -715,8 +736,8 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
715	736	bfloat16x4x2_t __ret; \
716	737	__builtin_neon_vld1_bf16_x2(&__ret, __p0, 11); \
717	738	\
718		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
719		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	739	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	740	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
720	741	__ret; \
721	742	})
722	743	#endif
...	...	@@ -732,9 +753,9 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
732	753	bfloat16x8x3_t __ret; \
733	754	__builtin_neon_vld1q_bf16_x3(&__ret, __p0, 43); \
734	755	\
735		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
736		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
737		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	756	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	757	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	758	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
738	759	__ret; \
739	760	})
740	761	#endif
...	...	@@ -750,9 +771,9 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
750	771	bfloat16x4x3_t __ret; \
751	772	__builtin_neon_vld1_bf16_x3(&__ret, __p0, 11); \
752	773	\
753		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
754		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
755		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	774	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	775	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	776	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
756	777	__ret; \
757	778	})
758	779	#endif
...	...	@@ -768,10 +789,10 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
768	789	bfloat16x8x4_t __ret; \
769	790	__builtin_neon_vld1q_bf16_x4(&__ret, __p0, 43); \
770	791	\
771		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
772		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
773		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
774		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	792	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	793	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	794	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
	795	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_16); \
775	796	__ret; \
776	797	})
777	798	#endif
...	...	@@ -787,10 +808,10 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
787	808	bfloat16x4x4_t __ret; \
788	809	__builtin_neon_vld1_bf16_x4(&__ret, __p0, 11); \
789	810	\
790		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
791		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
792		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
793		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	811	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	812	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	813	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
	814	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_16); \
794	815	__ret; \
795	816	})
796	817	#endif
...	...	@@ -806,8 +827,8 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
806	827	bfloat16x8x2_t __ret; \
807	828	__builtin_neon_vld2q_bf16(&__ret, __p0, 43); \
808	829	\
809		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
810		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	830	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	831	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
811	832	__ret; \
812	833	})
813	834	#endif
...	...	@@ -823,8 +844,8 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
823	844	bfloat16x4x2_t __ret; \
824	845	__builtin_neon_vld2_bf16(&__ret, __p0, 11); \
825	846	\
826		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
827		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	847	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	848	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
828	849	__ret; \
829	850	})
830	851	#endif
...	...	@@ -840,8 +861,8 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
840	861	bfloat16x8x2_t __ret; \
841	862	__builtin_neon_vld2q_dup_bf16(&__ret, __p0, 43); \
842	863	\
843		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
844		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	864	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	865	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
845	866	__ret; \
846	867	})
847	868	#endif
...	...	@@ -857,8 +878,8 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
857	878	bfloat16x4x2_t __ret; \
858	879	__builtin_neon_vld2_dup_bf16(&__ret, __p0, 11); \
859	880	\
860		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
861		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	881	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	882	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
862	883	__ret; \
863	884	})
864	885	#endif
...	...	@@ -867,7 +888,7 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
867	888	#define vld2q_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
868	889	bfloat16x8x2_t __ret; \
869	890	bfloat16x8x2_t __s1 = __p1; \
870		__builtin_neon_vld2q_lane_bf16(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 43); \
	891	__builtin_neon_vld2q_lane_bf16(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 43); \
871	892	__ret; \
872	893	})
873	894	#else
...	...	@@ -875,12 +896,12 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
875	896	bfloat16x8x2_t __ret; \
876	897	bfloat16x8x2_t __s1 = __p1; \
877	898	bfloat16x8x2_t __rev1; \
878		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
879		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
880		__builtin_neon_vld2q_lane_bf16(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 43); \
	899	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	900	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	901	__builtin_neon_vld2q_lane_bf16(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 43); \
881	902	\
882		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
883		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	903	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	904	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
884	905	__ret; \
885	906	})
886	907	#endif
...	...	@@ -889,7 +910,7 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
889	910	#define vld2_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
890	911	bfloat16x4x2_t __ret; \
891	912	bfloat16x4x2_t __s1 = __p1; \
892		__builtin_neon_vld2_lane_bf16(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 11); \
	913	__builtin_neon_vld2_lane_bf16(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 11); \
893	914	__ret; \
894	915	})
895	916	#else
...	...	@@ -897,12 +918,12 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
897	918	bfloat16x4x2_t __ret; \
898	919	bfloat16x4x2_t __s1 = __p1; \
899	920	bfloat16x4x2_t __rev1; \
900		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
901		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
902		__builtin_neon_vld2_lane_bf16(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 11); \
	921	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	922	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	923	__builtin_neon_vld2_lane_bf16(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __p2, 11); \
903	924	\
904		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
905		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	925	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	926	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
906	927	__ret; \
907	928	})
908	929	#endif
...	...	@@ -918,9 +939,9 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
918	939	bfloat16x8x3_t __ret; \
919	940	__builtin_neon_vld3q_bf16(&__ret, __p0, 43); \
920	941	\
921		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
922		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
923		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	942	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	943	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	944	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
924	945	__ret; \
925	946	})
926	947	#endif
...	...	@@ -936,9 +957,9 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
936	957	bfloat16x4x3_t __ret; \
937	958	__builtin_neon_vld3_bf16(&__ret, __p0, 11); \
938	959	\
939		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
940		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
941		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	960	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	961	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	962	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
942	963	__ret; \
943	964	})
944	965	#endif
...	...	@@ -954,9 +975,9 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
954	975	bfloat16x8x3_t __ret; \
955	976	__builtin_neon_vld3q_dup_bf16(&__ret, __p0, 43); \
956	977	\
957		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
958		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
959		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	978	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	979	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	980	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
960	981	__ret; \
961	982	})
962	983	#endif
...	...	@@ -972,9 +993,9 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
972	993	bfloat16x4x3_t __ret; \
973	994	__builtin_neon_vld3_dup_bf16(&__ret, __p0, 11); \
974	995	\
975		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
976		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
977		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	996	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	997	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	998	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
978	999	__ret; \
979	1000	})
980	1001	#endif
...	...	@@ -983,7 +1004,7 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
983	1004	#define vld3q_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
984	1005	bfloat16x8x3_t __ret; \
985	1006	bfloat16x8x3_t __s1 = __p1; \
986		__builtin_neon_vld3q_lane_bf16(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 43); \
	1007	__builtin_neon_vld3q_lane_bf16(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 43); \
987	1008	__ret; \
988	1009	})
989	1010	#else
...	...	@@ -991,14 +1012,14 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
991	1012	bfloat16x8x3_t __ret; \
992	1013	bfloat16x8x3_t __s1 = __p1; \
993	1014	bfloat16x8x3_t __rev1; \
994		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
995		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
996		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
997		__builtin_neon_vld3q_lane_bf16(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 43); \
	1015	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	1016	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	1017	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	1018	__builtin_neon_vld3q_lane_bf16(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 43); \
998	1019	\
999		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
1000		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
1001		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	1020	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	1021	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	1022	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
1002	1023	__ret; \
1003	1024	})
1004	1025	#endif
...	...	@@ -1007,7 +1028,7 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
1007	1028	#define vld3_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
1008	1029	bfloat16x4x3_t __ret; \
1009	1030	bfloat16x4x3_t __s1 = __p1; \
1010		__builtin_neon_vld3_lane_bf16(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 11); \
	1031	__builtin_neon_vld3_lane_bf16(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 11); \
1011	1032	__ret; \
1012	1033	})
1013	1034	#else
...	...	@@ -1015,14 +1036,14 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
1015	1036	bfloat16x4x3_t __ret; \
1016	1037	bfloat16x4x3_t __s1 = __p1; \
1017	1038	bfloat16x4x3_t __rev1; \
1018		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
1019		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
1020		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
1021		__builtin_neon_vld3_lane_bf16(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 11); \
	1039	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	1040	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	1041	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	1042	__builtin_neon_vld3_lane_bf16(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __p2, 11); \
1022	1043	\
1023		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
1024		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
1025		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	1044	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	1045	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	1046	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
1026	1047	__ret; \
1027	1048	})
1028	1049	#endif
...	...	@@ -1038,10 +1059,10 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
1038	1059	bfloat16x8x4_t __ret; \
1039	1060	__builtin_neon_vld4q_bf16(&__ret, __p0, 43); \
1040	1061	\
1041		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
1042		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
1043		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
1044		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	1062	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	1063	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	1064	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
	1065	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_16); \
1045	1066	__ret; \
1046	1067	})
1047	1068	#endif
...	...	@@ -1057,10 +1078,10 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
1057	1078	bfloat16x4x4_t __ret; \
1058	1079	__builtin_neon_vld4_bf16(&__ret, __p0, 11); \
1059	1080	\
1060		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
1061		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
1062		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
1063		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	1081	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	1082	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	1083	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
	1084	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_16); \
1064	1085	__ret; \
1065	1086	})
1066	1087	#endif
...	...	@@ -1076,10 +1097,10 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
1076	1097	bfloat16x8x4_t __ret; \
1077	1098	__builtin_neon_vld4q_dup_bf16(&__ret, __p0, 43); \
1078	1099	\
1079		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
1080		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
1081		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
1082		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	1100	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	1101	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	1102	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
	1103	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_16); \
1083	1104	__ret; \
1084	1105	})
1085	1106	#endif
...	...	@@ -1095,10 +1116,10 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
1095	1116	bfloat16x4x4_t __ret; \
1096	1117	__builtin_neon_vld4_dup_bf16(&__ret, __p0, 11); \
1097	1118	\
1098		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
1099		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
1100		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
1101		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	1119	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	1120	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	1121	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
	1122	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_16); \
1102	1123	__ret; \
1103	1124	})
1104	1125	#endif
...	...	@@ -1107,7 +1128,7 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
1107	1128	#define vld4q_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
1108	1129	bfloat16x8x4_t __ret; \
1109	1130	bfloat16x8x4_t __s1 = __p1; \
1110		__builtin_neon_vld4q_lane_bf16(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 43); \
	1131	__builtin_neon_vld4q_lane_bf16(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 43); \
1111	1132	__ret; \
1112	1133	})
1113	1134	#else
...	...	@@ -1115,16 +1136,16 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
1115	1136	bfloat16x8x4_t __ret; \
1116	1137	bfloat16x8x4_t __s1 = __p1; \
1117	1138	bfloat16x8x4_t __rev1; \
1118		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
1119		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
1120		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
1121		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
1122		__builtin_neon_vld4q_lane_bf16(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 43); \
	1139	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	1140	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	1141	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	1142	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_16); \
	1143	__builtin_neon_vld4q_lane_bf16(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 43); \
1123	1144	\
1124		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
1125		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
1126		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
1127		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	1145	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	1146	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	1147	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
	1148	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_16); \
1128	1149	__ret; \
1129	1150	})
1130	1151	#endif
...	...	@@ -1133,7 +1154,7 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
1133	1154	#define vld4_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
1134	1155	bfloat16x4x4_t __ret; \
1135	1156	bfloat16x4x4_t __s1 = __p1; \
1136		__builtin_neon_vld4_lane_bf16(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 11); \
	1157	__builtin_neon_vld4_lane_bf16(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 11); \
1137	1158	__ret; \
1138	1159	})
1139	1160	#else
...	...	@@ -1141,16 +1162,16 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
1141	1162	bfloat16x4x4_t __ret; \
1142	1163	bfloat16x4x4_t __s1 = __p1; \
1143	1164	bfloat16x4x4_t __rev1; \
1144		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
1145		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
1146		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
1147		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
1148		__builtin_neon_vld4_lane_bf16(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 11); \
	1165	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	1166	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	1167	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	1168	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_16); \
	1169	__builtin_neon_vld4_lane_bf16(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), __p2, 11); \
1149	1170	\
1150		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
1151		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
1152		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
1153		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	1171	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	1172	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	1173	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
	1174	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_16); \
1154	1175	__ret; \
1155	1176	})
1156	1177	#endif
...	...	@@ -1160,7 +1181,7 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
1160	1181	bfloat16x8_t __ret; \
1161	1182	bfloat16_t __s0 = __p0; \
1162	1183	bfloat16x8_t __s1 = __p1; \
1163		__ret = (bfloat16x8_t) __builtin_neon_vsetq_lane_bf16(__s0, (bfloat16x8_t)__s1, __p2); \
	1184	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vsetq_lane_bf16(__s0, __s1, __p2)); \
1164	1185	__ret; \
1165	1186	})
1166	1187	#else
...	...	@@ -1168,16 +1189,16 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
1168	1189	bfloat16x8_t __ret; \
1169	1190	bfloat16_t __s0 = __p0; \
1170	1191	bfloat16x8_t __s1 = __p1; \
1171		bfloat16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
1172		__ret = (bfloat16x8_t) __builtin_neon_vsetq_lane_bf16(__s0, (bfloat16x8_t)__rev1, __p2); \
1173		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	1192	bfloat16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	1193	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vsetq_lane_bf16(__s0, __rev1, __p2)); \
	1194	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
1174	1195	__ret; \
1175	1196	})
1176	1197	#define __noswap_vsetq_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
1177	1198	bfloat16x8_t __ret; \
1178	1199	bfloat16_t __s0 = __p0; \
1179	1200	bfloat16x8_t __s1 = __p1; \
1180		__ret = (bfloat16x8_t) __builtin_neon_vsetq_lane_bf16(__s0, (bfloat16x8_t)__s1, __p2); \
	1201	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vsetq_lane_bf16(__s0, __s1, __p2)); \
1181	1202	__ret; \
1182	1203	})
1183	1204	#endif
...	...	@@ -1187,7 +1208,7 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
1187	1208	bfloat16x4_t __ret; \
1188	1209	bfloat16_t __s0 = __p0; \
1189	1210	bfloat16x4_t __s1 = __p1; \
1190		__ret = (bfloat16x4_t) __builtin_neon_vset_lane_bf16(__s0, (bfloat16x4_t)__s1, __p2); \
	1211	__ret = __builtin_bit_cast(bfloat16x4_t, __builtin_neon_vset_lane_bf16(__s0, __s1, __p2)); \
1191	1212	__ret; \
1192	1213	})
1193	1214	#else
...	...	@@ -1195,16 +1216,16 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
1195	1216	bfloat16x4_t __ret; \
1196	1217	bfloat16_t __s0 = __p0; \
1197	1218	bfloat16x4_t __s1 = __p1; \
1198		bfloat16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
1199		__ret = (bfloat16x4_t) __builtin_neon_vset_lane_bf16(__s0, (bfloat16x4_t)__rev1, __p2); \
1200		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	1219	bfloat16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	1220	__ret = __builtin_bit_cast(bfloat16x4_t, __builtin_neon_vset_lane_bf16(__s0, __rev1, __p2)); \
	1221	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
1201	1222	__ret; \
1202	1223	})
1203	1224	#define __noswap_vset_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
1204	1225	bfloat16x4_t __ret; \
1205	1226	bfloat16_t __s0 = __p0; \
1206	1227	bfloat16x4_t __s1 = __p1; \
1207		__ret = (bfloat16x4_t) __builtin_neon_vset_lane_bf16(__s0, (bfloat16x4_t)__s1, __p2); \
	1228	__ret = __builtin_bit_cast(bfloat16x4_t, __builtin_neon_vset_lane_bf16(__s0, __s1, __p2)); \
1208	1229	__ret; \
1209	1230	})
1210	1231	#endif
...	...	@@ -1212,362 +1233,362 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap_vget_low_bf16(bf
1212	1233	#ifdef __LITTLE_ENDIAN__
1213	1234	#define vst1q_bf16(__p0, __p1) __extension__ ({ \
1214	1235	bfloat16x8_t __s1 = __p1; \
1215		__builtin_neon_vst1q_bf16(__p0, (int8x16_t)__s1, 43); \
	1236	__builtin_neon_vst1q_bf16(__p0, __builtin_bit_cast(int8x16_t, __s1), 43); \
1216	1237	})
1217	1238	#else
1218	1239	#define vst1q_bf16(__p0, __p1) __extension__ ({ \
1219	1240	bfloat16x8_t __s1 = __p1; \
1220		bfloat16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
1221		__builtin_neon_vst1q_bf16(__p0, (int8x16_t)__rev1, 43); \
	1241	bfloat16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	1242	__builtin_neon_vst1q_bf16(__p0, __builtin_bit_cast(int8x16_t, __rev1), 43); \
1222	1243	})
1223	1244	#endif
1224	1245
1225	1246	#ifdef __LITTLE_ENDIAN__
1226	1247	#define vst1_bf16(__p0, __p1) __extension__ ({ \
1227	1248	bfloat16x4_t __s1 = __p1; \
1228		__builtin_neon_vst1_bf16(__p0, (int8x8_t)__s1, 11); \
	1249	__builtin_neon_vst1_bf16(__p0, __builtin_bit_cast(int8x8_t, __s1), 11); \
1229	1250	})
1230	1251	#else
1231	1252	#define vst1_bf16(__p0, __p1) __extension__ ({ \
1232	1253	bfloat16x4_t __s1 = __p1; \
1233		bfloat16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
1234		__builtin_neon_vst1_bf16(__p0, (int8x8_t)__rev1, 11); \
	1254	bfloat16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	1255	__builtin_neon_vst1_bf16(__p0, __builtin_bit_cast(int8x8_t, __rev1), 11); \
1235	1256	})
1236	1257	#endif
1237	1258
1238	1259	#ifdef __LITTLE_ENDIAN__
1239	1260	#define vst1q_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
1240	1261	bfloat16x8_t __s1 = __p1; \
1241		__builtin_neon_vst1q_lane_bf16(__p0, (int8x16_t)__s1, __p2, 43); \
	1262	__builtin_neon_vst1q_lane_bf16(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 43); \
1242	1263	})
1243	1264	#else
1244	1265	#define vst1q_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
1245	1266	bfloat16x8_t __s1 = __p1; \
1246		bfloat16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
1247		__builtin_neon_vst1q_lane_bf16(__p0, (int8x16_t)__rev1, __p2, 43); \
	1267	bfloat16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	1268	__builtin_neon_vst1q_lane_bf16(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 43); \
1248	1269	})
1249	1270	#endif
1250	1271
1251	1272	#ifdef __LITTLE_ENDIAN__
1252	1273	#define vst1_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
1253	1274	bfloat16x4_t __s1 = __p1; \
1254		__builtin_neon_vst1_lane_bf16(__p0, (int8x8_t)__s1, __p2, 11); \
	1275	__builtin_neon_vst1_lane_bf16(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 11); \
1255	1276	})
1256	1277	#else
1257	1278	#define vst1_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
1258	1279	bfloat16x4_t __s1 = __p1; \
1259		bfloat16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
1260		__builtin_neon_vst1_lane_bf16(__p0, (int8x8_t)__rev1, __p2, 11); \
	1280	bfloat16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	1281	__builtin_neon_vst1_lane_bf16(__p0, __builtin_bit_cast(int8x8_t, __rev1), __p2, 11); \
1261	1282	})
1262	1283	#endif
1263	1284
1264	1285	#ifdef __LITTLE_ENDIAN__
1265	1286	#define vst1q_bf16_x2(__p0, __p1) __extension__ ({ \
1266	1287	bfloat16x8x2_t __s1 = __p1; \
1267		__builtin_neon_vst1q_bf16_x2(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 43); \
	1288	__builtin_neon_vst1q_bf16_x2(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 43); \
1268	1289	})
1269	1290	#else
1270	1291	#define vst1q_bf16_x2(__p0, __p1) __extension__ ({ \
1271	1292	bfloat16x8x2_t __s1 = __p1; \
1272	1293	bfloat16x8x2_t __rev1; \
1273		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
1274		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
1275		__builtin_neon_vst1q_bf16_x2(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 43); \
	1294	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	1295	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	1296	__builtin_neon_vst1q_bf16_x2(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 43); \
1276	1297	})
1277	1298	#endif
1278	1299
1279	1300	#ifdef __LITTLE_ENDIAN__
1280	1301	#define vst1_bf16_x2(__p0, __p1) __extension__ ({ \
1281	1302	bfloat16x4x2_t __s1 = __p1; \
1282		__builtin_neon_vst1_bf16_x2(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 11); \
	1303	__builtin_neon_vst1_bf16_x2(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 11); \
1283	1304	})
1284	1305	#else
1285	1306	#define vst1_bf16_x2(__p0, __p1) __extension__ ({ \
1286	1307	bfloat16x4x2_t __s1 = __p1; \
1287	1308	bfloat16x4x2_t __rev1; \
1288		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
1289		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
1290		__builtin_neon_vst1_bf16_x2(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 11); \
	1309	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	1310	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	1311	__builtin_neon_vst1_bf16_x2(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), 11); \
1291	1312	})
1292	1313	#endif
1293	1314
1294	1315	#ifdef __LITTLE_ENDIAN__
1295	1316	#define vst1q_bf16_x3(__p0, __p1) __extension__ ({ \
1296	1317	bfloat16x8x3_t __s1 = __p1; \
1297		__builtin_neon_vst1q_bf16_x3(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 43); \
	1318	__builtin_neon_vst1q_bf16_x3(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 43); \
1298	1319	})
1299	1320	#else
1300	1321	#define vst1q_bf16_x3(__p0, __p1) __extension__ ({ \
1301	1322	bfloat16x8x3_t __s1 = __p1; \
1302	1323	bfloat16x8x3_t __rev1; \
1303		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
1304		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
1305		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
1306		__builtin_neon_vst1q_bf16_x3(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 43); \
	1324	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	1325	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	1326	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	1327	__builtin_neon_vst1q_bf16_x3(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 43); \
1307	1328	})
1308	1329	#endif
1309	1330
1310	1331	#ifdef __LITTLE_ENDIAN__
1311	1332	#define vst1_bf16_x3(__p0, __p1) __extension__ ({ \
1312	1333	bfloat16x4x3_t __s1 = __p1; \
1313		__builtin_neon_vst1_bf16_x3(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 11); \
	1334	__builtin_neon_vst1_bf16_x3(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 11); \
1314	1335	})
1315	1336	#else
1316	1337	#define vst1_bf16_x3(__p0, __p1) __extension__ ({ \
1317	1338	bfloat16x4x3_t __s1 = __p1; \
1318	1339	bfloat16x4x3_t __rev1; \
1319		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
1320		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
1321		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
1322		__builtin_neon_vst1_bf16_x3(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 11); \
	1340	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	1341	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	1342	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	1343	__builtin_neon_vst1_bf16_x3(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), 11); \
1323	1344	})
1324	1345	#endif
1325	1346
1326	1347	#ifdef __LITTLE_ENDIAN__
1327	1348	#define vst1q_bf16_x4(__p0, __p1) __extension__ ({ \
1328	1349	bfloat16x8x4_t __s1 = __p1; \
1329		__builtin_neon_vst1q_bf16_x4(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 43); \
	1350	__builtin_neon_vst1q_bf16_x4(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 43); \
1330	1351	})
1331	1352	#else
1332	1353	#define vst1q_bf16_x4(__p0, __p1) __extension__ ({ \
1333	1354	bfloat16x8x4_t __s1 = __p1; \
1334	1355	bfloat16x8x4_t __rev1; \
1335		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
1336		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
1337		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
1338		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
1339		__builtin_neon_vst1q_bf16_x4(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 43); \
	1356	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	1357	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	1358	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	1359	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_16); \
	1360	__builtin_neon_vst1q_bf16_x4(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 43); \
1340	1361	})
1341	1362	#endif
1342	1363
1343	1364	#ifdef __LITTLE_ENDIAN__
1344	1365	#define vst1_bf16_x4(__p0, __p1) __extension__ ({ \
1345	1366	bfloat16x4x4_t __s1 = __p1; \
1346		__builtin_neon_vst1_bf16_x4(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 11); \
	1367	__builtin_neon_vst1_bf16_x4(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 11); \
1347	1368	})
1348	1369	#else
1349	1370	#define vst1_bf16_x4(__p0, __p1) __extension__ ({ \
1350	1371	bfloat16x4x4_t __s1 = __p1; \
1351	1372	bfloat16x4x4_t __rev1; \
1352		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
1353		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
1354		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
1355		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
1356		__builtin_neon_vst1_bf16_x4(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 11); \
	1373	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	1374	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	1375	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	1376	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_16); \
	1377	__builtin_neon_vst1_bf16_x4(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), 11); \
1357	1378	})
1358	1379	#endif
1359	1380
1360	1381	#ifdef __LITTLE_ENDIAN__
1361	1382	#define vst2q_bf16(__p0, __p1) __extension__ ({ \
1362	1383	bfloat16x8x2_t __s1 = __p1; \
1363		__builtin_neon_vst2q_bf16(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 43); \
	1384	__builtin_neon_vst2q_bf16(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 43); \
1364	1385	})
1365	1386	#else
1366	1387	#define vst2q_bf16(__p0, __p1) __extension__ ({ \
1367	1388	bfloat16x8x2_t __s1 = __p1; \
1368	1389	bfloat16x8x2_t __rev1; \
1369		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
1370		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
1371		__builtin_neon_vst2q_bf16(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 43); \
	1390	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	1391	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	1392	__builtin_neon_vst2q_bf16(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 43); \
1372	1393	})
1373	1394	#endif
1374	1395
1375	1396	#ifdef __LITTLE_ENDIAN__
1376	1397	#define vst2_bf16(__p0, __p1) __extension__ ({ \
1377	1398	bfloat16x4x2_t __s1 = __p1; \
1378		__builtin_neon_vst2_bf16(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 11); \
	1399	__builtin_neon_vst2_bf16(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 11); \
1379	1400	})
1380	1401	#else
1381	1402	#define vst2_bf16(__p0, __p1) __extension__ ({ \
1382	1403	bfloat16x4x2_t __s1 = __p1; \
1383	1404	bfloat16x4x2_t __rev1; \
1384		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
1385		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
1386		__builtin_neon_vst2_bf16(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 11); \
	1405	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	1406	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	1407	__builtin_neon_vst2_bf16(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), 11); \
1387	1408	})
1388	1409	#endif
1389	1410
1390	1411	#ifdef __LITTLE_ENDIAN__
1391	1412	#define vst2q_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
1392	1413	bfloat16x8x2_t __s1 = __p1; \
1393		__builtin_neon_vst2q_lane_bf16(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 43); \
	1414	__builtin_neon_vst2q_lane_bf16(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 43); \
1394	1415	})
1395	1416	#else
1396	1417	#define vst2q_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
1397	1418	bfloat16x8x2_t __s1 = __p1; \
1398	1419	bfloat16x8x2_t __rev1; \
1399		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
1400		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
1401		__builtin_neon_vst2q_lane_bf16(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 43); \
	1420	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	1421	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	1422	__builtin_neon_vst2q_lane_bf16(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 43); \
1402	1423	})
1403	1424	#endif
1404	1425
1405	1426	#ifdef __LITTLE_ENDIAN__
1406	1427	#define vst2_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
1407	1428	bfloat16x4x2_t __s1 = __p1; \
1408		__builtin_neon_vst2_lane_bf16(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 11); \
	1429	__builtin_neon_vst2_lane_bf16(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 11); \
1409	1430	})
1410	1431	#else
1411	1432	#define vst2_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
1412	1433	bfloat16x4x2_t __s1 = __p1; \
1413	1434	bfloat16x4x2_t __rev1; \
1414		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
1415		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
1416		__builtin_neon_vst2_lane_bf16(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 11); \
	1435	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	1436	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	1437	__builtin_neon_vst2_lane_bf16(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __p2, 11); \
1417	1438	})
1418	1439	#endif
1419	1440
1420	1441	#ifdef __LITTLE_ENDIAN__
1421	1442	#define vst3q_bf16(__p0, __p1) __extension__ ({ \
1422	1443	bfloat16x8x3_t __s1 = __p1; \
1423		__builtin_neon_vst3q_bf16(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 43); \
	1444	__builtin_neon_vst3q_bf16(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 43); \
1424	1445	})
1425	1446	#else
1426	1447	#define vst3q_bf16(__p0, __p1) __extension__ ({ \
1427	1448	bfloat16x8x3_t __s1 = __p1; \
1428	1449	bfloat16x8x3_t __rev1; \
1429		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
1430		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
1431		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
1432		__builtin_neon_vst3q_bf16(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 43); \
	1450	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	1451	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	1452	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	1453	__builtin_neon_vst3q_bf16(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 43); \
1433	1454	})
1434	1455	#endif
1435	1456
1436	1457	#ifdef __LITTLE_ENDIAN__
1437	1458	#define vst3_bf16(__p0, __p1) __extension__ ({ \
1438	1459	bfloat16x4x3_t __s1 = __p1; \
1439		__builtin_neon_vst3_bf16(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 11); \
	1460	__builtin_neon_vst3_bf16(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 11); \
1440	1461	})
1441	1462	#else
1442	1463	#define vst3_bf16(__p0, __p1) __extension__ ({ \
1443	1464	bfloat16x4x3_t __s1 = __p1; \
1444	1465	bfloat16x4x3_t __rev1; \
1445		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
1446		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
1447		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
1448		__builtin_neon_vst3_bf16(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 11); \
	1466	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	1467	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	1468	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	1469	__builtin_neon_vst3_bf16(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), 11); \
1449	1470	})
1450	1471	#endif
1451	1472
1452	1473	#ifdef __LITTLE_ENDIAN__
1453	1474	#define vst3q_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
1454	1475	bfloat16x8x3_t __s1 = __p1; \
1455		__builtin_neon_vst3q_lane_bf16(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 43); \
	1476	__builtin_neon_vst3q_lane_bf16(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 43); \
1456	1477	})
1457	1478	#else
1458	1479	#define vst3q_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
1459	1480	bfloat16x8x3_t __s1 = __p1; \
1460	1481	bfloat16x8x3_t __rev1; \
1461		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
1462		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
1463		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
1464		__builtin_neon_vst3q_lane_bf16(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 43); \
	1482	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	1483	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	1484	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	1485	__builtin_neon_vst3q_lane_bf16(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 43); \
1465	1486	})
1466	1487	#endif
1467	1488
1468	1489	#ifdef __LITTLE_ENDIAN__
1469	1490	#define vst3_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
1470	1491	bfloat16x4x3_t __s1 = __p1; \
1471		__builtin_neon_vst3_lane_bf16(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 11); \
	1492	__builtin_neon_vst3_lane_bf16(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 11); \
1472	1493	})
1473	1494	#else
1474	1495	#define vst3_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
1475	1496	bfloat16x4x3_t __s1 = __p1; \
1476	1497	bfloat16x4x3_t __rev1; \
1477		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
1478		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
1479		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
1480		__builtin_neon_vst3_lane_bf16(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 11); \
	1498	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	1499	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	1500	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	1501	__builtin_neon_vst3_lane_bf16(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __p2, 11); \
1481	1502	})
1482	1503	#endif
1483	1504
1484	1505	#ifdef __LITTLE_ENDIAN__
1485	1506	#define vst4q_bf16(__p0, __p1) __extension__ ({ \
1486	1507	bfloat16x8x4_t __s1 = __p1; \
1487		__builtin_neon_vst4q_bf16(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 43); \
	1508	__builtin_neon_vst4q_bf16(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 43); \
1488	1509	})
1489	1510	#else
1490	1511	#define vst4q_bf16(__p0, __p1) __extension__ ({ \
1491	1512	bfloat16x8x4_t __s1 = __p1; \
1492	1513	bfloat16x8x4_t __rev1; \
1493		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
1494		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
1495		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
1496		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
1497		__builtin_neon_vst4q_bf16(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 43); \
	1514	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	1515	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	1516	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	1517	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_16); \
	1518	__builtin_neon_vst4q_bf16(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 43); \
1498	1519	})
1499	1520	#endif
1500	1521
1501	1522	#ifdef __LITTLE_ENDIAN__
1502	1523	#define vst4_bf16(__p0, __p1) __extension__ ({ \
1503	1524	bfloat16x4x4_t __s1 = __p1; \
1504		__builtin_neon_vst4_bf16(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 11); \
	1525	__builtin_neon_vst4_bf16(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 11); \
1505	1526	})
1506	1527	#else
1507	1528	#define vst4_bf16(__p0, __p1) __extension__ ({ \
1508	1529	bfloat16x4x4_t __s1 = __p1; \
1509	1530	bfloat16x4x4_t __rev1; \
1510		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
1511		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
1512		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
1513		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
1514		__builtin_neon_vst4_bf16(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 11); \
	1531	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	1532	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	1533	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	1534	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_16); \
	1535	__builtin_neon_vst4_bf16(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), 11); \
1515	1536	})
1516	1537	#endif
1517	1538
1518	1539	#ifdef __LITTLE_ENDIAN__
1519	1540	#define vst4q_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
1520	1541	bfloat16x8x4_t __s1 = __p1; \
1521		__builtin_neon_vst4q_lane_bf16(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 43); \
	1542	__builtin_neon_vst4q_lane_bf16(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 43); \
1522	1543	})
1523	1544	#else
1524	1545	#define vst4q_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
1525	1546	bfloat16x8x4_t __s1 = __p1; \
1526	1547	bfloat16x8x4_t __rev1; \
1527		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
1528		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
1529		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
1530		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
1531		__builtin_neon_vst4q_lane_bf16(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 43); \
	1548	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	1549	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	1550	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	1551	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_16); \
	1552	__builtin_neon_vst4q_lane_bf16(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 43); \
1532	1553	})
1533	1554	#endif
1534	1555
1535	1556	#ifdef __LITTLE_ENDIAN__
1536	1557	#define vst4_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
1537	1558	bfloat16x4x4_t __s1 = __p1; \
1538		__builtin_neon_vst4_lane_bf16(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 11); \
	1559	__builtin_neon_vst4_lane_bf16(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 11); \
1539	1560	})
1540	1561	#else
1541	1562	#define vst4_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
1542	1563	bfloat16x4x4_t __s1 = __p1; \
1543	1564	bfloat16x4x4_t __rev1; \
1544		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
1545		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
1546		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
1547		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
1548		__builtin_neon_vst4_lane_bf16(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 11); \
	1565	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	1566	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	1567	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	1568	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_16); \
	1569	__builtin_neon_vst4_lane_bf16(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), __p2, 11); \
1549	1570	})
1550	1571	#endif
1551	1572
1552	1573	#ifdef __LITTLE_ENDIAN__
1553	1574	__ai __attribute__((target("dotprod,neon"))) uint32x4_t vdotq_u32(uint32x4_t __p0, uint8x16_t __p1, uint8x16_t __p2) {
1554	1575	uint32x4_t __ret;
1555		__ret = (uint32x4_t) __builtin_neon_vdotq_u32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 50);
	1576	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vdotq_u32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 50));
1556	1577	return __ret;
1557	1578	}
1558	1579	#else
1559	1580	__ai __attribute__((target("dotprod,neon"))) uint32x4_t vdotq_u32(uint32x4_t __p0, uint8x16_t __p1, uint8x16_t __p2) {
1560	1581	uint32x4_t __ret;
1561		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
1562		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
1563		uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
1564		__ret = (uint32x4_t) __builtin_neon_vdotq_u32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 50);
1565		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	1582	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	1583	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	1584	uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	1585	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vdotq_u32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 50));
	1586	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
1566	1587	return __ret;
1567	1588	}
1568	1589	__ai __attribute__((target("dotprod,neon"))) uint32x4_t __noswap_vdotq_u32(uint32x4_t __p0, uint8x16_t __p1, uint8x16_t __p2) {
1569	1590	uint32x4_t __ret;
1570		__ret = (uint32x4_t) __builtin_neon_vdotq_u32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 50);
	1591	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vdotq_u32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 50));
1571	1592	return __ret;
1572	1593	}
1573	1594	#endif
...	...	@@ -1575,22 +1596,22 @@ __ai __attribute__((target("dotprod,neon"))) uint32x4_t __noswap_vdotq_u32(uint3
1575	1596	#ifdef __LITTLE_ENDIAN__
1576	1597	__ai __attribute__((target("dotprod,neon"))) int32x4_t vdotq_s32(int32x4_t __p0, int8x16_t __p1, int8x16_t __p2) {
1577	1598	int32x4_t __ret;
1578		__ret = (int32x4_t) __builtin_neon_vdotq_s32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 34);
	1599	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vdotq_s32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 34));
1579	1600	return __ret;
1580	1601	}
1581	1602	#else
1582	1603	__ai __attribute__((target("dotprod,neon"))) int32x4_t vdotq_s32(int32x4_t __p0, int8x16_t __p1, int8x16_t __p2) {
1583	1604	int32x4_t __ret;
1584		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
1585		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
1586		int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
1587		__ret = (int32x4_t) __builtin_neon_vdotq_s32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 34);
1588		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	1605	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	1606	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	1607	int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	1608	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vdotq_s32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 34));
	1609	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
1589	1610	return __ret;
1590	1611	}
1591	1612	__ai __attribute__((target("dotprod,neon"))) int32x4_t __noswap_vdotq_s32(int32x4_t __p0, int8x16_t __p1, int8x16_t __p2) {
1592	1613	int32x4_t __ret;
1593		__ret = (int32x4_t) __builtin_neon_vdotq_s32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 34);
	1614	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vdotq_s32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 34));
1594	1615	return __ret;
1595	1616	}
1596	1617	#endif
...	...	@@ -1598,22 +1619,22 @@ __ai __attribute__((target("dotprod,neon"))) int32x4_t __noswap_vdotq_s32(int32x
1598	1619	#ifdef __LITTLE_ENDIAN__
1599	1620	__ai __attribute__((target("dotprod,neon"))) uint32x2_t vdot_u32(uint32x2_t __p0, uint8x8_t __p1, uint8x8_t __p2) {
1600	1621	uint32x2_t __ret;
1601		__ret = (uint32x2_t) __builtin_neon_vdot_u32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 18);
	1622	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vdot_u32(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 18));
1602	1623	return __ret;
1603	1624	}
1604	1625	#else
1605	1626	__ai __attribute__((target("dotprod,neon"))) uint32x2_t vdot_u32(uint32x2_t __p0, uint8x8_t __p1, uint8x8_t __p2) {
1606	1627	uint32x2_t __ret;
1607		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
1608		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
1609		uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
1610		__ret = (uint32x2_t) __builtin_neon_vdot_u32((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 18);
1611		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	1628	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	1629	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	1630	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	1631	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vdot_u32(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 18));
	1632	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
1612	1633	return __ret;
1613	1634	}
1614	1635	__ai __attribute__((target("dotprod,neon"))) uint32x2_t __noswap_vdot_u32(uint32x2_t __p0, uint8x8_t __p1, uint8x8_t __p2) {
1615	1636	uint32x2_t __ret;
1616		__ret = (uint32x2_t) __builtin_neon_vdot_u32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 18);
	1637	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vdot_u32(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 18));
1617	1638	return __ret;
1618	1639	}
1619	1640	#endif
...	...	@@ -1621,22 +1642,22 @@ __ai __attribute__((target("dotprod,neon"))) uint32x2_t __noswap_vdot_u32(uint32
1621	1642	#ifdef __LITTLE_ENDIAN__
1622	1643	__ai __attribute__((target("dotprod,neon"))) int32x2_t vdot_s32(int32x2_t __p0, int8x8_t __p1, int8x8_t __p2) {
1623	1644	int32x2_t __ret;
1624		__ret = (int32x2_t) __builtin_neon_vdot_s32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 2);
	1645	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vdot_s32(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 2));
1625	1646	return __ret;
1626	1647	}
1627	1648	#else
1628	1649	__ai __attribute__((target("dotprod,neon"))) int32x2_t vdot_s32(int32x2_t __p0, int8x8_t __p1, int8x8_t __p2) {
1629	1650	int32x2_t __ret;
1630		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
1631		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
1632		int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
1633		__ret = (int32x2_t) __builtin_neon_vdot_s32((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 2);
1634		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	1651	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	1652	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	1653	int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	1654	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vdot_s32(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 2));
	1655	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
1635	1656	return __ret;
1636	1657	}
1637	1658	__ai __attribute__((target("dotprod,neon"))) int32x2_t __noswap_vdot_s32(int32x2_t __p0, int8x8_t __p1, int8x8_t __p2) {
1638	1659	int32x2_t __ret;
1639		__ret = (int32x2_t) __builtin_neon_vdot_s32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 2);
	1660	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vdot_s32(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 2));
1640	1661	return __ret;
1641	1662	}
1642	1663	#endif
...	...	@@ -1644,16 +1665,16 @@ __ai __attribute__((target("dotprod,neon"))) int32x2_t __noswap_vdot_s32(int32x2
1644	1665	#ifdef __LITTLE_ENDIAN__
1645	1666	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vabdq_f16(float16x8_t __p0, float16x8_t __p1) {
1646	1667	float16x8_t __ret;
1647		__ret = (float16x8_t) __builtin_neon_vabdq_f16((int8x16_t)__p0, (int8x16_t)__p1, 40);
	1668	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vabdq_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 40));
1648	1669	return __ret;
1649	1670	}
1650	1671	#else
1651	1672	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vabdq_f16(float16x8_t __p0, float16x8_t __p1) {
1652	1673	float16x8_t __ret;
1653		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
1654		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
1655		__ret = (float16x8_t) __builtin_neon_vabdq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 40);
1656		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	1674	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	1675	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	1676	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vabdq_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 40));
	1677	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
1657	1678	return __ret;
1658	1679	}
1659	1680	#endif
...	...	@@ -1661,16 +1682,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vabdq_f16(float16x8_t
1661	1682	#ifdef __LITTLE_ENDIAN__
1662	1683	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vabd_f16(float16x4_t __p0, float16x4_t __p1) {
1663	1684	float16x4_t __ret;
1664		__ret = (float16x4_t) __builtin_neon_vabd_f16((int8x8_t)__p0, (int8x8_t)__p1, 8);
	1685	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vabd_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 8));
1665	1686	return __ret;
1666	1687	}
1667	1688	#else
1668	1689	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vabd_f16(float16x4_t __p0, float16x4_t __p1) {
1669	1690	float16x4_t __ret;
1670		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
1671		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
1672		__ret = (float16x4_t) __builtin_neon_vabd_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 8);
1673		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	1691	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	1692	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	1693	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vabd_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 8));
	1694	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
1674	1695	return __ret;
1675	1696	}
1676	1697	#endif
...	...	@@ -1678,15 +1699,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vabd_f16(float16x4_t _
1678	1699	#ifdef __LITTLE_ENDIAN__
1679	1700	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vabsq_f16(float16x8_t __p0) {
1680	1701	float16x8_t __ret;
1681		__ret = (float16x8_t) __builtin_neon_vabsq_f16((int8x16_t)__p0, 40);
	1702	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vabsq_f16(__builtin_bit_cast(int8x16_t, __p0), 40));
1682	1703	return __ret;
1683	1704	}
1684	1705	#else
1685	1706	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vabsq_f16(float16x8_t __p0) {
1686	1707	float16x8_t __ret;
1687		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
1688		__ret = (float16x8_t) __builtin_neon_vabsq_f16((int8x16_t)__rev0, 40);
1689		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	1708	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	1709	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vabsq_f16(__builtin_bit_cast(int8x16_t, __rev0), 40));
	1710	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
1690	1711	return __ret;
1691	1712	}
1692	1713	#endif
...	...	@@ -1694,15 +1715,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vabsq_f16(float16x8_t
1694	1715	#ifdef __LITTLE_ENDIAN__
1695	1716	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vabs_f16(float16x4_t __p0) {
1696	1717	float16x4_t __ret;
1697		__ret = (float16x4_t) __builtin_neon_vabs_f16((int8x8_t)__p0, 8);
	1718	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vabs_f16(__builtin_bit_cast(int8x8_t, __p0), 8));
1698	1719	return __ret;
1699	1720	}
1700	1721	#else
1701	1722	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vabs_f16(float16x4_t __p0) {
1702	1723	float16x4_t __ret;
1703		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
1704		__ret = (float16x4_t) __builtin_neon_vabs_f16((int8x8_t)__rev0, 8);
1705		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	1724	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	1725	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vabs_f16(__builtin_bit_cast(int8x8_t, __rev0), 8));
	1726	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
1706	1727	return __ret;
1707	1728	}
1708	1729	#endif
...	...	@@ -1716,10 +1737,10 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vaddq_f16(float16x8_t
1716	1737	#else
1717	1738	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vaddq_f16(float16x8_t __p0, float16x8_t __p1) {
1718	1739	float16x8_t __ret;
1719		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
1720		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	1740	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	1741	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
1721	1742	__ret = __rev0 + __rev1;
1722		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	1743	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
1723	1744	return __ret;
1724	1745	}
1725	1746	#endif
...	...	@@ -1733,10 +1754,10 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vadd_f16(float16x4_t _
1733	1754	#else
1734	1755	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vadd_f16(float16x4_t __p0, float16x4_t __p1) {
1735	1756	float16x4_t __ret;
1736		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
1737		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	1757	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	1758	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
1738	1759	__ret = __rev0 + __rev1;
1739		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	1760	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
1740	1761	return __ret;
1741	1762	}
1742	1763	#endif
...	...	@@ -1744,16 +1765,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vadd_f16(float16x4_t _
1744	1765	#ifdef __LITTLE_ENDIAN__
1745	1766	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcageq_f16(float16x8_t __p0, float16x8_t __p1) {
1746	1767	uint16x8_t __ret;
1747		__ret = (uint16x8_t) __builtin_neon_vcageq_f16((int8x16_t)__p0, (int8x16_t)__p1, 49);
	1768	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcageq_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 49));
1748	1769	return __ret;
1749	1770	}
1750	1771	#else
1751	1772	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcageq_f16(float16x8_t __p0, float16x8_t __p1) {
1752	1773	uint16x8_t __ret;
1753		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
1754		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
1755		__ret = (uint16x8_t) __builtin_neon_vcageq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
1756		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	1774	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	1775	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	1776	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcageq_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 49));
	1777	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
1757	1778	return __ret;
1758	1779	}
1759	1780	#endif
...	...	@@ -1761,16 +1782,16 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcageq_f16(float16x8_t
1761	1782	#ifdef __LITTLE_ENDIAN__
1762	1783	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcage_f16(float16x4_t __p0, float16x4_t __p1) {
1763	1784	uint16x4_t __ret;
1764		__ret = (uint16x4_t) __builtin_neon_vcage_f16((int8x8_t)__p0, (int8x8_t)__p1, 17);
	1785	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcage_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 17));
1765	1786	return __ret;
1766	1787	}
1767	1788	#else
1768	1789	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcage_f16(float16x4_t __p0, float16x4_t __p1) {
1769	1790	uint16x4_t __ret;
1770		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
1771		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
1772		__ret = (uint16x4_t) __builtin_neon_vcage_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
1773		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	1791	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	1792	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	1793	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcage_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 17));
	1794	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
1774	1795	return __ret;
1775	1796	}
1776	1797	#endif
...	...	@@ -1778,16 +1799,16 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcage_f16(float16x4_t _
1778	1799	#ifdef __LITTLE_ENDIAN__
1779	1800	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcagtq_f16(float16x8_t __p0, float16x8_t __p1) {
1780	1801	uint16x8_t __ret;
1781		__ret = (uint16x8_t) __builtin_neon_vcagtq_f16((int8x16_t)__p0, (int8x16_t)__p1, 49);
	1802	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcagtq_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 49));
1782	1803	return __ret;
1783	1804	}
1784	1805	#else
1785	1806	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcagtq_f16(float16x8_t __p0, float16x8_t __p1) {
1786	1807	uint16x8_t __ret;
1787		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
1788		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
1789		__ret = (uint16x8_t) __builtin_neon_vcagtq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
1790		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	1808	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	1809	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	1810	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcagtq_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 49));
	1811	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
1791	1812	return __ret;
1792	1813	}
1793	1814	#endif
...	...	@@ -1795,16 +1816,16 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcagtq_f16(float16x8_t
1795	1816	#ifdef __LITTLE_ENDIAN__
1796	1817	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcagt_f16(float16x4_t __p0, float16x4_t __p1) {
1797	1818	uint16x4_t __ret;
1798		__ret = (uint16x4_t) __builtin_neon_vcagt_f16((int8x8_t)__p0, (int8x8_t)__p1, 17);
	1819	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcagt_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 17));
1799	1820	return __ret;
1800	1821	}
1801	1822	#else
1802	1823	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcagt_f16(float16x4_t __p0, float16x4_t __p1) {
1803	1824	uint16x4_t __ret;
1804		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
1805		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
1806		__ret = (uint16x4_t) __builtin_neon_vcagt_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
1807		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	1825	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	1826	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	1827	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcagt_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 17));
	1828	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
1808	1829	return __ret;
1809	1830	}
1810	1831	#endif
...	...	@@ -1812,16 +1833,16 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcagt_f16(float16x4_t _
1812	1833	#ifdef __LITTLE_ENDIAN__
1813	1834	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcaleq_f16(float16x8_t __p0, float16x8_t __p1) {
1814	1835	uint16x8_t __ret;
1815		__ret = (uint16x8_t) __builtin_neon_vcaleq_f16((int8x16_t)__p0, (int8x16_t)__p1, 49);
	1836	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcaleq_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 49));
1816	1837	return __ret;
1817	1838	}
1818	1839	#else
1819	1840	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcaleq_f16(float16x8_t __p0, float16x8_t __p1) {
1820	1841	uint16x8_t __ret;
1821		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
1822		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
1823		__ret = (uint16x8_t) __builtin_neon_vcaleq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
1824		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	1842	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	1843	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	1844	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcaleq_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 49));
	1845	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
1825	1846	return __ret;
1826	1847	}
1827	1848	#endif
...	...	@@ -1829,16 +1850,16 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcaleq_f16(float16x8_t
1829	1850	#ifdef __LITTLE_ENDIAN__
1830	1851	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcale_f16(float16x4_t __p0, float16x4_t __p1) {
1831	1852	uint16x4_t __ret;
1832		__ret = (uint16x4_t) __builtin_neon_vcale_f16((int8x8_t)__p0, (int8x8_t)__p1, 17);
	1853	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcale_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 17));
1833	1854	return __ret;
1834	1855	}
1835	1856	#else
1836	1857	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcale_f16(float16x4_t __p0, float16x4_t __p1) {
1837	1858	uint16x4_t __ret;
1838		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
1839		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
1840		__ret = (uint16x4_t) __builtin_neon_vcale_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
1841		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	1859	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	1860	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	1861	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcale_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 17));
	1862	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
1842	1863	return __ret;
1843	1864	}
1844	1865	#endif
...	...	@@ -1846,16 +1867,16 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcale_f16(float16x4_t _
1846	1867	#ifdef __LITTLE_ENDIAN__
1847	1868	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcaltq_f16(float16x8_t __p0, float16x8_t __p1) {
1848	1869	uint16x8_t __ret;
1849		__ret = (uint16x8_t) __builtin_neon_vcaltq_f16((int8x16_t)__p0, (int8x16_t)__p1, 49);
	1870	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcaltq_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 49));
1850	1871	return __ret;
1851	1872	}
1852	1873	#else
1853	1874	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcaltq_f16(float16x8_t __p0, float16x8_t __p1) {
1854	1875	uint16x8_t __ret;
1855		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
1856		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
1857		__ret = (uint16x8_t) __builtin_neon_vcaltq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
1858		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	1876	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	1877	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	1878	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcaltq_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 49));
	1879	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
1859	1880	return __ret;
1860	1881	}
1861	1882	#endif
...	...	@@ -1863,16 +1884,16 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcaltq_f16(float16x8_t
1863	1884	#ifdef __LITTLE_ENDIAN__
1864	1885	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcalt_f16(float16x4_t __p0, float16x4_t __p1) {
1865	1886	uint16x4_t __ret;
1866		__ret = (uint16x4_t) __builtin_neon_vcalt_f16((int8x8_t)__p0, (int8x8_t)__p1, 17);
	1887	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcalt_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 17));
1867	1888	return __ret;
1868	1889	}
1869	1890	#else
1870	1891	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcalt_f16(float16x4_t __p0, float16x4_t __p1) {
1871	1892	uint16x4_t __ret;
1872		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
1873		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
1874		__ret = (uint16x4_t) __builtin_neon_vcalt_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
1875		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	1893	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	1894	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	1895	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcalt_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 17));
	1896	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
1876	1897	return __ret;
1877	1898	}
1878	1899	#endif
...	...	@@ -1880,16 +1901,16 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcalt_f16(float16x4_t _
1880	1901	#ifdef __LITTLE_ENDIAN__
1881	1902	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vceqq_f16(float16x8_t __p0, float16x8_t __p1) {
1882	1903	uint16x8_t __ret;
1883		__ret = (uint16x8_t)(__p0 == __p1);
	1904	__ret = __builtin_bit_cast(uint16x8_t, __p0 == __p1);
1884	1905	return __ret;
1885	1906	}
1886	1907	#else
1887	1908	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vceqq_f16(float16x8_t __p0, float16x8_t __p1) {
1888	1909	uint16x8_t __ret;
1889		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
1890		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
1891		__ret = (uint16x8_t)(__rev0 == __rev1);
1892		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	1910	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	1911	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	1912	__ret = __builtin_bit_cast(uint16x8_t, __rev0 == __rev1);
	1913	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
1893	1914	return __ret;
1894	1915	}
1895	1916	#endif
...	...	@@ -1897,16 +1918,16 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x8_t vceqq_f16(float16x8_t _
1897	1918	#ifdef __LITTLE_ENDIAN__
1898	1919	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vceq_f16(float16x4_t __p0, float16x4_t __p1) {
1899	1920	uint16x4_t __ret;
1900		__ret = (uint16x4_t)(__p0 == __p1);
	1921	__ret = __builtin_bit_cast(uint16x4_t, __p0 == __p1);
1901	1922	return __ret;
1902	1923	}
1903	1924	#else
1904	1925	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vceq_f16(float16x4_t __p0, float16x4_t __p1) {
1905	1926	uint16x4_t __ret;
1906		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
1907		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
1908		__ret = (uint16x4_t)(__rev0 == __rev1);
1909		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	1927	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	1928	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	1929	__ret = __builtin_bit_cast(uint16x4_t, __rev0 == __rev1);
	1930	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
1910	1931	return __ret;
1911	1932	}
1912	1933	#endif
...	...	@@ -1914,15 +1935,15 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x4_t vceq_f16(float16x4_t __
1914	1935	#ifdef __LITTLE_ENDIAN__
1915	1936	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vceqzq_f16(float16x8_t __p0) {
1916	1937	uint16x8_t __ret;
1917		__ret = (uint16x8_t) __builtin_neon_vceqzq_f16((int8x16_t)__p0, 49);
	1938	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vceqzq_f16(__builtin_bit_cast(int8x16_t, __p0), 40));
1918	1939	return __ret;
1919	1940	}
1920	1941	#else
1921	1942	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vceqzq_f16(float16x8_t __p0) {
1922	1943	uint16x8_t __ret;
1923		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
1924		__ret = (uint16x8_t) __builtin_neon_vceqzq_f16((int8x16_t)__rev0, 49);
1925		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	1944	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	1945	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vceqzq_f16(__builtin_bit_cast(int8x16_t, __rev0), 40));
	1946	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
1926	1947	return __ret;
1927	1948	}
1928	1949	#endif
...	...	@@ -1930,15 +1951,15 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x8_t vceqzq_f16(float16x8_t
1930	1951	#ifdef __LITTLE_ENDIAN__
1931	1952	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vceqz_f16(float16x4_t __p0) {
1932	1953	uint16x4_t __ret;
1933		__ret = (uint16x4_t) __builtin_neon_vceqz_f16((int8x8_t)__p0, 17);
	1954	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vceqz_f16(__builtin_bit_cast(int8x8_t, __p0), 8));
1934	1955	return __ret;
1935	1956	}
1936	1957	#else
1937	1958	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vceqz_f16(float16x4_t __p0) {
1938	1959	uint16x4_t __ret;
1939		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
1940		__ret = (uint16x4_t) __builtin_neon_vceqz_f16((int8x8_t)__rev0, 17);
1941		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	1960	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	1961	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vceqz_f16(__builtin_bit_cast(int8x8_t, __rev0), 8));
	1962	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
1942	1963	return __ret;
1943	1964	}
1944	1965	#endif
...	...	@@ -1946,16 +1967,16 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x4_t vceqz_f16(float16x4_t _
1946	1967	#ifdef __LITTLE_ENDIAN__
1947	1968	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcgeq_f16(float16x8_t __p0, float16x8_t __p1) {
1948	1969	uint16x8_t __ret;
1949		__ret = (uint16x8_t)(__p0 >= __p1);
	1970	__ret = __builtin_bit_cast(uint16x8_t, __p0 >= __p1);
1950	1971	return __ret;
1951	1972	}
1952	1973	#else
1953	1974	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcgeq_f16(float16x8_t __p0, float16x8_t __p1) {
1954	1975	uint16x8_t __ret;
1955		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
1956		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
1957		__ret = (uint16x8_t)(__rev0 >= __rev1);
1958		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	1976	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	1977	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	1978	__ret = __builtin_bit_cast(uint16x8_t, __rev0 >= __rev1);
	1979	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
1959	1980	return __ret;
1960	1981	}
1961	1982	#endif
...	...	@@ -1963,16 +1984,16 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcgeq_f16(float16x8_t _
1963	1984	#ifdef __LITTLE_ENDIAN__
1964	1985	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcge_f16(float16x4_t __p0, float16x4_t __p1) {
1965	1986	uint16x4_t __ret;
1966		__ret = (uint16x4_t)(__p0 >= __p1);
	1987	__ret = __builtin_bit_cast(uint16x4_t, __p0 >= __p1);
1967	1988	return __ret;
1968	1989	}
1969	1990	#else
1970	1991	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcge_f16(float16x4_t __p0, float16x4_t __p1) {
1971	1992	uint16x4_t __ret;
1972		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
1973		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
1974		__ret = (uint16x4_t)(__rev0 >= __rev1);
1975		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	1993	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	1994	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	1995	__ret = __builtin_bit_cast(uint16x4_t, __rev0 >= __rev1);
	1996	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
1976	1997	return __ret;
1977	1998	}
1978	1999	#endif
...	...	@@ -1980,15 +2001,15 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcge_f16(float16x4_t __
1980	2001	#ifdef __LITTLE_ENDIAN__
1981	2002	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcgezq_f16(float16x8_t __p0) {
1982	2003	uint16x8_t __ret;
1983		__ret = (uint16x8_t) __builtin_neon_vcgezq_f16((int8x16_t)__p0, 49);
	2004	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcgezq_f16(__builtin_bit_cast(int8x16_t, __p0), 40));
1984	2005	return __ret;
1985	2006	}
1986	2007	#else
1987	2008	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcgezq_f16(float16x8_t __p0) {
1988	2009	uint16x8_t __ret;
1989		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
1990		__ret = (uint16x8_t) __builtin_neon_vcgezq_f16((int8x16_t)__rev0, 49);
1991		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	2010	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	2011	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcgezq_f16(__builtin_bit_cast(int8x16_t, __rev0), 40));
	2012	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
1992	2013	return __ret;
1993	2014	}
1994	2015	#endif
...	...	@@ -1996,15 +2017,15 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcgezq_f16(float16x8_t
1996	2017	#ifdef __LITTLE_ENDIAN__
1997	2018	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcgez_f16(float16x4_t __p0) {
1998	2019	uint16x4_t __ret;
1999		__ret = (uint16x4_t) __builtin_neon_vcgez_f16((int8x8_t)__p0, 17);
	2020	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcgez_f16(__builtin_bit_cast(int8x8_t, __p0), 8));
2000	2021	return __ret;
2001	2022	}
2002	2023	#else
2003	2024	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcgez_f16(float16x4_t __p0) {
2004	2025	uint16x4_t __ret;
2005		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2006		__ret = (uint16x4_t) __builtin_neon_vcgez_f16((int8x8_t)__rev0, 17);
2007		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	2026	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	2027	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcgez_f16(__builtin_bit_cast(int8x8_t, __rev0), 8));
	2028	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
2008	2029	return __ret;
2009	2030	}
2010	2031	#endif
...	...	@@ -2012,16 +2033,16 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcgez_f16(float16x4_t _
2012	2033	#ifdef __LITTLE_ENDIAN__
2013	2034	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcgtq_f16(float16x8_t __p0, float16x8_t __p1) {
2014	2035	uint16x8_t __ret;
2015		__ret = (uint16x8_t)(__p0 > __p1);
	2036	__ret = __builtin_bit_cast(uint16x8_t, __p0 > __p1);
2016	2037	return __ret;
2017	2038	}
2018	2039	#else
2019	2040	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcgtq_f16(float16x8_t __p0, float16x8_t __p1) {
2020	2041	uint16x8_t __ret;
2021		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2022		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
2023		__ret = (uint16x8_t)(__rev0 > __rev1);
2024		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	2042	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	2043	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	2044	__ret = __builtin_bit_cast(uint16x8_t, __rev0 > __rev1);
	2045	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
2025	2046	return __ret;
2026	2047	}
2027	2048	#endif
...	...	@@ -2029,16 +2050,16 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcgtq_f16(float16x8_t _
2029	2050	#ifdef __LITTLE_ENDIAN__
2030	2051	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcgt_f16(float16x4_t __p0, float16x4_t __p1) {
2031	2052	uint16x4_t __ret;
2032		__ret = (uint16x4_t)(__p0 > __p1);
	2053	__ret = __builtin_bit_cast(uint16x4_t, __p0 > __p1);
2033	2054	return __ret;
2034	2055	}
2035	2056	#else
2036	2057	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcgt_f16(float16x4_t __p0, float16x4_t __p1) {
2037	2058	uint16x4_t __ret;
2038		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2039		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
2040		__ret = (uint16x4_t)(__rev0 > __rev1);
2041		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	2059	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	2060	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	2061	__ret = __builtin_bit_cast(uint16x4_t, __rev0 > __rev1);
	2062	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
2042	2063	return __ret;
2043	2064	}
2044	2065	#endif
...	...	@@ -2046,15 +2067,15 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcgt_f16(float16x4_t __
2046	2067	#ifdef __LITTLE_ENDIAN__
2047	2068	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcgtzq_f16(float16x8_t __p0) {
2048	2069	uint16x8_t __ret;
2049		__ret = (uint16x8_t) __builtin_neon_vcgtzq_f16((int8x16_t)__p0, 49);
	2070	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcgtzq_f16(__builtin_bit_cast(int8x16_t, __p0), 40));
2050	2071	return __ret;
2051	2072	}
2052	2073	#else
2053	2074	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcgtzq_f16(float16x8_t __p0) {
2054	2075	uint16x8_t __ret;
2055		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2056		__ret = (uint16x8_t) __builtin_neon_vcgtzq_f16((int8x16_t)__rev0, 49);
2057		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	2076	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	2077	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcgtzq_f16(__builtin_bit_cast(int8x16_t, __rev0), 40));
	2078	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
2058	2079	return __ret;
2059	2080	}
2060	2081	#endif
...	...	@@ -2062,15 +2083,15 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcgtzq_f16(float16x8_t
2062	2083	#ifdef __LITTLE_ENDIAN__
2063	2084	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcgtz_f16(float16x4_t __p0) {
2064	2085	uint16x4_t __ret;
2065		__ret = (uint16x4_t) __builtin_neon_vcgtz_f16((int8x8_t)__p0, 17);
	2086	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcgtz_f16(__builtin_bit_cast(int8x8_t, __p0), 8));
2066	2087	return __ret;
2067	2088	}
2068	2089	#else
2069	2090	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcgtz_f16(float16x4_t __p0) {
2070	2091	uint16x4_t __ret;
2071		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2072		__ret = (uint16x4_t) __builtin_neon_vcgtz_f16((int8x8_t)__rev0, 17);
2073		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	2092	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	2093	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcgtz_f16(__builtin_bit_cast(int8x8_t, __rev0), 8));
	2094	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
2074	2095	return __ret;
2075	2096	}
2076	2097	#endif
...	...	@@ -2078,16 +2099,16 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcgtz_f16(float16x4_t _
2078	2099	#ifdef __LITTLE_ENDIAN__
2079	2100	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcleq_f16(float16x8_t __p0, float16x8_t __p1) {
2080	2101	uint16x8_t __ret;
2081		__ret = (uint16x8_t)(__p0 <= __p1);
	2102	__ret = __builtin_bit_cast(uint16x8_t, __p0 <= __p1);
2082	2103	return __ret;
2083	2104	}
2084	2105	#else
2085	2106	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcleq_f16(float16x8_t __p0, float16x8_t __p1) {
2086	2107	uint16x8_t __ret;
2087		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2088		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
2089		__ret = (uint16x8_t)(__rev0 <= __rev1);
2090		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	2108	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	2109	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	2110	__ret = __builtin_bit_cast(uint16x8_t, __rev0 <= __rev1);
	2111	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
2091	2112	return __ret;
2092	2113	}
2093	2114	#endif
...	...	@@ -2095,16 +2116,16 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcleq_f16(float16x8_t _
2095	2116	#ifdef __LITTLE_ENDIAN__
2096	2117	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcle_f16(float16x4_t __p0, float16x4_t __p1) {
2097	2118	uint16x4_t __ret;
2098		__ret = (uint16x4_t)(__p0 <= __p1);
	2119	__ret = __builtin_bit_cast(uint16x4_t, __p0 <= __p1);
2099	2120	return __ret;
2100	2121	}
2101	2122	#else
2102	2123	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcle_f16(float16x4_t __p0, float16x4_t __p1) {
2103	2124	uint16x4_t __ret;
2104		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2105		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
2106		__ret = (uint16x4_t)(__rev0 <= __rev1);
2107		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	2125	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	2126	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	2127	__ret = __builtin_bit_cast(uint16x4_t, __rev0 <= __rev1);
	2128	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
2108	2129	return __ret;
2109	2130	}
2110	2131	#endif
...	...	@@ -2112,15 +2133,15 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcle_f16(float16x4_t __
2112	2133	#ifdef __LITTLE_ENDIAN__
2113	2134	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vclezq_f16(float16x8_t __p0) {
2114	2135	uint16x8_t __ret;
2115		__ret = (uint16x8_t) __builtin_neon_vclezq_f16((int8x16_t)__p0, 49);
	2136	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vclezq_f16(__builtin_bit_cast(int8x16_t, __p0), 40));
2116	2137	return __ret;
2117	2138	}
2118	2139	#else
2119	2140	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vclezq_f16(float16x8_t __p0) {
2120	2141	uint16x8_t __ret;
2121		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2122		__ret = (uint16x8_t) __builtin_neon_vclezq_f16((int8x16_t)__rev0, 49);
2123		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	2142	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	2143	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vclezq_f16(__builtin_bit_cast(int8x16_t, __rev0), 40));
	2144	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
2124	2145	return __ret;
2125	2146	}
2126	2147	#endif
...	...	@@ -2128,15 +2149,15 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x8_t vclezq_f16(float16x8_t
2128	2149	#ifdef __LITTLE_ENDIAN__
2129	2150	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vclez_f16(float16x4_t __p0) {
2130	2151	uint16x4_t __ret;
2131		__ret = (uint16x4_t) __builtin_neon_vclez_f16((int8x8_t)__p0, 17);
	2152	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vclez_f16(__builtin_bit_cast(int8x8_t, __p0), 8));
2132	2153	return __ret;
2133	2154	}
2134	2155	#else
2135	2156	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vclez_f16(float16x4_t __p0) {
2136	2157	uint16x4_t __ret;
2137		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2138		__ret = (uint16x4_t) __builtin_neon_vclez_f16((int8x8_t)__rev0, 17);
2139		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	2158	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	2159	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vclez_f16(__builtin_bit_cast(int8x8_t, __rev0), 8));
	2160	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
2140	2161	return __ret;
2141	2162	}
2142	2163	#endif
...	...	@@ -2144,16 +2165,16 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x4_t vclez_f16(float16x4_t _
2144	2165	#ifdef __LITTLE_ENDIAN__
2145	2166	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcltq_f16(float16x8_t __p0, float16x8_t __p1) {
2146	2167	uint16x8_t __ret;
2147		__ret = (uint16x8_t)(__p0 < __p1);
	2168	__ret = __builtin_bit_cast(uint16x8_t, __p0 < __p1);
2148	2169	return __ret;
2149	2170	}
2150	2171	#else
2151	2172	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcltq_f16(float16x8_t __p0, float16x8_t __p1) {
2152	2173	uint16x8_t __ret;
2153		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2154		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
2155		__ret = (uint16x8_t)(__rev0 < __rev1);
2156		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	2174	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	2175	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	2176	__ret = __builtin_bit_cast(uint16x8_t, __rev0 < __rev1);
	2177	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
2157	2178	return __ret;
2158	2179	}
2159	2180	#endif
...	...	@@ -2161,16 +2182,16 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcltq_f16(float16x8_t _
2161	2182	#ifdef __LITTLE_ENDIAN__
2162	2183	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vclt_f16(float16x4_t __p0, float16x4_t __p1) {
2163	2184	uint16x4_t __ret;
2164		__ret = (uint16x4_t)(__p0 < __p1);
	2185	__ret = __builtin_bit_cast(uint16x4_t, __p0 < __p1);
2165	2186	return __ret;
2166	2187	}
2167	2188	#else
2168	2189	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vclt_f16(float16x4_t __p0, float16x4_t __p1) {
2169	2190	uint16x4_t __ret;
2170		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2171		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
2172		__ret = (uint16x4_t)(__rev0 < __rev1);
2173		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	2191	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	2192	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	2193	__ret = __builtin_bit_cast(uint16x4_t, __rev0 < __rev1);
	2194	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
2174	2195	return __ret;
2175	2196	}
2176	2197	#endif
...	...	@@ -2178,15 +2199,15 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x4_t vclt_f16(float16x4_t __
2178	2199	#ifdef __LITTLE_ENDIAN__
2179	2200	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcltzq_f16(float16x8_t __p0) {
2180	2201	uint16x8_t __ret;
2181		__ret = (uint16x8_t) __builtin_neon_vcltzq_f16((int8x16_t)__p0, 49);
	2202	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcltzq_f16(__builtin_bit_cast(int8x16_t, __p0), 40));
2182	2203	return __ret;
2183	2204	}
2184	2205	#else
2185	2206	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcltzq_f16(float16x8_t __p0) {
2186	2207	uint16x8_t __ret;
2187		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2188		__ret = (uint16x8_t) __builtin_neon_vcltzq_f16((int8x16_t)__rev0, 49);
2189		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	2208	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	2209	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcltzq_f16(__builtin_bit_cast(int8x16_t, __rev0), 40));
	2210	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
2190	2211	return __ret;
2191	2212	}
2192	2213	#endif
...	...	@@ -2194,15 +2215,15 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcltzq_f16(float16x8_t
2194	2215	#ifdef __LITTLE_ENDIAN__
2195	2216	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcltz_f16(float16x4_t __p0) {
2196	2217	uint16x4_t __ret;
2197		__ret = (uint16x4_t) __builtin_neon_vcltz_f16((int8x8_t)__p0, 17);
	2218	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcltz_f16(__builtin_bit_cast(int8x8_t, __p0), 8));
2198	2219	return __ret;
2199	2220	}
2200	2221	#else
2201	2222	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcltz_f16(float16x4_t __p0) {
2202	2223	uint16x4_t __ret;
2203		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2204		__ret = (uint16x4_t) __builtin_neon_vcltz_f16((int8x8_t)__rev0, 17);
2205		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	2224	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	2225	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcltz_f16(__builtin_bit_cast(int8x8_t, __rev0), 8));
	2226	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
2206	2227	return __ret;
2207	2228	}
2208	2229	#endif
...	...	@@ -2210,15 +2231,15 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcltz_f16(float16x4_t _
2210	2231	#ifdef __LITTLE_ENDIAN__
2211	2232	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vcvtq_f16_u16(uint16x8_t __p0) {
2212	2233	float16x8_t __ret;
2213		__ret = (float16x8_t) __builtin_neon_vcvtq_f16_u16((int8x16_t)__p0, 49);
	2234	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcvtq_f16_u16(__builtin_bit_cast(int8x16_t, __p0), 49));
2214	2235	return __ret;
2215	2236	}
2216	2237	#else
2217	2238	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vcvtq_f16_u16(uint16x8_t __p0) {
2218	2239	float16x8_t __ret;
2219		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2220		__ret = (float16x8_t) __builtin_neon_vcvtq_f16_u16((int8x16_t)__rev0, 49);
2221		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	2240	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	2241	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcvtq_f16_u16(__builtin_bit_cast(int8x16_t, __rev0), 49));
	2242	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
2222	2243	return __ret;
2223	2244	}
2224	2245	#endif
...	...	@@ -2226,15 +2247,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vcvtq_f16_u16(uint16x8
2226	2247	#ifdef __LITTLE_ENDIAN__
2227	2248	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vcvtq_f16_s16(int16x8_t __p0) {
2228	2249	float16x8_t __ret;
2229		__ret = (float16x8_t) __builtin_neon_vcvtq_f16_s16((int8x16_t)__p0, 33);
	2250	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcvtq_f16_s16(__builtin_bit_cast(int8x16_t, __p0), 33));
2230	2251	return __ret;
2231	2252	}
2232	2253	#else
2233	2254	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vcvtq_f16_s16(int16x8_t __p0) {
2234	2255	float16x8_t __ret;
2235		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2236		__ret = (float16x8_t) __builtin_neon_vcvtq_f16_s16((int8x16_t)__rev0, 33);
2237		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	2256	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	2257	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcvtq_f16_s16(__builtin_bit_cast(int8x16_t, __rev0), 33));
	2258	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
2238	2259	return __ret;
2239	2260	}
2240	2261	#endif
...	...	@@ -2242,15 +2263,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vcvtq_f16_s16(int16x8_
2242	2263	#ifdef __LITTLE_ENDIAN__
2243	2264	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vcvt_f16_u16(uint16x4_t __p0) {
2244	2265	float16x4_t __ret;
2245		__ret = (float16x4_t) __builtin_neon_vcvt_f16_u16((int8x8_t)__p0, 17);
	2266	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vcvt_f16_u16(__builtin_bit_cast(int8x8_t, __p0), 17));
2246	2267	return __ret;
2247	2268	}
2248	2269	#else
2249	2270	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vcvt_f16_u16(uint16x4_t __p0) {
2250	2271	float16x4_t __ret;
2251		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2252		__ret = (float16x4_t) __builtin_neon_vcvt_f16_u16((int8x8_t)__rev0, 17);
2253		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	2272	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	2273	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vcvt_f16_u16(__builtin_bit_cast(int8x8_t, __rev0), 17));
	2274	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
2254	2275	return __ret;
2255	2276	}
2256	2277	#endif
...	...	@@ -2258,15 +2279,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vcvt_f16_u16(uint16x4_
2258	2279	#ifdef __LITTLE_ENDIAN__
2259	2280	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vcvt_f16_s16(int16x4_t __p0) {
2260	2281	float16x4_t __ret;
2261		__ret = (float16x4_t) __builtin_neon_vcvt_f16_s16((int8x8_t)__p0, 1);
	2282	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vcvt_f16_s16(__builtin_bit_cast(int8x8_t, __p0), 1));
2262	2283	return __ret;
2263	2284	}
2264	2285	#else
2265	2286	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vcvt_f16_s16(int16x4_t __p0) {
2266	2287	float16x4_t __ret;
2267		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2268		__ret = (float16x4_t) __builtin_neon_vcvt_f16_s16((int8x8_t)__rev0, 1);
2269		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	2288	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	2289	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vcvt_f16_s16(__builtin_bit_cast(int8x8_t, __rev0), 1));
	2290	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
2270	2291	return __ret;
2271	2292	}
2272	2293	#endif
...	...	@@ -2275,16 +2296,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vcvt_f16_s16(int16x4_t
2275	2296	#define vcvtq_n_f16_u16(__p0, __p1) __extension__ ({ \
2276	2297	float16x8_t __ret; \
2277	2298	uint16x8_t __s0 = __p0; \
2278		__ret = (float16x8_t) __builtin_neon_vcvtq_n_f16_u16((int8x16_t)__s0, __p1, 49); \
	2299	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcvtq_n_f16_u16(__builtin_bit_cast(int8x16_t, __s0), __p1, 49)); \
2279	2300	__ret; \
2280	2301	})
2281	2302	#else
2282	2303	#define vcvtq_n_f16_u16(__p0, __p1) __extension__ ({ \
2283	2304	float16x8_t __ret; \
2284	2305	uint16x8_t __s0 = __p0; \
2285		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
2286		__ret = (float16x8_t) __builtin_neon_vcvtq_n_f16_u16((int8x16_t)__rev0, __p1, 49); \
2287		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	2306	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	2307	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcvtq_n_f16_u16(__builtin_bit_cast(int8x16_t, __rev0), __p1, 49)); \
	2308	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
2288	2309	__ret; \
2289	2310	})
2290	2311	#endif
...	...	@@ -2293,16 +2314,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vcvt_f16_s16(int16x4_t
2293	2314	#define vcvtq_n_f16_s16(__p0, __p1) __extension__ ({ \
2294	2315	float16x8_t __ret; \
2295	2316	int16x8_t __s0 = __p0; \
2296		__ret = (float16x8_t) __builtin_neon_vcvtq_n_f16_s16((int8x16_t)__s0, __p1, 33); \
	2317	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcvtq_n_f16_s16(__builtin_bit_cast(int8x16_t, __s0), __p1, 33)); \
2297	2318	__ret; \
2298	2319	})
2299	2320	#else
2300	2321	#define vcvtq_n_f16_s16(__p0, __p1) __extension__ ({ \
2301	2322	float16x8_t __ret; \
2302	2323	int16x8_t __s0 = __p0; \
2303		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
2304		__ret = (float16x8_t) __builtin_neon_vcvtq_n_f16_s16((int8x16_t)__rev0, __p1, 33); \
2305		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	2324	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	2325	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcvtq_n_f16_s16(__builtin_bit_cast(int8x16_t, __rev0), __p1, 33)); \
	2326	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
2306	2327	__ret; \
2307	2328	})
2308	2329	#endif
...	...	@@ -2311,16 +2332,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vcvt_f16_s16(int16x4_t
2311	2332	#define vcvt_n_f16_u16(__p0, __p1) __extension__ ({ \
2312	2333	float16x4_t __ret; \
2313	2334	uint16x4_t __s0 = __p0; \
2314		__ret = (float16x4_t) __builtin_neon_vcvt_n_f16_u16((int8x8_t)__s0, __p1, 17); \
	2335	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vcvt_n_f16_u16(__builtin_bit_cast(int8x8_t, __s0), __p1, 17)); \
2315	2336	__ret; \
2316	2337	})
2317	2338	#else
2318	2339	#define vcvt_n_f16_u16(__p0, __p1) __extension__ ({ \
2319	2340	float16x4_t __ret; \
2320	2341	uint16x4_t __s0 = __p0; \
2321		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
2322		__ret = (float16x4_t) __builtin_neon_vcvt_n_f16_u16((int8x8_t)__rev0, __p1, 17); \
2323		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	2342	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	2343	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vcvt_n_f16_u16(__builtin_bit_cast(int8x8_t, __rev0), __p1, 17)); \
	2344	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
2324	2345	__ret; \
2325	2346	})
2326	2347	#endif
...	...	@@ -2329,16 +2350,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vcvt_f16_s16(int16x4_t
2329	2350	#define vcvt_n_f16_s16(__p0, __p1) __extension__ ({ \
2330	2351	float16x4_t __ret; \
2331	2352	int16x4_t __s0 = __p0; \
2332		__ret = (float16x4_t) __builtin_neon_vcvt_n_f16_s16((int8x8_t)__s0, __p1, 1); \
	2353	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vcvt_n_f16_s16(__builtin_bit_cast(int8x8_t, __s0), __p1, 1)); \
2333	2354	__ret; \
2334	2355	})
2335	2356	#else
2336	2357	#define vcvt_n_f16_s16(__p0, __p1) __extension__ ({ \
2337	2358	float16x4_t __ret; \
2338	2359	int16x4_t __s0 = __p0; \
2339		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
2340		__ret = (float16x4_t) __builtin_neon_vcvt_n_f16_s16((int8x8_t)__rev0, __p1, 1); \
2341		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	2360	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	2361	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vcvt_n_f16_s16(__builtin_bit_cast(int8x8_t, __rev0), __p1, 1)); \
	2362	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
2342	2363	__ret; \
2343	2364	})
2344	2365	#endif
...	...	@@ -2347,16 +2368,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vcvt_f16_s16(int16x4_t
2347	2368	#define vcvtq_n_s16_f16(__p0, __p1) __extension__ ({ \
2348	2369	int16x8_t __ret; \
2349	2370	float16x8_t __s0 = __p0; \
2350		__ret = (int16x8_t) __builtin_neon_vcvtq_n_s16_f16((int8x16_t)__s0, __p1, 33); \
	2371	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vcvtq_n_s16_f16(__builtin_bit_cast(int8x16_t, __s0), __p1, 33)); \
2351	2372	__ret; \
2352	2373	})
2353	2374	#else
2354	2375	#define vcvtq_n_s16_f16(__p0, __p1) __extension__ ({ \
2355	2376	int16x8_t __ret; \
2356	2377	float16x8_t __s0 = __p0; \
2357		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
2358		__ret = (int16x8_t) __builtin_neon_vcvtq_n_s16_f16((int8x16_t)__rev0, __p1, 33); \
2359		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	2378	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	2379	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vcvtq_n_s16_f16(__builtin_bit_cast(int8x16_t, __rev0), __p1, 33)); \
	2380	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
2360	2381	__ret; \
2361	2382	})
2362	2383	#endif
...	...	@@ -2365,16 +2386,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vcvt_f16_s16(int16x4_t
2365	2386	#define vcvt_n_s16_f16(__p0, __p1) __extension__ ({ \
2366	2387	int16x4_t __ret; \
2367	2388	float16x4_t __s0 = __p0; \
2368		__ret = (int16x4_t) __builtin_neon_vcvt_n_s16_f16((int8x8_t)__s0, __p1, 1); \
	2389	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vcvt_n_s16_f16(__builtin_bit_cast(int8x8_t, __s0), __p1, 1)); \
2369	2390	__ret; \
2370	2391	})
2371	2392	#else
2372	2393	#define vcvt_n_s16_f16(__p0, __p1) __extension__ ({ \
2373	2394	int16x4_t __ret; \
2374	2395	float16x4_t __s0 = __p0; \
2375		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
2376		__ret = (int16x4_t) __builtin_neon_vcvt_n_s16_f16((int8x8_t)__rev0, __p1, 1); \
2377		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	2396	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	2397	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vcvt_n_s16_f16(__builtin_bit_cast(int8x8_t, __rev0), __p1, 1)); \
	2398	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
2378	2399	__ret; \
2379	2400	})
2380	2401	#endif
...	...	@@ -2383,16 +2404,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vcvt_f16_s16(int16x4_t
2383	2404	#define vcvtq_n_u16_f16(__p0, __p1) __extension__ ({ \
2384	2405	uint16x8_t __ret; \
2385	2406	float16x8_t __s0 = __p0; \
2386		__ret = (uint16x8_t) __builtin_neon_vcvtq_n_u16_f16((int8x16_t)__s0, __p1, 49); \
	2407	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcvtq_n_u16_f16(__builtin_bit_cast(int8x16_t, __s0), __p1, 49)); \
2387	2408	__ret; \
2388	2409	})
2389	2410	#else
2390	2411	#define vcvtq_n_u16_f16(__p0, __p1) __extension__ ({ \
2391	2412	uint16x8_t __ret; \
2392	2413	float16x8_t __s0 = __p0; \
2393		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
2394		__ret = (uint16x8_t) __builtin_neon_vcvtq_n_u16_f16((int8x16_t)__rev0, __p1, 49); \
2395		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	2414	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	2415	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcvtq_n_u16_f16(__builtin_bit_cast(int8x16_t, __rev0), __p1, 49)); \
	2416	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
2396	2417	__ret; \
2397	2418	})
2398	2419	#endif
...	...	@@ -2401,16 +2422,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vcvt_f16_s16(int16x4_t
2401	2422	#define vcvt_n_u16_f16(__p0, __p1) __extension__ ({ \
2402	2423	uint16x4_t __ret; \
2403	2424	float16x4_t __s0 = __p0; \
2404		__ret = (uint16x4_t) __builtin_neon_vcvt_n_u16_f16((int8x8_t)__s0, __p1, 17); \
	2425	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcvt_n_u16_f16(__builtin_bit_cast(int8x8_t, __s0), __p1, 17)); \
2405	2426	__ret; \
2406	2427	})
2407	2428	#else
2408	2429	#define vcvt_n_u16_f16(__p0, __p1) __extension__ ({ \
2409	2430	uint16x4_t __ret; \
2410	2431	float16x4_t __s0 = __p0; \
2411		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
2412		__ret = (uint16x4_t) __builtin_neon_vcvt_n_u16_f16((int8x8_t)__rev0, __p1, 17); \
2413		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	2432	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	2433	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcvt_n_u16_f16(__builtin_bit_cast(int8x8_t, __rev0), __p1, 17)); \
	2434	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
2414	2435	__ret; \
2415	2436	})
2416	2437	#endif
...	...	@@ -2418,15 +2439,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vcvt_f16_s16(int16x4_t
2418	2439	#ifdef __LITTLE_ENDIAN__
2419	2440	__ai __attribute__((target("fullfp16,neon"))) int16x8_t vcvtq_s16_f16(float16x8_t __p0) {
2420	2441	int16x8_t __ret;
2421		__ret = (int16x8_t) __builtin_neon_vcvtq_s16_f16((int8x16_t)__p0, 33);
	2442	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vcvtq_s16_f16(__builtin_bit_cast(int8x16_t, __p0), 33));
2422	2443	return __ret;
2423	2444	}
2424	2445	#else
2425	2446	__ai __attribute__((target("fullfp16,neon"))) int16x8_t vcvtq_s16_f16(float16x8_t __p0) {
2426	2447	int16x8_t __ret;
2427		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2428		__ret = (int16x8_t) __builtin_neon_vcvtq_s16_f16((int8x16_t)__rev0, 33);
2429		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	2448	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	2449	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vcvtq_s16_f16(__builtin_bit_cast(int8x16_t, __rev0), 33));
	2450	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
2430	2451	return __ret;
2431	2452	}
2432	2453	#endif
...	...	@@ -2434,15 +2455,15 @@ __ai __attribute__((target("fullfp16,neon"))) int16x8_t vcvtq_s16_f16(float16x8_
2434	2455	#ifdef __LITTLE_ENDIAN__
2435	2456	__ai __attribute__((target("fullfp16,neon"))) int16x4_t vcvt_s16_f16(float16x4_t __p0) {
2436	2457	int16x4_t __ret;
2437		__ret = (int16x4_t) __builtin_neon_vcvt_s16_f16((int8x8_t)__p0, 1);
	2458	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vcvt_s16_f16(__builtin_bit_cast(int8x8_t, __p0), 1));
2438	2459	return __ret;
2439	2460	}
2440	2461	#else
2441	2462	__ai __attribute__((target("fullfp16,neon"))) int16x4_t vcvt_s16_f16(float16x4_t __p0) {
2442	2463	int16x4_t __ret;
2443		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2444		__ret = (int16x4_t) __builtin_neon_vcvt_s16_f16((int8x8_t)__rev0, 1);
2445		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	2464	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	2465	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vcvt_s16_f16(__builtin_bit_cast(int8x8_t, __rev0), 1));
	2466	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
2446	2467	return __ret;
2447	2468	}
2448	2469	#endif
...	...	@@ -2450,15 +2471,15 @@ __ai __attribute__((target("fullfp16,neon"))) int16x4_t vcvt_s16_f16(float16x4_t
2450	2471	#ifdef __LITTLE_ENDIAN__
2451	2472	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcvtq_u16_f16(float16x8_t __p0) {
2452	2473	uint16x8_t __ret;
2453		__ret = (uint16x8_t) __builtin_neon_vcvtq_u16_f16((int8x16_t)__p0, 49);
	2474	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcvtq_u16_f16(__builtin_bit_cast(int8x16_t, __p0), 49));
2454	2475	return __ret;
2455	2476	}
2456	2477	#else
2457	2478	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcvtq_u16_f16(float16x8_t __p0) {
2458	2479	uint16x8_t __ret;
2459		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2460		__ret = (uint16x8_t) __builtin_neon_vcvtq_u16_f16((int8x16_t)__rev0, 49);
2461		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	2480	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	2481	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcvtq_u16_f16(__builtin_bit_cast(int8x16_t, __rev0), 49));
	2482	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
2462	2483	return __ret;
2463	2484	}
2464	2485	#endif
...	...	@@ -2466,15 +2487,15 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcvtq_u16_f16(float16x8
2466	2487	#ifdef __LITTLE_ENDIAN__
2467	2488	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcvt_u16_f16(float16x4_t __p0) {
2468	2489	uint16x4_t __ret;
2469		__ret = (uint16x4_t) __builtin_neon_vcvt_u16_f16((int8x8_t)__p0, 17);
	2490	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcvt_u16_f16(__builtin_bit_cast(int8x8_t, __p0), 17));
2470	2491	return __ret;
2471	2492	}
2472	2493	#else
2473	2494	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcvt_u16_f16(float16x4_t __p0) {
2474	2495	uint16x4_t __ret;
2475		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2476		__ret = (uint16x4_t) __builtin_neon_vcvt_u16_f16((int8x8_t)__rev0, 17);
2477		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	2496	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	2497	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcvt_u16_f16(__builtin_bit_cast(int8x8_t, __rev0), 17));
	2498	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
2478	2499	return __ret;
2479	2500	}
2480	2501	#endif
...	...	@@ -2482,15 +2503,15 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcvt_u16_f16(float16x4_
2482	2503	#ifdef __LITTLE_ENDIAN__
2483	2504	__ai __attribute__((target("fullfp16,neon"))) int16x8_t vcvtaq_s16_f16(float16x8_t __p0) {
2484	2505	int16x8_t __ret;
2485		__ret = (int16x8_t) __builtin_neon_vcvtaq_s16_f16((int8x16_t)__p0, 33);
	2506	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vcvtaq_s16_f16(__builtin_bit_cast(int8x16_t, __p0), 33));
2486	2507	return __ret;
2487	2508	}
2488	2509	#else
2489	2510	__ai __attribute__((target("fullfp16,neon"))) int16x8_t vcvtaq_s16_f16(float16x8_t __p0) {
2490	2511	int16x8_t __ret;
2491		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2492		__ret = (int16x8_t) __builtin_neon_vcvtaq_s16_f16((int8x16_t)__rev0, 33);
2493		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	2512	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	2513	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vcvtaq_s16_f16(__builtin_bit_cast(int8x16_t, __rev0), 33));
	2514	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
2494	2515	return __ret;
2495	2516	}
2496	2517	#endif
...	...	@@ -2498,15 +2519,15 @@ __ai __attribute__((target("fullfp16,neon"))) int16x8_t vcvtaq_s16_f16(float16x8
2498	2519	#ifdef __LITTLE_ENDIAN__
2499	2520	__ai __attribute__((target("fullfp16,neon"))) int16x4_t vcvta_s16_f16(float16x4_t __p0) {
2500	2521	int16x4_t __ret;
2501		__ret = (int16x4_t) __builtin_neon_vcvta_s16_f16((int8x8_t)__p0, 1);
	2522	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vcvta_s16_f16(__builtin_bit_cast(int8x8_t, __p0), 1));
2502	2523	return __ret;
2503	2524	}
2504	2525	#else
2505	2526	__ai __attribute__((target("fullfp16,neon"))) int16x4_t vcvta_s16_f16(float16x4_t __p0) {
2506	2527	int16x4_t __ret;
2507		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2508		__ret = (int16x4_t) __builtin_neon_vcvta_s16_f16((int8x8_t)__rev0, 1);
2509		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	2528	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	2529	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vcvta_s16_f16(__builtin_bit_cast(int8x8_t, __rev0), 1));
	2530	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
2510	2531	return __ret;
2511	2532	}
2512	2533	#endif
...	...	@@ -2514,15 +2535,15 @@ __ai __attribute__((target("fullfp16,neon"))) int16x4_t vcvta_s16_f16(float16x4_
2514	2535	#ifdef __LITTLE_ENDIAN__
2515	2536	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcvtaq_u16_f16(float16x8_t __p0) {
2516	2537	uint16x8_t __ret;
2517		__ret = (uint16x8_t) __builtin_neon_vcvtaq_u16_f16((int8x16_t)__p0, 49);
	2538	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcvtaq_u16_f16(__builtin_bit_cast(int8x16_t, __p0), 49));
2518	2539	return __ret;
2519	2540	}
2520	2541	#else
2521	2542	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcvtaq_u16_f16(float16x8_t __p0) {
2522	2543	uint16x8_t __ret;
2523		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2524		__ret = (uint16x8_t) __builtin_neon_vcvtaq_u16_f16((int8x16_t)__rev0, 49);
2525		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	2544	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	2545	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcvtaq_u16_f16(__builtin_bit_cast(int8x16_t, __rev0), 49));
	2546	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
2526	2547	return __ret;
2527	2548	}
2528	2549	#endif
...	...	@@ -2530,15 +2551,15 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcvtaq_u16_f16(float16x
2530	2551	#ifdef __LITTLE_ENDIAN__
2531	2552	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcvta_u16_f16(float16x4_t __p0) {
2532	2553	uint16x4_t __ret;
2533		__ret = (uint16x4_t) __builtin_neon_vcvta_u16_f16((int8x8_t)__p0, 17);
	2554	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcvta_u16_f16(__builtin_bit_cast(int8x8_t, __p0), 17));
2534	2555	return __ret;
2535	2556	}
2536	2557	#else
2537	2558	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcvta_u16_f16(float16x4_t __p0) {
2538	2559	uint16x4_t __ret;
2539		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2540		__ret = (uint16x4_t) __builtin_neon_vcvta_u16_f16((int8x8_t)__rev0, 17);
2541		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	2560	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	2561	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcvta_u16_f16(__builtin_bit_cast(int8x8_t, __rev0), 17));
	2562	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
2542	2563	return __ret;
2543	2564	}
2544	2565	#endif
...	...	@@ -2546,15 +2567,15 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcvta_u16_f16(float16x4
2546	2567	#ifdef __LITTLE_ENDIAN__
2547	2568	__ai __attribute__((target("fullfp16,neon"))) int16x8_t vcvtmq_s16_f16(float16x8_t __p0) {
2548	2569	int16x8_t __ret;
2549		__ret = (int16x8_t) __builtin_neon_vcvtmq_s16_f16((int8x16_t)__p0, 33);
	2570	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vcvtmq_s16_f16(__builtin_bit_cast(int8x16_t, __p0), 33));
2550	2571	return __ret;
2551	2572	}
2552	2573	#else
2553	2574	__ai __attribute__((target("fullfp16,neon"))) int16x8_t vcvtmq_s16_f16(float16x8_t __p0) {
2554	2575	int16x8_t __ret;
2555		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2556		__ret = (int16x8_t) __builtin_neon_vcvtmq_s16_f16((int8x16_t)__rev0, 33);
2557		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	2576	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	2577	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vcvtmq_s16_f16(__builtin_bit_cast(int8x16_t, __rev0), 33));
	2578	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
2558	2579	return __ret;
2559	2580	}
2560	2581	#endif
...	...	@@ -2562,15 +2583,15 @@ __ai __attribute__((target("fullfp16,neon"))) int16x8_t vcvtmq_s16_f16(float16x8
2562	2583	#ifdef __LITTLE_ENDIAN__
2563	2584	__ai __attribute__((target("fullfp16,neon"))) int16x4_t vcvtm_s16_f16(float16x4_t __p0) {
2564	2585	int16x4_t __ret;
2565		__ret = (int16x4_t) __builtin_neon_vcvtm_s16_f16((int8x8_t)__p0, 1);
	2586	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vcvtm_s16_f16(__builtin_bit_cast(int8x8_t, __p0), 1));
2566	2587	return __ret;
2567	2588	}
2568	2589	#else
2569	2590	__ai __attribute__((target("fullfp16,neon"))) int16x4_t vcvtm_s16_f16(float16x4_t __p0) {
2570	2591	int16x4_t __ret;
2571		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2572		__ret = (int16x4_t) __builtin_neon_vcvtm_s16_f16((int8x8_t)__rev0, 1);
2573		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	2592	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	2593	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vcvtm_s16_f16(__builtin_bit_cast(int8x8_t, __rev0), 1));
	2594	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
2574	2595	return __ret;
2575	2596	}
2576	2597	#endif
...	...	@@ -2578,15 +2599,15 @@ __ai __attribute__((target("fullfp16,neon"))) int16x4_t vcvtm_s16_f16(float16x4_
2578	2599	#ifdef __LITTLE_ENDIAN__
2579	2600	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcvtmq_u16_f16(float16x8_t __p0) {
2580	2601	uint16x8_t __ret;
2581		__ret = (uint16x8_t) __builtin_neon_vcvtmq_u16_f16((int8x16_t)__p0, 49);
	2602	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcvtmq_u16_f16(__builtin_bit_cast(int8x16_t, __p0), 49));
2582	2603	return __ret;
2583	2604	}
2584	2605	#else
2585	2606	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcvtmq_u16_f16(float16x8_t __p0) {
2586	2607	uint16x8_t __ret;
2587		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2588		__ret = (uint16x8_t) __builtin_neon_vcvtmq_u16_f16((int8x16_t)__rev0, 49);
2589		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	2608	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	2609	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcvtmq_u16_f16(__builtin_bit_cast(int8x16_t, __rev0), 49));
	2610	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
2590	2611	return __ret;
2591	2612	}
2592	2613	#endif
...	...	@@ -2594,15 +2615,15 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcvtmq_u16_f16(float16x
2594	2615	#ifdef __LITTLE_ENDIAN__
2595	2616	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcvtm_u16_f16(float16x4_t __p0) {
2596	2617	uint16x4_t __ret;
2597		__ret = (uint16x4_t) __builtin_neon_vcvtm_u16_f16((int8x8_t)__p0, 17);
	2618	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcvtm_u16_f16(__builtin_bit_cast(int8x8_t, __p0), 17));
2598	2619	return __ret;
2599	2620	}
2600	2621	#else
2601	2622	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcvtm_u16_f16(float16x4_t __p0) {
2602	2623	uint16x4_t __ret;
2603		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2604		__ret = (uint16x4_t) __builtin_neon_vcvtm_u16_f16((int8x8_t)__rev0, 17);
2605		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	2624	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	2625	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcvtm_u16_f16(__builtin_bit_cast(int8x8_t, __rev0), 17));
	2626	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
2606	2627	return __ret;
2607	2628	}
2608	2629	#endif
...	...	@@ -2610,15 +2631,15 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcvtm_u16_f16(float16x4
2610	2631	#ifdef __LITTLE_ENDIAN__
2611	2632	__ai __attribute__((target("fullfp16,neon"))) int16x8_t vcvtnq_s16_f16(float16x8_t __p0) {
2612	2633	int16x8_t __ret;
2613		__ret = (int16x8_t) __builtin_neon_vcvtnq_s16_f16((int8x16_t)__p0, 33);
	2634	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vcvtnq_s16_f16(__builtin_bit_cast(int8x16_t, __p0), 33));
2614	2635	return __ret;
2615	2636	}
2616	2637	#else
2617	2638	__ai __attribute__((target("fullfp16,neon"))) int16x8_t vcvtnq_s16_f16(float16x8_t __p0) {
2618	2639	int16x8_t __ret;
2619		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2620		__ret = (int16x8_t) __builtin_neon_vcvtnq_s16_f16((int8x16_t)__rev0, 33);
2621		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	2640	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	2641	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vcvtnq_s16_f16(__builtin_bit_cast(int8x16_t, __rev0), 33));
	2642	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
2622	2643	return __ret;
2623	2644	}
2624	2645	#endif
...	...	@@ -2626,15 +2647,15 @@ __ai __attribute__((target("fullfp16,neon"))) int16x8_t vcvtnq_s16_f16(float16x8
2626	2647	#ifdef __LITTLE_ENDIAN__
2627	2648	__ai __attribute__((target("fullfp16,neon"))) int16x4_t vcvtn_s16_f16(float16x4_t __p0) {
2628	2649	int16x4_t __ret;
2629		__ret = (int16x4_t) __builtin_neon_vcvtn_s16_f16((int8x8_t)__p0, 1);
	2650	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vcvtn_s16_f16(__builtin_bit_cast(int8x8_t, __p0), 1));
2630	2651	return __ret;
2631	2652	}
2632	2653	#else
2633	2654	__ai __attribute__((target("fullfp16,neon"))) int16x4_t vcvtn_s16_f16(float16x4_t __p0) {
2634	2655	int16x4_t __ret;
2635		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2636		__ret = (int16x4_t) __builtin_neon_vcvtn_s16_f16((int8x8_t)__rev0, 1);
2637		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	2656	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	2657	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vcvtn_s16_f16(__builtin_bit_cast(int8x8_t, __rev0), 1));
	2658	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
2638	2659	return __ret;
2639	2660	}
2640	2661	#endif
...	...	@@ -2642,15 +2663,15 @@ __ai __attribute__((target("fullfp16,neon"))) int16x4_t vcvtn_s16_f16(float16x4_
2642	2663	#ifdef __LITTLE_ENDIAN__
2643	2664	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcvtnq_u16_f16(float16x8_t __p0) {
2644	2665	uint16x8_t __ret;
2645		__ret = (uint16x8_t) __builtin_neon_vcvtnq_u16_f16((int8x16_t)__p0, 49);
	2666	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcvtnq_u16_f16(__builtin_bit_cast(int8x16_t, __p0), 49));
2646	2667	return __ret;
2647	2668	}
2648	2669	#else
2649	2670	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcvtnq_u16_f16(float16x8_t __p0) {
2650	2671	uint16x8_t __ret;
2651		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2652		__ret = (uint16x8_t) __builtin_neon_vcvtnq_u16_f16((int8x16_t)__rev0, 49);
2653		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	2672	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	2673	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcvtnq_u16_f16(__builtin_bit_cast(int8x16_t, __rev0), 49));
	2674	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
2654	2675	return __ret;
2655	2676	}
2656	2677	#endif
...	...	@@ -2658,15 +2679,15 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcvtnq_u16_f16(float16x
2658	2679	#ifdef __LITTLE_ENDIAN__
2659	2680	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcvtn_u16_f16(float16x4_t __p0) {
2660	2681	uint16x4_t __ret;
2661		__ret = (uint16x4_t) __builtin_neon_vcvtn_u16_f16((int8x8_t)__p0, 17);
	2682	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcvtn_u16_f16(__builtin_bit_cast(int8x8_t, __p0), 17));
2662	2683	return __ret;
2663	2684	}
2664	2685	#else
2665	2686	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcvtn_u16_f16(float16x4_t __p0) {
2666	2687	uint16x4_t __ret;
2667		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2668		__ret = (uint16x4_t) __builtin_neon_vcvtn_u16_f16((int8x8_t)__rev0, 17);
2669		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	2688	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	2689	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcvtn_u16_f16(__builtin_bit_cast(int8x8_t, __rev0), 17));
	2690	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
2670	2691	return __ret;
2671	2692	}
2672	2693	#endif
...	...	@@ -2674,15 +2695,15 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcvtn_u16_f16(float16x4
2674	2695	#ifdef __LITTLE_ENDIAN__
2675	2696	__ai __attribute__((target("fullfp16,neon"))) int16x8_t vcvtpq_s16_f16(float16x8_t __p0) {
2676	2697	int16x8_t __ret;
2677		__ret = (int16x8_t) __builtin_neon_vcvtpq_s16_f16((int8x16_t)__p0, 33);
	2698	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vcvtpq_s16_f16(__builtin_bit_cast(int8x16_t, __p0), 33));
2678	2699	return __ret;
2679	2700	}
2680	2701	#else
2681	2702	__ai __attribute__((target("fullfp16,neon"))) int16x8_t vcvtpq_s16_f16(float16x8_t __p0) {
2682	2703	int16x8_t __ret;
2683		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2684		__ret = (int16x8_t) __builtin_neon_vcvtpq_s16_f16((int8x16_t)__rev0, 33);
2685		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	2704	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	2705	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vcvtpq_s16_f16(__builtin_bit_cast(int8x16_t, __rev0), 33));
	2706	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
2686	2707	return __ret;
2687	2708	}
2688	2709	#endif
...	...	@@ -2690,15 +2711,15 @@ __ai __attribute__((target("fullfp16,neon"))) int16x8_t vcvtpq_s16_f16(float16x8
2690	2711	#ifdef __LITTLE_ENDIAN__
2691	2712	__ai __attribute__((target("fullfp16,neon"))) int16x4_t vcvtp_s16_f16(float16x4_t __p0) {
2692	2713	int16x4_t __ret;
2693		__ret = (int16x4_t) __builtin_neon_vcvtp_s16_f16((int8x8_t)__p0, 1);
	2714	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vcvtp_s16_f16(__builtin_bit_cast(int8x8_t, __p0), 1));
2694	2715	return __ret;
2695	2716	}
2696	2717	#else
2697	2718	__ai __attribute__((target("fullfp16,neon"))) int16x4_t vcvtp_s16_f16(float16x4_t __p0) {
2698	2719	int16x4_t __ret;
2699		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2700		__ret = (int16x4_t) __builtin_neon_vcvtp_s16_f16((int8x8_t)__rev0, 1);
2701		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	2720	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	2721	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vcvtp_s16_f16(__builtin_bit_cast(int8x8_t, __rev0), 1));
	2722	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
2702	2723	return __ret;
2703	2724	}
2704	2725	#endif
...	...	@@ -2706,15 +2727,15 @@ __ai __attribute__((target("fullfp16,neon"))) int16x4_t vcvtp_s16_f16(float16x4_
2706	2727	#ifdef __LITTLE_ENDIAN__
2707	2728	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcvtpq_u16_f16(float16x8_t __p0) {
2708	2729	uint16x8_t __ret;
2709		__ret = (uint16x8_t) __builtin_neon_vcvtpq_u16_f16((int8x16_t)__p0, 49);
	2730	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcvtpq_u16_f16(__builtin_bit_cast(int8x16_t, __p0), 49));
2710	2731	return __ret;
2711	2732	}
2712	2733	#else
2713	2734	__ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcvtpq_u16_f16(float16x8_t __p0) {
2714	2735	uint16x8_t __ret;
2715		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2716		__ret = (uint16x8_t) __builtin_neon_vcvtpq_u16_f16((int8x16_t)__rev0, 49);
2717		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	2736	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	2737	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcvtpq_u16_f16(__builtin_bit_cast(int8x16_t, __rev0), 49));
	2738	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
2718	2739	return __ret;
2719	2740	}
2720	2741	#endif
...	...	@@ -2722,15 +2743,15 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x8_t vcvtpq_u16_f16(float16x
2722	2743	#ifdef __LITTLE_ENDIAN__
2723	2744	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcvtp_u16_f16(float16x4_t __p0) {
2724	2745	uint16x4_t __ret;
2725		__ret = (uint16x4_t) __builtin_neon_vcvtp_u16_f16((int8x8_t)__p0, 17);
	2746	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcvtp_u16_f16(__builtin_bit_cast(int8x8_t, __p0), 17));
2726	2747	return __ret;
2727	2748	}
2728	2749	#else
2729	2750	__ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcvtp_u16_f16(float16x4_t __p0) {
2730	2751	uint16x4_t __ret;
2731		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2732		__ret = (uint16x4_t) __builtin_neon_vcvtp_u16_f16((int8x8_t)__rev0, 17);
2733		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	2752	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	2753	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcvtp_u16_f16(__builtin_bit_cast(int8x8_t, __rev0), 17));
	2754	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
2734	2755	return __ret;
2735	2756	}
2736	2757	#endif
...	...	@@ -2738,22 +2759,22 @@ __ai __attribute__((target("fullfp16,neon"))) uint16x4_t vcvtp_u16_f16(float16x4
2738	2759	#ifdef __LITTLE_ENDIAN__
2739	2760	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vfmaq_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
2740	2761	float16x8_t __ret;
2741		__ret = (float16x8_t) __builtin_neon_vfmaq_f16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 40);
	2762	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vfmaq_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 40));
2742	2763	return __ret;
2743	2764	}
2744	2765	#else
2745	2766	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vfmaq_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
2746	2767	float16x8_t __ret;
2747		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2748		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
2749		float16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
2750		__ret = (float16x8_t) __builtin_neon_vfmaq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 40);
2751		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	2768	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	2769	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	2770	float16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
	2771	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vfmaq_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 40));
	2772	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
2752	2773	return __ret;
2753	2774	}
2754	2775	__ai __attribute__((target("fullfp16,neon"))) float16x8_t __noswap_vfmaq_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
2755	2776	float16x8_t __ret;
2756		__ret = (float16x8_t) __builtin_neon_vfmaq_f16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 40);
	2777	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vfmaq_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 40));
2757	2778	return __ret;
2758	2779	}
2759	2780	#endif
...	...	@@ -2761,22 +2782,22 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t __noswap_vfmaq_f16(flo
2761	2782	#ifdef __LITTLE_ENDIAN__
2762	2783	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vfma_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
2763	2784	float16x4_t __ret;
2764		__ret = (float16x4_t) __builtin_neon_vfma_f16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 8);
	2785	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vfma_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 8));
2765	2786	return __ret;
2766	2787	}
2767	2788	#else
2768	2789	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vfma_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
2769	2790	float16x4_t __ret;
2770		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2771		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
2772		float16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
2773		__ret = (float16x4_t) __builtin_neon_vfma_f16((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 8);
2774		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	2791	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	2792	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	2793	float16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
	2794	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vfma_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 8));
	2795	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
2775	2796	return __ret;
2776	2797	}
2777	2798	__ai __attribute__((target("fullfp16,neon"))) float16x4_t __noswap_vfma_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
2778	2799	float16x4_t __ret;
2779		__ret = (float16x4_t) __builtin_neon_vfma_f16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 8);
	2800	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vfma_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 8));
2780	2801	return __ret;
2781	2802	}
2782	2803	#endif
...	...	@@ -2790,11 +2811,11 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vfmsq_f16(float16x8_t
2790	2811	#else
2791	2812	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vfmsq_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
2792	2813	float16x8_t __ret;
2793		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2794		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
2795		float16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	2814	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	2815	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	2816	float16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
2796	2817	__ret = __noswap_vfmaq_f16(__rev0, -__rev1, __rev2);
2797		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	2818	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
2798	2819	return __ret;
2799	2820	}
2800	2821	#endif
...	...	@@ -2808,11 +2829,11 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vfms_f16(float16x4_t _
2808	2829	#else
2809	2830	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vfms_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
2810	2831	float16x4_t __ret;
2811		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2812		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
2813		float16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	2832	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	2833	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	2834	float16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
2814	2835	__ret = __noswap_vfma_f16(__rev0, -__rev1, __rev2);
2815		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	2836	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
2816	2837	return __ret;
2817	2838	}
2818	2839	#endif
...	...	@@ -2820,16 +2841,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vfms_f16(float16x4_t _
2820	2841	#ifdef __LITTLE_ENDIAN__
2821	2842	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vmaxq_f16(float16x8_t __p0, float16x8_t __p1) {
2822	2843	float16x8_t __ret;
2823		__ret = (float16x8_t) __builtin_neon_vmaxq_f16((int8x16_t)__p0, (int8x16_t)__p1, 40);
	2844	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vmaxq_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 40));
2824	2845	return __ret;
2825	2846	}
2826	2847	#else
2827	2848	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vmaxq_f16(float16x8_t __p0, float16x8_t __p1) {
2828	2849	float16x8_t __ret;
2829		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2830		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
2831		__ret = (float16x8_t) __builtin_neon_vmaxq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 40);
2832		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	2850	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	2851	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	2852	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vmaxq_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 40));
	2853	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
2833	2854	return __ret;
2834	2855	}
2835	2856	#endif
...	...	@@ -2837,16 +2858,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vmaxq_f16(float16x8_t
2837	2858	#ifdef __LITTLE_ENDIAN__
2838	2859	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vmax_f16(float16x4_t __p0, float16x4_t __p1) {
2839	2860	float16x4_t __ret;
2840		__ret = (float16x4_t) __builtin_neon_vmax_f16((int8x8_t)__p0, (int8x8_t)__p1, 8);
	2861	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vmax_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 8));
2841	2862	return __ret;
2842	2863	}
2843	2864	#else
2844	2865	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vmax_f16(float16x4_t __p0, float16x4_t __p1) {
2845	2866	float16x4_t __ret;
2846		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2847		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
2848		__ret = (float16x4_t) __builtin_neon_vmax_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 8);
2849		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	2867	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	2868	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	2869	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vmax_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 8));
	2870	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
2850	2871	return __ret;
2851	2872	}
2852	2873	#endif
...	...	@@ -2854,16 +2875,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vmax_f16(float16x4_t _
2854	2875	#ifdef __LITTLE_ENDIAN__
2855	2876	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vminq_f16(float16x8_t __p0, float16x8_t __p1) {
2856	2877	float16x8_t __ret;
2857		__ret = (float16x8_t) __builtin_neon_vminq_f16((int8x16_t)__p0, (int8x16_t)__p1, 40);
	2878	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vminq_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 40));
2858	2879	return __ret;
2859	2880	}
2860	2881	#else
2861	2882	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vminq_f16(float16x8_t __p0, float16x8_t __p1) {
2862	2883	float16x8_t __ret;
2863		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2864		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
2865		__ret = (float16x8_t) __builtin_neon_vminq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 40);
2866		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	2884	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	2885	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	2886	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vminq_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 40));
	2887	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
2867	2888	return __ret;
2868	2889	}
2869	2890	#endif
...	...	@@ -2871,16 +2892,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vminq_f16(float16x8_t
2871	2892	#ifdef __LITTLE_ENDIAN__
2872	2893	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vmin_f16(float16x4_t __p0, float16x4_t __p1) {
2873	2894	float16x4_t __ret;
2874		__ret = (float16x4_t) __builtin_neon_vmin_f16((int8x8_t)__p0, (int8x8_t)__p1, 8);
	2895	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vmin_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 8));
2875	2896	return __ret;
2876	2897	}
2877	2898	#else
2878	2899	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vmin_f16(float16x4_t __p0, float16x4_t __p1) {
2879	2900	float16x4_t __ret;
2880		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2881		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
2882		__ret = (float16x4_t) __builtin_neon_vmin_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 8);
2883		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	2901	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	2902	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	2903	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vmin_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 8));
	2904	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
2884	2905	return __ret;
2885	2906	}
2886	2907	#endif
...	...	@@ -2894,10 +2915,10 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vmulq_f16(float16x8_t
2894	2915	#else
2895	2916	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vmulq_f16(float16x8_t __p0, float16x8_t __p1) {
2896	2917	float16x8_t __ret;
2897		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2898		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	2918	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	2919	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
2899	2920	__ret = __rev0 * __rev1;
2900		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	2921	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
2901	2922	return __ret;
2902	2923	}
2903	2924	#endif
...	...	@@ -2911,10 +2932,10 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vmul_f16(float16x4_t _
2911	2932	#else
2912	2933	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vmul_f16(float16x4_t __p0, float16x4_t __p1) {
2913	2934	float16x4_t __ret;
2914		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2915		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	2935	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	2936	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
2916	2937	__ret = __rev0 * __rev1;
2917		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	2938	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
2918	2939	return __ret;
2919	2940	}
2920	2941	#endif
...	...	@@ -2932,9 +2953,9 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vmul_f16(float16x4_t _
2932	2953	float16x8_t __ret; \
2933	2954	float16x8_t __s0 = __p0; \
2934	2955	float16_t __s1 = __p1; \
2935		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
	2956	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
2936	2957	__ret = __rev0 * (float16x8_t) {__s1, __s1, __s1, __s1, __s1, __s1, __s1, __s1}; \
2937		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	2958	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
2938	2959	__ret; \
2939	2960	})
2940	2961	#endif
...	...	@@ -2952,9 +2973,9 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vmul_f16(float16x4_t _
2952	2973	float16x4_t __ret; \
2953	2974	float16x4_t __s0 = __p0; \
2954	2975	float16_t __s1 = __p1; \
2955		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
	2976	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
2956	2977	__ret = __rev0 * (float16x4_t) {__s1, __s1, __s1, __s1}; \
2957		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	2978	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
2958	2979	__ret; \
2959	2980	})
2960	2981	#endif
...	...	@@ -2968,9 +2989,9 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vnegq_f16(float16x8_t
2968	2989	#else
2969	2990	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vnegq_f16(float16x8_t __p0) {
2970	2991	float16x8_t __ret;
2971		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	2992	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
2972	2993	__ret = -__rev0;
2973		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	2994	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
2974	2995	return __ret;
2975	2996	}
2976	2997	#endif
...	...	@@ -2984,9 +3005,9 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vneg_f16(float16x4_t _
2984	3005	#else
2985	3006	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vneg_f16(float16x4_t __p0) {
2986	3007	float16x4_t __ret;
2987		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	3008	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
2988	3009	__ret = -__rev0;
2989		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	3010	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
2990	3011	return __ret;
2991	3012	}
2992	3013	#endif
...	...	@@ -2994,16 +3015,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vneg_f16(float16x4_t _
2994	3015	#ifdef __LITTLE_ENDIAN__
2995	3016	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vpadd_f16(float16x4_t __p0, float16x4_t __p1) {
2996	3017	float16x4_t __ret;
2997		__ret = (float16x4_t) __builtin_neon_vpadd_f16((int8x8_t)__p0, (int8x8_t)__p1, 8);
	3018	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vpadd_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 8));
2998	3019	return __ret;
2999	3020	}
3000	3021	#else
3001	3022	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vpadd_f16(float16x4_t __p0, float16x4_t __p1) {
3002	3023	float16x4_t __ret;
3003		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3004		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
3005		__ret = (float16x4_t) __builtin_neon_vpadd_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 8);
3006		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	3024	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	3025	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	3026	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vpadd_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 8));
	3027	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
3007	3028	return __ret;
3008	3029	}
3009	3030	#endif
...	...	@@ -3011,16 +3032,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vpadd_f16(float16x4_t
3011	3032	#ifdef __LITTLE_ENDIAN__
3012	3033	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vpmax_f16(float16x4_t __p0, float16x4_t __p1) {
3013	3034	float16x4_t __ret;
3014		__ret = (float16x4_t) __builtin_neon_vpmax_f16((int8x8_t)__p0, (int8x8_t)__p1, 8);
	3035	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vpmax_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 8));
3015	3036	return __ret;
3016	3037	}
3017	3038	#else
3018	3039	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vpmax_f16(float16x4_t __p0, float16x4_t __p1) {
3019	3040	float16x4_t __ret;
3020		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3021		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
3022		__ret = (float16x4_t) __builtin_neon_vpmax_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 8);
3023		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	3041	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	3042	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	3043	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vpmax_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 8));
	3044	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
3024	3045	return __ret;
3025	3046	}
3026	3047	#endif
...	...	@@ -3028,16 +3049,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vpmax_f16(float16x4_t
3028	3049	#ifdef __LITTLE_ENDIAN__
3029	3050	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vpmin_f16(float16x4_t __p0, float16x4_t __p1) {
3030	3051	float16x4_t __ret;
3031		__ret = (float16x4_t) __builtin_neon_vpmin_f16((int8x8_t)__p0, (int8x8_t)__p1, 8);
	3052	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vpmin_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 8));
3032	3053	return __ret;
3033	3054	}
3034	3055	#else
3035	3056	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vpmin_f16(float16x4_t __p0, float16x4_t __p1) {
3036	3057	float16x4_t __ret;
3037		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3038		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
3039		__ret = (float16x4_t) __builtin_neon_vpmin_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 8);
3040		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	3058	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	3059	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	3060	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vpmin_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 8));
	3061	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
3041	3062	return __ret;
3042	3063	}
3043	3064	#endif
...	...	@@ -3045,15 +3066,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vpmin_f16(float16x4_t
3045	3066	#ifdef __LITTLE_ENDIAN__
3046	3067	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vrecpeq_f16(float16x8_t __p0) {
3047	3068	float16x8_t __ret;
3048		__ret = (float16x8_t) __builtin_neon_vrecpeq_f16((int8x16_t)__p0, 40);
	3069	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vrecpeq_f16(__builtin_bit_cast(int8x16_t, __p0), 40));
3049	3070	return __ret;
3050	3071	}
3051	3072	#else
3052	3073	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vrecpeq_f16(float16x8_t __p0) {
3053	3074	float16x8_t __ret;
3054		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
3055		__ret = (float16x8_t) __builtin_neon_vrecpeq_f16((int8x16_t)__rev0, 40);
3056		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	3075	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	3076	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vrecpeq_f16(__builtin_bit_cast(int8x16_t, __rev0), 40));
	3077	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
3057	3078	return __ret;
3058	3079	}
3059	3080	#endif
...	...	@@ -3061,15 +3082,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vrecpeq_f16(float16x8_
3061	3082	#ifdef __LITTLE_ENDIAN__
3062	3083	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vrecpe_f16(float16x4_t __p0) {
3063	3084	float16x4_t __ret;
3064		__ret = (float16x4_t) __builtin_neon_vrecpe_f16((int8x8_t)__p0, 8);
	3085	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vrecpe_f16(__builtin_bit_cast(int8x8_t, __p0), 8));
3065	3086	return __ret;
3066	3087	}
3067	3088	#else
3068	3089	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vrecpe_f16(float16x4_t __p0) {
3069	3090	float16x4_t __ret;
3070		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3071		__ret = (float16x4_t) __builtin_neon_vrecpe_f16((int8x8_t)__rev0, 8);
3072		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	3091	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	3092	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vrecpe_f16(__builtin_bit_cast(int8x8_t, __rev0), 8));
	3093	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
3073	3094	return __ret;
3074	3095	}
3075	3096	#endif
...	...	@@ -3077,16 +3098,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vrecpe_f16(float16x4_t
3077	3098	#ifdef __LITTLE_ENDIAN__
3078	3099	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vrecpsq_f16(float16x8_t __p0, float16x8_t __p1) {
3079	3100	float16x8_t __ret;
3080		__ret = (float16x8_t) __builtin_neon_vrecpsq_f16((int8x16_t)__p0, (int8x16_t)__p1, 40);
	3101	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vrecpsq_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 40));
3081	3102	return __ret;
3082	3103	}
3083	3104	#else
3084	3105	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vrecpsq_f16(float16x8_t __p0, float16x8_t __p1) {
3085	3106	float16x8_t __ret;
3086		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
3087		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
3088		__ret = (float16x8_t) __builtin_neon_vrecpsq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 40);
3089		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	3107	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	3108	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	3109	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vrecpsq_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 40));
	3110	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
3090	3111	return __ret;
3091	3112	}
3092	3113	#endif
...	...	@@ -3094,16 +3115,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vrecpsq_f16(float16x8_
3094	3115	#ifdef __LITTLE_ENDIAN__
3095	3116	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vrecps_f16(float16x4_t __p0, float16x4_t __p1) {
3096	3117	float16x4_t __ret;
3097		__ret = (float16x4_t) __builtin_neon_vrecps_f16((int8x8_t)__p0, (int8x8_t)__p1, 8);
	3118	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vrecps_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 8));
3098	3119	return __ret;
3099	3120	}
3100	3121	#else
3101	3122	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vrecps_f16(float16x4_t __p0, float16x4_t __p1) {
3102	3123	float16x4_t __ret;
3103		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3104		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
3105		__ret = (float16x4_t) __builtin_neon_vrecps_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 8);
3106		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	3124	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	3125	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	3126	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vrecps_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 8));
	3127	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
3107	3128	return __ret;
3108	3129	}
3109	3130	#endif
...	...	@@ -3111,15 +3132,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vrecps_f16(float16x4_t
3111	3132	#ifdef __LITTLE_ENDIAN__
3112	3133	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vrsqrteq_f16(float16x8_t __p0) {
3113	3134	float16x8_t __ret;
3114		__ret = (float16x8_t) __builtin_neon_vrsqrteq_f16((int8x16_t)__p0, 40);
	3135	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vrsqrteq_f16(__builtin_bit_cast(int8x16_t, __p0), 40));
3115	3136	return __ret;
3116	3137	}
3117	3138	#else
3118	3139	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vrsqrteq_f16(float16x8_t __p0) {
3119	3140	float16x8_t __ret;
3120		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
3121		__ret = (float16x8_t) __builtin_neon_vrsqrteq_f16((int8x16_t)__rev0, 40);
3122		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	3141	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	3142	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vrsqrteq_f16(__builtin_bit_cast(int8x16_t, __rev0), 40));
	3143	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
3123	3144	return __ret;
3124	3145	}
3125	3146	#endif
...	...	@@ -3127,15 +3148,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vrsqrteq_f16(float16x8
3127	3148	#ifdef __LITTLE_ENDIAN__
3128	3149	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vrsqrte_f16(float16x4_t __p0) {
3129	3150	float16x4_t __ret;
3130		__ret = (float16x4_t) __builtin_neon_vrsqrte_f16((int8x8_t)__p0, 8);
	3151	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vrsqrte_f16(__builtin_bit_cast(int8x8_t, __p0), 8));
3131	3152	return __ret;
3132	3153	}
3133	3154	#else
3134	3155	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vrsqrte_f16(float16x4_t __p0) {
3135	3156	float16x4_t __ret;
3136		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3137		__ret = (float16x4_t) __builtin_neon_vrsqrte_f16((int8x8_t)__rev0, 8);
3138		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	3157	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	3158	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vrsqrte_f16(__builtin_bit_cast(int8x8_t, __rev0), 8));
	3159	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
3139	3160	return __ret;
3140	3161	}
3141	3162	#endif
...	...	@@ -3143,16 +3164,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vrsqrte_f16(float16x4_
3143	3164	#ifdef __LITTLE_ENDIAN__
3144	3165	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vrsqrtsq_f16(float16x8_t __p0, float16x8_t __p1) {
3145	3166	float16x8_t __ret;
3146		__ret = (float16x8_t) __builtin_neon_vrsqrtsq_f16((int8x16_t)__p0, (int8x16_t)__p1, 40);
	3167	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vrsqrtsq_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 40));
3147	3168	return __ret;
3148	3169	}
3149	3170	#else
3150	3171	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vrsqrtsq_f16(float16x8_t __p0, float16x8_t __p1) {
3151	3172	float16x8_t __ret;
3152		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
3153		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
3154		__ret = (float16x8_t) __builtin_neon_vrsqrtsq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 40);
3155		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	3173	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	3174	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	3175	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vrsqrtsq_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 40));
	3176	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
3156	3177	return __ret;
3157	3178	}
3158	3179	#endif
...	...	@@ -3160,16 +3181,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vrsqrtsq_f16(float16x8
3160	3181	#ifdef __LITTLE_ENDIAN__
3161	3182	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vrsqrts_f16(float16x4_t __p0, float16x4_t __p1) {
3162	3183	float16x4_t __ret;
3163		__ret = (float16x4_t) __builtin_neon_vrsqrts_f16((int8x8_t)__p0, (int8x8_t)__p1, 8);
	3184	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vrsqrts_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 8));
3164	3185	return __ret;
3165	3186	}
3166	3187	#else
3167	3188	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vrsqrts_f16(float16x4_t __p0, float16x4_t __p1) {
3168	3189	float16x4_t __ret;
3169		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3170		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
3171		__ret = (float16x4_t) __builtin_neon_vrsqrts_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 8);
3172		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	3190	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	3191	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	3192	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vrsqrts_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 8));
	3193	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
3173	3194	return __ret;
3174	3195	}
3175	3196	#endif
...	...	@@ -3183,10 +3204,10 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vsubq_f16(float16x8_t
3183	3204	#else
3184	3205	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vsubq_f16(float16x8_t __p0, float16x8_t __p1) {
3185	3206	float16x8_t __ret;
3186		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
3187		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	3207	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	3208	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
3188	3209	__ret = __rev0 - __rev1;
3189		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	3210	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
3190	3211	return __ret;
3191	3212	}
3192	3213	#endif
...	...	@@ -3200,10 +3221,10 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vsub_f16(float16x4_t _
3200	3221	#else
3201	3222	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vsub_f16(float16x4_t __p0, float16x4_t __p1) {
3202	3223	float16x4_t __ret;
3203		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3204		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	3224	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	3225	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
3205	3226	__ret = __rev0 - __rev1;
3206		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	3227	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
3207	3228	return __ret;
3208	3229	}
3209	3230	#endif
...	...	@@ -3211,17 +3232,17 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vsub_f16(float16x4_t _
3211	3232	#ifdef __LITTLE_ENDIAN__
3212	3233	__ai __attribute__((target("i8mm,neon"))) uint32x4_t vmmlaq_u32(uint32x4_t __p0, uint8x16_t __p1, uint8x16_t __p2) {
3213	3234	uint32x4_t __ret;
3214		__ret = (uint32x4_t) __builtin_neon_vmmlaq_u32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 50);
	3235	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vmmlaq_u32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 50));
3215	3236	return __ret;
3216	3237	}
3217	3238	#else
3218	3239	__ai __attribute__((target("i8mm,neon"))) uint32x4_t vmmlaq_u32(uint32x4_t __p0, uint8x16_t __p1, uint8x16_t __p2) {
3219	3240	uint32x4_t __ret;
3220		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3221		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
3222		uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
3223		__ret = (uint32x4_t) __builtin_neon_vmmlaq_u32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 50);
3224		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	3241	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	3242	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	3243	uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	3244	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vmmlaq_u32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 50));
	3245	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
3225	3246	return __ret;
3226	3247	}
3227	3248	#endif
...	...	@@ -3229,17 +3250,17 @@ __ai __attribute__((target("i8mm,neon"))) uint32x4_t vmmlaq_u32(uint32x4_t __p0,
3229	3250	#ifdef __LITTLE_ENDIAN__
3230	3251	__ai __attribute__((target("i8mm,neon"))) int32x4_t vmmlaq_s32(int32x4_t __p0, int8x16_t __p1, int8x16_t __p2) {
3231	3252	int32x4_t __ret;
3232		__ret = (int32x4_t) __builtin_neon_vmmlaq_s32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 34);
	3253	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vmmlaq_s32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 34));
3233	3254	return __ret;
3234	3255	}
3235	3256	#else
3236	3257	__ai __attribute__((target("i8mm,neon"))) int32x4_t vmmlaq_s32(int32x4_t __p0, int8x16_t __p1, int8x16_t __p2) {
3237	3258	int32x4_t __ret;
3238		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3239		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
3240		int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
3241		__ret = (int32x4_t) __builtin_neon_vmmlaq_s32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 34);
3242		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	3259	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	3260	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	3261	int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	3262	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vmmlaq_s32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 34));
	3263	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
3243	3264	return __ret;
3244	3265	}
3245	3266	#endif
...	...	@@ -3247,22 +3268,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vmmlaq_s32(int32x4_t __p0, i
3247	3268	#ifdef __LITTLE_ENDIAN__
3248	3269	__ai __attribute__((target("i8mm,neon"))) int32x4_t vusdotq_s32(int32x4_t __p0, uint8x16_t __p1, int8x16_t __p2) {
3249	3270	int32x4_t __ret;
3250		__ret = (int32x4_t) __builtin_neon_vusdotq_s32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 34);
	3271	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vusdotq_s32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 34));
3251	3272	return __ret;
3252	3273	}
3253	3274	#else
3254	3275	__ai __attribute__((target("i8mm,neon"))) int32x4_t vusdotq_s32(int32x4_t __p0, uint8x16_t __p1, int8x16_t __p2) {
3255	3276	int32x4_t __ret;
3256		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3257		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
3258		int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
3259		__ret = (int32x4_t) __builtin_neon_vusdotq_s32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 34);
3260		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	3277	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	3278	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	3279	int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	3280	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vusdotq_s32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 34));
	3281	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
3261	3282	return __ret;
3262	3283	}
3263	3284	__ai __attribute__((target("i8mm,neon"))) int32x4_t __noswap_vusdotq_s32(int32x4_t __p0, uint8x16_t __p1, int8x16_t __p2) {
3264	3285	int32x4_t __ret;
3265		__ret = (int32x4_t) __builtin_neon_vusdotq_s32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 34);
	3286	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vusdotq_s32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 34));
3266	3287	return __ret;
3267	3288	}
3268	3289	#endif
...	...	@@ -3270,22 +3291,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t __noswap_vusdotq_s32(int32x4
3270	3291	#ifdef __LITTLE_ENDIAN__
3271	3292	__ai __attribute__((target("i8mm,neon"))) int32x2_t vusdot_s32(int32x2_t __p0, uint8x8_t __p1, int8x8_t __p2) {
3272	3293	int32x2_t __ret;
3273		__ret = (int32x2_t) __builtin_neon_vusdot_s32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 2);
	3294	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vusdot_s32(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 2));
3274	3295	return __ret;
3275	3296	}
3276	3297	#else
3277	3298	__ai __attribute__((target("i8mm,neon"))) int32x2_t vusdot_s32(int32x2_t __p0, uint8x8_t __p1, int8x8_t __p2) {
3278	3299	int32x2_t __ret;
3279		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
3280		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
3281		int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
3282		__ret = (int32x2_t) __builtin_neon_vusdot_s32((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 2);
3283		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	3300	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	3301	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	3302	int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	3303	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vusdot_s32(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 2));
	3304	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
3284	3305	return __ret;
3285	3306	}
3286	3307	__ai __attribute__((target("i8mm,neon"))) int32x2_t __noswap_vusdot_s32(int32x2_t __p0, uint8x8_t __p1, int8x8_t __p2) {
3287	3308	int32x2_t __ret;
3288		__ret = (int32x2_t) __builtin_neon_vusdot_s32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 2);
	3309	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vusdot_s32(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 2));
3289	3310	return __ret;
3290	3311	}
3291	3312	#endif
...	...	@@ -3293,17 +3314,17 @@ __ai __attribute__((target("i8mm,neon"))) int32x2_t __noswap_vusdot_s32(int32x2_
3293	3314	#ifdef __LITTLE_ENDIAN__
3294	3315	__ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0, uint8x16_t __p1, int8x16_t __p2) {
3295	3316	int32x4_t __ret;
3296		__ret = (int32x4_t) __builtin_neon_vusmmlaq_s32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 34);
	3317	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vusmmlaq_s32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 34));
3297	3318	return __ret;
3298	3319	}
3299	3320	#else
3300	3321	__ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0, uint8x16_t __p1, int8x16_t __p2) {
3301	3322	int32x4_t __ret;
3302		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3303		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
3304		int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
3305		__ret = (int32x4_t) __builtin_neon_vusmmlaq_s32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 34);
3306		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	3323	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	3324	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	3325	int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	3326	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vusmmlaq_s32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 34));
	3327	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
3307	3328	return __ret;
3308	3329	}
3309	3330	#endif
...	...	@@ -3312,22 +3333,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
3312	3333	#define splat_lane_p8(__p0, __p1) __extension__ ({ \
3313	3334	poly8x8_t __ret; \
3314	3335	poly8x8_t __s0 = __p0; \
3315		__ret = (poly8x8_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 4); \
	3336	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 4)); \
3316	3337	__ret; \
3317	3338	})
3318	3339	#else
3319	3340	#define splat_lane_p8(__p0, __p1) __extension__ ({ \
3320	3341	poly8x8_t __ret; \
3321	3342	poly8x8_t __s0 = __p0; \
3322		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
3323		__ret = (poly8x8_t) __builtin_neon_splat_lane_v((int8x8_t)__rev0, __p1, 4); \
3324		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	3343	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	3344	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 4)); \
	3345	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
3325	3346	__ret; \
3326	3347	})
3327	3348	#define __noswap_splat_lane_p8(__p0, __p1) __extension__ ({ \
3328	3349	poly8x8_t __ret; \
3329	3350	poly8x8_t __s0 = __p0; \
3330		__ret = (poly8x8_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 4); \
	3351	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 4)); \
3331	3352	__ret; \
3332	3353	})
3333	3354	#endif
...	...	@@ -3335,29 +3356,29 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
3335	3356	#define splat_lane_p64(__p0, __p1) __extension__ ({ \
3336	3357	poly64x1_t __ret; \
3337	3358	poly64x1_t __s0 = __p0; \
3338		__ret = (poly64x1_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 6); \
	3359	__ret = __builtin_bit_cast(poly64x1_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 6)); \
3339	3360	__ret; \
3340	3361	})
3341	3362	#ifdef __LITTLE_ENDIAN__
3342	3363	#define splat_lane_p16(__p0, __p1) __extension__ ({ \
3343	3364	poly16x4_t __ret; \
3344	3365	poly16x4_t __s0 = __p0; \
3345		__ret = (poly16x4_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 5); \
	3366	__ret = __builtin_bit_cast(poly16x4_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 5)); \
3346	3367	__ret; \
3347	3368	})
3348	3369	#else
3349	3370	#define splat_lane_p16(__p0, __p1) __extension__ ({ \
3350	3371	poly16x4_t __ret; \
3351	3372	poly16x4_t __s0 = __p0; \
3352		poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
3353		__ret = (poly16x4_t) __builtin_neon_splat_lane_v((int8x8_t)__rev0, __p1, 5); \
3354		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	3373	poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	3374	__ret = __builtin_bit_cast(poly16x4_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 5)); \
	3375	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
3355	3376	__ret; \
3356	3377	})
3357	3378	#define __noswap_splat_lane_p16(__p0, __p1) __extension__ ({ \
3358	3379	poly16x4_t __ret; \
3359	3380	poly16x4_t __s0 = __p0; \
3360		__ret = (poly16x4_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 5); \
	3381	__ret = __builtin_bit_cast(poly16x4_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 5)); \
3361	3382	__ret; \
3362	3383	})
3363	3384	#endif
...	...	@@ -3366,22 +3387,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
3366	3387	#define splatq_lane_p8(__p0, __p1) __extension__ ({ \
3367	3388	poly8x16_t __ret; \
3368	3389	poly8x8_t __s0 = __p0; \
3369		__ret = (poly8x16_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 4); \
	3390	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 4)); \
3370	3391	__ret; \
3371	3392	})
3372	3393	#else
3373	3394	#define splatq_lane_p8(__p0, __p1) __extension__ ({ \
3374	3395	poly8x16_t __ret; \
3375	3396	poly8x8_t __s0 = __p0; \
3376		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
3377		__ret = (poly8x16_t) __builtin_neon_splatq_lane_v((int8x8_t)__rev0, __p1, 4); \
3378		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	3397	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	3398	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 4)); \
	3399	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
3379	3400	__ret; \
3380	3401	})
3381	3402	#define __noswap_splatq_lane_p8(__p0, __p1) __extension__ ({ \
3382	3403	poly8x16_t __ret; \
3383	3404	poly8x8_t __s0 = __p0; \
3384		__ret = (poly8x16_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 4); \
	3405	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 4)); \
3385	3406	__ret; \
3386	3407	})
3387	3408	#endif
...	...	@@ -3390,21 +3411,21 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
3390	3411	#define splatq_lane_p64(__p0, __p1) __extension__ ({ \
3391	3412	poly64x2_t __ret; \
3392	3413	poly64x1_t __s0 = __p0; \
3393		__ret = (poly64x2_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 6); \
	3414	__ret = __builtin_bit_cast(poly64x2_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 6)); \
3394	3415	__ret; \
3395	3416	})
3396	3417	#else
3397	3418	#define splatq_lane_p64(__p0, __p1) __extension__ ({ \
3398	3419	poly64x2_t __ret; \
3399	3420	poly64x1_t __s0 = __p0; \
3400		__ret = (poly64x2_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 6); \
3401		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	3421	__ret = __builtin_bit_cast(poly64x2_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 6)); \
	3422	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
3402	3423	__ret; \
3403	3424	})
3404	3425	#define __noswap_splatq_lane_p64(__p0, __p1) __extension__ ({ \
3405	3426	poly64x2_t __ret; \
3406	3427	poly64x1_t __s0 = __p0; \
3407		__ret = (poly64x2_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 6); \
	3428	__ret = __builtin_bit_cast(poly64x2_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 6)); \
3408	3429	__ret; \
3409	3430	})
3410	3431	#endif
...	...	@@ -3413,22 +3434,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
3413	3434	#define splatq_lane_p16(__p0, __p1) __extension__ ({ \
3414	3435	poly16x8_t __ret; \
3415	3436	poly16x4_t __s0 = __p0; \
3416		__ret = (poly16x8_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 5); \
	3437	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 5)); \
3417	3438	__ret; \
3418	3439	})
3419	3440	#else
3420	3441	#define splatq_lane_p16(__p0, __p1) __extension__ ({ \
3421	3442	poly16x8_t __ret; \
3422	3443	poly16x4_t __s0 = __p0; \
3423		poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
3424		__ret = (poly16x8_t) __builtin_neon_splatq_lane_v((int8x8_t)__rev0, __p1, 5); \
3425		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	3444	poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	3445	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 5)); \
	3446	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
3426	3447	__ret; \
3427	3448	})
3428	3449	#define __noswap_splatq_lane_p16(__p0, __p1) __extension__ ({ \
3429	3450	poly16x8_t __ret; \
3430	3451	poly16x4_t __s0 = __p0; \
3431		__ret = (poly16x8_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 5); \
	3452	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 5)); \
3432	3453	__ret; \
3433	3454	})
3434	3455	#endif
...	...	@@ -3437,22 +3458,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
3437	3458	#define splatq_lane_u8(__p0, __p1) __extension__ ({ \
3438	3459	uint8x16_t __ret; \
3439	3460	uint8x8_t __s0 = __p0; \
3440		__ret = (uint8x16_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 16); \
	3461	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 16)); \
3441	3462	__ret; \
3442	3463	})
3443	3464	#else
3444	3465	#define splatq_lane_u8(__p0, __p1) __extension__ ({ \
3445	3466	uint8x16_t __ret; \
3446	3467	uint8x8_t __s0 = __p0; \
3447		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
3448		__ret = (uint8x16_t) __builtin_neon_splatq_lane_v((int8x8_t)__rev0, __p1, 16); \
3449		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	3468	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	3469	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 16)); \
	3470	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
3450	3471	__ret; \
3451	3472	})
3452	3473	#define __noswap_splatq_lane_u8(__p0, __p1) __extension__ ({ \
3453	3474	uint8x16_t __ret; \
3454	3475	uint8x8_t __s0 = __p0; \
3455		__ret = (uint8x16_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 16); \
	3476	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 16)); \
3456	3477	__ret; \
3457	3478	})
3458	3479	#endif
...	...	@@ -3461,22 +3482,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
3461	3482	#define splatq_lane_u32(__p0, __p1) __extension__ ({ \
3462	3483	uint32x4_t __ret; \
3463	3484	uint32x2_t __s0 = __p0; \
3464		__ret = (uint32x4_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 18); \
	3485	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 18)); \
3465	3486	__ret; \
3466	3487	})
3467	3488	#else
3468	3489	#define splatq_lane_u32(__p0, __p1) __extension__ ({ \
3469	3490	uint32x4_t __ret; \
3470	3491	uint32x2_t __s0 = __p0; \
3471		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
3472		__ret = (uint32x4_t) __builtin_neon_splatq_lane_v((int8x8_t)__rev0, __p1, 18); \
3473		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	3492	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	3493	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 18)); \
	3494	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
3474	3495	__ret; \
3475	3496	})
3476	3497	#define __noswap_splatq_lane_u32(__p0, __p1) __extension__ ({ \
3477	3498	uint32x4_t __ret; \
3478	3499	uint32x2_t __s0 = __p0; \
3479		__ret = (uint32x4_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 18); \
	3500	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 18)); \
3480	3501	__ret; \
3481	3502	})
3482	3503	#endif
...	...	@@ -3485,21 +3506,21 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
3485	3506	#define splatq_lane_u64(__p0, __p1) __extension__ ({ \
3486	3507	uint64x2_t __ret; \
3487	3508	uint64x1_t __s0 = __p0; \
3488		__ret = (uint64x2_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 19); \
	3509	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 19)); \
3489	3510	__ret; \
3490	3511	})
3491	3512	#else
3492	3513	#define splatq_lane_u64(__p0, __p1) __extension__ ({ \
3493	3514	uint64x2_t __ret; \
3494	3515	uint64x1_t __s0 = __p0; \
3495		__ret = (uint64x2_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 19); \
3496		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	3516	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 19)); \
	3517	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
3497	3518	__ret; \
3498	3519	})
3499	3520	#define __noswap_splatq_lane_u64(__p0, __p1) __extension__ ({ \
3500	3521	uint64x2_t __ret; \
3501	3522	uint64x1_t __s0 = __p0; \
3502		__ret = (uint64x2_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 19); \
	3523	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 19)); \
3503	3524	__ret; \
3504	3525	})
3505	3526	#endif
...	...	@@ -3508,22 +3529,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
3508	3529	#define splatq_lane_u16(__p0, __p1) __extension__ ({ \
3509	3530	uint16x8_t __ret; \
3510	3531	uint16x4_t __s0 = __p0; \
3511		__ret = (uint16x8_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 17); \
	3532	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 17)); \
3512	3533	__ret; \
3513	3534	})
3514	3535	#else
3515	3536	#define splatq_lane_u16(__p0, __p1) __extension__ ({ \
3516	3537	uint16x8_t __ret; \
3517	3538	uint16x4_t __s0 = __p0; \
3518		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
3519		__ret = (uint16x8_t) __builtin_neon_splatq_lane_v((int8x8_t)__rev0, __p1, 17); \
3520		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	3539	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	3540	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 17)); \
	3541	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
3521	3542	__ret; \
3522	3543	})
3523	3544	#define __noswap_splatq_lane_u16(__p0, __p1) __extension__ ({ \
3524	3545	uint16x8_t __ret; \
3525	3546	uint16x4_t __s0 = __p0; \
3526		__ret = (uint16x8_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 17); \
	3547	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 17)); \
3527	3548	__ret; \
3528	3549	})
3529	3550	#endif
...	...	@@ -3532,22 +3553,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
3532	3553	#define splatq_lane_s8(__p0, __p1) __extension__ ({ \
3533	3554	int8x16_t __ret; \
3534	3555	int8x8_t __s0 = __p0; \
3535		__ret = (int8x16_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 0); \
	3556	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 0)); \
3536	3557	__ret; \
3537	3558	})
3538	3559	#else
3539	3560	#define splatq_lane_s8(__p0, __p1) __extension__ ({ \
3540	3561	int8x16_t __ret; \
3541	3562	int8x8_t __s0 = __p0; \
3542		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
3543		__ret = (int8x16_t) __builtin_neon_splatq_lane_v((int8x8_t)__rev0, __p1, 0); \
3544		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	3563	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	3564	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 0)); \
	3565	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
3545	3566	__ret; \
3546	3567	})
3547	3568	#define __noswap_splatq_lane_s8(__p0, __p1) __extension__ ({ \
3548	3569	int8x16_t __ret; \
3549	3570	int8x8_t __s0 = __p0; \
3550		__ret = (int8x16_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 0); \
	3571	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 0)); \
3551	3572	__ret; \
3552	3573	})
3553	3574	#endif
...	...	@@ -3556,21 +3577,21 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
3556	3577	#define splatq_lane_f64(__p0, __p1) __extension__ ({ \
3557	3578	float64x2_t __ret; \
3558	3579	float64x1_t __s0 = __p0; \
3559		__ret = (float64x2_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 10); \
	3580	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 10)); \
3560	3581	__ret; \
3561	3582	})
3562	3583	#else
3563	3584	#define splatq_lane_f64(__p0, __p1) __extension__ ({ \
3564	3585	float64x2_t __ret; \
3565	3586	float64x1_t __s0 = __p0; \
3566		__ret = (float64x2_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 10); \
3567		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	3587	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 10)); \
	3588	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
3568	3589	__ret; \
3569	3590	})
3570	3591	#define __noswap_splatq_lane_f64(__p0, __p1) __extension__ ({ \
3571	3592	float64x2_t __ret; \
3572	3593	float64x1_t __s0 = __p0; \
3573		__ret = (float64x2_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 10); \
	3594	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 10)); \
3574	3595	__ret; \
3575	3596	})
3576	3597	#endif
...	...	@@ -3579,22 +3600,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
3579	3600	#define splatq_lane_f32(__p0, __p1) __extension__ ({ \
3580	3601	float32x4_t __ret; \
3581	3602	float32x2_t __s0 = __p0; \
3582		__ret = (float32x4_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 9); \
	3603	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 9)); \
3583	3604	__ret; \
3584	3605	})
3585	3606	#else
3586	3607	#define splatq_lane_f32(__p0, __p1) __extension__ ({ \
3587	3608	float32x4_t __ret; \
3588	3609	float32x2_t __s0 = __p0; \
3589		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
3590		__ret = (float32x4_t) __builtin_neon_splatq_lane_v((int8x8_t)__rev0, __p1, 9); \
3591		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	3610	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	3611	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 9)); \
	3612	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
3592	3613	__ret; \
3593	3614	})
3594	3615	#define __noswap_splatq_lane_f32(__p0, __p1) __extension__ ({ \
3595	3616	float32x4_t __ret; \
3596	3617	float32x2_t __s0 = __p0; \
3597		__ret = (float32x4_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 9); \
	3618	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 9)); \
3598	3619	__ret; \
3599	3620	})
3600	3621	#endif
...	...	@@ -3603,22 +3624,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
3603	3624	#define splatq_lane_f16(__p0, __p1) __extension__ ({ \
3604	3625	float16x8_t __ret; \
3605	3626	float16x4_t __s0 = __p0; \
3606		__ret = (float16x8_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 8); \
	3627	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 8)); \
3607	3628	__ret; \
3608	3629	})
3609	3630	#else
3610	3631	#define splatq_lane_f16(__p0, __p1) __extension__ ({ \
3611	3632	float16x8_t __ret; \
3612	3633	float16x4_t __s0 = __p0; \
3613		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
3614		__ret = (float16x8_t) __builtin_neon_splatq_lane_v((int8x8_t)__rev0, __p1, 8); \
3615		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	3634	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	3635	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 8)); \
	3636	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
3616	3637	__ret; \
3617	3638	})
3618	3639	#define __noswap_splatq_lane_f16(__p0, __p1) __extension__ ({ \
3619	3640	float16x8_t __ret; \
3620	3641	float16x4_t __s0 = __p0; \
3621		__ret = (float16x8_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 8); \
	3642	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 8)); \
3622	3643	__ret; \
3623	3644	})
3624	3645	#endif
...	...	@@ -3627,22 +3648,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
3627	3648	#define splatq_lane_s32(__p0, __p1) __extension__ ({ \
3628	3649	int32x4_t __ret; \
3629	3650	int32x2_t __s0 = __p0; \
3630		__ret = (int32x4_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 2); \
	3651	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 2)); \
3631	3652	__ret; \
3632	3653	})
3633	3654	#else
3634	3655	#define splatq_lane_s32(__p0, __p1) __extension__ ({ \
3635	3656	int32x4_t __ret; \
3636	3657	int32x2_t __s0 = __p0; \
3637		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
3638		__ret = (int32x4_t) __builtin_neon_splatq_lane_v((int8x8_t)__rev0, __p1, 2); \
3639		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	3658	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	3659	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 2)); \
	3660	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
3640	3661	__ret; \
3641	3662	})
3642	3663	#define __noswap_splatq_lane_s32(__p0, __p1) __extension__ ({ \
3643	3664	int32x4_t __ret; \
3644	3665	int32x2_t __s0 = __p0; \
3645		__ret = (int32x4_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 2); \
	3666	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 2)); \
3646	3667	__ret; \
3647	3668	})
3648	3669	#endif
...	...	@@ -3651,21 +3672,21 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
3651	3672	#define splatq_lane_s64(__p0, __p1) __extension__ ({ \
3652	3673	int64x2_t __ret; \
3653	3674	int64x1_t __s0 = __p0; \
3654		__ret = (int64x2_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 3); \
	3675	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 3)); \
3655	3676	__ret; \
3656	3677	})
3657	3678	#else
3658	3679	#define splatq_lane_s64(__p0, __p1) __extension__ ({ \
3659	3680	int64x2_t __ret; \
3660	3681	int64x1_t __s0 = __p0; \
3661		__ret = (int64x2_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 3); \
3662		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	3682	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 3)); \
	3683	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
3663	3684	__ret; \
3664	3685	})
3665	3686	#define __noswap_splatq_lane_s64(__p0, __p1) __extension__ ({ \
3666	3687	int64x2_t __ret; \
3667	3688	int64x1_t __s0 = __p0; \
3668		__ret = (int64x2_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 3); \
	3689	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 3)); \
3669	3690	__ret; \
3670	3691	})
3671	3692	#endif
...	...	@@ -3674,22 +3695,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
3674	3695	#define splatq_lane_s16(__p0, __p1) __extension__ ({ \
3675	3696	int16x8_t __ret; \
3676	3697	int16x4_t __s0 = __p0; \
3677		__ret = (int16x8_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 1); \
	3698	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 1)); \
3678	3699	__ret; \
3679	3700	})
3680	3701	#else
3681	3702	#define splatq_lane_s16(__p0, __p1) __extension__ ({ \
3682	3703	int16x8_t __ret; \
3683	3704	int16x4_t __s0 = __p0; \
3684		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
3685		__ret = (int16x8_t) __builtin_neon_splatq_lane_v((int8x8_t)__rev0, __p1, 1); \
3686		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	3705	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	3706	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 1)); \
	3707	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
3687	3708	__ret; \
3688	3709	})
3689	3710	#define __noswap_splatq_lane_s16(__p0, __p1) __extension__ ({ \
3690	3711	int16x8_t __ret; \
3691	3712	int16x4_t __s0 = __p0; \
3692		__ret = (int16x8_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 1); \
	3713	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 1)); \
3693	3714	__ret; \
3694	3715	})
3695	3716	#endif
...	...	@@ -3698,22 +3719,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
3698	3719	#define splat_lane_u8(__p0, __p1) __extension__ ({ \
3699	3720	uint8x8_t __ret; \
3700	3721	uint8x8_t __s0 = __p0; \
3701		__ret = (uint8x8_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 16); \
	3722	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 16)); \
3702	3723	__ret; \
3703	3724	})
3704	3725	#else
3705	3726	#define splat_lane_u8(__p0, __p1) __extension__ ({ \
3706	3727	uint8x8_t __ret; \
3707	3728	uint8x8_t __s0 = __p0; \
3708		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
3709		__ret = (uint8x8_t) __builtin_neon_splat_lane_v((int8x8_t)__rev0, __p1, 16); \
3710		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	3729	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	3730	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 16)); \
	3731	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
3711	3732	__ret; \
3712	3733	})
3713	3734	#define __noswap_splat_lane_u8(__p0, __p1) __extension__ ({ \
3714	3735	uint8x8_t __ret; \
3715	3736	uint8x8_t __s0 = __p0; \
3716		__ret = (uint8x8_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 16); \
	3737	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 16)); \
3717	3738	__ret; \
3718	3739	})
3719	3740	#endif
...	...	@@ -3722,22 +3743,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
3722	3743	#define splat_lane_u32(__p0, __p1) __extension__ ({ \
3723	3744	uint32x2_t __ret; \
3724	3745	uint32x2_t __s0 = __p0; \
3725		__ret = (uint32x2_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 18); \
	3746	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 18)); \
3726	3747	__ret; \
3727	3748	})
3728	3749	#else
3729	3750	#define splat_lane_u32(__p0, __p1) __extension__ ({ \
3730	3751	uint32x2_t __ret; \
3731	3752	uint32x2_t __s0 = __p0; \
3732		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
3733		__ret = (uint32x2_t) __builtin_neon_splat_lane_v((int8x8_t)__rev0, __p1, 18); \
3734		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	3753	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	3754	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 18)); \
	3755	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
3735	3756	__ret; \
3736	3757	})
3737	3758	#define __noswap_splat_lane_u32(__p0, __p1) __extension__ ({ \
3738	3759	uint32x2_t __ret; \
3739	3760	uint32x2_t __s0 = __p0; \
3740		__ret = (uint32x2_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 18); \
	3761	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 18)); \
3741	3762	__ret; \
3742	3763	})
3743	3764	#endif
...	...	@@ -3745,29 +3766,29 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
3745	3766	#define splat_lane_u64(__p0, __p1) __extension__ ({ \
3746	3767	uint64x1_t __ret; \
3747	3768	uint64x1_t __s0 = __p0; \
3748		__ret = (uint64x1_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 19); \
	3769	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 19)); \
3749	3770	__ret; \
3750	3771	})
3751	3772	#ifdef __LITTLE_ENDIAN__
3752	3773	#define splat_lane_u16(__p0, __p1) __extension__ ({ \
3753	3774	uint16x4_t __ret; \
3754	3775	uint16x4_t __s0 = __p0; \
3755		__ret = (uint16x4_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 17); \
	3776	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 17)); \
3756	3777	__ret; \
3757	3778	})
3758	3779	#else
3759	3780	#define splat_lane_u16(__p0, __p1) __extension__ ({ \
3760	3781	uint16x4_t __ret; \
3761	3782	uint16x4_t __s0 = __p0; \
3762		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
3763		__ret = (uint16x4_t) __builtin_neon_splat_lane_v((int8x8_t)__rev0, __p1, 17); \
3764		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	3783	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	3784	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 17)); \
	3785	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
3765	3786	__ret; \
3766	3787	})
3767	3788	#define __noswap_splat_lane_u16(__p0, __p1) __extension__ ({ \
3768	3789	uint16x4_t __ret; \
3769	3790	uint16x4_t __s0 = __p0; \
3770		__ret = (uint16x4_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 17); \
	3791	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 17)); \
3771	3792	__ret; \
3772	3793	})
3773	3794	#endif
...	...	@@ -3776,22 +3797,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
3776	3797	#define splat_lane_s8(__p0, __p1) __extension__ ({ \
3777	3798	int8x8_t __ret; \
3778	3799	int8x8_t __s0 = __p0; \
3779		__ret = (int8x8_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 0); \
	3800	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 0)); \
3780	3801	__ret; \
3781	3802	})
3782	3803	#else
3783	3804	#define splat_lane_s8(__p0, __p1) __extension__ ({ \
3784	3805	int8x8_t __ret; \
3785	3806	int8x8_t __s0 = __p0; \
3786		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
3787		__ret = (int8x8_t) __builtin_neon_splat_lane_v((int8x8_t)__rev0, __p1, 0); \
3788		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	3807	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	3808	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 0)); \
	3809	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
3789	3810	__ret; \
3790	3811	})
3791	3812	#define __noswap_splat_lane_s8(__p0, __p1) __extension__ ({ \
3792	3813	int8x8_t __ret; \
3793	3814	int8x8_t __s0 = __p0; \
3794		__ret = (int8x8_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 0); \
	3815	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 0)); \
3795	3816	__ret; \
3796	3817	})
3797	3818	#endif
...	...	@@ -3799,29 +3820,29 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
3799	3820	#define splat_lane_f64(__p0, __p1) __extension__ ({ \
3800	3821	float64x1_t __ret; \
3801	3822	float64x1_t __s0 = __p0; \
3802		__ret = (float64x1_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 10); \
	3823	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 10)); \
3803	3824	__ret; \
3804	3825	})
3805	3826	#ifdef __LITTLE_ENDIAN__
3806	3827	#define splat_lane_f32(__p0, __p1) __extension__ ({ \
3807	3828	float32x2_t __ret; \
3808	3829	float32x2_t __s0 = __p0; \
3809		__ret = (float32x2_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 9); \
	3830	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 9)); \
3810	3831	__ret; \
3811	3832	})
3812	3833	#else
3813	3834	#define splat_lane_f32(__p0, __p1) __extension__ ({ \
3814	3835	float32x2_t __ret; \
3815	3836	float32x2_t __s0 = __p0; \
3816		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
3817		__ret = (float32x2_t) __builtin_neon_splat_lane_v((int8x8_t)__rev0, __p1, 9); \
3818		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	3837	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	3838	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 9)); \
	3839	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
3819	3840	__ret; \
3820	3841	})
3821	3842	#define __noswap_splat_lane_f32(__p0, __p1) __extension__ ({ \
3822	3843	float32x2_t __ret; \
3823	3844	float32x2_t __s0 = __p0; \
3824		__ret = (float32x2_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 9); \
	3845	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 9)); \
3825	3846	__ret; \
3826	3847	})
3827	3848	#endif
...	...	@@ -3830,22 +3851,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
3830	3851	#define splat_lane_f16(__p0, __p1) __extension__ ({ \
3831	3852	float16x4_t __ret; \
3832	3853	float16x4_t __s0 = __p0; \
3833		__ret = (float16x4_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 8); \
	3854	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 8)); \
3834	3855	__ret; \
3835	3856	})
3836	3857	#else
3837	3858	#define splat_lane_f16(__p0, __p1) __extension__ ({ \
3838	3859	float16x4_t __ret; \
3839	3860	float16x4_t __s0 = __p0; \
3840		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
3841		__ret = (float16x4_t) __builtin_neon_splat_lane_v((int8x8_t)__rev0, __p1, 8); \
3842		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	3861	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	3862	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 8)); \
	3863	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
3843	3864	__ret; \
3844	3865	})
3845	3866	#define __noswap_splat_lane_f16(__p0, __p1) __extension__ ({ \
3846	3867	float16x4_t __ret; \
3847	3868	float16x4_t __s0 = __p0; \
3848		__ret = (float16x4_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 8); \
	3869	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 8)); \
3849	3870	__ret; \
3850	3871	})
3851	3872	#endif
...	...	@@ -3854,22 +3875,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
3854	3875	#define splat_lane_s32(__p0, __p1) __extension__ ({ \
3855	3876	int32x2_t __ret; \
3856	3877	int32x2_t __s0 = __p0; \
3857		__ret = (int32x2_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 2); \
	3878	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 2)); \
3858	3879	__ret; \
3859	3880	})
3860	3881	#else
3861	3882	#define splat_lane_s32(__p0, __p1) __extension__ ({ \
3862	3883	int32x2_t __ret; \
3863	3884	int32x2_t __s0 = __p0; \
3864		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
3865		__ret = (int32x2_t) __builtin_neon_splat_lane_v((int8x8_t)__rev0, __p1, 2); \
3866		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	3885	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	3886	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 2)); \
	3887	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
3867	3888	__ret; \
3868	3889	})
3869	3890	#define __noswap_splat_lane_s32(__p0, __p1) __extension__ ({ \
3870	3891	int32x2_t __ret; \
3871	3892	int32x2_t __s0 = __p0; \
3872		__ret = (int32x2_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 2); \
	3893	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 2)); \
3873	3894	__ret; \
3874	3895	})
3875	3896	#endif
...	...	@@ -3877,29 +3898,29 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
3877	3898	#define splat_lane_s64(__p0, __p1) __extension__ ({ \
3878	3899	int64x1_t __ret; \
3879	3900	int64x1_t __s0 = __p0; \
3880		__ret = (int64x1_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 3); \
	3901	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 3)); \
3881	3902	__ret; \
3882	3903	})
3883	3904	#ifdef __LITTLE_ENDIAN__
3884	3905	#define splat_lane_s16(__p0, __p1) __extension__ ({ \
3885	3906	int16x4_t __ret; \
3886	3907	int16x4_t __s0 = __p0; \
3887		__ret = (int16x4_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 1); \
	3908	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 1)); \
3888	3909	__ret; \
3889	3910	})
3890	3911	#else
3891	3912	#define splat_lane_s16(__p0, __p1) __extension__ ({ \
3892	3913	int16x4_t __ret; \
3893	3914	int16x4_t __s0 = __p0; \
3894		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
3895		__ret = (int16x4_t) __builtin_neon_splat_lane_v((int8x8_t)__rev0, __p1, 1); \
3896		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	3915	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	3916	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 1)); \
	3917	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
3897	3918	__ret; \
3898	3919	})
3899	3920	#define __noswap_splat_lane_s16(__p0, __p1) __extension__ ({ \
3900	3921	int16x4_t __ret; \
3901	3922	int16x4_t __s0 = __p0; \
3902		__ret = (int16x4_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 1); \
	3923	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 1)); \
3903	3924	__ret; \
3904	3925	})
3905	3926	#endif
...	...	@@ -3908,22 +3929,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
3908	3929	#define splat_laneq_p8(__p0, __p1) __extension__ ({ \
3909	3930	poly8x8_t __ret; \
3910	3931	poly8x16_t __s0 = __p0; \
3911		__ret = (poly8x8_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 36); \
	3932	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 36)); \
3912	3933	__ret; \
3913	3934	})
3914	3935	#else
3915	3936	#define splat_laneq_p8(__p0, __p1) __extension__ ({ \
3916	3937	poly8x8_t __ret; \
3917	3938	poly8x16_t __s0 = __p0; \
3918		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
3919		__ret = (poly8x8_t) __builtin_neon_splat_laneq_v((int8x16_t)__rev0, __p1, 36); \
3920		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	3939	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	3940	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 36)); \
	3941	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
3921	3942	__ret; \
3922	3943	})
3923	3944	#define __noswap_splat_laneq_p8(__p0, __p1) __extension__ ({ \
3924	3945	poly8x8_t __ret; \
3925	3946	poly8x16_t __s0 = __p0; \
3926		__ret = (poly8x8_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 36); \
	3947	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 36)); \
3927	3948	__ret; \
3928	3949	})
3929	3950	#endif
...	...	@@ -3932,21 +3953,21 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
3932	3953	#define splat_laneq_p64(__p0, __p1) __extension__ ({ \
3933	3954	poly64x1_t __ret; \
3934	3955	poly64x2_t __s0 = __p0; \
3935		__ret = (poly64x1_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 38); \
	3956	__ret = __builtin_bit_cast(poly64x1_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 38)); \
3936	3957	__ret; \
3937	3958	})
3938	3959	#else
3939	3960	#define splat_laneq_p64(__p0, __p1) __extension__ ({ \
3940	3961	poly64x1_t __ret; \
3941	3962	poly64x2_t __s0 = __p0; \
3942		poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
3943		__ret = (poly64x1_t) __builtin_neon_splat_laneq_v((int8x16_t)__rev0, __p1, 38); \
	3963	poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	3964	__ret = __builtin_bit_cast(poly64x1_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 38)); \
3944	3965	__ret; \
3945	3966	})
3946	3967	#define __noswap_splat_laneq_p64(__p0, __p1) __extension__ ({ \
3947	3968	poly64x1_t __ret; \
3948	3969	poly64x2_t __s0 = __p0; \
3949		__ret = (poly64x1_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 38); \
	3970	__ret = __builtin_bit_cast(poly64x1_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 38)); \
3950	3971	__ret; \
3951	3972	})
3952	3973	#endif
...	...	@@ -3955,22 +3976,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
3955	3976	#define splat_laneq_p16(__p0, __p1) __extension__ ({ \
3956	3977	poly16x4_t __ret; \
3957	3978	poly16x8_t __s0 = __p0; \
3958		__ret = (poly16x4_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 37); \
	3979	__ret = __builtin_bit_cast(poly16x4_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 37)); \
3959	3980	__ret; \
3960	3981	})
3961	3982	#else
3962	3983	#define splat_laneq_p16(__p0, __p1) __extension__ ({ \
3963	3984	poly16x4_t __ret; \
3964	3985	poly16x8_t __s0 = __p0; \
3965		poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
3966		__ret = (poly16x4_t) __builtin_neon_splat_laneq_v((int8x16_t)__rev0, __p1, 37); \
3967		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	3986	poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	3987	__ret = __builtin_bit_cast(poly16x4_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 37)); \
	3988	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
3968	3989	__ret; \
3969	3990	})
3970	3991	#define __noswap_splat_laneq_p16(__p0, __p1) __extension__ ({ \
3971	3992	poly16x4_t __ret; \
3972	3993	poly16x8_t __s0 = __p0; \
3973		__ret = (poly16x4_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 37); \
	3994	__ret = __builtin_bit_cast(poly16x4_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 37)); \
3974	3995	__ret; \
3975	3996	})
3976	3997	#endif
...	...	@@ -3979,22 +4000,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
3979	4000	#define splatq_laneq_p8(__p0, __p1) __extension__ ({ \
3980	4001	poly8x16_t __ret; \
3981	4002	poly8x16_t __s0 = __p0; \
3982		__ret = (poly8x16_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 36); \
	4003	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 36)); \
3983	4004	__ret; \
3984	4005	})
3985	4006	#else
3986	4007	#define splatq_laneq_p8(__p0, __p1) __extension__ ({ \
3987	4008	poly8x16_t __ret; \
3988	4009	poly8x16_t __s0 = __p0; \
3989		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
3990		__ret = (poly8x16_t) __builtin_neon_splatq_laneq_v((int8x16_t)__rev0, __p1, 36); \
3991		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	4010	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	4011	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 36)); \
	4012	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
3992	4013	__ret; \
3993	4014	})
3994	4015	#define __noswap_splatq_laneq_p8(__p0, __p1) __extension__ ({ \
3995	4016	poly8x16_t __ret; \
3996	4017	poly8x16_t __s0 = __p0; \
3997		__ret = (poly8x16_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 36); \
	4018	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 36)); \
3998	4019	__ret; \
3999	4020	})
4000	4021	#endif
...	...	@@ -4003,22 +4024,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
4003	4024	#define splatq_laneq_p64(__p0, __p1) __extension__ ({ \
4004	4025	poly64x2_t __ret; \
4005	4026	poly64x2_t __s0 = __p0; \
4006		__ret = (poly64x2_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 38); \
	4027	__ret = __builtin_bit_cast(poly64x2_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 38)); \
4007	4028	__ret; \
4008	4029	})
4009	4030	#else
4010	4031	#define splatq_laneq_p64(__p0, __p1) __extension__ ({ \
4011	4032	poly64x2_t __ret; \
4012	4033	poly64x2_t __s0 = __p0; \
4013		poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
4014		__ret = (poly64x2_t) __builtin_neon_splatq_laneq_v((int8x16_t)__rev0, __p1, 38); \
4015		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	4034	poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	4035	__ret = __builtin_bit_cast(poly64x2_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 38)); \
	4036	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
4016	4037	__ret; \
4017	4038	})
4018	4039	#define __noswap_splatq_laneq_p64(__p0, __p1) __extension__ ({ \
4019	4040	poly64x2_t __ret; \
4020	4041	poly64x2_t __s0 = __p0; \
4021		__ret = (poly64x2_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 38); \
	4042	__ret = __builtin_bit_cast(poly64x2_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 38)); \
4022	4043	__ret; \
4023	4044	})
4024	4045	#endif
...	...	@@ -4027,22 +4048,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
4027	4048	#define splatq_laneq_p16(__p0, __p1) __extension__ ({ \
4028	4049	poly16x8_t __ret; \
4029	4050	poly16x8_t __s0 = __p0; \
4030		__ret = (poly16x8_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 37); \
	4051	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 37)); \
4031	4052	__ret; \
4032	4053	})
4033	4054	#else
4034	4055	#define splatq_laneq_p16(__p0, __p1) __extension__ ({ \
4035	4056	poly16x8_t __ret; \
4036	4057	poly16x8_t __s0 = __p0; \
4037		poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
4038		__ret = (poly16x8_t) __builtin_neon_splatq_laneq_v((int8x16_t)__rev0, __p1, 37); \
4039		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	4058	poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	4059	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 37)); \
	4060	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
4040	4061	__ret; \
4041	4062	})
4042	4063	#define __noswap_splatq_laneq_p16(__p0, __p1) __extension__ ({ \
4043	4064	poly16x8_t __ret; \
4044	4065	poly16x8_t __s0 = __p0; \
4045		__ret = (poly16x8_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 37); \
	4066	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 37)); \
4046	4067	__ret; \
4047	4068	})
4048	4069	#endif
...	...	@@ -4051,22 +4072,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
4051	4072	#define splatq_laneq_u8(__p0, __p1) __extension__ ({ \
4052	4073	uint8x16_t __ret; \
4053	4074	uint8x16_t __s0 = __p0; \
4054		__ret = (uint8x16_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 48); \
	4075	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 48)); \
4055	4076	__ret; \
4056	4077	})
4057	4078	#else
4058	4079	#define splatq_laneq_u8(__p0, __p1) __extension__ ({ \
4059	4080	uint8x16_t __ret; \
4060	4081	uint8x16_t __s0 = __p0; \
4061		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
4062		__ret = (uint8x16_t) __builtin_neon_splatq_laneq_v((int8x16_t)__rev0, __p1, 48); \
4063		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	4082	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	4083	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 48)); \
	4084	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
4064	4085	__ret; \
4065	4086	})
4066	4087	#define __noswap_splatq_laneq_u8(__p0, __p1) __extension__ ({ \
4067	4088	uint8x16_t __ret; \
4068	4089	uint8x16_t __s0 = __p0; \
4069		__ret = (uint8x16_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 48); \
	4090	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 48)); \
4070	4091	__ret; \
4071	4092	})
4072	4093	#endif
...	...	@@ -4075,22 +4096,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
4075	4096	#define splatq_laneq_u32(__p0, __p1) __extension__ ({ \
4076	4097	uint32x4_t __ret; \
4077	4098	uint32x4_t __s0 = __p0; \
4078		__ret = (uint32x4_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 50); \
	4099	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 50)); \
4079	4100	__ret; \
4080	4101	})
4081	4102	#else
4082	4103	#define splatq_laneq_u32(__p0, __p1) __extension__ ({ \
4083	4104	uint32x4_t __ret; \
4084	4105	uint32x4_t __s0 = __p0; \
4085		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
4086		__ret = (uint32x4_t) __builtin_neon_splatq_laneq_v((int8x16_t)__rev0, __p1, 50); \
4087		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	4106	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	4107	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 50)); \
	4108	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
4088	4109	__ret; \
4089	4110	})
4090	4111	#define __noswap_splatq_laneq_u32(__p0, __p1) __extension__ ({ \
4091	4112	uint32x4_t __ret; \
4092	4113	uint32x4_t __s0 = __p0; \
4093		__ret = (uint32x4_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 50); \
	4114	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 50)); \
4094	4115	__ret; \
4095	4116	})
4096	4117	#endif
...	...	@@ -4099,22 +4120,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
4099	4120	#define splatq_laneq_u64(__p0, __p1) __extension__ ({ \
4100	4121	uint64x2_t __ret; \
4101	4122	uint64x2_t __s0 = __p0; \
4102		__ret = (uint64x2_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 51); \
	4123	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 51)); \
4103	4124	__ret; \
4104	4125	})
4105	4126	#else
4106	4127	#define splatq_laneq_u64(__p0, __p1) __extension__ ({ \
4107	4128	uint64x2_t __ret; \
4108	4129	uint64x2_t __s0 = __p0; \
4109		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
4110		__ret = (uint64x2_t) __builtin_neon_splatq_laneq_v((int8x16_t)__rev0, __p1, 51); \
4111		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	4130	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	4131	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 51)); \
	4132	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
4112	4133	__ret; \
4113	4134	})
4114	4135	#define __noswap_splatq_laneq_u64(__p0, __p1) __extension__ ({ \
4115	4136	uint64x2_t __ret; \
4116	4137	uint64x2_t __s0 = __p0; \
4117		__ret = (uint64x2_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 51); \
	4138	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 51)); \
4118	4139	__ret; \
4119	4140	})
4120	4141	#endif
...	...	@@ -4123,22 +4144,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
4123	4144	#define splatq_laneq_u16(__p0, __p1) __extension__ ({ \
4124	4145	uint16x8_t __ret; \
4125	4146	uint16x8_t __s0 = __p0; \
4126		__ret = (uint16x8_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 49); \
	4147	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 49)); \
4127	4148	__ret; \
4128	4149	})
4129	4150	#else
4130	4151	#define splatq_laneq_u16(__p0, __p1) __extension__ ({ \
4131	4152	uint16x8_t __ret; \
4132	4153	uint16x8_t __s0 = __p0; \
4133		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
4134		__ret = (uint16x8_t) __builtin_neon_splatq_laneq_v((int8x16_t)__rev0, __p1, 49); \
4135		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	4154	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	4155	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 49)); \
	4156	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
4136	4157	__ret; \
4137	4158	})
4138	4159	#define __noswap_splatq_laneq_u16(__p0, __p1) __extension__ ({ \
4139	4160	uint16x8_t __ret; \
4140	4161	uint16x8_t __s0 = __p0; \
4141		__ret = (uint16x8_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 49); \
	4162	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 49)); \
4142	4163	__ret; \
4143	4164	})
4144	4165	#endif
...	...	@@ -4147,22 +4168,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
4147	4168	#define splatq_laneq_s8(__p0, __p1) __extension__ ({ \
4148	4169	int8x16_t __ret; \
4149	4170	int8x16_t __s0 = __p0; \
4150		__ret = (int8x16_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 32); \
	4171	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 32)); \
4151	4172	__ret; \
4152	4173	})
4153	4174	#else
4154	4175	#define splatq_laneq_s8(__p0, __p1) __extension__ ({ \
4155	4176	int8x16_t __ret; \
4156	4177	int8x16_t __s0 = __p0; \
4157		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
4158		__ret = (int8x16_t) __builtin_neon_splatq_laneq_v((int8x16_t)__rev0, __p1, 32); \
4159		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	4178	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	4179	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 32)); \
	4180	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
4160	4181	__ret; \
4161	4182	})
4162	4183	#define __noswap_splatq_laneq_s8(__p0, __p1) __extension__ ({ \
4163	4184	int8x16_t __ret; \
4164	4185	int8x16_t __s0 = __p0; \
4165		__ret = (int8x16_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 32); \
	4186	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 32)); \
4166	4187	__ret; \
4167	4188	})
4168	4189	#endif
...	...	@@ -4171,22 +4192,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
4171	4192	#define splatq_laneq_f64(__p0, __p1) __extension__ ({ \
4172	4193	float64x2_t __ret; \
4173	4194	float64x2_t __s0 = __p0; \
4174		__ret = (float64x2_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 42); \
	4195	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 42)); \
4175	4196	__ret; \
4176	4197	})
4177	4198	#else
4178	4199	#define splatq_laneq_f64(__p0, __p1) __extension__ ({ \
4179	4200	float64x2_t __ret; \
4180	4201	float64x2_t __s0 = __p0; \
4181		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
4182		__ret = (float64x2_t) __builtin_neon_splatq_laneq_v((int8x16_t)__rev0, __p1, 42); \
4183		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	4202	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	4203	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 42)); \
	4204	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
4184	4205	__ret; \
4185	4206	})
4186	4207	#define __noswap_splatq_laneq_f64(__p0, __p1) __extension__ ({ \
4187	4208	float64x2_t __ret; \
4188	4209	float64x2_t __s0 = __p0; \
4189		__ret = (float64x2_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 42); \
	4210	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 42)); \
4190	4211	__ret; \
4191	4212	})
4192	4213	#endif
...	...	@@ -4195,22 +4216,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
4195	4216	#define splatq_laneq_f32(__p0, __p1) __extension__ ({ \
4196	4217	float32x4_t __ret; \
4197	4218	float32x4_t __s0 = __p0; \
4198		__ret = (float32x4_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 41); \
	4219	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 41)); \
4199	4220	__ret; \
4200	4221	})
4201	4222	#else
4202	4223	#define splatq_laneq_f32(__p0, __p1) __extension__ ({ \
4203	4224	float32x4_t __ret; \
4204	4225	float32x4_t __s0 = __p0; \
4205		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
4206		__ret = (float32x4_t) __builtin_neon_splatq_laneq_v((int8x16_t)__rev0, __p1, 41); \
4207		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	4226	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	4227	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 41)); \
	4228	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
4208	4229	__ret; \
4209	4230	})
4210	4231	#define __noswap_splatq_laneq_f32(__p0, __p1) __extension__ ({ \
4211	4232	float32x4_t __ret; \
4212	4233	float32x4_t __s0 = __p0; \
4213		__ret = (float32x4_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 41); \
	4234	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 41)); \
4214	4235	__ret; \
4215	4236	})
4216	4237	#endif
...	...	@@ -4219,22 +4240,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
4219	4240	#define splatq_laneq_f16(__p0, __p1) __extension__ ({ \
4220	4241	float16x8_t __ret; \
4221	4242	float16x8_t __s0 = __p0; \
4222		__ret = (float16x8_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 40); \
	4243	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 40)); \
4223	4244	__ret; \
4224	4245	})
4225	4246	#else
4226	4247	#define splatq_laneq_f16(__p0, __p1) __extension__ ({ \
4227	4248	float16x8_t __ret; \
4228	4249	float16x8_t __s0 = __p0; \
4229		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
4230		__ret = (float16x8_t) __builtin_neon_splatq_laneq_v((int8x16_t)__rev0, __p1, 40); \
4231		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	4250	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	4251	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 40)); \
	4252	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
4232	4253	__ret; \
4233	4254	})
4234	4255	#define __noswap_splatq_laneq_f16(__p0, __p1) __extension__ ({ \
4235	4256	float16x8_t __ret; \
4236	4257	float16x8_t __s0 = __p0; \
4237		__ret = (float16x8_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 40); \
	4258	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 40)); \
4238	4259	__ret; \
4239	4260	})
4240	4261	#endif
...	...	@@ -4243,22 +4264,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
4243	4264	#define splatq_laneq_s32(__p0, __p1) __extension__ ({ \
4244	4265	int32x4_t __ret; \
4245	4266	int32x4_t __s0 = __p0; \
4246		__ret = (int32x4_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 34); \
	4267	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 34)); \
4247	4268	__ret; \
4248	4269	})
4249	4270	#else
4250	4271	#define splatq_laneq_s32(__p0, __p1) __extension__ ({ \
4251	4272	int32x4_t __ret; \
4252	4273	int32x4_t __s0 = __p0; \
4253		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
4254		__ret = (int32x4_t) __builtin_neon_splatq_laneq_v((int8x16_t)__rev0, __p1, 34); \
4255		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	4274	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	4275	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 34)); \
	4276	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
4256	4277	__ret; \
4257	4278	})
4258	4279	#define __noswap_splatq_laneq_s32(__p0, __p1) __extension__ ({ \
4259	4280	int32x4_t __ret; \
4260	4281	int32x4_t __s0 = __p0; \
4261		__ret = (int32x4_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 34); \
	4282	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 34)); \
4262	4283	__ret; \
4263	4284	})
4264	4285	#endif
...	...	@@ -4267,22 +4288,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
4267	4288	#define splatq_laneq_s64(__p0, __p1) __extension__ ({ \
4268	4289	int64x2_t __ret; \
4269	4290	int64x2_t __s0 = __p0; \
4270		__ret = (int64x2_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 35); \
	4291	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 35)); \
4271	4292	__ret; \
4272	4293	})
4273	4294	#else
4274	4295	#define splatq_laneq_s64(__p0, __p1) __extension__ ({ \
4275	4296	int64x2_t __ret; \
4276	4297	int64x2_t __s0 = __p0; \
4277		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
4278		__ret = (int64x2_t) __builtin_neon_splatq_laneq_v((int8x16_t)__rev0, __p1, 35); \
4279		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	4298	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	4299	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 35)); \
	4300	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
4280	4301	__ret; \
4281	4302	})
4282	4303	#define __noswap_splatq_laneq_s64(__p0, __p1) __extension__ ({ \
4283	4304	int64x2_t __ret; \
4284	4305	int64x2_t __s0 = __p0; \
4285		__ret = (int64x2_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 35); \
	4306	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 35)); \
4286	4307	__ret; \
4287	4308	})
4288	4309	#endif
...	...	@@ -4291,22 +4312,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
4291	4312	#define splatq_laneq_s16(__p0, __p1) __extension__ ({ \
4292	4313	int16x8_t __ret; \
4293	4314	int16x8_t __s0 = __p0; \
4294		__ret = (int16x8_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 33); \
	4315	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 33)); \
4295	4316	__ret; \
4296	4317	})
4297	4318	#else
4298	4319	#define splatq_laneq_s16(__p0, __p1) __extension__ ({ \
4299	4320	int16x8_t __ret; \
4300	4321	int16x8_t __s0 = __p0; \
4301		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
4302		__ret = (int16x8_t) __builtin_neon_splatq_laneq_v((int8x16_t)__rev0, __p1, 33); \
4303		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	4322	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	4323	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 33)); \
	4324	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
4304	4325	__ret; \
4305	4326	})
4306	4327	#define __noswap_splatq_laneq_s16(__p0, __p1) __extension__ ({ \
4307	4328	int16x8_t __ret; \
4308	4329	int16x8_t __s0 = __p0; \
4309		__ret = (int16x8_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 33); \
	4330	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 33)); \
4310	4331	__ret; \
4311	4332	})
4312	4333	#endif
...	...	@@ -4315,22 +4336,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
4315	4336	#define splat_laneq_u8(__p0, __p1) __extension__ ({ \
4316	4337	uint8x8_t __ret; \
4317	4338	uint8x16_t __s0 = __p0; \
4318		__ret = (uint8x8_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 48); \
	4339	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 48)); \
4319	4340	__ret; \
4320	4341	})
4321	4342	#else
4322	4343	#define splat_laneq_u8(__p0, __p1) __extension__ ({ \
4323	4344	uint8x8_t __ret; \
4324	4345	uint8x16_t __s0 = __p0; \
4325		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
4326		__ret = (uint8x8_t) __builtin_neon_splat_laneq_v((int8x16_t)__rev0, __p1, 48); \
4327		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	4346	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	4347	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 48)); \
	4348	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
4328	4349	__ret; \
4329	4350	})
4330	4351	#define __noswap_splat_laneq_u8(__p0, __p1) __extension__ ({ \
4331	4352	uint8x8_t __ret; \
4332	4353	uint8x16_t __s0 = __p0; \
4333		__ret = (uint8x8_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 48); \
	4354	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 48)); \
4334	4355	__ret; \
4335	4356	})
4336	4357	#endif
...	...	@@ -4339,22 +4360,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
4339	4360	#define splat_laneq_u32(__p0, __p1) __extension__ ({ \
4340	4361	uint32x2_t __ret; \
4341	4362	uint32x4_t __s0 = __p0; \
4342		__ret = (uint32x2_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 50); \
	4363	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 50)); \
4343	4364	__ret; \
4344	4365	})
4345	4366	#else
4346	4367	#define splat_laneq_u32(__p0, __p1) __extension__ ({ \
4347	4368	uint32x2_t __ret; \
4348	4369	uint32x4_t __s0 = __p0; \
4349		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
4350		__ret = (uint32x2_t) __builtin_neon_splat_laneq_v((int8x16_t)__rev0, __p1, 50); \
4351		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	4370	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	4371	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 50)); \
	4372	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
4352	4373	__ret; \
4353	4374	})
4354	4375	#define __noswap_splat_laneq_u32(__p0, __p1) __extension__ ({ \
4355	4376	uint32x2_t __ret; \
4356	4377	uint32x4_t __s0 = __p0; \
4357		__ret = (uint32x2_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 50); \
	4378	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 50)); \
4358	4379	__ret; \
4359	4380	})
4360	4381	#endif
...	...	@@ -4363,21 +4384,21 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
4363	4384	#define splat_laneq_u64(__p0, __p1) __extension__ ({ \
4364	4385	uint64x1_t __ret; \
4365	4386	uint64x2_t __s0 = __p0; \
4366		__ret = (uint64x1_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 51); \
	4387	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 51)); \
4367	4388	__ret; \
4368	4389	})
4369	4390	#else
4370	4391	#define splat_laneq_u64(__p0, __p1) __extension__ ({ \
4371	4392	uint64x1_t __ret; \
4372	4393	uint64x2_t __s0 = __p0; \
4373		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
4374		__ret = (uint64x1_t) __builtin_neon_splat_laneq_v((int8x16_t)__rev0, __p1, 51); \
	4394	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	4395	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 51)); \
4375	4396	__ret; \
4376	4397	})
4377	4398	#define __noswap_splat_laneq_u64(__p0, __p1) __extension__ ({ \
4378	4399	uint64x1_t __ret; \
4379	4400	uint64x2_t __s0 = __p0; \
4380		__ret = (uint64x1_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 51); \
	4401	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 51)); \
4381	4402	__ret; \
4382	4403	})
4383	4404	#endif
...	...	@@ -4386,22 +4407,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
4386	4407	#define splat_laneq_u16(__p0, __p1) __extension__ ({ \
4387	4408	uint16x4_t __ret; \
4388	4409	uint16x8_t __s0 = __p0; \
4389		__ret = (uint16x4_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 49); \
	4410	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 49)); \
4390	4411	__ret; \
4391	4412	})
4392	4413	#else
4393	4414	#define splat_laneq_u16(__p0, __p1) __extension__ ({ \
4394	4415	uint16x4_t __ret; \
4395	4416	uint16x8_t __s0 = __p0; \
4396		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
4397		__ret = (uint16x4_t) __builtin_neon_splat_laneq_v((int8x16_t)__rev0, __p1, 49); \
4398		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	4417	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	4418	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 49)); \
	4419	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
4399	4420	__ret; \
4400	4421	})
4401	4422	#define __noswap_splat_laneq_u16(__p0, __p1) __extension__ ({ \
4402	4423	uint16x4_t __ret; \
4403	4424	uint16x8_t __s0 = __p0; \
4404		__ret = (uint16x4_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 49); \
	4425	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 49)); \
4405	4426	__ret; \
4406	4427	})
4407	4428	#endif
...	...	@@ -4410,22 +4431,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
4410	4431	#define splat_laneq_s8(__p0, __p1) __extension__ ({ \
4411	4432	int8x8_t __ret; \
4412	4433	int8x16_t __s0 = __p0; \
4413		__ret = (int8x8_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 32); \
	4434	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 32)); \
4414	4435	__ret; \
4415	4436	})
4416	4437	#else
4417	4438	#define splat_laneq_s8(__p0, __p1) __extension__ ({ \
4418	4439	int8x8_t __ret; \
4419	4440	int8x16_t __s0 = __p0; \
4420		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
4421		__ret = (int8x8_t) __builtin_neon_splat_laneq_v((int8x16_t)__rev0, __p1, 32); \
4422		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	4441	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	4442	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 32)); \
	4443	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
4423	4444	__ret; \
4424	4445	})
4425	4446	#define __noswap_splat_laneq_s8(__p0, __p1) __extension__ ({ \
4426	4447	int8x8_t __ret; \
4427	4448	int8x16_t __s0 = __p0; \
4428		__ret = (int8x8_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 32); \
	4449	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 32)); \
4429	4450	__ret; \
4430	4451	})
4431	4452	#endif
...	...	@@ -4434,21 +4455,21 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
4434	4455	#define splat_laneq_f64(__p0, __p1) __extension__ ({ \
4435	4456	float64x1_t __ret; \
4436	4457	float64x2_t __s0 = __p0; \
4437		__ret = (float64x1_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 42); \
	4458	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 42)); \
4438	4459	__ret; \
4439	4460	})
4440	4461	#else
4441	4462	#define splat_laneq_f64(__p0, __p1) __extension__ ({ \
4442	4463	float64x1_t __ret; \
4443	4464	float64x2_t __s0 = __p0; \
4444		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
4445		__ret = (float64x1_t) __builtin_neon_splat_laneq_v((int8x16_t)__rev0, __p1, 42); \
	4465	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	4466	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 42)); \
4446	4467	__ret; \
4447	4468	})
4448	4469	#define __noswap_splat_laneq_f64(__p0, __p1) __extension__ ({ \
4449	4470	float64x1_t __ret; \
4450	4471	float64x2_t __s0 = __p0; \
4451		__ret = (float64x1_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 42); \
	4472	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 42)); \
4452	4473	__ret; \
4453	4474	})
4454	4475	#endif
...	...	@@ -4457,22 +4478,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
4457	4478	#define splat_laneq_f32(__p0, __p1) __extension__ ({ \
4458	4479	float32x2_t __ret; \
4459	4480	float32x4_t __s0 = __p0; \
4460		__ret = (float32x2_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 41); \
	4481	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 41)); \
4461	4482	__ret; \
4462	4483	})
4463	4484	#else
4464	4485	#define splat_laneq_f32(__p0, __p1) __extension__ ({ \
4465	4486	float32x2_t __ret; \
4466	4487	float32x4_t __s0 = __p0; \
4467		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
4468		__ret = (float32x2_t) __builtin_neon_splat_laneq_v((int8x16_t)__rev0, __p1, 41); \
4469		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	4488	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	4489	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 41)); \
	4490	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
4470	4491	__ret; \
4471	4492	})
4472	4493	#define __noswap_splat_laneq_f32(__p0, __p1) __extension__ ({ \
4473	4494	float32x2_t __ret; \
4474	4495	float32x4_t __s0 = __p0; \
4475		__ret = (float32x2_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 41); \
	4496	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 41)); \
4476	4497	__ret; \
4477	4498	})
4478	4499	#endif
...	...	@@ -4481,22 +4502,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
4481	4502	#define splat_laneq_f16(__p0, __p1) __extension__ ({ \
4482	4503	float16x4_t __ret; \
4483	4504	float16x8_t __s0 = __p0; \
4484		__ret = (float16x4_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 40); \
	4505	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 40)); \
4485	4506	__ret; \
4486	4507	})
4487	4508	#else
4488	4509	#define splat_laneq_f16(__p0, __p1) __extension__ ({ \
4489	4510	float16x4_t __ret; \
4490	4511	float16x8_t __s0 = __p0; \
4491		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
4492		__ret = (float16x4_t) __builtin_neon_splat_laneq_v((int8x16_t)__rev0, __p1, 40); \
4493		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	4512	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	4513	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 40)); \
	4514	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
4494	4515	__ret; \
4495	4516	})
4496	4517	#define __noswap_splat_laneq_f16(__p0, __p1) __extension__ ({ \
4497	4518	float16x4_t __ret; \
4498	4519	float16x8_t __s0 = __p0; \
4499		__ret = (float16x4_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 40); \
	4520	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 40)); \
4500	4521	__ret; \
4501	4522	})
4502	4523	#endif
...	...	@@ -4505,22 +4526,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
4505	4526	#define splat_laneq_s32(__p0, __p1) __extension__ ({ \
4506	4527	int32x2_t __ret; \
4507	4528	int32x4_t __s0 = __p0; \
4508		__ret = (int32x2_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 34); \
	4529	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 34)); \
4509	4530	__ret; \
4510	4531	})
4511	4532	#else
4512	4533	#define splat_laneq_s32(__p0, __p1) __extension__ ({ \
4513	4534	int32x2_t __ret; \
4514	4535	int32x4_t __s0 = __p0; \
4515		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
4516		__ret = (int32x2_t) __builtin_neon_splat_laneq_v((int8x16_t)__rev0, __p1, 34); \
4517		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	4536	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	4537	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 34)); \
	4538	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
4518	4539	__ret; \
4519	4540	})
4520	4541	#define __noswap_splat_laneq_s32(__p0, __p1) __extension__ ({ \
4521	4542	int32x2_t __ret; \
4522	4543	int32x4_t __s0 = __p0; \
4523		__ret = (int32x2_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 34); \
	4544	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 34)); \
4524	4545	__ret; \
4525	4546	})
4526	4547	#endif
...	...	@@ -4529,21 +4550,21 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
4529	4550	#define splat_laneq_s64(__p0, __p1) __extension__ ({ \
4530	4551	int64x1_t __ret; \
4531	4552	int64x2_t __s0 = __p0; \
4532		__ret = (int64x1_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 35); \
	4553	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 35)); \
4533	4554	__ret; \
4534	4555	})
4535	4556	#else
4536	4557	#define splat_laneq_s64(__p0, __p1) __extension__ ({ \
4537	4558	int64x1_t __ret; \
4538	4559	int64x2_t __s0 = __p0; \
4539		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
4540		__ret = (int64x1_t) __builtin_neon_splat_laneq_v((int8x16_t)__rev0, __p1, 35); \
	4560	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	4561	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 35)); \
4541	4562	__ret; \
4542	4563	})
4543	4564	#define __noswap_splat_laneq_s64(__p0, __p1) __extension__ ({ \
4544	4565	int64x1_t __ret; \
4545	4566	int64x2_t __s0 = __p0; \
4546		__ret = (int64x1_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 35); \
	4567	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 35)); \
4547	4568	__ret; \
4548	4569	})
4549	4570	#endif
...	...	@@ -4552,22 +4573,22 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
4552	4573	#define splat_laneq_s16(__p0, __p1) __extension__ ({ \
4553	4574	int16x4_t __ret; \
4554	4575	int16x8_t __s0 = __p0; \
4555		__ret = (int16x4_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 33); \
	4576	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 33)); \
4556	4577	__ret; \
4557	4578	})
4558	4579	#else
4559	4580	#define splat_laneq_s16(__p0, __p1) __extension__ ({ \
4560	4581	int16x4_t __ret; \
4561	4582	int16x8_t __s0 = __p0; \
4562		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
4563		__ret = (int16x4_t) __builtin_neon_splat_laneq_v((int8x16_t)__rev0, __p1, 33); \
4564		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	4583	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	4584	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 33)); \
	4585	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
4565	4586	__ret; \
4566	4587	})
4567	4588	#define __noswap_splat_laneq_s16(__p0, __p1) __extension__ ({ \
4568	4589	int16x4_t __ret; \
4569	4590	int16x8_t __s0 = __p0; \
4570		__ret = (int16x4_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 33); \
	4591	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 33)); \
4571	4592	__ret; \
4572	4593	})
4573	4594	#endif
...	...	@@ -4575,21 +4596,21 @@ __ai __attribute__((target("i8mm,neon"))) int32x4_t vusmmlaq_s32(int32x4_t __p0,
4575	4596	#ifdef __LITTLE_ENDIAN__
4576	4597	__ai __attribute__((target("neon"))) uint8x16_t vabdq_u8(uint8x16_t __p0, uint8x16_t __p1) {
4577	4598	uint8x16_t __ret;
4578		__ret = (uint8x16_t) __builtin_neon_vabdq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
	4599	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vabdq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 48));
4579	4600	return __ret;
4580	4601	}
4581	4602	#else
4582	4603	__ai __attribute__((target("neon"))) uint8x16_t vabdq_u8(uint8x16_t __p0, uint8x16_t __p1) {
4583	4604	uint8x16_t __ret;
4584		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4585		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4586		__ret = (uint8x16_t) __builtin_neon_vabdq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
4587		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	4605	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	4606	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	4607	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vabdq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 48));
	4608	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
4588	4609	return __ret;
4589	4610	}
4590	4611	__ai __attribute__((target("neon"))) uint8x16_t __noswap_vabdq_u8(uint8x16_t __p0, uint8x16_t __p1) {
4591	4612	uint8x16_t __ret;
4592		__ret = (uint8x16_t) __builtin_neon_vabdq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
	4613	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vabdq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 48));
4593	4614	return __ret;
4594	4615	}
4595	4616	#endif
...	...	@@ -4597,21 +4618,21 @@ __ai __attribute__((target("neon"))) uint8x16_t __noswap_vabdq_u8(uint8x16_t __p
4597	4618	#ifdef __LITTLE_ENDIAN__
4598	4619	__ai __attribute__((target("neon"))) uint32x4_t vabdq_u32(uint32x4_t __p0, uint32x4_t __p1) {
4599	4620	uint32x4_t __ret;
4600		__ret = (uint32x4_t) __builtin_neon_vabdq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
	4621	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vabdq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50));
4601	4622	return __ret;
4602	4623	}
4603	4624	#else
4604	4625	__ai __attribute__((target("neon"))) uint32x4_t vabdq_u32(uint32x4_t __p0, uint32x4_t __p1) {
4605	4626	uint32x4_t __ret;
4606		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
4607		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
4608		__ret = (uint32x4_t) __builtin_neon_vabdq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
4609		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	4627	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	4628	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	4629	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vabdq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 50));
	4630	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
4610	4631	return __ret;
4611	4632	}
4612	4633	__ai __attribute__((target("neon"))) uint32x4_t __noswap_vabdq_u32(uint32x4_t __p0, uint32x4_t __p1) {
4613	4634	uint32x4_t __ret;
4614		__ret = (uint32x4_t) __builtin_neon_vabdq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
	4635	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vabdq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50));
4615	4636	return __ret;
4616	4637	}
4617	4638	#endif
...	...	@@ -4619,21 +4640,21 @@ __ai __attribute__((target("neon"))) uint32x4_t __noswap_vabdq_u32(uint32x4_t __
4619	4640	#ifdef __LITTLE_ENDIAN__
4620	4641	__ai __attribute__((target("neon"))) uint16x8_t vabdq_u16(uint16x8_t __p0, uint16x8_t __p1) {
4621	4642	uint16x8_t __ret;
4622		__ret = (uint16x8_t) __builtin_neon_vabdq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
	4643	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vabdq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 49));
4623	4644	return __ret;
4624	4645	}
4625	4646	#else
4626	4647	__ai __attribute__((target("neon"))) uint16x8_t vabdq_u16(uint16x8_t __p0, uint16x8_t __p1) {
4627	4648	uint16x8_t __ret;
4628		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
4629		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
4630		__ret = (uint16x8_t) __builtin_neon_vabdq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
4631		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	4649	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	4650	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	4651	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vabdq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 49));
	4652	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
4632	4653	return __ret;
4633	4654	}
4634	4655	__ai __attribute__((target("neon"))) uint16x8_t __noswap_vabdq_u16(uint16x8_t __p0, uint16x8_t __p1) {
4635	4656	uint16x8_t __ret;
4636		__ret = (uint16x8_t) __builtin_neon_vabdq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
	4657	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vabdq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 49));
4637	4658	return __ret;
4638	4659	}
4639	4660	#endif
...	...	@@ -4641,21 +4662,21 @@ __ai __attribute__((target("neon"))) uint16x8_t __noswap_vabdq_u16(uint16x8_t __
4641	4662	#ifdef __LITTLE_ENDIAN__
4642	4663	__ai __attribute__((target("neon"))) int8x16_t vabdq_s8(int8x16_t __p0, int8x16_t __p1) {
4643	4664	int8x16_t __ret;
4644		__ret = (int8x16_t) __builtin_neon_vabdq_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
	4665	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vabdq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 32));
4645	4666	return __ret;
4646	4667	}
4647	4668	#else
4648	4669	__ai __attribute__((target("neon"))) int8x16_t vabdq_s8(int8x16_t __p0, int8x16_t __p1) {
4649	4670	int8x16_t __ret;
4650		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4651		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4652		__ret = (int8x16_t) __builtin_neon_vabdq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 32);
4653		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	4671	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	4672	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	4673	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vabdq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 32));
	4674	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
4654	4675	return __ret;
4655	4676	}
4656	4677	__ai __attribute__((target("neon"))) int8x16_t __noswap_vabdq_s8(int8x16_t __p0, int8x16_t __p1) {
4657	4678	int8x16_t __ret;
4658		__ret = (int8x16_t) __builtin_neon_vabdq_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
	4679	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vabdq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 32));
4659	4680	return __ret;
4660	4681	}
4661	4682	#endif
...	...	@@ -4663,16 +4684,16 @@ __ai __attribute__((target("neon"))) int8x16_t __noswap_vabdq_s8(int8x16_t __p0,
4663	4684	#ifdef __LITTLE_ENDIAN__
4664	4685	__ai __attribute__((target("neon"))) float32x4_t vabdq_f32(float32x4_t __p0, float32x4_t __p1) {
4665	4686	float32x4_t __ret;
4666		__ret = (float32x4_t) __builtin_neon_vabdq_v((int8x16_t)__p0, (int8x16_t)__p1, 41);
	4687	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vabdq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 41));
4667	4688	return __ret;
4668	4689	}
4669	4690	#else
4670	4691	__ai __attribute__((target("neon"))) float32x4_t vabdq_f32(float32x4_t __p0, float32x4_t __p1) {
4671	4692	float32x4_t __ret;
4672		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
4673		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
4674		__ret = (float32x4_t) __builtin_neon_vabdq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 41);
4675		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	4693	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	4694	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	4695	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vabdq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 41));
	4696	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
4676	4697	return __ret;
4677	4698	}
4678	4699	#endif
...	...	@@ -4680,21 +4701,21 @@ __ai __attribute__((target("neon"))) float32x4_t vabdq_f32(float32x4_t __p0, flo
4680	4701	#ifdef __LITTLE_ENDIAN__
4681	4702	__ai __attribute__((target("neon"))) int32x4_t vabdq_s32(int32x4_t __p0, int32x4_t __p1) {
4682	4703	int32x4_t __ret;
4683		__ret = (int32x4_t) __builtin_neon_vabdq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
	4704	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vabdq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 34));
4684	4705	return __ret;
4685	4706	}
4686	4707	#else
4687	4708	__ai __attribute__((target("neon"))) int32x4_t vabdq_s32(int32x4_t __p0, int32x4_t __p1) {
4688	4709	int32x4_t __ret;
4689		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
4690		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
4691		__ret = (int32x4_t) __builtin_neon_vabdq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
4692		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	4710	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	4711	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	4712	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vabdq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 34));
	4713	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
4693	4714	return __ret;
4694	4715	}
4695	4716	__ai __attribute__((target("neon"))) int32x4_t __noswap_vabdq_s32(int32x4_t __p0, int32x4_t __p1) {
4696	4717	int32x4_t __ret;
4697		__ret = (int32x4_t) __builtin_neon_vabdq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
	4718	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vabdq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 34));
4698	4719	return __ret;
4699	4720	}
4700	4721	#endif
...	...	@@ -4702,21 +4723,21 @@ __ai __attribute__((target("neon"))) int32x4_t __noswap_vabdq_s32(int32x4_t __p0
4702	4723	#ifdef __LITTLE_ENDIAN__
4703	4724	__ai __attribute__((target("neon"))) int16x8_t vabdq_s16(int16x8_t __p0, int16x8_t __p1) {
4704	4725	int16x8_t __ret;
4705		__ret = (int16x8_t) __builtin_neon_vabdq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
	4726	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vabdq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 33));
4706	4727	return __ret;
4707	4728	}
4708	4729	#else
4709	4730	__ai __attribute__((target("neon"))) int16x8_t vabdq_s16(int16x8_t __p0, int16x8_t __p1) {
4710	4731	int16x8_t __ret;
4711		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
4712		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
4713		__ret = (int16x8_t) __builtin_neon_vabdq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
4714		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	4732	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	4733	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	4734	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vabdq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 33));
	4735	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
4715	4736	return __ret;
4716	4737	}
4717	4738	__ai __attribute__((target("neon"))) int16x8_t __noswap_vabdq_s16(int16x8_t __p0, int16x8_t __p1) {
4718	4739	int16x8_t __ret;
4719		__ret = (int16x8_t) __builtin_neon_vabdq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
	4740	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vabdq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 33));
4720	4741	return __ret;
4721	4742	}
4722	4743	#endif
...	...	@@ -4724,21 +4745,21 @@ __ai __attribute__((target("neon"))) int16x8_t __noswap_vabdq_s16(int16x8_t __p0
4724	4745	#ifdef __LITTLE_ENDIAN__
4725	4746	__ai __attribute__((target("neon"))) uint8x8_t vabd_u8(uint8x8_t __p0, uint8x8_t __p1) {
4726	4747	uint8x8_t __ret;
4727		__ret = (uint8x8_t) __builtin_neon_vabd_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
	4748	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vabd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 16));
4728	4749	return __ret;
4729	4750	}
4730	4751	#else
4731	4752	__ai __attribute__((target("neon"))) uint8x8_t vabd_u8(uint8x8_t __p0, uint8x8_t __p1) {
4732	4753	uint8x8_t __ret;
4733		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
4734		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
4735		__ret = (uint8x8_t) __builtin_neon_vabd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
4736		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	4754	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	4755	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	4756	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vabd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 16));
	4757	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
4737	4758	return __ret;
4738	4759	}
4739	4760	__ai __attribute__((target("neon"))) uint8x8_t __noswap_vabd_u8(uint8x8_t __p0, uint8x8_t __p1) {
4740	4761	uint8x8_t __ret;
4741		__ret = (uint8x8_t) __builtin_neon_vabd_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
	4762	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vabd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 16));
4742	4763	return __ret;
4743	4764	}
4744	4765	#endif
...	...	@@ -4746,21 +4767,21 @@ __ai __attribute__((target("neon"))) uint8x8_t __noswap_vabd_u8(uint8x8_t __p0,
4746	4767	#ifdef __LITTLE_ENDIAN__
4747	4768	__ai __attribute__((target("neon"))) uint32x2_t vabd_u32(uint32x2_t __p0, uint32x2_t __p1) {
4748	4769	uint32x2_t __ret;
4749		__ret = (uint32x2_t) __builtin_neon_vabd_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
	4770	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vabd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 18));
4750	4771	return __ret;
4751	4772	}
4752	4773	#else
4753	4774	__ai __attribute__((target("neon"))) uint32x2_t vabd_u32(uint32x2_t __p0, uint32x2_t __p1) {
4754	4775	uint32x2_t __ret;
4755		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
4756		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
4757		__ret = (uint32x2_t) __builtin_neon_vabd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
4758		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	4776	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	4777	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	4778	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vabd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 18));
	4779	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
4759	4780	return __ret;
4760	4781	}
4761	4782	__ai __attribute__((target("neon"))) uint32x2_t __noswap_vabd_u32(uint32x2_t __p0, uint32x2_t __p1) {
4762	4783	uint32x2_t __ret;
4763		__ret = (uint32x2_t) __builtin_neon_vabd_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
	4784	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vabd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 18));
4764	4785	return __ret;
4765	4786	}
4766	4787	#endif
...	...	@@ -4768,21 +4789,21 @@ __ai __attribute__((target("neon"))) uint32x2_t __noswap_vabd_u32(uint32x2_t __p
4768	4789	#ifdef __LITTLE_ENDIAN__
4769	4790	__ai __attribute__((target("neon"))) uint16x4_t vabd_u16(uint16x4_t __p0, uint16x4_t __p1) {
4770	4791	uint16x4_t __ret;
4771		__ret = (uint16x4_t) __builtin_neon_vabd_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
	4792	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vabd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 17));
4772	4793	return __ret;
4773	4794	}
4774	4795	#else
4775	4796	__ai __attribute__((target("neon"))) uint16x4_t vabd_u16(uint16x4_t __p0, uint16x4_t __p1) {
4776	4797	uint16x4_t __ret;
4777		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
4778		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
4779		__ret = (uint16x4_t) __builtin_neon_vabd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
4780		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	4798	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	4799	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	4800	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vabd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 17));
	4801	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
4781	4802	return __ret;
4782	4803	}
4783	4804	__ai __attribute__((target("neon"))) uint16x4_t __noswap_vabd_u16(uint16x4_t __p0, uint16x4_t __p1) {
4784	4805	uint16x4_t __ret;
4785		__ret = (uint16x4_t) __builtin_neon_vabd_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
	4806	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vabd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 17));
4786	4807	return __ret;
4787	4808	}
4788	4809	#endif
...	...	@@ -4790,21 +4811,21 @@ __ai __attribute__((target("neon"))) uint16x4_t __noswap_vabd_u16(uint16x4_t __p
4790	4811	#ifdef __LITTLE_ENDIAN__
4791	4812	__ai __attribute__((target("neon"))) int8x8_t vabd_s8(int8x8_t __p0, int8x8_t __p1) {
4792	4813	int8x8_t __ret;
4793		__ret = (int8x8_t) __builtin_neon_vabd_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
	4814	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vabd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 0));
4794	4815	return __ret;
4795	4816	}
4796	4817	#else
4797	4818	__ai __attribute__((target("neon"))) int8x8_t vabd_s8(int8x8_t __p0, int8x8_t __p1) {
4798	4819	int8x8_t __ret;
4799		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
4800		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
4801		__ret = (int8x8_t) __builtin_neon_vabd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
4802		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	4820	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	4821	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	4822	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vabd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 0));
	4823	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
4803	4824	return __ret;
4804	4825	}
4805	4826	__ai __attribute__((target("neon"))) int8x8_t __noswap_vabd_s8(int8x8_t __p0, int8x8_t __p1) {
4806	4827	int8x8_t __ret;
4807		__ret = (int8x8_t) __builtin_neon_vabd_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
	4828	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vabd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 0));
4808	4829	return __ret;
4809	4830	}
4810	4831	#endif
...	...	@@ -4812,16 +4833,16 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vabd_s8(int8x8_t __p0, in
4812	4833	#ifdef __LITTLE_ENDIAN__
4813	4834	__ai __attribute__((target("neon"))) float32x2_t vabd_f32(float32x2_t __p0, float32x2_t __p1) {
4814	4835	float32x2_t __ret;
4815		__ret = (float32x2_t) __builtin_neon_vabd_v((int8x8_t)__p0, (int8x8_t)__p1, 9);
	4836	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vabd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 9));
4816	4837	return __ret;
4817	4838	}
4818	4839	#else
4819	4840	__ai __attribute__((target("neon"))) float32x2_t vabd_f32(float32x2_t __p0, float32x2_t __p1) {
4820	4841	float32x2_t __ret;
4821		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
4822		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
4823		__ret = (float32x2_t) __builtin_neon_vabd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 9);
4824		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	4842	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	4843	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	4844	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vabd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 9));
	4845	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
4825	4846	return __ret;
4826	4847	}
4827	4848	#endif
...	...	@@ -4829,21 +4850,21 @@ __ai __attribute__((target("neon"))) float32x2_t vabd_f32(float32x2_t __p0, floa
4829	4850	#ifdef __LITTLE_ENDIAN__
4830	4851	__ai __attribute__((target("neon"))) int32x2_t vabd_s32(int32x2_t __p0, int32x2_t __p1) {
4831	4852	int32x2_t __ret;
4832		__ret = (int32x2_t) __builtin_neon_vabd_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
	4853	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vabd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 2));
4833	4854	return __ret;
4834	4855	}
4835	4856	#else
4836	4857	__ai __attribute__((target("neon"))) int32x2_t vabd_s32(int32x2_t __p0, int32x2_t __p1) {
4837	4858	int32x2_t __ret;
4838		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
4839		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
4840		__ret = (int32x2_t) __builtin_neon_vabd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
4841		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	4859	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	4860	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	4861	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vabd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 2));
	4862	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
4842	4863	return __ret;
4843	4864	}
4844	4865	__ai __attribute__((target("neon"))) int32x2_t __noswap_vabd_s32(int32x2_t __p0, int32x2_t __p1) {
4845	4866	int32x2_t __ret;
4846		__ret = (int32x2_t) __builtin_neon_vabd_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
	4867	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vabd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 2));
4847	4868	return __ret;
4848	4869	}
4849	4870	#endif
...	...	@@ -4851,21 +4872,21 @@ __ai __attribute__((target("neon"))) int32x2_t __noswap_vabd_s32(int32x2_t __p0,
4851	4872	#ifdef __LITTLE_ENDIAN__
4852	4873	__ai __attribute__((target("neon"))) int16x4_t vabd_s16(int16x4_t __p0, int16x4_t __p1) {
4853	4874	int16x4_t __ret;
4854		__ret = (int16x4_t) __builtin_neon_vabd_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
	4875	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vabd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 1));
4855	4876	return __ret;
4856	4877	}
4857	4878	#else
4858	4879	__ai __attribute__((target("neon"))) int16x4_t vabd_s16(int16x4_t __p0, int16x4_t __p1) {
4859	4880	int16x4_t __ret;
4860		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
4861		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
4862		__ret = (int16x4_t) __builtin_neon_vabd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
4863		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	4881	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	4882	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	4883	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vabd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 1));
	4884	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
4864	4885	return __ret;
4865	4886	}
4866	4887	__ai __attribute__((target("neon"))) int16x4_t __noswap_vabd_s16(int16x4_t __p0, int16x4_t __p1) {
4867	4888	int16x4_t __ret;
4868		__ret = (int16x4_t) __builtin_neon_vabd_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
	4889	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vabd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 1));
4869	4890	return __ret;
4870	4891	}
4871	4892	#endif
...	...	@@ -4873,15 +4894,15 @@ __ai __attribute__((target("neon"))) int16x4_t __noswap_vabd_s16(int16x4_t __p0,
4873	4894	#ifdef __LITTLE_ENDIAN__
4874	4895	__ai __attribute__((target("neon"))) int8x16_t vabsq_s8(int8x16_t __p0) {
4875	4896	int8x16_t __ret;
4876		__ret = (int8x16_t) __builtin_neon_vabsq_v((int8x16_t)__p0, 32);
	4897	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vabsq_v(__builtin_bit_cast(int8x16_t, __p0), 32));
4877	4898	return __ret;
4878	4899	}
4879	4900	#else
4880	4901	__ai __attribute__((target("neon"))) int8x16_t vabsq_s8(int8x16_t __p0) {
4881	4902	int8x16_t __ret;
4882		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4883		__ret = (int8x16_t) __builtin_neon_vabsq_v((int8x16_t)__rev0, 32);
4884		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	4903	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	4904	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vabsq_v(__builtin_bit_cast(int8x16_t, __rev0), 32));
	4905	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
4885	4906	return __ret;
4886	4907	}
4887	4908	#endif
...	...	@@ -4889,15 +4910,15 @@ __ai __attribute__((target("neon"))) int8x16_t vabsq_s8(int8x16_t __p0) {
4889	4910	#ifdef __LITTLE_ENDIAN__
4890	4911	__ai __attribute__((target("neon"))) float32x4_t vabsq_f32(float32x4_t __p0) {
4891	4912	float32x4_t __ret;
4892		__ret = (float32x4_t) __builtin_neon_vabsq_v((int8x16_t)__p0, 41);
	4913	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vabsq_v(__builtin_bit_cast(int8x16_t, __p0), 41));
4893	4914	return __ret;
4894	4915	}
4895	4916	#else
4896	4917	__ai __attribute__((target("neon"))) float32x4_t vabsq_f32(float32x4_t __p0) {
4897	4918	float32x4_t __ret;
4898		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
4899		__ret = (float32x4_t) __builtin_neon_vabsq_v((int8x16_t)__rev0, 41);
4900		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	4919	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	4920	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vabsq_v(__builtin_bit_cast(int8x16_t, __rev0), 41));
	4921	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
4901	4922	return __ret;
4902	4923	}
4903	4924	#endif
...	...	@@ -4905,15 +4926,15 @@ __ai __attribute__((target("neon"))) float32x4_t vabsq_f32(float32x4_t __p0) {
4905	4926	#ifdef __LITTLE_ENDIAN__
4906	4927	__ai __attribute__((target("neon"))) int32x4_t vabsq_s32(int32x4_t __p0) {
4907	4928	int32x4_t __ret;
4908		__ret = (int32x4_t) __builtin_neon_vabsq_v((int8x16_t)__p0, 34);
	4929	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vabsq_v(__builtin_bit_cast(int8x16_t, __p0), 34));
4909	4930	return __ret;
4910	4931	}
4911	4932	#else
4912	4933	__ai __attribute__((target("neon"))) int32x4_t vabsq_s32(int32x4_t __p0) {
4913	4934	int32x4_t __ret;
4914		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
4915		__ret = (int32x4_t) __builtin_neon_vabsq_v((int8x16_t)__rev0, 34);
4916		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	4935	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	4936	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vabsq_v(__builtin_bit_cast(int8x16_t, __rev0), 34));
	4937	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
4917	4938	return __ret;
4918	4939	}
4919	4940	#endif
...	...	@@ -4921,15 +4942,15 @@ __ai __attribute__((target("neon"))) int32x4_t vabsq_s32(int32x4_t __p0) {
4921	4942	#ifdef __LITTLE_ENDIAN__
4922	4943	__ai __attribute__((target("neon"))) int16x8_t vabsq_s16(int16x8_t __p0) {
4923	4944	int16x8_t __ret;
4924		__ret = (int16x8_t) __builtin_neon_vabsq_v((int8x16_t)__p0, 33);
	4945	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vabsq_v(__builtin_bit_cast(int8x16_t, __p0), 33));
4925	4946	return __ret;
4926	4947	}
4927	4948	#else
4928	4949	__ai __attribute__((target("neon"))) int16x8_t vabsq_s16(int16x8_t __p0) {
4929	4950	int16x8_t __ret;
4930		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
4931		__ret = (int16x8_t) __builtin_neon_vabsq_v((int8x16_t)__rev0, 33);
4932		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	4951	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	4952	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vabsq_v(__builtin_bit_cast(int8x16_t, __rev0), 33));
	4953	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
4933	4954	return __ret;
4934	4955	}
4935	4956	#endif
...	...	@@ -4937,15 +4958,15 @@ __ai __attribute__((target("neon"))) int16x8_t vabsq_s16(int16x8_t __p0) {
4937	4958	#ifdef __LITTLE_ENDIAN__
4938	4959	__ai __attribute__((target("neon"))) int8x8_t vabs_s8(int8x8_t __p0) {
4939	4960	int8x8_t __ret;
4940		__ret = (int8x8_t) __builtin_neon_vabs_v((int8x8_t)__p0, 0);
	4961	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vabs_v(__builtin_bit_cast(int8x8_t, __p0), 0));
4941	4962	return __ret;
4942	4963	}
4943	4964	#else
4944	4965	__ai __attribute__((target("neon"))) int8x8_t vabs_s8(int8x8_t __p0) {
4945	4966	int8x8_t __ret;
4946		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
4947		__ret = (int8x8_t) __builtin_neon_vabs_v((int8x8_t)__rev0, 0);
4948		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	4967	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	4968	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vabs_v(__builtin_bit_cast(int8x8_t, __rev0), 0));
	4969	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
4949	4970	return __ret;
4950	4971	}
4951	4972	#endif
...	...	@@ -4953,15 +4974,15 @@ __ai __attribute__((target("neon"))) int8x8_t vabs_s8(int8x8_t __p0) {
4953	4974	#ifdef __LITTLE_ENDIAN__
4954	4975	__ai __attribute__((target("neon"))) float32x2_t vabs_f32(float32x2_t __p0) {
4955	4976	float32x2_t __ret;
4956		__ret = (float32x2_t) __builtin_neon_vabs_v((int8x8_t)__p0, 9);
	4977	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vabs_v(__builtin_bit_cast(int8x8_t, __p0), 9));
4957	4978	return __ret;
4958	4979	}
4959	4980	#else
4960	4981	__ai __attribute__((target("neon"))) float32x2_t vabs_f32(float32x2_t __p0) {
4961	4982	float32x2_t __ret;
4962		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
4963		__ret = (float32x2_t) __builtin_neon_vabs_v((int8x8_t)__rev0, 9);
4964		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	4983	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	4984	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vabs_v(__builtin_bit_cast(int8x8_t, __rev0), 9));
	4985	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
4965	4986	return __ret;
4966	4987	}
4967	4988	#endif
...	...	@@ -4969,15 +4990,15 @@ __ai __attribute__((target("neon"))) float32x2_t vabs_f32(float32x2_t __p0) {
4969	4990	#ifdef __LITTLE_ENDIAN__
4970	4991	__ai __attribute__((target("neon"))) int32x2_t vabs_s32(int32x2_t __p0) {
4971	4992	int32x2_t __ret;
4972		__ret = (int32x2_t) __builtin_neon_vabs_v((int8x8_t)__p0, 2);
	4993	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vabs_v(__builtin_bit_cast(int8x8_t, __p0), 2));
4973	4994	return __ret;
4974	4995	}
4975	4996	#else
4976	4997	__ai __attribute__((target("neon"))) int32x2_t vabs_s32(int32x2_t __p0) {
4977	4998	int32x2_t __ret;
4978		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
4979		__ret = (int32x2_t) __builtin_neon_vabs_v((int8x8_t)__rev0, 2);
4980		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	4999	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	5000	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vabs_v(__builtin_bit_cast(int8x8_t, __rev0), 2));
	5001	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
4981	5002	return __ret;
4982	5003	}
4983	5004	#endif
...	...	@@ -4985,15 +5006,15 @@ __ai __attribute__((target("neon"))) int32x2_t vabs_s32(int32x2_t __p0) {
4985	5006	#ifdef __LITTLE_ENDIAN__
4986	5007	__ai __attribute__((target("neon"))) int16x4_t vabs_s16(int16x4_t __p0) {
4987	5008	int16x4_t __ret;
4988		__ret = (int16x4_t) __builtin_neon_vabs_v((int8x8_t)__p0, 1);
	5009	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vabs_v(__builtin_bit_cast(int8x8_t, __p0), 1));
4989	5010	return __ret;
4990	5011	}
4991	5012	#else
4992	5013	__ai __attribute__((target("neon"))) int16x4_t vabs_s16(int16x4_t __p0) {
4993	5014	int16x4_t __ret;
4994		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
4995		__ret = (int16x4_t) __builtin_neon_vabs_v((int8x8_t)__rev0, 1);
4996		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	5015	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	5016	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vabs_v(__builtin_bit_cast(int8x8_t, __rev0), 1));
	5017	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
4997	5018	return __ret;
4998	5019	}
4999	5020	#endif
...	...	@@ -5007,10 +5028,10 @@ __ai __attribute__((target("neon"))) uint8x16_t vaddq_u8(uint8x16_t __p0, uint8x
5007	5028	#else
5008	5029	__ai __attribute__((target("neon"))) uint8x16_t vaddq_u8(uint8x16_t __p0, uint8x16_t __p1) {
5009	5030	uint8x16_t __ret;
5010		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
5011		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	5031	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	5032	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
5012	5033	__ret = __rev0 + __rev1;
5013		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	5034	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
5014	5035	return __ret;
5015	5036	}
5016	5037	#endif
...	...	@@ -5024,10 +5045,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vaddq_u32(uint32x4_t __p0, uint3
5024	5045	#else
5025	5046	__ai __attribute__((target("neon"))) uint32x4_t vaddq_u32(uint32x4_t __p0, uint32x4_t __p1) {
5026	5047	uint32x4_t __ret;
5027		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
5028		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	5048	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	5049	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
5029	5050	__ret = __rev0 + __rev1;
5030		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	5051	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
5031	5052	return __ret;
5032	5053	}
5033	5054	#endif
...	...	@@ -5041,10 +5062,10 @@ __ai __attribute__((target("neon"))) uint64x2_t vaddq_u64(uint64x2_t __p0, uint6
5041	5062	#else
5042	5063	__ai __attribute__((target("neon"))) uint64x2_t vaddq_u64(uint64x2_t __p0, uint64x2_t __p1) {
5043	5064	uint64x2_t __ret;
5044		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
5045		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	5065	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	5066	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
5046	5067	__ret = __rev0 + __rev1;
5047		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	5068	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
5048	5069	return __ret;
5049	5070	}
5050	5071	#endif
...	...	@@ -5058,10 +5079,10 @@ __ai __attribute__((target("neon"))) uint16x8_t vaddq_u16(uint16x8_t __p0, uint1
5058	5079	#else
5059	5080	__ai __attribute__((target("neon"))) uint16x8_t vaddq_u16(uint16x8_t __p0, uint16x8_t __p1) {
5060	5081	uint16x8_t __ret;
5061		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
5062		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	5082	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	5083	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
5063	5084	__ret = __rev0 + __rev1;
5064		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	5085	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
5065	5086	return __ret;
5066	5087	}
5067	5088	#endif
...	...	@@ -5075,10 +5096,10 @@ __ai __attribute__((target("neon"))) int8x16_t vaddq_s8(int8x16_t __p0, int8x16_
5075	5096	#else
5076	5097	__ai __attribute__((target("neon"))) int8x16_t vaddq_s8(int8x16_t __p0, int8x16_t __p1) {
5077	5098	int8x16_t __ret;
5078		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
5079		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	5099	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	5100	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
5080	5101	__ret = __rev0 + __rev1;
5081		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	5102	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
5082	5103	return __ret;
5083	5104	}
5084	5105	#endif
...	...	@@ -5092,10 +5113,10 @@ __ai __attribute__((target("neon"))) float32x4_t vaddq_f32(float32x4_t __p0, flo
5092	5113	#else
5093	5114	__ai __attribute__((target("neon"))) float32x4_t vaddq_f32(float32x4_t __p0, float32x4_t __p1) {
5094	5115	float32x4_t __ret;
5095		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
5096		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	5116	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	5117	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
5097	5118	__ret = __rev0 + __rev1;
5098		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	5119	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
5099	5120	return __ret;
5100	5121	}
5101	5122	#endif
...	...	@@ -5109,10 +5130,10 @@ __ai __attribute__((target("neon"))) int32x4_t vaddq_s32(int32x4_t __p0, int32x4
5109	5130	#else
5110	5131	__ai __attribute__((target("neon"))) int32x4_t vaddq_s32(int32x4_t __p0, int32x4_t __p1) {
5111	5132	int32x4_t __ret;
5112		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
5113		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	5133	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	5134	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
5114	5135	__ret = __rev0 + __rev1;
5115		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	5136	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
5116	5137	return __ret;
5117	5138	}
5118	5139	#endif
...	...	@@ -5126,10 +5147,10 @@ __ai __attribute__((target("neon"))) int64x2_t vaddq_s64(int64x2_t __p0, int64x2
5126	5147	#else
5127	5148	__ai __attribute__((target("neon"))) int64x2_t vaddq_s64(int64x2_t __p0, int64x2_t __p1) {
5128	5149	int64x2_t __ret;
5129		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
5130		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	5150	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	5151	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
5131	5152	__ret = __rev0 + __rev1;
5132		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	5153	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
5133	5154	return __ret;
5134	5155	}
5135	5156	#endif
...	...	@@ -5143,10 +5164,10 @@ __ai __attribute__((target("neon"))) int16x8_t vaddq_s16(int16x8_t __p0, int16x8
5143	5164	#else
5144	5165	__ai __attribute__((target("neon"))) int16x8_t vaddq_s16(int16x8_t __p0, int16x8_t __p1) {
5145	5166	int16x8_t __ret;
5146		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
5147		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	5167	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	5168	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
5148	5169	__ret = __rev0 + __rev1;
5149		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	5170	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
5150	5171	return __ret;
5151	5172	}
5152	5173	#endif
...	...	@@ -5160,10 +5181,10 @@ __ai __attribute__((target("neon"))) uint8x8_t vadd_u8(uint8x8_t __p0, uint8x8_t
5160	5181	#else
5161	5182	__ai __attribute__((target("neon"))) uint8x8_t vadd_u8(uint8x8_t __p0, uint8x8_t __p1) {
5162	5183	uint8x8_t __ret;
5163		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
5164		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	5184	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	5185	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
5165	5186	__ret = __rev0 + __rev1;
5166		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	5187	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
5167	5188	return __ret;
5168	5189	}
5169	5190	#endif
...	...	@@ -5177,10 +5198,10 @@ __ai __attribute__((target("neon"))) uint32x2_t vadd_u32(uint32x2_t __p0, uint32
5177	5198	#else
5178	5199	__ai __attribute__((target("neon"))) uint32x2_t vadd_u32(uint32x2_t __p0, uint32x2_t __p1) {
5179	5200	uint32x2_t __ret;
5180		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
5181		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	5201	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	5202	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
5182	5203	__ret = __rev0 + __rev1;
5183		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	5204	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
5184	5205	return __ret;
5185	5206	}
5186	5207	#endif
...	...	@@ -5199,10 +5220,10 @@ __ai __attribute__((target("neon"))) uint16x4_t vadd_u16(uint16x4_t __p0, uint16
5199	5220	#else
5200	5221	__ai __attribute__((target("neon"))) uint16x4_t vadd_u16(uint16x4_t __p0, uint16x4_t __p1) {
5201	5222	uint16x4_t __ret;
5202		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
5203		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	5223	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	5224	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
5204	5225	__ret = __rev0 + __rev1;
5205		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	5226	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
5206	5227	return __ret;
5207	5228	}
5208	5229	#endif
...	...	@@ -5216,10 +5237,10 @@ __ai __attribute__((target("neon"))) int8x8_t vadd_s8(int8x8_t __p0, int8x8_t __
5216	5237	#else
5217	5238	__ai __attribute__((target("neon"))) int8x8_t vadd_s8(int8x8_t __p0, int8x8_t __p1) {
5218	5239	int8x8_t __ret;
5219		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
5220		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	5240	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	5241	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
5221	5242	__ret = __rev0 + __rev1;
5222		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	5243	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
5223	5244	return __ret;
5224	5245	}
5225	5246	#endif
...	...	@@ -5233,10 +5254,10 @@ __ai __attribute__((target("neon"))) float32x2_t vadd_f32(float32x2_t __p0, floa
5233	5254	#else
5234	5255	__ai __attribute__((target("neon"))) float32x2_t vadd_f32(float32x2_t __p0, float32x2_t __p1) {
5235	5256	float32x2_t __ret;
5236		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
5237		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	5257	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	5258	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
5238	5259	__ret = __rev0 + __rev1;
5239		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	5260	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
5240	5261	return __ret;
5241	5262	}
5242	5263	#endif
...	...	@@ -5250,10 +5271,10 @@ __ai __attribute__((target("neon"))) int32x2_t vadd_s32(int32x2_t __p0, int32x2_
5250	5271	#else
5251	5272	__ai __attribute__((target("neon"))) int32x2_t vadd_s32(int32x2_t __p0, int32x2_t __p1) {
5252	5273	int32x2_t __ret;
5253		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
5254		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	5274	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	5275	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
5255	5276	__ret = __rev0 + __rev1;
5256		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	5277	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
5257	5278	return __ret;
5258	5279	}
5259	5280	#endif
...	...	@@ -5272,10 +5293,10 @@ __ai __attribute__((target("neon"))) int16x4_t vadd_s16(int16x4_t __p0, int16x4_
5272	5293	#else
5273	5294	__ai __attribute__((target("neon"))) int16x4_t vadd_s16(int16x4_t __p0, int16x4_t __p1) {
5274	5295	int16x4_t __ret;
5275		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
5276		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	5296	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	5297	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
5277	5298	__ret = __rev0 + __rev1;
5278		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	5299	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
5279	5300	return __ret;
5280	5301	}
5281	5302	#endif
...	...	@@ -5283,38 +5304,38 @@ __ai __attribute__((target("neon"))) int16x4_t vadd_s16(int16x4_t __p0, int16x4_
5283	5304	#ifdef __LITTLE_ENDIAN__
5284	5305	__ai __attribute__((target("neon"))) poly8x8_t vadd_p8(poly8x8_t __p0, poly8x8_t __p1) {
5285	5306	poly8x8_t __ret;
5286		__ret = (poly8x8_t) __builtin_neon_vadd_v((int8x8_t)__p0, (int8x8_t)__p1, 4);
	5307	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 4));
5287	5308	return __ret;
5288	5309	}
5289	5310	#else
5290	5311	__ai __attribute__((target("neon"))) poly8x8_t vadd_p8(poly8x8_t __p0, poly8x8_t __p1) {
5291	5312	poly8x8_t __ret;
5292		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
5293		poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
5294		__ret = (poly8x8_t) __builtin_neon_vadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 4);
5295		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	5313	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	5314	poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	5315	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 4));
	5316	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
5296	5317	return __ret;
5297	5318	}
5298	5319	#endif
5299	5320
5300	5321	__ai __attribute__((target("neon"))) poly64x1_t vadd_p64(poly64x1_t __p0, poly64x1_t __p1) {
5301	5322	poly64x1_t __ret;
5302		__ret = (poly64x1_t) __builtin_neon_vadd_v((int8x8_t)__p0, (int8x8_t)__p1, 6);
	5323	__ret = __builtin_bit_cast(poly64x1_t, __builtin_neon_vadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 6));
5303	5324	return __ret;
5304	5325	}
5305	5326	#ifdef __LITTLE_ENDIAN__
5306	5327	__ai __attribute__((target("neon"))) poly16x4_t vadd_p16(poly16x4_t __p0, poly16x4_t __p1) {
5307	5328	poly16x4_t __ret;
5308		__ret = (poly16x4_t) __builtin_neon_vadd_v((int8x8_t)__p0, (int8x8_t)__p1, 5);
	5329	__ret = __builtin_bit_cast(poly16x4_t, __builtin_neon_vadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 5));
5309	5330	return __ret;
5310	5331	}
5311	5332	#else
5312	5333	__ai __attribute__((target("neon"))) poly16x4_t vadd_p16(poly16x4_t __p0, poly16x4_t __p1) {
5313	5334	poly16x4_t __ret;
5314		poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
5315		poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
5316		__ret = (poly16x4_t) __builtin_neon_vadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 5);
5317		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	5335	poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	5336	poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	5337	__ret = __builtin_bit_cast(poly16x4_t, __builtin_neon_vadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 5));
	5338	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
5318	5339	return __ret;
5319	5340	}
5320	5341	#endif
...	...	@@ -5322,16 +5343,16 @@ __ai __attribute__((target("neon"))) poly16x4_t vadd_p16(poly16x4_t __p0, poly16
5322	5343	#ifdef __LITTLE_ENDIAN__
5323	5344	__ai __attribute__((target("neon"))) poly8x16_t vaddq_p8(poly8x16_t __p0, poly8x16_t __p1) {
5324	5345	poly8x16_t __ret;
5325		__ret = (poly8x16_t) __builtin_neon_vaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 36);
	5346	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 36));
5326	5347	return __ret;
5327	5348	}
5328	5349	#else
5329	5350	__ai __attribute__((target("neon"))) poly8x16_t vaddq_p8(poly8x16_t __p0, poly8x16_t __p1) {
5330	5351	poly8x16_t __ret;
5331		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
5332		poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
5333		__ret = (poly8x16_t) __builtin_neon_vaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 36);
5334		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	5352	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	5353	poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	5354	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 36));
	5355	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
5335	5356	return __ret;
5336	5357	}
5337	5358	#endif
...	...	@@ -5339,16 +5360,16 @@ __ai __attribute__((target("neon"))) poly8x16_t vaddq_p8(poly8x16_t __p0, poly8x
5339	5360	#ifdef __LITTLE_ENDIAN__
5340	5361	__ai __attribute__((target("neon"))) poly64x2_t vaddq_p64(poly64x2_t __p0, poly64x2_t __p1) {
5341	5362	poly64x2_t __ret;
5342		__ret = (poly64x2_t) __builtin_neon_vaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 38);
	5363	__ret = __builtin_bit_cast(poly64x2_t, __builtin_neon_vaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 38));
5343	5364	return __ret;
5344	5365	}
5345	5366	#else
5346	5367	__ai __attribute__((target("neon"))) poly64x2_t vaddq_p64(poly64x2_t __p0, poly64x2_t __p1) {
5347	5368	poly64x2_t __ret;
5348		poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
5349		poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
5350		__ret = (poly64x2_t) __builtin_neon_vaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 38);
5351		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	5369	poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	5370	poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	5371	__ret = __builtin_bit_cast(poly64x2_t, __builtin_neon_vaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 38));
	5372	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
5352	5373	return __ret;
5353	5374	}
5354	5375	#endif
...	...	@@ -5356,16 +5377,16 @@ __ai __attribute__((target("neon"))) poly64x2_t vaddq_p64(poly64x2_t __p0, poly6
5356	5377	#ifdef __LITTLE_ENDIAN__
5357	5378	__ai __attribute__((target("neon"))) poly16x8_t vaddq_p16(poly16x8_t __p0, poly16x8_t __p1) {
5358	5379	poly16x8_t __ret;
5359		__ret = (poly16x8_t) __builtin_neon_vaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 37);
	5380	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 37));
5360	5381	return __ret;
5361	5382	}
5362	5383	#else
5363	5384	__ai __attribute__((target("neon"))) poly16x8_t vaddq_p16(poly16x8_t __p0, poly16x8_t __p1) {
5364	5385	poly16x8_t __ret;
5365		poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
5366		poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
5367		__ret = (poly16x8_t) __builtin_neon_vaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 37);
5368		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	5386	poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	5387	poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	5388	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 37));
	5389	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
5369	5390	return __ret;
5370	5391	}
5371	5392	#endif
...	...	@@ -5373,21 +5394,21 @@ __ai __attribute__((target("neon"))) poly16x8_t vaddq_p16(poly16x8_t __p0, poly1
5373	5394	#ifdef __LITTLE_ENDIAN__
5374	5395	__ai __attribute__((target("neon"))) uint16x4_t vaddhn_u32(uint32x4_t __p0, uint32x4_t __p1) {
5375	5396	uint16x4_t __ret;
5376		__ret = (uint16x4_t) __builtin_neon_vaddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 17);
	5397	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vaddhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 17));
5377	5398	return __ret;
5378	5399	}
5379	5400	#else
5380	5401	__ai __attribute__((target("neon"))) uint16x4_t vaddhn_u32(uint32x4_t __p0, uint32x4_t __p1) {
5381	5402	uint16x4_t __ret;
5382		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
5383		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
5384		__ret = (uint16x4_t) __builtin_neon_vaddhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 17);
5385		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	5403	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	5404	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	5405	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vaddhn_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 17));
	5406	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
5386	5407	return __ret;
5387	5408	}
5388	5409	__ai __attribute__((target("neon"))) uint16x4_t __noswap_vaddhn_u32(uint32x4_t __p0, uint32x4_t __p1) {
5389	5410	uint16x4_t __ret;
5390		__ret = (uint16x4_t) __builtin_neon_vaddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 17);
	5411	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vaddhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 17));
5391	5412	return __ret;
5392	5413	}
5393	5414	#endif
...	...	@@ -5395,21 +5416,21 @@ __ai __attribute__((target("neon"))) uint16x4_t __noswap_vaddhn_u32(uint32x4_t _
5395	5416	#ifdef __LITTLE_ENDIAN__
5396	5417	__ai __attribute__((target("neon"))) uint32x2_t vaddhn_u64(uint64x2_t __p0, uint64x2_t __p1) {
5397	5418	uint32x2_t __ret;
5398		__ret = (uint32x2_t) __builtin_neon_vaddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 18);
	5419	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vaddhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 18));
5399	5420	return __ret;
5400	5421	}
5401	5422	#else
5402	5423	__ai __attribute__((target("neon"))) uint32x2_t vaddhn_u64(uint64x2_t __p0, uint64x2_t __p1) {
5403	5424	uint32x2_t __ret;
5404		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
5405		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
5406		__ret = (uint32x2_t) __builtin_neon_vaddhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 18);
5407		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	5425	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	5426	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	5427	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vaddhn_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 18));
	5428	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
5408	5429	return __ret;
5409	5430	}
5410	5431	__ai __attribute__((target("neon"))) uint32x2_t __noswap_vaddhn_u64(uint64x2_t __p0, uint64x2_t __p1) {
5411	5432	uint32x2_t __ret;
5412		__ret = (uint32x2_t) __builtin_neon_vaddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 18);
	5433	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vaddhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 18));
5413	5434	return __ret;
5414	5435	}
5415	5436	#endif
...	...	@@ -5417,21 +5438,21 @@ __ai __attribute__((target("neon"))) uint32x2_t __noswap_vaddhn_u64(uint64x2_t _
5417	5438	#ifdef __LITTLE_ENDIAN__
5418	5439	__ai __attribute__((target("neon"))) uint8x8_t vaddhn_u16(uint16x8_t __p0, uint16x8_t __p1) {
5419	5440	uint8x8_t __ret;
5420		__ret = (uint8x8_t) __builtin_neon_vaddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 16);
	5441	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vaddhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 16));
5421	5442	return __ret;
5422	5443	}
5423	5444	#else
5424	5445	__ai __attribute__((target("neon"))) uint8x8_t vaddhn_u16(uint16x8_t __p0, uint16x8_t __p1) {
5425	5446	uint8x8_t __ret;
5426		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
5427		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
5428		__ret = (uint8x8_t) __builtin_neon_vaddhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 16);
5429		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	5447	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	5448	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	5449	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vaddhn_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 16));
	5450	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
5430	5451	return __ret;
5431	5452	}
5432	5453	__ai __attribute__((target("neon"))) uint8x8_t __noswap_vaddhn_u16(uint16x8_t __p0, uint16x8_t __p1) {
5433	5454	uint8x8_t __ret;
5434		__ret = (uint8x8_t) __builtin_neon_vaddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 16);
	5455	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vaddhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 16));
5435	5456	return __ret;
5436	5457	}
5437	5458	#endif
...	...	@@ -5439,21 +5460,21 @@ __ai __attribute__((target("neon"))) uint8x8_t __noswap_vaddhn_u16(uint16x8_t __
5439	5460	#ifdef __LITTLE_ENDIAN__
5440	5461	__ai __attribute__((target("neon"))) int16x4_t vaddhn_s32(int32x4_t __p0, int32x4_t __p1) {
5441	5462	int16x4_t __ret;
5442		__ret = (int16x4_t) __builtin_neon_vaddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 1);
	5463	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vaddhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 1));
5443	5464	return __ret;
5444	5465	}
5445	5466	#else
5446	5467	__ai __attribute__((target("neon"))) int16x4_t vaddhn_s32(int32x4_t __p0, int32x4_t __p1) {
5447	5468	int16x4_t __ret;
5448		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
5449		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
5450		__ret = (int16x4_t) __builtin_neon_vaddhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 1);
5451		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	5469	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	5470	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	5471	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vaddhn_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 1));
	5472	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
5452	5473	return __ret;
5453	5474	}
5454	5475	__ai __attribute__((target("neon"))) int16x4_t __noswap_vaddhn_s32(int32x4_t __p0, int32x4_t __p1) {
5455	5476	int16x4_t __ret;
5456		__ret = (int16x4_t) __builtin_neon_vaddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 1);
	5477	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vaddhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 1));
5457	5478	return __ret;
5458	5479	}
5459	5480	#endif
...	...	@@ -5461,21 +5482,21 @@ __ai __attribute__((target("neon"))) int16x4_t __noswap_vaddhn_s32(int32x4_t __p
5461	5482	#ifdef __LITTLE_ENDIAN__
5462	5483	__ai __attribute__((target("neon"))) int32x2_t vaddhn_s64(int64x2_t __p0, int64x2_t __p1) {
5463	5484	int32x2_t __ret;
5464		__ret = (int32x2_t) __builtin_neon_vaddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 2);
	5485	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vaddhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 2));
5465	5486	return __ret;
5466	5487	}
5467	5488	#else
5468	5489	__ai __attribute__((target("neon"))) int32x2_t vaddhn_s64(int64x2_t __p0, int64x2_t __p1) {
5469	5490	int32x2_t __ret;
5470		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
5471		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
5472		__ret = (int32x2_t) __builtin_neon_vaddhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 2);
5473		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	5491	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	5492	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	5493	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vaddhn_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 2));
	5494	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
5474	5495	return __ret;
5475	5496	}
5476	5497	__ai __attribute__((target("neon"))) int32x2_t __noswap_vaddhn_s64(int64x2_t __p0, int64x2_t __p1) {
5477	5498	int32x2_t __ret;
5478		__ret = (int32x2_t) __builtin_neon_vaddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 2);
	5499	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vaddhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 2));
5479	5500	return __ret;
5480	5501	}
5481	5502	#endif
...	...	@@ -5483,21 +5504,21 @@ __ai __attribute__((target("neon"))) int32x2_t __noswap_vaddhn_s64(int64x2_t __p
5483	5504	#ifdef __LITTLE_ENDIAN__
5484	5505	__ai __attribute__((target("neon"))) int8x8_t vaddhn_s16(int16x8_t __p0, int16x8_t __p1) {
5485	5506	int8x8_t __ret;
5486		__ret = (int8x8_t) __builtin_neon_vaddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 0);
	5507	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vaddhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 0));
5487	5508	return __ret;
5488	5509	}
5489	5510	#else
5490	5511	__ai __attribute__((target("neon"))) int8x8_t vaddhn_s16(int16x8_t __p0, int16x8_t __p1) {
5491	5512	int8x8_t __ret;
5492		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
5493		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
5494		__ret = (int8x8_t) __builtin_neon_vaddhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 0);
5495		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	5513	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	5514	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	5515	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vaddhn_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 0));
	5516	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
5496	5517	return __ret;
5497	5518	}
5498	5519	__ai __attribute__((target("neon"))) int8x8_t __noswap_vaddhn_s16(int16x8_t __p0, int16x8_t __p1) {
5499	5520	int8x8_t __ret;
5500		__ret = (int8x8_t) __builtin_neon_vaddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 0);
	5521	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vaddhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 0));
5501	5522	return __ret;
5502	5523	}
5503	5524	#endif
...	...	@@ -5511,10 +5532,10 @@ __ai __attribute__((target("neon"))) uint8x16_t vandq_u8(uint8x16_t __p0, uint8x
5511	5532	#else
5512	5533	__ai __attribute__((target("neon"))) uint8x16_t vandq_u8(uint8x16_t __p0, uint8x16_t __p1) {
5513	5534	uint8x16_t __ret;
5514		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
5515		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	5535	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	5536	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
5516	5537	__ret = __rev0 & __rev1;
5517		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	5538	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
5518	5539	return __ret;
5519	5540	}
5520	5541	#endif
...	...	@@ -5528,10 +5549,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vandq_u32(uint32x4_t __p0, uint3
5528	5549	#else
5529	5550	__ai __attribute__((target("neon"))) uint32x4_t vandq_u32(uint32x4_t __p0, uint32x4_t __p1) {
5530	5551	uint32x4_t __ret;
5531		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
5532		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	5552	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	5553	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
5533	5554	__ret = __rev0 & __rev1;
5534		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	5555	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
5535	5556	return __ret;
5536	5557	}
5537	5558	#endif
...	...	@@ -5545,10 +5566,10 @@ __ai __attribute__((target("neon"))) uint64x2_t vandq_u64(uint64x2_t __p0, uint6
5545	5566	#else
5546	5567	__ai __attribute__((target("neon"))) uint64x2_t vandq_u64(uint64x2_t __p0, uint64x2_t __p1) {
5547	5568	uint64x2_t __ret;
5548		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
5549		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	5569	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	5570	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
5550	5571	__ret = __rev0 & __rev1;
5551		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	5572	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
5552	5573	return __ret;
5553	5574	}
5554	5575	#endif
...	...	@@ -5562,10 +5583,10 @@ __ai __attribute__((target("neon"))) uint16x8_t vandq_u16(uint16x8_t __p0, uint1
5562	5583	#else
5563	5584	__ai __attribute__((target("neon"))) uint16x8_t vandq_u16(uint16x8_t __p0, uint16x8_t __p1) {
5564	5585	uint16x8_t __ret;
5565		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
5566		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	5586	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	5587	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
5567	5588	__ret = __rev0 & __rev1;
5568		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	5589	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
5569	5590	return __ret;
5570	5591	}
5571	5592	#endif
...	...	@@ -5579,10 +5600,10 @@ __ai __attribute__((target("neon"))) int8x16_t vandq_s8(int8x16_t __p0, int8x16_
5579	5600	#else
5580	5601	__ai __attribute__((target("neon"))) int8x16_t vandq_s8(int8x16_t __p0, int8x16_t __p1) {
5581	5602	int8x16_t __ret;
5582		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
5583		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	5603	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	5604	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
5584	5605	__ret = __rev0 & __rev1;
5585		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	5606	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
5586	5607	return __ret;
5587	5608	}
5588	5609	#endif
...	...	@@ -5596,10 +5617,10 @@ __ai __attribute__((target("neon"))) int32x4_t vandq_s32(int32x4_t __p0, int32x4
5596	5617	#else
5597	5618	__ai __attribute__((target("neon"))) int32x4_t vandq_s32(int32x4_t __p0, int32x4_t __p1) {
5598	5619	int32x4_t __ret;
5599		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
5600		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	5620	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	5621	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
5601	5622	__ret = __rev0 & __rev1;
5602		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	5623	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
5603	5624	return __ret;
5604	5625	}
5605	5626	#endif
...	...	@@ -5613,10 +5634,10 @@ __ai __attribute__((target("neon"))) int64x2_t vandq_s64(int64x2_t __p0, int64x2
5613	5634	#else
5614	5635	__ai __attribute__((target("neon"))) int64x2_t vandq_s64(int64x2_t __p0, int64x2_t __p1) {
5615	5636	int64x2_t __ret;
5616		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
5617		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	5637	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	5638	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
5618	5639	__ret = __rev0 & __rev1;
5619		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	5640	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
5620	5641	return __ret;
5621	5642	}
5622	5643	#endif
...	...	@@ -5630,10 +5651,10 @@ __ai __attribute__((target("neon"))) int16x8_t vandq_s16(int16x8_t __p0, int16x8
5630	5651	#else
5631	5652	__ai __attribute__((target("neon"))) int16x8_t vandq_s16(int16x8_t __p0, int16x8_t __p1) {
5632	5653	int16x8_t __ret;
5633		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
5634		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	5654	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	5655	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
5635	5656	__ret = __rev0 & __rev1;
5636		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	5657	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
5637	5658	return __ret;
5638	5659	}
5639	5660	#endif
...	...	@@ -5647,10 +5668,10 @@ __ai __attribute__((target("neon"))) uint8x8_t vand_u8(uint8x8_t __p0, uint8x8_t
5647	5668	#else
5648	5669	__ai __attribute__((target("neon"))) uint8x8_t vand_u8(uint8x8_t __p0, uint8x8_t __p1) {
5649	5670	uint8x8_t __ret;
5650		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
5651		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	5671	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	5672	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
5652	5673	__ret = __rev0 & __rev1;
5653		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	5674	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
5654	5675	return __ret;
5655	5676	}
5656	5677	#endif
...	...	@@ -5664,10 +5685,10 @@ __ai __attribute__((target("neon"))) uint32x2_t vand_u32(uint32x2_t __p0, uint32
5664	5685	#else
5665	5686	__ai __attribute__((target("neon"))) uint32x2_t vand_u32(uint32x2_t __p0, uint32x2_t __p1) {
5666	5687	uint32x2_t __ret;
5667		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
5668		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	5688	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	5689	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
5669	5690	__ret = __rev0 & __rev1;
5670		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	5691	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
5671	5692	return __ret;
5672	5693	}
5673	5694	#endif
...	...	@@ -5686,10 +5707,10 @@ __ai __attribute__((target("neon"))) uint16x4_t vand_u16(uint16x4_t __p0, uint16
5686	5707	#else
5687	5708	__ai __attribute__((target("neon"))) uint16x4_t vand_u16(uint16x4_t __p0, uint16x4_t __p1) {
5688	5709	uint16x4_t __ret;
5689		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
5690		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	5710	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	5711	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
5691	5712	__ret = __rev0 & __rev1;
5692		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	5713	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
5693	5714	return __ret;
5694	5715	}
5695	5716	#endif
...	...	@@ -5703,10 +5724,10 @@ __ai __attribute__((target("neon"))) int8x8_t vand_s8(int8x8_t __p0, int8x8_t __
5703	5724	#else
5704	5725	__ai __attribute__((target("neon"))) int8x8_t vand_s8(int8x8_t __p0, int8x8_t __p1) {
5705	5726	int8x8_t __ret;
5706		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
5707		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	5727	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	5728	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
5708	5729	__ret = __rev0 & __rev1;
5709		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	5730	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
5710	5731	return __ret;
5711	5732	}
5712	5733	#endif
...	...	@@ -5720,10 +5741,10 @@ __ai __attribute__((target("neon"))) int32x2_t vand_s32(int32x2_t __p0, int32x2_
5720	5741	#else
5721	5742	__ai __attribute__((target("neon"))) int32x2_t vand_s32(int32x2_t __p0, int32x2_t __p1) {
5722	5743	int32x2_t __ret;
5723		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
5724		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	5744	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	5745	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
5725	5746	__ret = __rev0 & __rev1;
5726		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	5747	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
5727	5748	return __ret;
5728	5749	}
5729	5750	#endif
...	...	@@ -5742,10 +5763,10 @@ __ai __attribute__((target("neon"))) int16x4_t vand_s16(int16x4_t __p0, int16x4_
5742	5763	#else
5743	5764	__ai __attribute__((target("neon"))) int16x4_t vand_s16(int16x4_t __p0, int16x4_t __p1) {
5744	5765	int16x4_t __ret;
5745		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
5746		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	5766	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	5767	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
5747	5768	__ret = __rev0 & __rev1;
5748		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	5769	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
5749	5770	return __ret;
5750	5771	}
5751	5772	#endif
...	...	@@ -5759,10 +5780,10 @@ __ai __attribute__((target("neon"))) uint8x16_t vbicq_u8(uint8x16_t __p0, uint8x
5759	5780	#else
5760	5781	__ai __attribute__((target("neon"))) uint8x16_t vbicq_u8(uint8x16_t __p0, uint8x16_t __p1) {
5761	5782	uint8x16_t __ret;
5762		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
5763		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	5783	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	5784	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
5764	5785	__ret = __rev0 & ~__rev1;
5765		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	5786	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
5766	5787	return __ret;
5767	5788	}
5768	5789	#endif
...	...	@@ -5776,10 +5797,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vbicq_u32(uint32x4_t __p0, uint3
5776	5797	#else
5777	5798	__ai __attribute__((target("neon"))) uint32x4_t vbicq_u32(uint32x4_t __p0, uint32x4_t __p1) {
5778	5799	uint32x4_t __ret;
5779		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
5780		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	5800	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	5801	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
5781	5802	__ret = __rev0 & ~__rev1;
5782		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	5803	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
5783	5804	return __ret;
5784	5805	}
5785	5806	#endif
...	...	@@ -5793,10 +5814,10 @@ __ai __attribute__((target("neon"))) uint64x2_t vbicq_u64(uint64x2_t __p0, uint6
5793	5814	#else
5794	5815	__ai __attribute__((target("neon"))) uint64x2_t vbicq_u64(uint64x2_t __p0, uint64x2_t __p1) {
5795	5816	uint64x2_t __ret;
5796		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
5797		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	5817	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	5818	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
5798	5819	__ret = __rev0 & ~__rev1;
5799		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	5820	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
5800	5821	return __ret;
5801	5822	}
5802	5823	#endif
...	...	@@ -5810,10 +5831,10 @@ __ai __attribute__((target("neon"))) uint16x8_t vbicq_u16(uint16x8_t __p0, uint1
5810	5831	#else
5811	5832	__ai __attribute__((target("neon"))) uint16x8_t vbicq_u16(uint16x8_t __p0, uint16x8_t __p1) {
5812	5833	uint16x8_t __ret;
5813		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
5814		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	5834	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	5835	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
5815	5836	__ret = __rev0 & ~__rev1;
5816		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	5837	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
5817	5838	return __ret;
5818	5839	}
5819	5840	#endif
...	...	@@ -5827,10 +5848,10 @@ __ai __attribute__((target("neon"))) int8x16_t vbicq_s8(int8x16_t __p0, int8x16_
5827	5848	#else
5828	5849	__ai __attribute__((target("neon"))) int8x16_t vbicq_s8(int8x16_t __p0, int8x16_t __p1) {
5829	5850	int8x16_t __ret;
5830		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
5831		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	5851	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	5852	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
5832	5853	__ret = __rev0 & ~__rev1;
5833		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	5854	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
5834	5855	return __ret;
5835	5856	}
5836	5857	#endif
...	...	@@ -5844,10 +5865,10 @@ __ai __attribute__((target("neon"))) int32x4_t vbicq_s32(int32x4_t __p0, int32x4
5844	5865	#else
5845	5866	__ai __attribute__((target("neon"))) int32x4_t vbicq_s32(int32x4_t __p0, int32x4_t __p1) {
5846	5867	int32x4_t __ret;
5847		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
5848		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	5868	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	5869	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
5849	5870	__ret = __rev0 & ~__rev1;
5850		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	5871	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
5851	5872	return __ret;
5852	5873	}
5853	5874	#endif
...	...	@@ -5861,10 +5882,10 @@ __ai __attribute__((target("neon"))) int64x2_t vbicq_s64(int64x2_t __p0, int64x2
5861	5882	#else
5862	5883	__ai __attribute__((target("neon"))) int64x2_t vbicq_s64(int64x2_t __p0, int64x2_t __p1) {
5863	5884	int64x2_t __ret;
5864		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
5865		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	5885	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	5886	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
5866	5887	__ret = __rev0 & ~__rev1;
5867		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	5888	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
5868	5889	return __ret;
5869	5890	}
5870	5891	#endif
...	...	@@ -5878,10 +5899,10 @@ __ai __attribute__((target("neon"))) int16x8_t vbicq_s16(int16x8_t __p0, int16x8
5878	5899	#else
5879	5900	__ai __attribute__((target("neon"))) int16x8_t vbicq_s16(int16x8_t __p0, int16x8_t __p1) {
5880	5901	int16x8_t __ret;
5881		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
5882		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	5902	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	5903	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
5883	5904	__ret = __rev0 & ~__rev1;
5884		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	5905	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
5885	5906	return __ret;
5886	5907	}
5887	5908	#endif
...	...	@@ -5895,10 +5916,10 @@ __ai __attribute__((target("neon"))) uint8x8_t vbic_u8(uint8x8_t __p0, uint8x8_t
5895	5916	#else
5896	5917	__ai __attribute__((target("neon"))) uint8x8_t vbic_u8(uint8x8_t __p0, uint8x8_t __p1) {
5897	5918	uint8x8_t __ret;
5898		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
5899		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	5919	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	5920	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
5900	5921	__ret = __rev0 & ~__rev1;
5901		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	5922	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
5902	5923	return __ret;
5903	5924	}
5904	5925	#endif
...	...	@@ -5912,10 +5933,10 @@ __ai __attribute__((target("neon"))) uint32x2_t vbic_u32(uint32x2_t __p0, uint32
5912	5933	#else
5913	5934	__ai __attribute__((target("neon"))) uint32x2_t vbic_u32(uint32x2_t __p0, uint32x2_t __p1) {
5914	5935	uint32x2_t __ret;
5915		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
5916		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	5936	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	5937	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
5917	5938	__ret = __rev0 & ~__rev1;
5918		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	5939	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
5919	5940	return __ret;
5920	5941	}
5921	5942	#endif
...	...	@@ -5934,10 +5955,10 @@ __ai __attribute__((target("neon"))) uint16x4_t vbic_u16(uint16x4_t __p0, uint16
5934	5955	#else
5935	5956	__ai __attribute__((target("neon"))) uint16x4_t vbic_u16(uint16x4_t __p0, uint16x4_t __p1) {
5936	5957	uint16x4_t __ret;
5937		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
5938		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	5958	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	5959	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
5939	5960	__ret = __rev0 & ~__rev1;
5940		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	5961	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
5941	5962	return __ret;
5942	5963	}
5943	5964	#endif
...	...	@@ -5951,10 +5972,10 @@ __ai __attribute__((target("neon"))) int8x8_t vbic_s8(int8x8_t __p0, int8x8_t __
5951	5972	#else
5952	5973	__ai __attribute__((target("neon"))) int8x8_t vbic_s8(int8x8_t __p0, int8x8_t __p1) {
5953	5974	int8x8_t __ret;
5954		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
5955		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	5975	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	5976	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
5956	5977	__ret = __rev0 & ~__rev1;
5957		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	5978	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
5958	5979	return __ret;
5959	5980	}
5960	5981	#endif
...	...	@@ -5968,10 +5989,10 @@ __ai __attribute__((target("neon"))) int32x2_t vbic_s32(int32x2_t __p0, int32x2_
5968	5989	#else
5969	5990	__ai __attribute__((target("neon"))) int32x2_t vbic_s32(int32x2_t __p0, int32x2_t __p1) {
5970	5991	int32x2_t __ret;
5971		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
5972		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	5992	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	5993	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
5973	5994	__ret = __rev0 & ~__rev1;
5974		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	5995	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
5975	5996	return __ret;
5976	5997	}
5977	5998	#endif
...	...	@@ -5990,10 +6011,10 @@ __ai __attribute__((target("neon"))) int16x4_t vbic_s16(int16x4_t __p0, int16x4_
5990	6011	#else
5991	6012	__ai __attribute__((target("neon"))) int16x4_t vbic_s16(int16x4_t __p0, int16x4_t __p1) {
5992	6013	int16x4_t __ret;
5993		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
5994		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	6014	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	6015	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
5995	6016	__ret = __rev0 & ~__rev1;
5996		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	6017	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
5997	6018	return __ret;
5998	6019	}
5999	6020	#endif
...	...	@@ -6001,17 +6022,17 @@ __ai __attribute__((target("neon"))) int16x4_t vbic_s16(int16x4_t __p0, int16x4_
6001	6022	#ifdef __LITTLE_ENDIAN__
6002	6023	__ai __attribute__((target("neon"))) poly8x8_t vbsl_p8(uint8x8_t __p0, poly8x8_t __p1, poly8x8_t __p2) {
6003	6024	poly8x8_t __ret;
6004		__ret = (poly8x8_t) __builtin_neon_vbsl_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 4);
	6025	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vbsl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 4));
6005	6026	return __ret;
6006	6027	}
6007	6028	#else
6008	6029	__ai __attribute__((target("neon"))) poly8x8_t vbsl_p8(uint8x8_t __p0, poly8x8_t __p1, poly8x8_t __p2) {
6009	6030	poly8x8_t __ret;
6010		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
6011		poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
6012		poly8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
6013		__ret = (poly8x8_t) __builtin_neon_vbsl_v((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 4);
6014		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	6031	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	6032	poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	6033	poly8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	6034	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vbsl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 4));
	6035	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
6015	6036	return __ret;
6016	6037	}
6017	6038	#endif
...	...	@@ -6019,17 +6040,17 @@ __ai __attribute__((target("neon"))) poly8x8_t vbsl_p8(uint8x8_t __p0, poly8x8_t
6019	6040	#ifdef __LITTLE_ENDIAN__
6020	6041	__ai __attribute__((target("neon"))) poly16x4_t vbsl_p16(uint16x4_t __p0, poly16x4_t __p1, poly16x4_t __p2) {
6021	6042	poly16x4_t __ret;
6022		__ret = (poly16x4_t) __builtin_neon_vbsl_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 5);
	6043	__ret = __builtin_bit_cast(poly16x4_t, __builtin_neon_vbsl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 5));
6023	6044	return __ret;
6024	6045	}
6025	6046	#else
6026	6047	__ai __attribute__((target("neon"))) poly16x4_t vbsl_p16(uint16x4_t __p0, poly16x4_t __p1, poly16x4_t __p2) {
6027	6048	poly16x4_t __ret;
6028		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
6029		poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
6030		poly16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
6031		__ret = (poly16x4_t) __builtin_neon_vbsl_v((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 5);
6032		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	6049	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	6050	poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	6051	poly16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
	6052	__ret = __builtin_bit_cast(poly16x4_t, __builtin_neon_vbsl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 5));
	6053	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
6033	6054	return __ret;
6034	6055	}
6035	6056	#endif
...	...	@@ -6037,17 +6058,17 @@ __ai __attribute__((target("neon"))) poly16x4_t vbsl_p16(uint16x4_t __p0, poly16
6037	6058	#ifdef __LITTLE_ENDIAN__
6038	6059	__ai __attribute__((target("neon"))) poly8x16_t vbslq_p8(uint8x16_t __p0, poly8x16_t __p1, poly8x16_t __p2) {
6039	6060	poly8x16_t __ret;
6040		__ret = (poly8x16_t) __builtin_neon_vbslq_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 36);
	6061	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vbslq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 36));
6041	6062	return __ret;
6042	6063	}
6043	6064	#else
6044	6065	__ai __attribute__((target("neon"))) poly8x16_t vbslq_p8(uint8x16_t __p0, poly8x16_t __p1, poly8x16_t __p2) {
6045	6066	poly8x16_t __ret;
6046		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
6047		poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
6048		poly8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
6049		__ret = (poly8x16_t) __builtin_neon_vbslq_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 36);
6050		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	6067	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	6068	poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	6069	poly8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	6070	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vbslq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 36));
	6071	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
6051	6072	return __ret;
6052	6073	}
6053	6074	#endif
...	...	@@ -6055,17 +6076,17 @@ __ai __attribute__((target("neon"))) poly8x16_t vbslq_p8(uint8x16_t __p0, poly8x
6055	6076	#ifdef __LITTLE_ENDIAN__
6056	6077	__ai __attribute__((target("neon"))) poly16x8_t vbslq_p16(uint16x8_t __p0, poly16x8_t __p1, poly16x8_t __p2) {
6057	6078	poly16x8_t __ret;
6058		__ret = (poly16x8_t) __builtin_neon_vbslq_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 37);
	6079	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vbslq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 37));
6059	6080	return __ret;
6060	6081	}
6061	6082	#else
6062	6083	__ai __attribute__((target("neon"))) poly16x8_t vbslq_p16(uint16x8_t __p0, poly16x8_t __p1, poly16x8_t __p2) {
6063	6084	poly16x8_t __ret;
6064		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
6065		poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
6066		poly16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
6067		__ret = (poly16x8_t) __builtin_neon_vbslq_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 37);
6068		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	6085	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	6086	poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	6087	poly16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
	6088	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vbslq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 37));
	6089	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
6069	6090	return __ret;
6070	6091	}
6071	6092	#endif
...	...	@@ -6073,17 +6094,17 @@ __ai __attribute__((target("neon"))) poly16x8_t vbslq_p16(uint16x8_t __p0, poly1
6073	6094	#ifdef __LITTLE_ENDIAN__
6074	6095	__ai __attribute__((target("neon"))) uint8x16_t vbslq_u8(uint8x16_t __p0, uint8x16_t __p1, uint8x16_t __p2) {
6075	6096	uint8x16_t __ret;
6076		__ret = (uint8x16_t) __builtin_neon_vbslq_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 48);
	6097	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vbslq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 48));
6077	6098	return __ret;
6078	6099	}
6079	6100	#else
6080	6101	__ai __attribute__((target("neon"))) uint8x16_t vbslq_u8(uint8x16_t __p0, uint8x16_t __p1, uint8x16_t __p2) {
6081	6102	uint8x16_t __ret;
6082		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
6083		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
6084		uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
6085		__ret = (uint8x16_t) __builtin_neon_vbslq_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 48);
6086		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	6103	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	6104	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	6105	uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	6106	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vbslq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 48));
	6107	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
6087	6108	return __ret;
6088	6109	}
6089	6110	#endif
...	...	@@ -6091,17 +6112,17 @@ __ai __attribute__((target("neon"))) uint8x16_t vbslq_u8(uint8x16_t __p0, uint8x
6091	6112	#ifdef __LITTLE_ENDIAN__
6092	6113	__ai __attribute__((target("neon"))) uint32x4_t vbslq_u32(uint32x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
6093	6114	uint32x4_t __ret;
6094		__ret = (uint32x4_t) __builtin_neon_vbslq_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 50);
	6115	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vbslq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 50));
6095	6116	return __ret;
6096	6117	}
6097	6118	#else
6098	6119	__ai __attribute__((target("neon"))) uint32x4_t vbslq_u32(uint32x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
6099	6120	uint32x4_t __ret;
6100		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
6101		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
6102		uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
6103		__ret = (uint32x4_t) __builtin_neon_vbslq_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 50);
6104		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	6121	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	6122	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	6123	uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
	6124	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vbslq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 50));
	6125	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
6105	6126	return __ret;
6106	6127	}
6107	6128	#endif
...	...	@@ -6109,17 +6130,17 @@ __ai __attribute__((target("neon"))) uint32x4_t vbslq_u32(uint32x4_t __p0, uint3
6109	6130	#ifdef __LITTLE_ENDIAN__
6110	6131	__ai __attribute__((target("neon"))) uint64x2_t vbslq_u64(uint64x2_t __p0, uint64x2_t __p1, uint64x2_t __p2) {
6111	6132	uint64x2_t __ret;
6112		__ret = (uint64x2_t) __builtin_neon_vbslq_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 51);
	6133	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vbslq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 51));
6113	6134	return __ret;
6114	6135	}
6115	6136	#else
6116	6137	__ai __attribute__((target("neon"))) uint64x2_t vbslq_u64(uint64x2_t __p0, uint64x2_t __p1, uint64x2_t __p2) {
6117	6138	uint64x2_t __ret;
6118		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
6119		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
6120		uint64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
6121		__ret = (uint64x2_t) __builtin_neon_vbslq_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 51);
6122		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	6139	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	6140	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	6141	uint64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_64);
	6142	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vbslq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 51));
	6143	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
6123	6144	return __ret;
6124	6145	}
6125	6146	#endif
...	...	@@ -6127,17 +6148,17 @@ __ai __attribute__((target("neon"))) uint64x2_t vbslq_u64(uint64x2_t __p0, uint6
6127	6148	#ifdef __LITTLE_ENDIAN__
6128	6149	__ai __attribute__((target("neon"))) uint16x8_t vbslq_u16(uint16x8_t __p0, uint16x8_t __p1, uint16x8_t __p2) {
6129	6150	uint16x8_t __ret;
6130		__ret = (uint16x8_t) __builtin_neon_vbslq_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 49);
	6151	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vbslq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 49));
6131	6152	return __ret;
6132	6153	}
6133	6154	#else
6134	6155	__ai __attribute__((target("neon"))) uint16x8_t vbslq_u16(uint16x8_t __p0, uint16x8_t __p1, uint16x8_t __p2) {
6135	6156	uint16x8_t __ret;
6136		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
6137		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
6138		uint16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
6139		__ret = (uint16x8_t) __builtin_neon_vbslq_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 49);
6140		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	6157	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	6158	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	6159	uint16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
	6160	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vbslq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 49));
	6161	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
6141	6162	return __ret;
6142	6163	}
6143	6164	#endif
...	...	@@ -6145,17 +6166,17 @@ __ai __attribute__((target("neon"))) uint16x8_t vbslq_u16(uint16x8_t __p0, uint1
6145	6166	#ifdef __LITTLE_ENDIAN__
6146	6167	__ai __attribute__((target("neon"))) int8x16_t vbslq_s8(uint8x16_t __p0, int8x16_t __p1, int8x16_t __p2) {
6147	6168	int8x16_t __ret;
6148		__ret = (int8x16_t) __builtin_neon_vbslq_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 32);
	6169	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vbslq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 32));
6149	6170	return __ret;
6150	6171	}
6151	6172	#else
6152	6173	__ai __attribute__((target("neon"))) int8x16_t vbslq_s8(uint8x16_t __p0, int8x16_t __p1, int8x16_t __p2) {
6153	6174	int8x16_t __ret;
6154		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
6155		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
6156		int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
6157		__ret = (int8x16_t) __builtin_neon_vbslq_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 32);
6158		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	6175	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	6176	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	6177	int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	6178	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vbslq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 32));
	6179	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
6159	6180	return __ret;
6160	6181	}
6161	6182	#endif
...	...	@@ -6163,17 +6184,17 @@ __ai __attribute__((target("neon"))) int8x16_t vbslq_s8(uint8x16_t __p0, int8x16
6163	6184	#ifdef __LITTLE_ENDIAN__
6164	6185	__ai __attribute__((target("neon"))) float32x4_t vbslq_f32(uint32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
6165	6186	float32x4_t __ret;
6166		__ret = (float32x4_t) __builtin_neon_vbslq_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
	6187	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vbslq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 41));
6167	6188	return __ret;
6168	6189	}
6169	6190	#else
6170	6191	__ai __attribute__((target("neon"))) float32x4_t vbslq_f32(uint32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
6171	6192	float32x4_t __ret;
6172		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
6173		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
6174		float32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
6175		__ret = (float32x4_t) __builtin_neon_vbslq_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 41);
6176		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	6193	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	6194	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	6195	float32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
	6196	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vbslq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 41));
	6197	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
6177	6198	return __ret;
6178	6199	}
6179	6200	#endif
...	...	@@ -6181,17 +6202,17 @@ __ai __attribute__((target("neon"))) float32x4_t vbslq_f32(uint32x4_t __p0, floa
6181	6202	#ifdef __LITTLE_ENDIAN__
6182	6203	__ai __attribute__((target("neon"))) int32x4_t vbslq_s32(uint32x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
6183	6204	int32x4_t __ret;
6184		__ret = (int32x4_t) __builtin_neon_vbslq_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 34);
	6205	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vbslq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 34));
6185	6206	return __ret;
6186	6207	}
6187	6208	#else
6188	6209	__ai __attribute__((target("neon"))) int32x4_t vbslq_s32(uint32x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
6189	6210	int32x4_t __ret;
6190		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
6191		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
6192		int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
6193		__ret = (int32x4_t) __builtin_neon_vbslq_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 34);
6194		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	6211	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	6212	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	6213	int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
	6214	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vbslq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 34));
	6215	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
6195	6216	return __ret;
6196	6217	}
6197	6218	#endif
...	...	@@ -6199,17 +6220,17 @@ __ai __attribute__((target("neon"))) int32x4_t vbslq_s32(uint32x4_t __p0, int32x
6199	6220	#ifdef __LITTLE_ENDIAN__
6200	6221	__ai __attribute__((target("neon"))) int64x2_t vbslq_s64(uint64x2_t __p0, int64x2_t __p1, int64x2_t __p2) {
6201	6222	int64x2_t __ret;
6202		__ret = (int64x2_t) __builtin_neon_vbslq_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 35);
	6223	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vbslq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 35));
6203	6224	return __ret;
6204	6225	}
6205	6226	#else
6206	6227	__ai __attribute__((target("neon"))) int64x2_t vbslq_s64(uint64x2_t __p0, int64x2_t __p1, int64x2_t __p2) {
6207	6228	int64x2_t __ret;
6208		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
6209		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
6210		int64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
6211		__ret = (int64x2_t) __builtin_neon_vbslq_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 35);
6212		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	6229	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	6230	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	6231	int64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_64);
	6232	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vbslq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 35));
	6233	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
6213	6234	return __ret;
6214	6235	}
6215	6236	#endif
...	...	@@ -6217,17 +6238,17 @@ __ai __attribute__((target("neon"))) int64x2_t vbslq_s64(uint64x2_t __p0, int64x
6217	6238	#ifdef __LITTLE_ENDIAN__
6218	6239	__ai __attribute__((target("neon"))) int16x8_t vbslq_s16(uint16x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
6219	6240	int16x8_t __ret;
6220		__ret = (int16x8_t) __builtin_neon_vbslq_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 33);
	6241	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vbslq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 33));
6221	6242	return __ret;
6222	6243	}
6223	6244	#else
6224	6245	__ai __attribute__((target("neon"))) int16x8_t vbslq_s16(uint16x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
6225	6246	int16x8_t __ret;
6226		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
6227		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
6228		int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
6229		__ret = (int16x8_t) __builtin_neon_vbslq_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 33);
6230		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	6247	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	6248	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	6249	int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
	6250	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vbslq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 33));
	6251	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
6231	6252	return __ret;
6232	6253	}
6233	6254	#endif
...	...	@@ -6235,17 +6256,17 @@ __ai __attribute__((target("neon"))) int16x8_t vbslq_s16(uint16x8_t __p0, int16x
6235	6256	#ifdef __LITTLE_ENDIAN__
6236	6257	__ai __attribute__((target("neon"))) uint8x8_t vbsl_u8(uint8x8_t __p0, uint8x8_t __p1, uint8x8_t __p2) {
6237	6258	uint8x8_t __ret;
6238		__ret = (uint8x8_t) __builtin_neon_vbsl_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 16);
	6259	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vbsl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 16));
6239	6260	return __ret;
6240	6261	}
6241	6262	#else
6242	6263	__ai __attribute__((target("neon"))) uint8x8_t vbsl_u8(uint8x8_t __p0, uint8x8_t __p1, uint8x8_t __p2) {
6243	6264	uint8x8_t __ret;
6244		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
6245		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
6246		uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
6247		__ret = (uint8x8_t) __builtin_neon_vbsl_v((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 16);
6248		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	6265	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	6266	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	6267	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	6268	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vbsl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 16));
	6269	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
6249	6270	return __ret;
6250	6271	}
6251	6272	#endif
...	...	@@ -6253,40 +6274,40 @@ __ai __attribute__((target("neon"))) uint8x8_t vbsl_u8(uint8x8_t __p0, uint8x8_t
6253	6274	#ifdef __LITTLE_ENDIAN__
6254	6275	__ai __attribute__((target("neon"))) uint32x2_t vbsl_u32(uint32x2_t __p0, uint32x2_t __p1, uint32x2_t __p2) {
6255	6276	uint32x2_t __ret;
6256		__ret = (uint32x2_t) __builtin_neon_vbsl_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 18);
	6277	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vbsl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 18));
6257	6278	return __ret;
6258	6279	}
6259	6280	#else
6260	6281	__ai __attribute__((target("neon"))) uint32x2_t vbsl_u32(uint32x2_t __p0, uint32x2_t __p1, uint32x2_t __p2) {
6261	6282	uint32x2_t __ret;
6262		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
6263		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
6264		uint32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
6265		__ret = (uint32x2_t) __builtin_neon_vbsl_v((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 18);
6266		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	6283	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	6284	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	6285	uint32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_32);
	6286	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vbsl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 18));
	6287	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
6267	6288	return __ret;
6268	6289	}
6269	6290	#endif
6270	6291
6271	6292	__ai __attribute__((target("neon"))) uint64x1_t vbsl_u64(uint64x1_t __p0, uint64x1_t __p1, uint64x1_t __p2) {
6272	6293	uint64x1_t __ret;
6273		__ret = (uint64x1_t) __builtin_neon_vbsl_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 19);
	6294	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vbsl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 19));
6274	6295	return __ret;
6275	6296	}
6276	6297	#ifdef __LITTLE_ENDIAN__
6277	6298	__ai __attribute__((target("neon"))) uint16x4_t vbsl_u16(uint16x4_t __p0, uint16x4_t __p1, uint16x4_t __p2) {
6278	6299	uint16x4_t __ret;
6279		__ret = (uint16x4_t) __builtin_neon_vbsl_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 17);
	6300	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vbsl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 17));
6280	6301	return __ret;
6281	6302	}
6282	6303	#else
6283	6304	__ai __attribute__((target("neon"))) uint16x4_t vbsl_u16(uint16x4_t __p0, uint16x4_t __p1, uint16x4_t __p2) {
6284	6305	uint16x4_t __ret;
6285		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
6286		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
6287		uint16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
6288		__ret = (uint16x4_t) __builtin_neon_vbsl_v((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 17);
6289		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	6306	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	6307	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	6308	uint16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
	6309	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vbsl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 17));
	6310	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
6290	6311	return __ret;
6291	6312	}
6292	6313	#endif
...	...	@@ -6294,17 +6315,17 @@ __ai __attribute__((target("neon"))) uint16x4_t vbsl_u16(uint16x4_t __p0, uint16
6294	6315	#ifdef __LITTLE_ENDIAN__
6295	6316	__ai __attribute__((target("neon"))) int8x8_t vbsl_s8(uint8x8_t __p0, int8x8_t __p1, int8x8_t __p2) {
6296	6317	int8x8_t __ret;
6297		__ret = (int8x8_t) __builtin_neon_vbsl_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 0);
	6318	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vbsl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 0));
6298	6319	return __ret;
6299	6320	}
6300	6321	#else
6301	6322	__ai __attribute__((target("neon"))) int8x8_t vbsl_s8(uint8x8_t __p0, int8x8_t __p1, int8x8_t __p2) {
6302	6323	int8x8_t __ret;
6303		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
6304		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
6305		int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
6306		__ret = (int8x8_t) __builtin_neon_vbsl_v((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 0);
6307		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	6324	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	6325	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	6326	int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	6327	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vbsl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 0));
	6328	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
6308	6329	return __ret;
6309	6330	}
6310	6331	#endif
...	...	@@ -6312,17 +6333,17 @@ __ai __attribute__((target("neon"))) int8x8_t vbsl_s8(uint8x8_t __p0, int8x8_t _
6312	6333	#ifdef __LITTLE_ENDIAN__
6313	6334	__ai __attribute__((target("neon"))) float32x2_t vbsl_f32(uint32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
6314	6335	float32x2_t __ret;
6315		__ret = (float32x2_t) __builtin_neon_vbsl_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
	6336	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vbsl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 9));
6316	6337	return __ret;
6317	6338	}
6318	6339	#else
6319	6340	__ai __attribute__((target("neon"))) float32x2_t vbsl_f32(uint32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
6320	6341	float32x2_t __ret;
6321		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
6322		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
6323		float32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
6324		__ret = (float32x2_t) __builtin_neon_vbsl_v((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 9);
6325		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	6342	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	6343	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	6344	float32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_32);
	6345	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vbsl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 9));
	6346	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
6326	6347	return __ret;
6327	6348	}
6328	6349	#endif
...	...	@@ -6330,40 +6351,40 @@ __ai __attribute__((target("neon"))) float32x2_t vbsl_f32(uint32x2_t __p0, float
6330	6351	#ifdef __LITTLE_ENDIAN__
6331	6352	__ai __attribute__((target("neon"))) int32x2_t vbsl_s32(uint32x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
6332	6353	int32x2_t __ret;
6333		__ret = (int32x2_t) __builtin_neon_vbsl_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 2);
	6354	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vbsl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 2));
6334	6355	return __ret;
6335	6356	}
6336	6357	#else
6337	6358	__ai __attribute__((target("neon"))) int32x2_t vbsl_s32(uint32x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
6338	6359	int32x2_t __ret;
6339		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
6340		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
6341		int32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
6342		__ret = (int32x2_t) __builtin_neon_vbsl_v((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 2);
6343		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	6360	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	6361	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	6362	int32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_32);
	6363	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vbsl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 2));
	6364	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
6344	6365	return __ret;
6345	6366	}
6346	6367	#endif
6347	6368
6348	6369	__ai __attribute__((target("neon"))) int64x1_t vbsl_s64(uint64x1_t __p0, int64x1_t __p1, int64x1_t __p2) {
6349	6370	int64x1_t __ret;
6350		__ret = (int64x1_t) __builtin_neon_vbsl_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 3);
	6371	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vbsl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 3));
6351	6372	return __ret;
6352	6373	}
6353	6374	#ifdef __LITTLE_ENDIAN__
6354	6375	__ai __attribute__((target("neon"))) int16x4_t vbsl_s16(uint16x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
6355	6376	int16x4_t __ret;
6356		__ret = (int16x4_t) __builtin_neon_vbsl_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 1);
	6377	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vbsl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 1));
6357	6378	return __ret;
6358	6379	}
6359	6380	#else
6360	6381	__ai __attribute__((target("neon"))) int16x4_t vbsl_s16(uint16x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
6361	6382	int16x4_t __ret;
6362		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
6363		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
6364		int16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
6365		__ret = (int16x4_t) __builtin_neon_vbsl_v((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 1);
6366		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	6383	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	6384	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	6385	int16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
	6386	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vbsl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 1));
	6387	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
6367	6388	return __ret;
6368	6389	}
6369	6390	#endif
...	...	@@ -6371,17 +6392,17 @@ __ai __attribute__((target("neon"))) int16x4_t vbsl_s16(uint16x4_t __p0, int16x4
6371	6392	#ifdef __LITTLE_ENDIAN__
6372	6393	__ai __attribute__((target("neon"))) float16x8_t vbslq_f16(uint16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
6373	6394	float16x8_t __ret;
6374		__ret = (float16x8_t) __builtin_neon_vbslq_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 40);
	6395	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vbslq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 40));
6375	6396	return __ret;
6376	6397	}
6377	6398	#else
6378	6399	__ai __attribute__((target("neon"))) float16x8_t vbslq_f16(uint16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
6379	6400	float16x8_t __ret;
6380		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
6381		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
6382		float16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
6383		__ret = (float16x8_t) __builtin_neon_vbslq_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 40);
6384		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	6401	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	6402	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	6403	float16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
	6404	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vbslq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 40));
	6405	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
6385	6406	return __ret;
6386	6407	}
6387	6408	#endif
...	...	@@ -6389,17 +6410,17 @@ __ai __attribute__((target("neon"))) float16x8_t vbslq_f16(uint16x8_t __p0, floa
6389	6410	#ifdef __LITTLE_ENDIAN__
6390	6411	__ai __attribute__((target("neon"))) float16x4_t vbsl_f16(uint16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
6391	6412	float16x4_t __ret;
6392		__ret = (float16x4_t) __builtin_neon_vbsl_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 8);
	6413	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vbsl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 8));
6393	6414	return __ret;
6394	6415	}
6395	6416	#else
6396	6417	__ai __attribute__((target("neon"))) float16x4_t vbsl_f16(uint16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
6397	6418	float16x4_t __ret;
6398		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
6399		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
6400		float16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
6401		__ret = (float16x4_t) __builtin_neon_vbsl_v((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 8);
6402		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	6419	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	6420	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	6421	float16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
	6422	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vbsl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 8));
	6423	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
6403	6424	return __ret;
6404	6425	}
6405	6426	#endif
...	...	@@ -6407,16 +6428,16 @@ __ai __attribute__((target("neon"))) float16x4_t vbsl_f16(uint16x4_t __p0, float
6407	6428	#ifdef __LITTLE_ENDIAN__
6408	6429	__ai __attribute__((target("neon"))) uint32x4_t vcageq_f32(float32x4_t __p0, float32x4_t __p1) {
6409	6430	uint32x4_t __ret;
6410		__ret = (uint32x4_t) __builtin_neon_vcageq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
	6431	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcageq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50));
6411	6432	return __ret;
6412	6433	}
6413	6434	#else
6414	6435	__ai __attribute__((target("neon"))) uint32x4_t vcageq_f32(float32x4_t __p0, float32x4_t __p1) {
6415	6436	uint32x4_t __ret;
6416		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
6417		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
6418		__ret = (uint32x4_t) __builtin_neon_vcageq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
6419		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	6437	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	6438	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	6439	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcageq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 50));
	6440	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
6420	6441	return __ret;
6421	6442	}
6422	6443	#endif
...	...	@@ -6424,16 +6445,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vcageq_f32(float32x4_t __p0, flo
6424	6445	#ifdef __LITTLE_ENDIAN__
6425	6446	__ai __attribute__((target("neon"))) uint32x2_t vcage_f32(float32x2_t __p0, float32x2_t __p1) {
6426	6447	uint32x2_t __ret;
6427		__ret = (uint32x2_t) __builtin_neon_vcage_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
	6448	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcage_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 18));
6428	6449	return __ret;
6429	6450	}
6430	6451	#else
6431	6452	__ai __attribute__((target("neon"))) uint32x2_t vcage_f32(float32x2_t __p0, float32x2_t __p1) {
6432	6453	uint32x2_t __ret;
6433		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
6434		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
6435		__ret = (uint32x2_t) __builtin_neon_vcage_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
6436		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	6454	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	6455	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	6456	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcage_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 18));
	6457	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
6437	6458	return __ret;
6438	6459	}
6439	6460	#endif
...	...	@@ -6441,16 +6462,16 @@ __ai __attribute__((target("neon"))) uint32x2_t vcage_f32(float32x2_t __p0, floa
6441	6462	#ifdef __LITTLE_ENDIAN__
6442	6463	__ai __attribute__((target("neon"))) uint32x4_t vcagtq_f32(float32x4_t __p0, float32x4_t __p1) {
6443	6464	uint32x4_t __ret;
6444		__ret = (uint32x4_t) __builtin_neon_vcagtq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
	6465	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcagtq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50));
6445	6466	return __ret;
6446	6467	}
6447	6468	#else
6448	6469	__ai __attribute__((target("neon"))) uint32x4_t vcagtq_f32(float32x4_t __p0, float32x4_t __p1) {
6449	6470	uint32x4_t __ret;
6450		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
6451		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
6452		__ret = (uint32x4_t) __builtin_neon_vcagtq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
6453		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	6471	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	6472	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	6473	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcagtq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 50));
	6474	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
6454	6475	return __ret;
6455	6476	}
6456	6477	#endif
...	...	@@ -6458,16 +6479,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vcagtq_f32(float32x4_t __p0, flo
6458	6479	#ifdef __LITTLE_ENDIAN__
6459	6480	__ai __attribute__((target("neon"))) uint32x2_t vcagt_f32(float32x2_t __p0, float32x2_t __p1) {
6460	6481	uint32x2_t __ret;
6461		__ret = (uint32x2_t) __builtin_neon_vcagt_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
	6482	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcagt_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 18));
6462	6483	return __ret;
6463	6484	}
6464	6485	#else
6465	6486	__ai __attribute__((target("neon"))) uint32x2_t vcagt_f32(float32x2_t __p0, float32x2_t __p1) {
6466	6487	uint32x2_t __ret;
6467		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
6468		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
6469		__ret = (uint32x2_t) __builtin_neon_vcagt_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
6470		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	6488	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	6489	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	6490	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcagt_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 18));
	6491	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
6471	6492	return __ret;
6472	6493	}
6473	6494	#endif
...	...	@@ -6475,16 +6496,16 @@ __ai __attribute__((target("neon"))) uint32x2_t vcagt_f32(float32x2_t __p0, floa
6475	6496	#ifdef __LITTLE_ENDIAN__
6476	6497	__ai __attribute__((target("neon"))) uint32x4_t vcaleq_f32(float32x4_t __p0, float32x4_t __p1) {
6477	6498	uint32x4_t __ret;
6478		__ret = (uint32x4_t) __builtin_neon_vcaleq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
	6499	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcaleq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50));
6479	6500	return __ret;
6480	6501	}
6481	6502	#else
6482	6503	__ai __attribute__((target("neon"))) uint32x4_t vcaleq_f32(float32x4_t __p0, float32x4_t __p1) {
6483	6504	uint32x4_t __ret;
6484		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
6485		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
6486		__ret = (uint32x4_t) __builtin_neon_vcaleq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
6487		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	6505	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	6506	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	6507	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcaleq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 50));
	6508	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
6488	6509	return __ret;
6489	6510	}
6490	6511	#endif
...	...	@@ -6492,16 +6513,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vcaleq_f32(float32x4_t __p0, flo
6492	6513	#ifdef __LITTLE_ENDIAN__
6493	6514	__ai __attribute__((target("neon"))) uint32x2_t vcale_f32(float32x2_t __p0, float32x2_t __p1) {
6494	6515	uint32x2_t __ret;
6495		__ret = (uint32x2_t) __builtin_neon_vcale_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
	6516	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcale_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 18));
6496	6517	return __ret;
6497	6518	}
6498	6519	#else
6499	6520	__ai __attribute__((target("neon"))) uint32x2_t vcale_f32(float32x2_t __p0, float32x2_t __p1) {
6500	6521	uint32x2_t __ret;
6501		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
6502		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
6503		__ret = (uint32x2_t) __builtin_neon_vcale_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
6504		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	6522	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	6523	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	6524	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcale_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 18));
	6525	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
6505	6526	return __ret;
6506	6527	}
6507	6528	#endif
...	...	@@ -6509,16 +6530,16 @@ __ai __attribute__((target("neon"))) uint32x2_t vcale_f32(float32x2_t __p0, floa
6509	6530	#ifdef __LITTLE_ENDIAN__
6510	6531	__ai __attribute__((target("neon"))) uint32x4_t vcaltq_f32(float32x4_t __p0, float32x4_t __p1) {
6511	6532	uint32x4_t __ret;
6512		__ret = (uint32x4_t) __builtin_neon_vcaltq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
	6533	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcaltq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50));
6513	6534	return __ret;
6514	6535	}
6515	6536	#else
6516	6537	__ai __attribute__((target("neon"))) uint32x4_t vcaltq_f32(float32x4_t __p0, float32x4_t __p1) {
6517	6538	uint32x4_t __ret;
6518		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
6519		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
6520		__ret = (uint32x4_t) __builtin_neon_vcaltq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
6521		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	6539	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	6540	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	6541	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcaltq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 50));
	6542	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
6522	6543	return __ret;
6523	6544	}
6524	6545	#endif
...	...	@@ -6526,16 +6547,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vcaltq_f32(float32x4_t __p0, flo
6526	6547	#ifdef __LITTLE_ENDIAN__
6527	6548	__ai __attribute__((target("neon"))) uint32x2_t vcalt_f32(float32x2_t __p0, float32x2_t __p1) {
6528	6549	uint32x2_t __ret;
6529		__ret = (uint32x2_t) __builtin_neon_vcalt_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
	6550	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcalt_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 18));
6530	6551	return __ret;
6531	6552	}
6532	6553	#else
6533	6554	__ai __attribute__((target("neon"))) uint32x2_t vcalt_f32(float32x2_t __p0, float32x2_t __p1) {
6534	6555	uint32x2_t __ret;
6535		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
6536		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
6537		__ret = (uint32x2_t) __builtin_neon_vcalt_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
6538		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	6556	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	6557	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	6558	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcalt_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 18));
	6559	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
6539	6560	return __ret;
6540	6561	}
6541	6562	#endif
...	...	@@ -6543,16 +6564,16 @@ __ai __attribute__((target("neon"))) uint32x2_t vcalt_f32(float32x2_t __p0, floa
6543	6564	#ifdef __LITTLE_ENDIAN__
6544	6565	__ai __attribute__((target("neon"))) uint8x8_t vceq_p8(poly8x8_t __p0, poly8x8_t __p1) {
6545	6566	uint8x8_t __ret;
6546		__ret = (uint8x8_t)(__p0 == __p1);
	6567	__ret = __builtin_bit_cast(uint8x8_t, __p0 == __p1);
6547	6568	return __ret;
6548	6569	}
6549	6570	#else
6550	6571	__ai __attribute__((target("neon"))) uint8x8_t vceq_p8(poly8x8_t __p0, poly8x8_t __p1) {
6551	6572	uint8x8_t __ret;
6552		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
6553		poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
6554		__ret = (uint8x8_t)(__rev0 == __rev1);
6555		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	6573	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	6574	poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	6575	__ret = __builtin_bit_cast(uint8x8_t, __rev0 == __rev1);
	6576	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
6556	6577	return __ret;
6557	6578	}
6558	6579	#endif
...	...	@@ -6560,16 +6581,16 @@ __ai __attribute__((target("neon"))) uint8x8_t vceq_p8(poly8x8_t __p0, poly8x8_t
6560	6581	#ifdef __LITTLE_ENDIAN__
6561	6582	__ai __attribute__((target("neon"))) uint8x16_t vceqq_p8(poly8x16_t __p0, poly8x16_t __p1) {
6562	6583	uint8x16_t __ret;
6563		__ret = (uint8x16_t)(__p0 == __p1);
	6584	__ret = __builtin_bit_cast(uint8x16_t, __p0 == __p1);
6564	6585	return __ret;
6565	6586	}
6566	6587	#else
6567	6588	__ai __attribute__((target("neon"))) uint8x16_t vceqq_p8(poly8x16_t __p0, poly8x16_t __p1) {
6568	6589	uint8x16_t __ret;
6569		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
6570		poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
6571		__ret = (uint8x16_t)(__rev0 == __rev1);
6572		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	6590	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	6591	poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	6592	__ret = __builtin_bit_cast(uint8x16_t, __rev0 == __rev1);
	6593	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
6573	6594	return __ret;
6574	6595	}
6575	6596	#endif
...	...	@@ -6577,16 +6598,16 @@ __ai __attribute__((target("neon"))) uint8x16_t vceqq_p8(poly8x16_t __p0, poly8x
6577	6598	#ifdef __LITTLE_ENDIAN__
6578	6599	__ai __attribute__((target("neon"))) uint8x16_t vceqq_u8(uint8x16_t __p0, uint8x16_t __p1) {
6579	6600	uint8x16_t __ret;
6580		__ret = (uint8x16_t)(__p0 == __p1);
	6601	__ret = __builtin_bit_cast(uint8x16_t, __p0 == __p1);
6581	6602	return __ret;
6582	6603	}
6583	6604	#else
6584	6605	__ai __attribute__((target("neon"))) uint8x16_t vceqq_u8(uint8x16_t __p0, uint8x16_t __p1) {
6585	6606	uint8x16_t __ret;
6586		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
6587		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
6588		__ret = (uint8x16_t)(__rev0 == __rev1);
6589		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	6607	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	6608	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	6609	__ret = __builtin_bit_cast(uint8x16_t, __rev0 == __rev1);
	6610	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
6590	6611	return __ret;
6591	6612	}
6592	6613	#endif
...	...	@@ -6594,16 +6615,16 @@ __ai __attribute__((target("neon"))) uint8x16_t vceqq_u8(uint8x16_t __p0, uint8x
6594	6615	#ifdef __LITTLE_ENDIAN__
6595	6616	__ai __attribute__((target("neon"))) uint32x4_t vceqq_u32(uint32x4_t __p0, uint32x4_t __p1) {
6596	6617	uint32x4_t __ret;
6597		__ret = (uint32x4_t)(__p0 == __p1);
	6618	__ret = __builtin_bit_cast(uint32x4_t, __p0 == __p1);
6598	6619	return __ret;
6599	6620	}
6600	6621	#else
6601	6622	__ai __attribute__((target("neon"))) uint32x4_t vceqq_u32(uint32x4_t __p0, uint32x4_t __p1) {
6602	6623	uint32x4_t __ret;
6603		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
6604		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
6605		__ret = (uint32x4_t)(__rev0 == __rev1);
6606		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	6624	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	6625	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	6626	__ret = __builtin_bit_cast(uint32x4_t, __rev0 == __rev1);
	6627	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
6607	6628	return __ret;
6608	6629	}
6609	6630	#endif
...	...	@@ -6611,16 +6632,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vceqq_u32(uint32x4_t __p0, uint3
6611	6632	#ifdef __LITTLE_ENDIAN__
6612	6633	__ai __attribute__((target("neon"))) uint16x8_t vceqq_u16(uint16x8_t __p0, uint16x8_t __p1) {
6613	6634	uint16x8_t __ret;
6614		__ret = (uint16x8_t)(__p0 == __p1);
	6635	__ret = __builtin_bit_cast(uint16x8_t, __p0 == __p1);
6615	6636	return __ret;
6616	6637	}
6617	6638	#else
6618	6639	__ai __attribute__((target("neon"))) uint16x8_t vceqq_u16(uint16x8_t __p0, uint16x8_t __p1) {
6619	6640	uint16x8_t __ret;
6620		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
6621		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
6622		__ret = (uint16x8_t)(__rev0 == __rev1);
6623		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	6641	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	6642	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	6643	__ret = __builtin_bit_cast(uint16x8_t, __rev0 == __rev1);
	6644	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
6624	6645	return __ret;
6625	6646	}
6626	6647	#endif
...	...	@@ -6628,16 +6649,16 @@ __ai __attribute__((target("neon"))) uint16x8_t vceqq_u16(uint16x8_t __p0, uint1
6628	6649	#ifdef __LITTLE_ENDIAN__
6629	6650	__ai __attribute__((target("neon"))) uint8x16_t vceqq_s8(int8x16_t __p0, int8x16_t __p1) {
6630	6651	uint8x16_t __ret;
6631		__ret = (uint8x16_t)(__p0 == __p1);
	6652	__ret = __builtin_bit_cast(uint8x16_t, __p0 == __p1);
6632	6653	return __ret;
6633	6654	}
6634	6655	#else
6635	6656	__ai __attribute__((target("neon"))) uint8x16_t vceqq_s8(int8x16_t __p0, int8x16_t __p1) {
6636	6657	uint8x16_t __ret;
6637		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
6638		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
6639		__ret = (uint8x16_t)(__rev0 == __rev1);
6640		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	6658	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	6659	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	6660	__ret = __builtin_bit_cast(uint8x16_t, __rev0 == __rev1);
	6661	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
6641	6662	return __ret;
6642	6663	}
6643	6664	#endif
...	...	@@ -6645,16 +6666,16 @@ __ai __attribute__((target("neon"))) uint8x16_t vceqq_s8(int8x16_t __p0, int8x16
6645	6666	#ifdef __LITTLE_ENDIAN__
6646	6667	__ai __attribute__((target("neon"))) uint32x4_t vceqq_f32(float32x4_t __p0, float32x4_t __p1) {
6647	6668	uint32x4_t __ret;
6648		__ret = (uint32x4_t)(__p0 == __p1);
	6669	__ret = __builtin_bit_cast(uint32x4_t, __p0 == __p1);
6649	6670	return __ret;
6650	6671	}
6651	6672	#else
6652	6673	__ai __attribute__((target("neon"))) uint32x4_t vceqq_f32(float32x4_t __p0, float32x4_t __p1) {
6653	6674	uint32x4_t __ret;
6654		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
6655		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
6656		__ret = (uint32x4_t)(__rev0 == __rev1);
6657		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	6675	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	6676	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	6677	__ret = __builtin_bit_cast(uint32x4_t, __rev0 == __rev1);
	6678	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
6658	6679	return __ret;
6659	6680	}
6660	6681	#endif
...	...	@@ -6662,16 +6683,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vceqq_f32(float32x4_t __p0, floa
6662	6683	#ifdef __LITTLE_ENDIAN__
6663	6684	__ai __attribute__((target("neon"))) uint32x4_t vceqq_s32(int32x4_t __p0, int32x4_t __p1) {
6664	6685	uint32x4_t __ret;
6665		__ret = (uint32x4_t)(__p0 == __p1);
	6686	__ret = __builtin_bit_cast(uint32x4_t, __p0 == __p1);
6666	6687	return __ret;
6667	6688	}
6668	6689	#else
6669	6690	__ai __attribute__((target("neon"))) uint32x4_t vceqq_s32(int32x4_t __p0, int32x4_t __p1) {
6670	6691	uint32x4_t __ret;
6671		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
6672		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
6673		__ret = (uint32x4_t)(__rev0 == __rev1);
6674		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	6692	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	6693	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	6694	__ret = __builtin_bit_cast(uint32x4_t, __rev0 == __rev1);
	6695	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
6675	6696	return __ret;
6676	6697	}
6677	6698	#endif
...	...	@@ -6679,16 +6700,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vceqq_s32(int32x4_t __p0, int32x
6679	6700	#ifdef __LITTLE_ENDIAN__
6680	6701	__ai __attribute__((target("neon"))) uint16x8_t vceqq_s16(int16x8_t __p0, int16x8_t __p1) {
6681	6702	uint16x8_t __ret;
6682		__ret = (uint16x8_t)(__p0 == __p1);
	6703	__ret = __builtin_bit_cast(uint16x8_t, __p0 == __p1);
6683	6704	return __ret;
6684	6705	}
6685	6706	#else
6686	6707	__ai __attribute__((target("neon"))) uint16x8_t vceqq_s16(int16x8_t __p0, int16x8_t __p1) {
6687	6708	uint16x8_t __ret;
6688		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
6689		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
6690		__ret = (uint16x8_t)(__rev0 == __rev1);
6691		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	6709	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	6710	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	6711	__ret = __builtin_bit_cast(uint16x8_t, __rev0 == __rev1);
	6712	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
6692	6713	return __ret;
6693	6714	}
6694	6715	#endif
...	...	@@ -6696,16 +6717,16 @@ __ai __attribute__((target("neon"))) uint16x8_t vceqq_s16(int16x8_t __p0, int16x
6696	6717	#ifdef __LITTLE_ENDIAN__
6697	6718	__ai __attribute__((target("neon"))) uint8x8_t vceq_u8(uint8x8_t __p0, uint8x8_t __p1) {
6698	6719	uint8x8_t __ret;
6699		__ret = (uint8x8_t)(__p0 == __p1);
	6720	__ret = __builtin_bit_cast(uint8x8_t, __p0 == __p1);
6700	6721	return __ret;
6701	6722	}
6702	6723	#else
6703	6724	__ai __attribute__((target("neon"))) uint8x8_t vceq_u8(uint8x8_t __p0, uint8x8_t __p1) {
6704	6725	uint8x8_t __ret;
6705		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
6706		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
6707		__ret = (uint8x8_t)(__rev0 == __rev1);
6708		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	6726	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	6727	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	6728	__ret = __builtin_bit_cast(uint8x8_t, __rev0 == __rev1);
	6729	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
6709	6730	return __ret;
6710	6731	}
6711	6732	#endif
...	...	@@ -6713,16 +6734,16 @@ __ai __attribute__((target("neon"))) uint8x8_t vceq_u8(uint8x8_t __p0, uint8x8_t
6713	6734	#ifdef __LITTLE_ENDIAN__
6714	6735	__ai __attribute__((target("neon"))) uint32x2_t vceq_u32(uint32x2_t __p0, uint32x2_t __p1) {
6715	6736	uint32x2_t __ret;
6716		__ret = (uint32x2_t)(__p0 == __p1);
	6737	__ret = __builtin_bit_cast(uint32x2_t, __p0 == __p1);
6717	6738	return __ret;
6718	6739	}
6719	6740	#else
6720	6741	__ai __attribute__((target("neon"))) uint32x2_t vceq_u32(uint32x2_t __p0, uint32x2_t __p1) {
6721	6742	uint32x2_t __ret;
6722		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
6723		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
6724		__ret = (uint32x2_t)(__rev0 == __rev1);
6725		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	6743	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	6744	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	6745	__ret = __builtin_bit_cast(uint32x2_t, __rev0 == __rev1);
	6746	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
6726	6747	return __ret;
6727	6748	}
6728	6749	#endif
...	...	@@ -6730,16 +6751,16 @@ __ai __attribute__((target("neon"))) uint32x2_t vceq_u32(uint32x2_t __p0, uint32
6730	6751	#ifdef __LITTLE_ENDIAN__
6731	6752	__ai __attribute__((target("neon"))) uint16x4_t vceq_u16(uint16x4_t __p0, uint16x4_t __p1) {
6732	6753	uint16x4_t __ret;
6733		__ret = (uint16x4_t)(__p0 == __p1);
	6754	__ret = __builtin_bit_cast(uint16x4_t, __p0 == __p1);
6734	6755	return __ret;
6735	6756	}
6736	6757	#else
6737	6758	__ai __attribute__((target("neon"))) uint16x4_t vceq_u16(uint16x4_t __p0, uint16x4_t __p1) {
6738	6759	uint16x4_t __ret;
6739		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
6740		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
6741		__ret = (uint16x4_t)(__rev0 == __rev1);
6742		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	6760	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	6761	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	6762	__ret = __builtin_bit_cast(uint16x4_t, __rev0 == __rev1);
	6763	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
6743	6764	return __ret;
6744	6765	}
6745	6766	#endif
...	...	@@ -6747,16 +6768,16 @@ __ai __attribute__((target("neon"))) uint16x4_t vceq_u16(uint16x4_t __p0, uint16
6747	6768	#ifdef __LITTLE_ENDIAN__
6748	6769	__ai __attribute__((target("neon"))) uint8x8_t vceq_s8(int8x8_t __p0, int8x8_t __p1) {
6749	6770	uint8x8_t __ret;
6750		__ret = (uint8x8_t)(__p0 == __p1);
	6771	__ret = __builtin_bit_cast(uint8x8_t, __p0 == __p1);
6751	6772	return __ret;
6752	6773	}
6753	6774	#else
6754	6775	__ai __attribute__((target("neon"))) uint8x8_t vceq_s8(int8x8_t __p0, int8x8_t __p1) {
6755	6776	uint8x8_t __ret;
6756		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
6757		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
6758		__ret = (uint8x8_t)(__rev0 == __rev1);
6759		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	6777	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	6778	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	6779	__ret = __builtin_bit_cast(uint8x8_t, __rev0 == __rev1);
	6780	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
6760	6781	return __ret;
6761	6782	}
6762	6783	#endif
...	...	@@ -6764,16 +6785,16 @@ __ai __attribute__((target("neon"))) uint8x8_t vceq_s8(int8x8_t __p0, int8x8_t _
6764	6785	#ifdef __LITTLE_ENDIAN__
6765	6786	__ai __attribute__((target("neon"))) uint32x2_t vceq_f32(float32x2_t __p0, float32x2_t __p1) {
6766	6787	uint32x2_t __ret;
6767		__ret = (uint32x2_t)(__p0 == __p1);
	6788	__ret = __builtin_bit_cast(uint32x2_t, __p0 == __p1);
6768	6789	return __ret;
6769	6790	}
6770	6791	#else
6771	6792	__ai __attribute__((target("neon"))) uint32x2_t vceq_f32(float32x2_t __p0, float32x2_t __p1) {
6772	6793	uint32x2_t __ret;
6773		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
6774		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
6775		__ret = (uint32x2_t)(__rev0 == __rev1);
6776		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	6794	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	6795	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	6796	__ret = __builtin_bit_cast(uint32x2_t, __rev0 == __rev1);
	6797	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
6777	6798	return __ret;
6778	6799	}
6779	6800	#endif
...	...	@@ -6781,16 +6802,16 @@ __ai __attribute__((target("neon"))) uint32x2_t vceq_f32(float32x2_t __p0, float
6781	6802	#ifdef __LITTLE_ENDIAN__
6782	6803	__ai __attribute__((target("neon"))) uint32x2_t vceq_s32(int32x2_t __p0, int32x2_t __p1) {
6783	6804	uint32x2_t __ret;
6784		__ret = (uint32x2_t)(__p0 == __p1);
	6805	__ret = __builtin_bit_cast(uint32x2_t, __p0 == __p1);
6785	6806	return __ret;
6786	6807	}
6787	6808	#else
6788	6809	__ai __attribute__((target("neon"))) uint32x2_t vceq_s32(int32x2_t __p0, int32x2_t __p1) {
6789	6810	uint32x2_t __ret;
6790		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
6791		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
6792		__ret = (uint32x2_t)(__rev0 == __rev1);
6793		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	6811	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	6812	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	6813	__ret = __builtin_bit_cast(uint32x2_t, __rev0 == __rev1);
	6814	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
6794	6815	return __ret;
6795	6816	}
6796	6817	#endif
...	...	@@ -6798,16 +6819,16 @@ __ai __attribute__((target("neon"))) uint32x2_t vceq_s32(int32x2_t __p0, int32x2
6798	6819	#ifdef __LITTLE_ENDIAN__
6799	6820	__ai __attribute__((target("neon"))) uint16x4_t vceq_s16(int16x4_t __p0, int16x4_t __p1) {
6800	6821	uint16x4_t __ret;
6801		__ret = (uint16x4_t)(__p0 == __p1);
	6822	__ret = __builtin_bit_cast(uint16x4_t, __p0 == __p1);
6802	6823	return __ret;
6803	6824	}
6804	6825	#else
6805	6826	__ai __attribute__((target("neon"))) uint16x4_t vceq_s16(int16x4_t __p0, int16x4_t __p1) {
6806	6827	uint16x4_t __ret;
6807		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
6808		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
6809		__ret = (uint16x4_t)(__rev0 == __rev1);
6810		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	6828	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	6829	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	6830	__ret = __builtin_bit_cast(uint16x4_t, __rev0 == __rev1);
	6831	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
6811	6832	return __ret;
6812	6833	}
6813	6834	#endif
...	...	@@ -6815,16 +6836,16 @@ __ai __attribute__((target("neon"))) uint16x4_t vceq_s16(int16x4_t __p0, int16x4
6815	6836	#ifdef __LITTLE_ENDIAN__
6816	6837	__ai __attribute__((target("neon"))) uint8x16_t vcgeq_u8(uint8x16_t __p0, uint8x16_t __p1) {
6817	6838	uint8x16_t __ret;
6818		__ret = (uint8x16_t)(__p0 >= __p1);
	6839	__ret = __builtin_bit_cast(uint8x16_t, __p0 >= __p1);
6819	6840	return __ret;
6820	6841	}
6821	6842	#else
6822	6843	__ai __attribute__((target("neon"))) uint8x16_t vcgeq_u8(uint8x16_t __p0, uint8x16_t __p1) {
6823	6844	uint8x16_t __ret;
6824		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
6825		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
6826		__ret = (uint8x16_t)(__rev0 >= __rev1);
6827		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	6845	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	6846	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	6847	__ret = __builtin_bit_cast(uint8x16_t, __rev0 >= __rev1);
	6848	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
6828	6849	return __ret;
6829	6850	}
6830	6851	#endif
...	...	@@ -6832,16 +6853,16 @@ __ai __attribute__((target("neon"))) uint8x16_t vcgeq_u8(uint8x16_t __p0, uint8x
6832	6853	#ifdef __LITTLE_ENDIAN__
6833	6854	__ai __attribute__((target("neon"))) uint32x4_t vcgeq_u32(uint32x4_t __p0, uint32x4_t __p1) {
6834	6855	uint32x4_t __ret;
6835		__ret = (uint32x4_t)(__p0 >= __p1);
	6856	__ret = __builtin_bit_cast(uint32x4_t, __p0 >= __p1);
6836	6857	return __ret;
6837	6858	}
6838	6859	#else
6839	6860	__ai __attribute__((target("neon"))) uint32x4_t vcgeq_u32(uint32x4_t __p0, uint32x4_t __p1) {
6840	6861	uint32x4_t __ret;
6841		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
6842		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
6843		__ret = (uint32x4_t)(__rev0 >= __rev1);
6844		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	6862	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	6863	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	6864	__ret = __builtin_bit_cast(uint32x4_t, __rev0 >= __rev1);
	6865	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
6845	6866	return __ret;
6846	6867	}
6847	6868	#endif
...	...	@@ -6849,16 +6870,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vcgeq_u32(uint32x4_t __p0, uint3
6849	6870	#ifdef __LITTLE_ENDIAN__
6850	6871	__ai __attribute__((target("neon"))) uint16x8_t vcgeq_u16(uint16x8_t __p0, uint16x8_t __p1) {
6851	6872	uint16x8_t __ret;
6852		__ret = (uint16x8_t)(__p0 >= __p1);
	6873	__ret = __builtin_bit_cast(uint16x8_t, __p0 >= __p1);
6853	6874	return __ret;
6854	6875	}
6855	6876	#else
6856	6877	__ai __attribute__((target("neon"))) uint16x8_t vcgeq_u16(uint16x8_t __p0, uint16x8_t __p1) {
6857	6878	uint16x8_t __ret;
6858		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
6859		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
6860		__ret = (uint16x8_t)(__rev0 >= __rev1);
6861		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	6879	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	6880	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	6881	__ret = __builtin_bit_cast(uint16x8_t, __rev0 >= __rev1);
	6882	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
6862	6883	return __ret;
6863	6884	}
6864	6885	#endif
...	...	@@ -6866,16 +6887,16 @@ __ai __attribute__((target("neon"))) uint16x8_t vcgeq_u16(uint16x8_t __p0, uint1
6866	6887	#ifdef __LITTLE_ENDIAN__
6867	6888	__ai __attribute__((target("neon"))) uint8x16_t vcgeq_s8(int8x16_t __p0, int8x16_t __p1) {
6868	6889	uint8x16_t __ret;
6869		__ret = (uint8x16_t)(__p0 >= __p1);
	6890	__ret = __builtin_bit_cast(uint8x16_t, __p0 >= __p1);
6870	6891	return __ret;
6871	6892	}
6872	6893	#else
6873	6894	__ai __attribute__((target("neon"))) uint8x16_t vcgeq_s8(int8x16_t __p0, int8x16_t __p1) {
6874	6895	uint8x16_t __ret;
6875		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
6876		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
6877		__ret = (uint8x16_t)(__rev0 >= __rev1);
6878		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	6896	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	6897	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	6898	__ret = __builtin_bit_cast(uint8x16_t, __rev0 >= __rev1);
	6899	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
6879	6900	return __ret;
6880	6901	}
6881	6902	#endif
...	...	@@ -6883,16 +6904,16 @@ __ai __attribute__((target("neon"))) uint8x16_t vcgeq_s8(int8x16_t __p0, int8x16
6883	6904	#ifdef __LITTLE_ENDIAN__
6884	6905	__ai __attribute__((target("neon"))) uint32x4_t vcgeq_f32(float32x4_t __p0, float32x4_t __p1) {
6885	6906	uint32x4_t __ret;
6886		__ret = (uint32x4_t)(__p0 >= __p1);
	6907	__ret = __builtin_bit_cast(uint32x4_t, __p0 >= __p1);
6887	6908	return __ret;
6888	6909	}
6889	6910	#else
6890	6911	__ai __attribute__((target("neon"))) uint32x4_t vcgeq_f32(float32x4_t __p0, float32x4_t __p1) {
6891	6912	uint32x4_t __ret;
6892		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
6893		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
6894		__ret = (uint32x4_t)(__rev0 >= __rev1);
6895		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	6913	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	6914	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	6915	__ret = __builtin_bit_cast(uint32x4_t, __rev0 >= __rev1);
	6916	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
6896	6917	return __ret;
6897	6918	}
6898	6919	#endif
...	...	@@ -6900,16 +6921,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vcgeq_f32(float32x4_t __p0, floa
6900	6921	#ifdef __LITTLE_ENDIAN__
6901	6922	__ai __attribute__((target("neon"))) uint32x4_t vcgeq_s32(int32x4_t __p0, int32x4_t __p1) {
6902	6923	uint32x4_t __ret;
6903		__ret = (uint32x4_t)(__p0 >= __p1);
	6924	__ret = __builtin_bit_cast(uint32x4_t, __p0 >= __p1);
6904	6925	return __ret;
6905	6926	}
6906	6927	#else
6907	6928	__ai __attribute__((target("neon"))) uint32x4_t vcgeq_s32(int32x4_t __p0, int32x4_t __p1) {
6908	6929	uint32x4_t __ret;
6909		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
6910		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
6911		__ret = (uint32x4_t)(__rev0 >= __rev1);
6912		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	6930	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	6931	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	6932	__ret = __builtin_bit_cast(uint32x4_t, __rev0 >= __rev1);
	6933	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
6913	6934	return __ret;
6914	6935	}
6915	6936	#endif
...	...	@@ -6917,16 +6938,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vcgeq_s32(int32x4_t __p0, int32x
6917	6938	#ifdef __LITTLE_ENDIAN__
6918	6939	__ai __attribute__((target("neon"))) uint16x8_t vcgeq_s16(int16x8_t __p0, int16x8_t __p1) {
6919	6940	uint16x8_t __ret;
6920		__ret = (uint16x8_t)(__p0 >= __p1);
	6941	__ret = __builtin_bit_cast(uint16x8_t, __p0 >= __p1);
6921	6942	return __ret;
6922	6943	}
6923	6944	#else
6924	6945	__ai __attribute__((target("neon"))) uint16x8_t vcgeq_s16(int16x8_t __p0, int16x8_t __p1) {
6925	6946	uint16x8_t __ret;
6926		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
6927		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
6928		__ret = (uint16x8_t)(__rev0 >= __rev1);
6929		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	6947	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	6948	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	6949	__ret = __builtin_bit_cast(uint16x8_t, __rev0 >= __rev1);
	6950	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
6930	6951	return __ret;
6931	6952	}
6932	6953	#endif
...	...	@@ -6934,16 +6955,16 @@ __ai __attribute__((target("neon"))) uint16x8_t vcgeq_s16(int16x8_t __p0, int16x
6934	6955	#ifdef __LITTLE_ENDIAN__
6935	6956	__ai __attribute__((target("neon"))) uint8x8_t vcge_u8(uint8x8_t __p0, uint8x8_t __p1) {
6936	6957	uint8x8_t __ret;
6937		__ret = (uint8x8_t)(__p0 >= __p1);
	6958	__ret = __builtin_bit_cast(uint8x8_t, __p0 >= __p1);
6938	6959	return __ret;
6939	6960	}
6940	6961	#else
6941	6962	__ai __attribute__((target("neon"))) uint8x8_t vcge_u8(uint8x8_t __p0, uint8x8_t __p1) {
6942	6963	uint8x8_t __ret;
6943		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
6944		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
6945		__ret = (uint8x8_t)(__rev0 >= __rev1);
6946		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	6964	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	6965	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	6966	__ret = __builtin_bit_cast(uint8x8_t, __rev0 >= __rev1);
	6967	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
6947	6968	return __ret;
6948	6969	}
6949	6970	#endif
...	...	@@ -6951,16 +6972,16 @@ __ai __attribute__((target("neon"))) uint8x8_t vcge_u8(uint8x8_t __p0, uint8x8_t
6951	6972	#ifdef __LITTLE_ENDIAN__
6952	6973	__ai __attribute__((target("neon"))) uint32x2_t vcge_u32(uint32x2_t __p0, uint32x2_t __p1) {
6953	6974	uint32x2_t __ret;
6954		__ret = (uint32x2_t)(__p0 >= __p1);
	6975	__ret = __builtin_bit_cast(uint32x2_t, __p0 >= __p1);
6955	6976	return __ret;
6956	6977	}
6957	6978	#else
6958	6979	__ai __attribute__((target("neon"))) uint32x2_t vcge_u32(uint32x2_t __p0, uint32x2_t __p1) {
6959	6980	uint32x2_t __ret;
6960		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
6961		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
6962		__ret = (uint32x2_t)(__rev0 >= __rev1);
6963		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	6981	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	6982	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	6983	__ret = __builtin_bit_cast(uint32x2_t, __rev0 >= __rev1);
	6984	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
6964	6985	return __ret;
6965	6986	}
6966	6987	#endif
...	...	@@ -6968,16 +6989,16 @@ __ai __attribute__((target("neon"))) uint32x2_t vcge_u32(uint32x2_t __p0, uint32
6968	6989	#ifdef __LITTLE_ENDIAN__
6969	6990	__ai __attribute__((target("neon"))) uint16x4_t vcge_u16(uint16x4_t __p0, uint16x4_t __p1) {
6970	6991	uint16x4_t __ret;
6971		__ret = (uint16x4_t)(__p0 >= __p1);
	6992	__ret = __builtin_bit_cast(uint16x4_t, __p0 >= __p1);
6972	6993	return __ret;
6973	6994	}
6974	6995	#else
6975	6996	__ai __attribute__((target("neon"))) uint16x4_t vcge_u16(uint16x4_t __p0, uint16x4_t __p1) {
6976	6997	uint16x4_t __ret;
6977		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
6978		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
6979		__ret = (uint16x4_t)(__rev0 >= __rev1);
6980		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	6998	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	6999	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	7000	__ret = __builtin_bit_cast(uint16x4_t, __rev0 >= __rev1);
	7001	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
6981	7002	return __ret;
6982	7003	}
6983	7004	#endif
...	...	@@ -6985,16 +7006,16 @@ __ai __attribute__((target("neon"))) uint16x4_t vcge_u16(uint16x4_t __p0, uint16
6985	7006	#ifdef __LITTLE_ENDIAN__
6986	7007	__ai __attribute__((target("neon"))) uint8x8_t vcge_s8(int8x8_t __p0, int8x8_t __p1) {
6987	7008	uint8x8_t __ret;
6988		__ret = (uint8x8_t)(__p0 >= __p1);
	7009	__ret = __builtin_bit_cast(uint8x8_t, __p0 >= __p1);
6989	7010	return __ret;
6990	7011	}
6991	7012	#else
6992	7013	__ai __attribute__((target("neon"))) uint8x8_t vcge_s8(int8x8_t __p0, int8x8_t __p1) {
6993	7014	uint8x8_t __ret;
6994		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
6995		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
6996		__ret = (uint8x8_t)(__rev0 >= __rev1);
6997		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	7015	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	7016	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	7017	__ret = __builtin_bit_cast(uint8x8_t, __rev0 >= __rev1);
	7018	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
6998	7019	return __ret;
6999	7020	}
7000	7021	#endif
...	...	@@ -7002,16 +7023,16 @@ __ai __attribute__((target("neon"))) uint8x8_t vcge_s8(int8x8_t __p0, int8x8_t _
7002	7023	#ifdef __LITTLE_ENDIAN__
7003	7024	__ai __attribute__((target("neon"))) uint32x2_t vcge_f32(float32x2_t __p0, float32x2_t __p1) {
7004	7025	uint32x2_t __ret;
7005		__ret = (uint32x2_t)(__p0 >= __p1);
	7026	__ret = __builtin_bit_cast(uint32x2_t, __p0 >= __p1);
7006	7027	return __ret;
7007	7028	}
7008	7029	#else
7009	7030	__ai __attribute__((target("neon"))) uint32x2_t vcge_f32(float32x2_t __p0, float32x2_t __p1) {
7010	7031	uint32x2_t __ret;
7011		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
7012		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
7013		__ret = (uint32x2_t)(__rev0 >= __rev1);
7014		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	7032	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	7033	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	7034	__ret = __builtin_bit_cast(uint32x2_t, __rev0 >= __rev1);
	7035	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
7015	7036	return __ret;
7016	7037	}
7017	7038	#endif
...	...	@@ -7019,16 +7040,16 @@ __ai __attribute__((target("neon"))) uint32x2_t vcge_f32(float32x2_t __p0, float
7019	7040	#ifdef __LITTLE_ENDIAN__
7020	7041	__ai __attribute__((target("neon"))) uint32x2_t vcge_s32(int32x2_t __p0, int32x2_t __p1) {
7021	7042	uint32x2_t __ret;
7022		__ret = (uint32x2_t)(__p0 >= __p1);
	7043	__ret = __builtin_bit_cast(uint32x2_t, __p0 >= __p1);
7023	7044	return __ret;
7024	7045	}
7025	7046	#else
7026	7047	__ai __attribute__((target("neon"))) uint32x2_t vcge_s32(int32x2_t __p0, int32x2_t __p1) {
7027	7048	uint32x2_t __ret;
7028		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
7029		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
7030		__ret = (uint32x2_t)(__rev0 >= __rev1);
7031		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	7049	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	7050	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	7051	__ret = __builtin_bit_cast(uint32x2_t, __rev0 >= __rev1);
	7052	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
7032	7053	return __ret;
7033	7054	}
7034	7055	#endif
...	...	@@ -7036,16 +7057,16 @@ __ai __attribute__((target("neon"))) uint32x2_t vcge_s32(int32x2_t __p0, int32x2
7036	7057	#ifdef __LITTLE_ENDIAN__
7037	7058	__ai __attribute__((target("neon"))) uint16x4_t vcge_s16(int16x4_t __p0, int16x4_t __p1) {
7038	7059	uint16x4_t __ret;
7039		__ret = (uint16x4_t)(__p0 >= __p1);
	7060	__ret = __builtin_bit_cast(uint16x4_t, __p0 >= __p1);
7040	7061	return __ret;
7041	7062	}
7042	7063	#else
7043	7064	__ai __attribute__((target("neon"))) uint16x4_t vcge_s16(int16x4_t __p0, int16x4_t __p1) {
7044	7065	uint16x4_t __ret;
7045		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
7046		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
7047		__ret = (uint16x4_t)(__rev0 >= __rev1);
7048		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	7066	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	7067	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	7068	__ret = __builtin_bit_cast(uint16x4_t, __rev0 >= __rev1);
	7069	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
7049	7070	return __ret;
7050	7071	}
7051	7072	#endif
...	...	@@ -7053,16 +7074,16 @@ __ai __attribute__((target("neon"))) uint16x4_t vcge_s16(int16x4_t __p0, int16x4
7053	7074	#ifdef __LITTLE_ENDIAN__
7054	7075	__ai __attribute__((target("neon"))) uint8x16_t vcgtq_u8(uint8x16_t __p0, uint8x16_t __p1) {
7055	7076	uint8x16_t __ret;
7056		__ret = (uint8x16_t)(__p0 > __p1);
	7077	__ret = __builtin_bit_cast(uint8x16_t, __p0 > __p1);
7057	7078	return __ret;
7058	7079	}
7059	7080	#else
7060	7081	__ai __attribute__((target("neon"))) uint8x16_t vcgtq_u8(uint8x16_t __p0, uint8x16_t __p1) {
7061	7082	uint8x16_t __ret;
7062		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
7063		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
7064		__ret = (uint8x16_t)(__rev0 > __rev1);
7065		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	7083	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	7084	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	7085	__ret = __builtin_bit_cast(uint8x16_t, __rev0 > __rev1);
	7086	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
7066	7087	return __ret;
7067	7088	}
7068	7089	#endif
...	...	@@ -7070,16 +7091,16 @@ __ai __attribute__((target("neon"))) uint8x16_t vcgtq_u8(uint8x16_t __p0, uint8x
7070	7091	#ifdef __LITTLE_ENDIAN__
7071	7092	__ai __attribute__((target("neon"))) uint32x4_t vcgtq_u32(uint32x4_t __p0, uint32x4_t __p1) {
7072	7093	uint32x4_t __ret;
7073		__ret = (uint32x4_t)(__p0 > __p1);
	7094	__ret = __builtin_bit_cast(uint32x4_t, __p0 > __p1);
7074	7095	return __ret;
7075	7096	}
7076	7097	#else
7077	7098	__ai __attribute__((target("neon"))) uint32x4_t vcgtq_u32(uint32x4_t __p0, uint32x4_t __p1) {
7078	7099	uint32x4_t __ret;
7079		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
7080		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
7081		__ret = (uint32x4_t)(__rev0 > __rev1);
7082		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	7100	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	7101	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	7102	__ret = __builtin_bit_cast(uint32x4_t, __rev0 > __rev1);
	7103	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
7083	7104	return __ret;
7084	7105	}
7085	7106	#endif
...	...	@@ -7087,16 +7108,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vcgtq_u32(uint32x4_t __p0, uint3
7087	7108	#ifdef __LITTLE_ENDIAN__
7088	7109	__ai __attribute__((target("neon"))) uint16x8_t vcgtq_u16(uint16x8_t __p0, uint16x8_t __p1) {
7089	7110	uint16x8_t __ret;
7090		__ret = (uint16x8_t)(__p0 > __p1);
	7111	__ret = __builtin_bit_cast(uint16x8_t, __p0 > __p1);
7091	7112	return __ret;
7092	7113	}
7093	7114	#else
7094	7115	__ai __attribute__((target("neon"))) uint16x8_t vcgtq_u16(uint16x8_t __p0, uint16x8_t __p1) {
7095	7116	uint16x8_t __ret;
7096		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
7097		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
7098		__ret = (uint16x8_t)(__rev0 > __rev1);
7099		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	7117	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	7118	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	7119	__ret = __builtin_bit_cast(uint16x8_t, __rev0 > __rev1);
	7120	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
7100	7121	return __ret;
7101	7122	}
7102	7123	#endif
...	...	@@ -7104,16 +7125,16 @@ __ai __attribute__((target("neon"))) uint16x8_t vcgtq_u16(uint16x8_t __p0, uint1
7104	7125	#ifdef __LITTLE_ENDIAN__
7105	7126	__ai __attribute__((target("neon"))) uint8x16_t vcgtq_s8(int8x16_t __p0, int8x16_t __p1) {
7106	7127	uint8x16_t __ret;
7107		__ret = (uint8x16_t)(__p0 > __p1);
	7128	__ret = __builtin_bit_cast(uint8x16_t, __p0 > __p1);
7108	7129	return __ret;
7109	7130	}
7110	7131	#else
7111	7132	__ai __attribute__((target("neon"))) uint8x16_t vcgtq_s8(int8x16_t __p0, int8x16_t __p1) {
7112	7133	uint8x16_t __ret;
7113		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
7114		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
7115		__ret = (uint8x16_t)(__rev0 > __rev1);
7116		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	7134	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	7135	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	7136	__ret = __builtin_bit_cast(uint8x16_t, __rev0 > __rev1);
	7137	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
7117	7138	return __ret;
7118	7139	}
7119	7140	#endif
...	...	@@ -7121,16 +7142,16 @@ __ai __attribute__((target("neon"))) uint8x16_t vcgtq_s8(int8x16_t __p0, int8x16
7121	7142	#ifdef __LITTLE_ENDIAN__
7122	7143	__ai __attribute__((target("neon"))) uint32x4_t vcgtq_f32(float32x4_t __p0, float32x4_t __p1) {
7123	7144	uint32x4_t __ret;
7124		__ret = (uint32x4_t)(__p0 > __p1);
	7145	__ret = __builtin_bit_cast(uint32x4_t, __p0 > __p1);
7125	7146	return __ret;
7126	7147	}
7127	7148	#else
7128	7149	__ai __attribute__((target("neon"))) uint32x4_t vcgtq_f32(float32x4_t __p0, float32x4_t __p1) {
7129	7150	uint32x4_t __ret;
7130		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
7131		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
7132		__ret = (uint32x4_t)(__rev0 > __rev1);
7133		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	7151	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	7152	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	7153	__ret = __builtin_bit_cast(uint32x4_t, __rev0 > __rev1);
	7154	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
7134	7155	return __ret;
7135	7156	}
7136	7157	#endif
...	...	@@ -7138,16 +7159,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vcgtq_f32(float32x4_t __p0, floa
7138	7159	#ifdef __LITTLE_ENDIAN__
7139	7160	__ai __attribute__((target("neon"))) uint32x4_t vcgtq_s32(int32x4_t __p0, int32x4_t __p1) {
7140	7161	uint32x4_t __ret;
7141		__ret = (uint32x4_t)(__p0 > __p1);
	7162	__ret = __builtin_bit_cast(uint32x4_t, __p0 > __p1);
7142	7163	return __ret;
7143	7164	}
7144	7165	#else
7145	7166	__ai __attribute__((target("neon"))) uint32x4_t vcgtq_s32(int32x4_t __p0, int32x4_t __p1) {
7146	7167	uint32x4_t __ret;
7147		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
7148		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
7149		__ret = (uint32x4_t)(__rev0 > __rev1);
7150		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	7168	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	7169	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	7170	__ret = __builtin_bit_cast(uint32x4_t, __rev0 > __rev1);
	7171	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
7151	7172	return __ret;
7152	7173	}
7153	7174	#endif
...	...	@@ -7155,16 +7176,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vcgtq_s32(int32x4_t __p0, int32x
7155	7176	#ifdef __LITTLE_ENDIAN__
7156	7177	__ai __attribute__((target("neon"))) uint16x8_t vcgtq_s16(int16x8_t __p0, int16x8_t __p1) {
7157	7178	uint16x8_t __ret;
7158		__ret = (uint16x8_t)(__p0 > __p1);
	7179	__ret = __builtin_bit_cast(uint16x8_t, __p0 > __p1);
7159	7180	return __ret;
7160	7181	}
7161	7182	#else
7162	7183	__ai __attribute__((target("neon"))) uint16x8_t vcgtq_s16(int16x8_t __p0, int16x8_t __p1) {
7163	7184	uint16x8_t __ret;
7164		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
7165		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
7166		__ret = (uint16x8_t)(__rev0 > __rev1);
7167		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	7185	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	7186	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	7187	__ret = __builtin_bit_cast(uint16x8_t, __rev0 > __rev1);
	7188	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
7168	7189	return __ret;
7169	7190	}
7170	7191	#endif
...	...	@@ -7172,16 +7193,16 @@ __ai __attribute__((target("neon"))) uint16x8_t vcgtq_s16(int16x8_t __p0, int16x
7172	7193	#ifdef __LITTLE_ENDIAN__
7173	7194	__ai __attribute__((target("neon"))) uint8x8_t vcgt_u8(uint8x8_t __p0, uint8x8_t __p1) {
7174	7195	uint8x8_t __ret;
7175		__ret = (uint8x8_t)(__p0 > __p1);
	7196	__ret = __builtin_bit_cast(uint8x8_t, __p0 > __p1);
7176	7197	return __ret;
7177	7198	}
7178	7199	#else
7179	7200	__ai __attribute__((target("neon"))) uint8x8_t vcgt_u8(uint8x8_t __p0, uint8x8_t __p1) {
7180	7201	uint8x8_t __ret;
7181		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
7182		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
7183		__ret = (uint8x8_t)(__rev0 > __rev1);
7184		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	7202	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	7203	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	7204	__ret = __builtin_bit_cast(uint8x8_t, __rev0 > __rev1);
	7205	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
7185	7206	return __ret;
7186	7207	}
7187	7208	#endif
...	...	@@ -7189,16 +7210,16 @@ __ai __attribute__((target("neon"))) uint8x8_t vcgt_u8(uint8x8_t __p0, uint8x8_t
7189	7210	#ifdef __LITTLE_ENDIAN__
7190	7211	__ai __attribute__((target("neon"))) uint32x2_t vcgt_u32(uint32x2_t __p0, uint32x2_t __p1) {
7191	7212	uint32x2_t __ret;
7192		__ret = (uint32x2_t)(__p0 > __p1);
	7213	__ret = __builtin_bit_cast(uint32x2_t, __p0 > __p1);
7193	7214	return __ret;
7194	7215	}
7195	7216	#else
7196	7217	__ai __attribute__((target("neon"))) uint32x2_t vcgt_u32(uint32x2_t __p0, uint32x2_t __p1) {
7197	7218	uint32x2_t __ret;
7198		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
7199		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
7200		__ret = (uint32x2_t)(__rev0 > __rev1);
7201		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	7219	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	7220	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	7221	__ret = __builtin_bit_cast(uint32x2_t, __rev0 > __rev1);
	7222	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
7202	7223	return __ret;
7203	7224	}
7204	7225	#endif
...	...	@@ -7206,16 +7227,16 @@ __ai __attribute__((target("neon"))) uint32x2_t vcgt_u32(uint32x2_t __p0, uint32
7206	7227	#ifdef __LITTLE_ENDIAN__
7207	7228	__ai __attribute__((target("neon"))) uint16x4_t vcgt_u16(uint16x4_t __p0, uint16x4_t __p1) {
7208	7229	uint16x4_t __ret;
7209		__ret = (uint16x4_t)(__p0 > __p1);
	7230	__ret = __builtin_bit_cast(uint16x4_t, __p0 > __p1);
7210	7231	return __ret;
7211	7232	}
7212	7233	#else
7213	7234	__ai __attribute__((target("neon"))) uint16x4_t vcgt_u16(uint16x4_t __p0, uint16x4_t __p1) {
7214	7235	uint16x4_t __ret;
7215		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
7216		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
7217		__ret = (uint16x4_t)(__rev0 > __rev1);
7218		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	7236	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	7237	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	7238	__ret = __builtin_bit_cast(uint16x4_t, __rev0 > __rev1);
	7239	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
7219	7240	return __ret;
7220	7241	}
7221	7242	#endif
...	...	@@ -7223,16 +7244,16 @@ __ai __attribute__((target("neon"))) uint16x4_t vcgt_u16(uint16x4_t __p0, uint16
7223	7244	#ifdef __LITTLE_ENDIAN__
7224	7245	__ai __attribute__((target("neon"))) uint8x8_t vcgt_s8(int8x8_t __p0, int8x8_t __p1) {
7225	7246	uint8x8_t __ret;
7226		__ret = (uint8x8_t)(__p0 > __p1);
	7247	__ret = __builtin_bit_cast(uint8x8_t, __p0 > __p1);
7227	7248	return __ret;
7228	7249	}
7229	7250	#else
7230	7251	__ai __attribute__((target("neon"))) uint8x8_t vcgt_s8(int8x8_t __p0, int8x8_t __p1) {
7231	7252	uint8x8_t __ret;
7232		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
7233		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
7234		__ret = (uint8x8_t)(__rev0 > __rev1);
7235		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	7253	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	7254	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	7255	__ret = __builtin_bit_cast(uint8x8_t, __rev0 > __rev1);
	7256	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
7236	7257	return __ret;
7237	7258	}
7238	7259	#endif
...	...	@@ -7240,16 +7261,16 @@ __ai __attribute__((target("neon"))) uint8x8_t vcgt_s8(int8x8_t __p0, int8x8_t _
7240	7261	#ifdef __LITTLE_ENDIAN__
7241	7262	__ai __attribute__((target("neon"))) uint32x2_t vcgt_f32(float32x2_t __p0, float32x2_t __p1) {
7242	7263	uint32x2_t __ret;
7243		__ret = (uint32x2_t)(__p0 > __p1);
	7264	__ret = __builtin_bit_cast(uint32x2_t, __p0 > __p1);
7244	7265	return __ret;
7245	7266	}
7246	7267	#else
7247	7268	__ai __attribute__((target("neon"))) uint32x2_t vcgt_f32(float32x2_t __p0, float32x2_t __p1) {
7248	7269	uint32x2_t __ret;
7249		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
7250		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
7251		__ret = (uint32x2_t)(__rev0 > __rev1);
7252		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	7270	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	7271	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	7272	__ret = __builtin_bit_cast(uint32x2_t, __rev0 > __rev1);
	7273	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
7253	7274	return __ret;
7254	7275	}
7255	7276	#endif
...	...	@@ -7257,16 +7278,16 @@ __ai __attribute__((target("neon"))) uint32x2_t vcgt_f32(float32x2_t __p0, float
7257	7278	#ifdef __LITTLE_ENDIAN__
7258	7279	__ai __attribute__((target("neon"))) uint32x2_t vcgt_s32(int32x2_t __p0, int32x2_t __p1) {
7259	7280	uint32x2_t __ret;
7260		__ret = (uint32x2_t)(__p0 > __p1);
	7281	__ret = __builtin_bit_cast(uint32x2_t, __p0 > __p1);
7261	7282	return __ret;
7262	7283	}
7263	7284	#else
7264	7285	__ai __attribute__((target("neon"))) uint32x2_t vcgt_s32(int32x2_t __p0, int32x2_t __p1) {
7265	7286	uint32x2_t __ret;
7266		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
7267		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
7268		__ret = (uint32x2_t)(__rev0 > __rev1);
7269		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	7287	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	7288	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	7289	__ret = __builtin_bit_cast(uint32x2_t, __rev0 > __rev1);
	7290	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
7270	7291	return __ret;
7271	7292	}
7272	7293	#endif
...	...	@@ -7274,16 +7295,16 @@ __ai __attribute__((target("neon"))) uint32x2_t vcgt_s32(int32x2_t __p0, int32x2
7274	7295	#ifdef __LITTLE_ENDIAN__
7275	7296	__ai __attribute__((target("neon"))) uint16x4_t vcgt_s16(int16x4_t __p0, int16x4_t __p1) {
7276	7297	uint16x4_t __ret;
7277		__ret = (uint16x4_t)(__p0 > __p1);
	7298	__ret = __builtin_bit_cast(uint16x4_t, __p0 > __p1);
7278	7299	return __ret;
7279	7300	}
7280	7301	#else
7281	7302	__ai __attribute__((target("neon"))) uint16x4_t vcgt_s16(int16x4_t __p0, int16x4_t __p1) {
7282	7303	uint16x4_t __ret;
7283		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
7284		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
7285		__ret = (uint16x4_t)(__rev0 > __rev1);
7286		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	7304	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	7305	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	7306	__ret = __builtin_bit_cast(uint16x4_t, __rev0 > __rev1);
	7307	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
7287	7308	return __ret;
7288	7309	}
7289	7310	#endif
...	...	@@ -7291,16 +7312,16 @@ __ai __attribute__((target("neon"))) uint16x4_t vcgt_s16(int16x4_t __p0, int16x4
7291	7312	#ifdef __LITTLE_ENDIAN__
7292	7313	__ai __attribute__((target("neon"))) uint8x16_t vcleq_u8(uint8x16_t __p0, uint8x16_t __p1) {
7293	7314	uint8x16_t __ret;
7294		__ret = (uint8x16_t)(__p0 <= __p1);
	7315	__ret = __builtin_bit_cast(uint8x16_t, __p0 <= __p1);
7295	7316	return __ret;
7296	7317	}
7297	7318	#else
7298	7319	__ai __attribute__((target("neon"))) uint8x16_t vcleq_u8(uint8x16_t __p0, uint8x16_t __p1) {
7299	7320	uint8x16_t __ret;
7300		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
7301		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
7302		__ret = (uint8x16_t)(__rev0 <= __rev1);
7303		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	7321	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	7322	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	7323	__ret = __builtin_bit_cast(uint8x16_t, __rev0 <= __rev1);
	7324	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
7304	7325	return __ret;
7305	7326	}
7306	7327	#endif
...	...	@@ -7308,16 +7329,16 @@ __ai __attribute__((target("neon"))) uint8x16_t vcleq_u8(uint8x16_t __p0, uint8x
7308	7329	#ifdef __LITTLE_ENDIAN__
7309	7330	__ai __attribute__((target("neon"))) uint32x4_t vcleq_u32(uint32x4_t __p0, uint32x4_t __p1) {
7310	7331	uint32x4_t __ret;
7311		__ret = (uint32x4_t)(__p0 <= __p1);
	7332	__ret = __builtin_bit_cast(uint32x4_t, __p0 <= __p1);
7312	7333	return __ret;
7313	7334	}
7314	7335	#else
7315	7336	__ai __attribute__((target("neon"))) uint32x4_t vcleq_u32(uint32x4_t __p0, uint32x4_t __p1) {
7316	7337	uint32x4_t __ret;
7317		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
7318		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
7319		__ret = (uint32x4_t)(__rev0 <= __rev1);
7320		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	7338	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	7339	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	7340	__ret = __builtin_bit_cast(uint32x4_t, __rev0 <= __rev1);
	7341	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
7321	7342	return __ret;
7322	7343	}
7323	7344	#endif
...	...	@@ -7325,16 +7346,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vcleq_u32(uint32x4_t __p0, uint3
7325	7346	#ifdef __LITTLE_ENDIAN__
7326	7347	__ai __attribute__((target("neon"))) uint16x8_t vcleq_u16(uint16x8_t __p0, uint16x8_t __p1) {
7327	7348	uint16x8_t __ret;
7328		__ret = (uint16x8_t)(__p0 <= __p1);
	7349	__ret = __builtin_bit_cast(uint16x8_t, __p0 <= __p1);
7329	7350	return __ret;
7330	7351	}
7331	7352	#else
7332	7353	__ai __attribute__((target("neon"))) uint16x8_t vcleq_u16(uint16x8_t __p0, uint16x8_t __p1) {
7333	7354	uint16x8_t __ret;
7334		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
7335		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
7336		__ret = (uint16x8_t)(__rev0 <= __rev1);
7337		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	7355	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	7356	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	7357	__ret = __builtin_bit_cast(uint16x8_t, __rev0 <= __rev1);
	7358	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
7338	7359	return __ret;
7339	7360	}
7340	7361	#endif
...	...	@@ -7342,16 +7363,16 @@ __ai __attribute__((target("neon"))) uint16x8_t vcleq_u16(uint16x8_t __p0, uint1
7342	7363	#ifdef __LITTLE_ENDIAN__
7343	7364	__ai __attribute__((target("neon"))) uint8x16_t vcleq_s8(int8x16_t __p0, int8x16_t __p1) {
7344	7365	uint8x16_t __ret;
7345		__ret = (uint8x16_t)(__p0 <= __p1);
	7366	__ret = __builtin_bit_cast(uint8x16_t, __p0 <= __p1);
7346	7367	return __ret;
7347	7368	}
7348	7369	#else
7349	7370	__ai __attribute__((target("neon"))) uint8x16_t vcleq_s8(int8x16_t __p0, int8x16_t __p1) {
7350	7371	uint8x16_t __ret;
7351		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
7352		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
7353		__ret = (uint8x16_t)(__rev0 <= __rev1);
7354		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	7372	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	7373	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	7374	__ret = __builtin_bit_cast(uint8x16_t, __rev0 <= __rev1);
	7375	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
7355	7376	return __ret;
7356	7377	}
7357	7378	#endif
...	...	@@ -7359,16 +7380,16 @@ __ai __attribute__((target("neon"))) uint8x16_t vcleq_s8(int8x16_t __p0, int8x16
7359	7380	#ifdef __LITTLE_ENDIAN__
7360	7381	__ai __attribute__((target("neon"))) uint32x4_t vcleq_f32(float32x4_t __p0, float32x4_t __p1) {
7361	7382	uint32x4_t __ret;
7362		__ret = (uint32x4_t)(__p0 <= __p1);
	7383	__ret = __builtin_bit_cast(uint32x4_t, __p0 <= __p1);
7363	7384	return __ret;
7364	7385	}
7365	7386	#else
7366	7387	__ai __attribute__((target("neon"))) uint32x4_t vcleq_f32(float32x4_t __p0, float32x4_t __p1) {
7367	7388	uint32x4_t __ret;
7368		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
7369		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
7370		__ret = (uint32x4_t)(__rev0 <= __rev1);
7371		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	7389	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	7390	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	7391	__ret = __builtin_bit_cast(uint32x4_t, __rev0 <= __rev1);
	7392	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
7372	7393	return __ret;
7373	7394	}
7374	7395	#endif
...	...	@@ -7376,16 +7397,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vcleq_f32(float32x4_t __p0, floa
7376	7397	#ifdef __LITTLE_ENDIAN__
7377	7398	__ai __attribute__((target("neon"))) uint32x4_t vcleq_s32(int32x4_t __p0, int32x4_t __p1) {
7378	7399	uint32x4_t __ret;
7379		__ret = (uint32x4_t)(__p0 <= __p1);
	7400	__ret = __builtin_bit_cast(uint32x4_t, __p0 <= __p1);
7380	7401	return __ret;
7381	7402	}
7382	7403	#else
7383	7404	__ai __attribute__((target("neon"))) uint32x4_t vcleq_s32(int32x4_t __p0, int32x4_t __p1) {
7384	7405	uint32x4_t __ret;
7385		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
7386		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
7387		__ret = (uint32x4_t)(__rev0 <= __rev1);
7388		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	7406	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	7407	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	7408	__ret = __builtin_bit_cast(uint32x4_t, __rev0 <= __rev1);
	7409	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
7389	7410	return __ret;
7390	7411	}
7391	7412	#endif
...	...	@@ -7393,16 +7414,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vcleq_s32(int32x4_t __p0, int32x
7393	7414	#ifdef __LITTLE_ENDIAN__
7394	7415	__ai __attribute__((target("neon"))) uint16x8_t vcleq_s16(int16x8_t __p0, int16x8_t __p1) {
7395	7416	uint16x8_t __ret;
7396		__ret = (uint16x8_t)(__p0 <= __p1);
	7417	__ret = __builtin_bit_cast(uint16x8_t, __p0 <= __p1);
7397	7418	return __ret;
7398	7419	}
7399	7420	#else
7400	7421	__ai __attribute__((target("neon"))) uint16x8_t vcleq_s16(int16x8_t __p0, int16x8_t __p1) {
7401	7422	uint16x8_t __ret;
7402		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
7403		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
7404		__ret = (uint16x8_t)(__rev0 <= __rev1);
7405		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	7423	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	7424	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	7425	__ret = __builtin_bit_cast(uint16x8_t, __rev0 <= __rev1);
	7426	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
7406	7427	return __ret;
7407	7428	}
7408	7429	#endif
...	...	@@ -7410,16 +7431,16 @@ __ai __attribute__((target("neon"))) uint16x8_t vcleq_s16(int16x8_t __p0, int16x
7410	7431	#ifdef __LITTLE_ENDIAN__
7411	7432	__ai __attribute__((target("neon"))) uint8x8_t vcle_u8(uint8x8_t __p0, uint8x8_t __p1) {
7412	7433	uint8x8_t __ret;
7413		__ret = (uint8x8_t)(__p0 <= __p1);
	7434	__ret = __builtin_bit_cast(uint8x8_t, __p0 <= __p1);
7414	7435	return __ret;
7415	7436	}
7416	7437	#else
7417	7438	__ai __attribute__((target("neon"))) uint8x8_t vcle_u8(uint8x8_t __p0, uint8x8_t __p1) {
7418	7439	uint8x8_t __ret;
7419		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
7420		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
7421		__ret = (uint8x8_t)(__rev0 <= __rev1);
7422		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	7440	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	7441	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	7442	__ret = __builtin_bit_cast(uint8x8_t, __rev0 <= __rev1);
	7443	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
7423	7444	return __ret;
7424	7445	}
7425	7446	#endif
...	...	@@ -7427,16 +7448,16 @@ __ai __attribute__((target("neon"))) uint8x8_t vcle_u8(uint8x8_t __p0, uint8x8_t
7427	7448	#ifdef __LITTLE_ENDIAN__
7428	7449	__ai __attribute__((target("neon"))) uint32x2_t vcle_u32(uint32x2_t __p0, uint32x2_t __p1) {
7429	7450	uint32x2_t __ret;
7430		__ret = (uint32x2_t)(__p0 <= __p1);
	7451	__ret = __builtin_bit_cast(uint32x2_t, __p0 <= __p1);
7431	7452	return __ret;
7432	7453	}
7433	7454	#else
7434	7455	__ai __attribute__((target("neon"))) uint32x2_t vcle_u32(uint32x2_t __p0, uint32x2_t __p1) {
7435	7456	uint32x2_t __ret;
7436		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
7437		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
7438		__ret = (uint32x2_t)(__rev0 <= __rev1);
7439		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	7457	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	7458	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	7459	__ret = __builtin_bit_cast(uint32x2_t, __rev0 <= __rev1);
	7460	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
7440	7461	return __ret;
7441	7462	}
7442	7463	#endif
...	...	@@ -7444,16 +7465,16 @@ __ai __attribute__((target("neon"))) uint32x2_t vcle_u32(uint32x2_t __p0, uint32
7444	7465	#ifdef __LITTLE_ENDIAN__
7445	7466	__ai __attribute__((target("neon"))) uint16x4_t vcle_u16(uint16x4_t __p0, uint16x4_t __p1) {
7446	7467	uint16x4_t __ret;
7447		__ret = (uint16x4_t)(__p0 <= __p1);
	7468	__ret = __builtin_bit_cast(uint16x4_t, __p0 <= __p1);
7448	7469	return __ret;
7449	7470	}
7450	7471	#else
7451	7472	__ai __attribute__((target("neon"))) uint16x4_t vcle_u16(uint16x4_t __p0, uint16x4_t __p1) {
7452	7473	uint16x4_t __ret;
7453		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
7454		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
7455		__ret = (uint16x4_t)(__rev0 <= __rev1);
7456		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	7474	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	7475	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	7476	__ret = __builtin_bit_cast(uint16x4_t, __rev0 <= __rev1);
	7477	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
7457	7478	return __ret;
7458	7479	}
7459	7480	#endif
...	...	@@ -7461,16 +7482,16 @@ __ai __attribute__((target("neon"))) uint16x4_t vcle_u16(uint16x4_t __p0, uint16
7461	7482	#ifdef __LITTLE_ENDIAN__
7462	7483	__ai __attribute__((target("neon"))) uint8x8_t vcle_s8(int8x8_t __p0, int8x8_t __p1) {
7463	7484	uint8x8_t __ret;
7464		__ret = (uint8x8_t)(__p0 <= __p1);
	7485	__ret = __builtin_bit_cast(uint8x8_t, __p0 <= __p1);
7465	7486	return __ret;
7466	7487	}
7467	7488	#else
7468	7489	__ai __attribute__((target("neon"))) uint8x8_t vcle_s8(int8x8_t __p0, int8x8_t __p1) {
7469	7490	uint8x8_t __ret;
7470		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
7471		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
7472		__ret = (uint8x8_t)(__rev0 <= __rev1);
7473		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	7491	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	7492	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	7493	__ret = __builtin_bit_cast(uint8x8_t, __rev0 <= __rev1);
	7494	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
7474	7495	return __ret;
7475	7496	}
7476	7497	#endif
...	...	@@ -7478,16 +7499,16 @@ __ai __attribute__((target("neon"))) uint8x8_t vcle_s8(int8x8_t __p0, int8x8_t _
7478	7499	#ifdef __LITTLE_ENDIAN__
7479	7500	__ai __attribute__((target("neon"))) uint32x2_t vcle_f32(float32x2_t __p0, float32x2_t __p1) {
7480	7501	uint32x2_t __ret;
7481		__ret = (uint32x2_t)(__p0 <= __p1);
	7502	__ret = __builtin_bit_cast(uint32x2_t, __p0 <= __p1);
7482	7503	return __ret;
7483	7504	}
7484	7505	#else
7485	7506	__ai __attribute__((target("neon"))) uint32x2_t vcle_f32(float32x2_t __p0, float32x2_t __p1) {
7486	7507	uint32x2_t __ret;
7487		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
7488		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
7489		__ret = (uint32x2_t)(__rev0 <= __rev1);
7490		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	7508	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	7509	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	7510	__ret = __builtin_bit_cast(uint32x2_t, __rev0 <= __rev1);
	7511	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
7491	7512	return __ret;
7492	7513	}
7493	7514	#endif
...	...	@@ -7495,16 +7516,16 @@ __ai __attribute__((target("neon"))) uint32x2_t vcle_f32(float32x2_t __p0, float
7495	7516	#ifdef __LITTLE_ENDIAN__
7496	7517	__ai __attribute__((target("neon"))) uint32x2_t vcle_s32(int32x2_t __p0, int32x2_t __p1) {
7497	7518	uint32x2_t __ret;
7498		__ret = (uint32x2_t)(__p0 <= __p1);
	7519	__ret = __builtin_bit_cast(uint32x2_t, __p0 <= __p1);
7499	7520	return __ret;
7500	7521	}
7501	7522	#else
7502	7523	__ai __attribute__((target("neon"))) uint32x2_t vcle_s32(int32x2_t __p0, int32x2_t __p1) {
7503	7524	uint32x2_t __ret;
7504		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
7505		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
7506		__ret = (uint32x2_t)(__rev0 <= __rev1);
7507		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	7525	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	7526	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	7527	__ret = __builtin_bit_cast(uint32x2_t, __rev0 <= __rev1);
	7528	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
7508	7529	return __ret;
7509	7530	}
7510	7531	#endif
...	...	@@ -7512,16 +7533,16 @@ __ai __attribute__((target("neon"))) uint32x2_t vcle_s32(int32x2_t __p0, int32x2
7512	7533	#ifdef __LITTLE_ENDIAN__
7513	7534	__ai __attribute__((target("neon"))) uint16x4_t vcle_s16(int16x4_t __p0, int16x4_t __p1) {
7514	7535	uint16x4_t __ret;
7515		__ret = (uint16x4_t)(__p0 <= __p1);
	7536	__ret = __builtin_bit_cast(uint16x4_t, __p0 <= __p1);
7516	7537	return __ret;
7517	7538	}
7518	7539	#else
7519	7540	__ai __attribute__((target("neon"))) uint16x4_t vcle_s16(int16x4_t __p0, int16x4_t __p1) {
7520	7541	uint16x4_t __ret;
7521		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
7522		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
7523		__ret = (uint16x4_t)(__rev0 <= __rev1);
7524		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	7542	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	7543	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	7544	__ret = __builtin_bit_cast(uint16x4_t, __rev0 <= __rev1);
	7545	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
7525	7546	return __ret;
7526	7547	}
7527	7548	#endif
...	...	@@ -7529,15 +7550,15 @@ __ai __attribute__((target("neon"))) uint16x4_t vcle_s16(int16x4_t __p0, int16x4
7529	7550	#ifdef __LITTLE_ENDIAN__
7530	7551	__ai __attribute__((target("neon"))) int8x16_t vclsq_u8(uint8x16_t __p0) {
7531	7552	int8x16_t __ret;
7532		__ret = (int8x16_t) __builtin_neon_vclsq_v((int8x16_t)__p0, 32);
	7553	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vclsq_v(__builtin_bit_cast(int8x16_t, __p0), 32));
7533	7554	return __ret;
7534	7555	}
7535	7556	#else
7536	7557	__ai __attribute__((target("neon"))) int8x16_t vclsq_u8(uint8x16_t __p0) {
7537	7558	int8x16_t __ret;
7538		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
7539		__ret = (int8x16_t) __builtin_neon_vclsq_v((int8x16_t)__rev0, 32);
7540		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	7559	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	7560	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vclsq_v(__builtin_bit_cast(int8x16_t, __rev0), 32));
	7561	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
7541	7562	return __ret;
7542	7563	}
7543	7564	#endif
...	...	@@ -7545,15 +7566,15 @@ __ai __attribute__((target("neon"))) int8x16_t vclsq_u8(uint8x16_t __p0) {
7545	7566	#ifdef __LITTLE_ENDIAN__
7546	7567	__ai __attribute__((target("neon"))) int32x4_t vclsq_u32(uint32x4_t __p0) {
7547	7568	int32x4_t __ret;
7548		__ret = (int32x4_t) __builtin_neon_vclsq_v((int8x16_t)__p0, 34);
	7569	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vclsq_v(__builtin_bit_cast(int8x16_t, __p0), 34));
7549	7570	return __ret;
7550	7571	}
7551	7572	#else
7552	7573	__ai __attribute__((target("neon"))) int32x4_t vclsq_u32(uint32x4_t __p0) {
7553	7574	int32x4_t __ret;
7554		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
7555		__ret = (int32x4_t) __builtin_neon_vclsq_v((int8x16_t)__rev0, 34);
7556		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	7575	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	7576	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vclsq_v(__builtin_bit_cast(int8x16_t, __rev0), 34));
	7577	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
7557	7578	return __ret;
7558	7579	}
7559	7580	#endif
...	...	@@ -7561,15 +7582,15 @@ __ai __attribute__((target("neon"))) int32x4_t vclsq_u32(uint32x4_t __p0) {
7561	7582	#ifdef __LITTLE_ENDIAN__
7562	7583	__ai __attribute__((target("neon"))) int16x8_t vclsq_u16(uint16x8_t __p0) {
7563	7584	int16x8_t __ret;
7564		__ret = (int16x8_t) __builtin_neon_vclsq_v((int8x16_t)__p0, 33);
	7585	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vclsq_v(__builtin_bit_cast(int8x16_t, __p0), 33));
7565	7586	return __ret;
7566	7587	}
7567	7588	#else
7568	7589	__ai __attribute__((target("neon"))) int16x8_t vclsq_u16(uint16x8_t __p0) {
7569	7590	int16x8_t __ret;
7570		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
7571		__ret = (int16x8_t) __builtin_neon_vclsq_v((int8x16_t)__rev0, 33);
7572		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	7591	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	7592	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vclsq_v(__builtin_bit_cast(int8x16_t, __rev0), 33));
	7593	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
7573	7594	return __ret;
7574	7595	}
7575	7596	#endif
...	...	@@ -7577,15 +7598,15 @@ __ai __attribute__((target("neon"))) int16x8_t vclsq_u16(uint16x8_t __p0) {
7577	7598	#ifdef __LITTLE_ENDIAN__
7578	7599	__ai __attribute__((target("neon"))) int8x16_t vclsq_s8(int8x16_t __p0) {
7579	7600	int8x16_t __ret;
7580		__ret = (int8x16_t) __builtin_neon_vclsq_v((int8x16_t)__p0, 32);
	7601	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vclsq_v(__builtin_bit_cast(int8x16_t, __p0), 32));
7581	7602	return __ret;
7582	7603	}
7583	7604	#else
7584	7605	__ai __attribute__((target("neon"))) int8x16_t vclsq_s8(int8x16_t __p0) {
7585	7606	int8x16_t __ret;
7586		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
7587		__ret = (int8x16_t) __builtin_neon_vclsq_v((int8x16_t)__rev0, 32);
7588		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	7607	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	7608	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vclsq_v(__builtin_bit_cast(int8x16_t, __rev0), 32));
	7609	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
7589	7610	return __ret;
7590	7611	}
7591	7612	#endif
...	...	@@ -7593,15 +7614,15 @@ __ai __attribute__((target("neon"))) int8x16_t vclsq_s8(int8x16_t __p0) {
7593	7614	#ifdef __LITTLE_ENDIAN__
7594	7615	__ai __attribute__((target("neon"))) int32x4_t vclsq_s32(int32x4_t __p0) {
7595	7616	int32x4_t __ret;
7596		__ret = (int32x4_t) __builtin_neon_vclsq_v((int8x16_t)__p0, 34);
	7617	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vclsq_v(__builtin_bit_cast(int8x16_t, __p0), 34));
7597	7618	return __ret;
7598	7619	}
7599	7620	#else
7600	7621	__ai __attribute__((target("neon"))) int32x4_t vclsq_s32(int32x4_t __p0) {
7601	7622	int32x4_t __ret;
7602		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
7603		__ret = (int32x4_t) __builtin_neon_vclsq_v((int8x16_t)__rev0, 34);
7604		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	7623	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	7624	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vclsq_v(__builtin_bit_cast(int8x16_t, __rev0), 34));
	7625	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
7605	7626	return __ret;
7606	7627	}
7607	7628	#endif
...	...	@@ -7609,15 +7630,15 @@ __ai __attribute__((target("neon"))) int32x4_t vclsq_s32(int32x4_t __p0) {
7609	7630	#ifdef __LITTLE_ENDIAN__
7610	7631	__ai __attribute__((target("neon"))) int16x8_t vclsq_s16(int16x8_t __p0) {
7611	7632	int16x8_t __ret;
7612		__ret = (int16x8_t) __builtin_neon_vclsq_v((int8x16_t)__p0, 33);
	7633	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vclsq_v(__builtin_bit_cast(int8x16_t, __p0), 33));
7613	7634	return __ret;
7614	7635	}
7615	7636	#else
7616	7637	__ai __attribute__((target("neon"))) int16x8_t vclsq_s16(int16x8_t __p0) {
7617	7638	int16x8_t __ret;
7618		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
7619		__ret = (int16x8_t) __builtin_neon_vclsq_v((int8x16_t)__rev0, 33);
7620		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	7639	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	7640	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vclsq_v(__builtin_bit_cast(int8x16_t, __rev0), 33));
	7641	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
7621	7642	return __ret;
7622	7643	}
7623	7644	#endif
...	...	@@ -7625,15 +7646,15 @@ __ai __attribute__((target("neon"))) int16x8_t vclsq_s16(int16x8_t __p0) {
7625	7646	#ifdef __LITTLE_ENDIAN__
7626	7647	__ai __attribute__((target("neon"))) int8x8_t vcls_u8(uint8x8_t __p0) {
7627	7648	int8x8_t __ret;
7628		__ret = (int8x8_t) __builtin_neon_vcls_v((int8x8_t)__p0, 0);
	7649	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vcls_v(__builtin_bit_cast(int8x8_t, __p0), 0));
7629	7650	return __ret;
7630	7651	}
7631	7652	#else
7632	7653	__ai __attribute__((target("neon"))) int8x8_t vcls_u8(uint8x8_t __p0) {
7633	7654	int8x8_t __ret;
7634		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
7635		__ret = (int8x8_t) __builtin_neon_vcls_v((int8x8_t)__rev0, 0);
7636		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	7655	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	7656	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vcls_v(__builtin_bit_cast(int8x8_t, __rev0), 0));
	7657	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
7637	7658	return __ret;
7638	7659	}
7639	7660	#endif
...	...	@@ -7641,15 +7662,15 @@ __ai __attribute__((target("neon"))) int8x8_t vcls_u8(uint8x8_t __p0) {
7641	7662	#ifdef __LITTLE_ENDIAN__
7642	7663	__ai __attribute__((target("neon"))) int32x2_t vcls_u32(uint32x2_t __p0) {
7643	7664	int32x2_t __ret;
7644		__ret = (int32x2_t) __builtin_neon_vcls_v((int8x8_t)__p0, 2);
	7665	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vcls_v(__builtin_bit_cast(int8x8_t, __p0), 2));
7645	7666	return __ret;
7646	7667	}
7647	7668	#else
7648	7669	__ai __attribute__((target("neon"))) int32x2_t vcls_u32(uint32x2_t __p0) {
7649	7670	int32x2_t __ret;
7650		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
7651		__ret = (int32x2_t) __builtin_neon_vcls_v((int8x8_t)__rev0, 2);
7652		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	7671	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	7672	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vcls_v(__builtin_bit_cast(int8x8_t, __rev0), 2));
	7673	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
7653	7674	return __ret;
7654	7675	}
7655	7676	#endif
...	...	@@ -7657,15 +7678,15 @@ __ai __attribute__((target("neon"))) int32x2_t vcls_u32(uint32x2_t __p0) {
7657	7678	#ifdef __LITTLE_ENDIAN__
7658	7679	__ai __attribute__((target("neon"))) int16x4_t vcls_u16(uint16x4_t __p0) {
7659	7680	int16x4_t __ret;
7660		__ret = (int16x4_t) __builtin_neon_vcls_v((int8x8_t)__p0, 1);
	7681	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vcls_v(__builtin_bit_cast(int8x8_t, __p0), 1));
7661	7682	return __ret;
7662	7683	}
7663	7684	#else
7664	7685	__ai __attribute__((target("neon"))) int16x4_t vcls_u16(uint16x4_t __p0) {
7665	7686	int16x4_t __ret;
7666		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
7667		__ret = (int16x4_t) __builtin_neon_vcls_v((int8x8_t)__rev0, 1);
7668		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	7687	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	7688	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vcls_v(__builtin_bit_cast(int8x8_t, __rev0), 1));
	7689	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
7669	7690	return __ret;
7670	7691	}
7671	7692	#endif
...	...	@@ -7673,15 +7694,15 @@ __ai __attribute__((target("neon"))) int16x4_t vcls_u16(uint16x4_t __p0) {
7673	7694	#ifdef __LITTLE_ENDIAN__
7674	7695	__ai __attribute__((target("neon"))) int8x8_t vcls_s8(int8x8_t __p0) {
7675	7696	int8x8_t __ret;
7676		__ret = (int8x8_t) __builtin_neon_vcls_v((int8x8_t)__p0, 0);
	7697	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vcls_v(__builtin_bit_cast(int8x8_t, __p0), 0));
7677	7698	return __ret;
7678	7699	}
7679	7700	#else
7680	7701	__ai __attribute__((target("neon"))) int8x8_t vcls_s8(int8x8_t __p0) {
7681	7702	int8x8_t __ret;
7682		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
7683		__ret = (int8x8_t) __builtin_neon_vcls_v((int8x8_t)__rev0, 0);
7684		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	7703	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	7704	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vcls_v(__builtin_bit_cast(int8x8_t, __rev0), 0));
	7705	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
7685	7706	return __ret;
7686	7707	}
7687	7708	#endif
...	...	@@ -7689,15 +7710,15 @@ __ai __attribute__((target("neon"))) int8x8_t vcls_s8(int8x8_t __p0) {
7689	7710	#ifdef __LITTLE_ENDIAN__
7690	7711	__ai __attribute__((target("neon"))) int32x2_t vcls_s32(int32x2_t __p0) {
7691	7712	int32x2_t __ret;
7692		__ret = (int32x2_t) __builtin_neon_vcls_v((int8x8_t)__p0, 2);
	7713	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vcls_v(__builtin_bit_cast(int8x8_t, __p0), 2));
7693	7714	return __ret;
7694	7715	}
7695	7716	#else
7696	7717	__ai __attribute__((target("neon"))) int32x2_t vcls_s32(int32x2_t __p0) {
7697	7718	int32x2_t __ret;
7698		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
7699		__ret = (int32x2_t) __builtin_neon_vcls_v((int8x8_t)__rev0, 2);
7700		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	7719	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	7720	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vcls_v(__builtin_bit_cast(int8x8_t, __rev0), 2));
	7721	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
7701	7722	return __ret;
7702	7723	}
7703	7724	#endif
...	...	@@ -7705,15 +7726,15 @@ __ai __attribute__((target("neon"))) int32x2_t vcls_s32(int32x2_t __p0) {
7705	7726	#ifdef __LITTLE_ENDIAN__
7706	7727	__ai __attribute__((target("neon"))) int16x4_t vcls_s16(int16x4_t __p0) {
7707	7728	int16x4_t __ret;
7708		__ret = (int16x4_t) __builtin_neon_vcls_v((int8x8_t)__p0, 1);
	7729	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vcls_v(__builtin_bit_cast(int8x8_t, __p0), 1));
7709	7730	return __ret;
7710	7731	}
7711	7732	#else
7712	7733	__ai __attribute__((target("neon"))) int16x4_t vcls_s16(int16x4_t __p0) {
7713	7734	int16x4_t __ret;
7714		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
7715		__ret = (int16x4_t) __builtin_neon_vcls_v((int8x8_t)__rev0, 1);
7716		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	7735	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	7736	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vcls_v(__builtin_bit_cast(int8x8_t, __rev0), 1));
	7737	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
7717	7738	return __ret;
7718	7739	}
7719	7740	#endif
...	...	@@ -7721,16 +7742,16 @@ __ai __attribute__((target("neon"))) int16x4_t vcls_s16(int16x4_t __p0) {
7721	7742	#ifdef __LITTLE_ENDIAN__
7722	7743	__ai __attribute__((target("neon"))) uint8x16_t vcltq_u8(uint8x16_t __p0, uint8x16_t __p1) {
7723	7744	uint8x16_t __ret;
7724		__ret = (uint8x16_t)(__p0 < __p1);
	7745	__ret = __builtin_bit_cast(uint8x16_t, __p0 < __p1);
7725	7746	return __ret;
7726	7747	}
7727	7748	#else
7728	7749	__ai __attribute__((target("neon"))) uint8x16_t vcltq_u8(uint8x16_t __p0, uint8x16_t __p1) {
7729	7750	uint8x16_t __ret;
7730		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
7731		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
7732		__ret = (uint8x16_t)(__rev0 < __rev1);
7733		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	7751	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	7752	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	7753	__ret = __builtin_bit_cast(uint8x16_t, __rev0 < __rev1);
	7754	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
7734	7755	return __ret;
7735	7756	}
7736	7757	#endif
...	...	@@ -7738,16 +7759,16 @@ __ai __attribute__((target("neon"))) uint8x16_t vcltq_u8(uint8x16_t __p0, uint8x
7738	7759	#ifdef __LITTLE_ENDIAN__
7739	7760	__ai __attribute__((target("neon"))) uint32x4_t vcltq_u32(uint32x4_t __p0, uint32x4_t __p1) {
7740	7761	uint32x4_t __ret;
7741		__ret = (uint32x4_t)(__p0 < __p1);
	7762	__ret = __builtin_bit_cast(uint32x4_t, __p0 < __p1);
7742	7763	return __ret;
7743	7764	}
7744	7765	#else
7745	7766	__ai __attribute__((target("neon"))) uint32x4_t vcltq_u32(uint32x4_t __p0, uint32x4_t __p1) {
7746	7767	uint32x4_t __ret;
7747		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
7748		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
7749		__ret = (uint32x4_t)(__rev0 < __rev1);
7750		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	7768	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	7769	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	7770	__ret = __builtin_bit_cast(uint32x4_t, __rev0 < __rev1);
	7771	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
7751	7772	return __ret;
7752	7773	}
7753	7774	#endif
...	...	@@ -7755,16 +7776,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vcltq_u32(uint32x4_t __p0, uint3
7755	7776	#ifdef __LITTLE_ENDIAN__
7756	7777	__ai __attribute__((target("neon"))) uint16x8_t vcltq_u16(uint16x8_t __p0, uint16x8_t __p1) {
7757	7778	uint16x8_t __ret;
7758		__ret = (uint16x8_t)(__p0 < __p1);
	7779	__ret = __builtin_bit_cast(uint16x8_t, __p0 < __p1);
7759	7780	return __ret;
7760	7781	}
7761	7782	#else
7762	7783	__ai __attribute__((target("neon"))) uint16x8_t vcltq_u16(uint16x8_t __p0, uint16x8_t __p1) {
7763	7784	uint16x8_t __ret;
7764		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
7765		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
7766		__ret = (uint16x8_t)(__rev0 < __rev1);
7767		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	7785	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	7786	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	7787	__ret = __builtin_bit_cast(uint16x8_t, __rev0 < __rev1);
	7788	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
7768	7789	return __ret;
7769	7790	}
7770	7791	#endif
...	...	@@ -7772,16 +7793,16 @@ __ai __attribute__((target("neon"))) uint16x8_t vcltq_u16(uint16x8_t __p0, uint1
7772	7793	#ifdef __LITTLE_ENDIAN__
7773	7794	__ai __attribute__((target("neon"))) uint8x16_t vcltq_s8(int8x16_t __p0, int8x16_t __p1) {
7774	7795	uint8x16_t __ret;
7775		__ret = (uint8x16_t)(__p0 < __p1);
	7796	__ret = __builtin_bit_cast(uint8x16_t, __p0 < __p1);
7776	7797	return __ret;
7777	7798	}
7778	7799	#else
7779	7800	__ai __attribute__((target("neon"))) uint8x16_t vcltq_s8(int8x16_t __p0, int8x16_t __p1) {
7780	7801	uint8x16_t __ret;
7781		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
7782		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
7783		__ret = (uint8x16_t)(__rev0 < __rev1);
7784		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	7802	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	7803	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	7804	__ret = __builtin_bit_cast(uint8x16_t, __rev0 < __rev1);
	7805	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
7785	7806	return __ret;
7786	7807	}
7787	7808	#endif
...	...	@@ -7789,16 +7810,16 @@ __ai __attribute__((target("neon"))) uint8x16_t vcltq_s8(int8x16_t __p0, int8x16
7789	7810	#ifdef __LITTLE_ENDIAN__
7790	7811	__ai __attribute__((target("neon"))) uint32x4_t vcltq_f32(float32x4_t __p0, float32x4_t __p1) {
7791	7812	uint32x4_t __ret;
7792		__ret = (uint32x4_t)(__p0 < __p1);
	7813	__ret = __builtin_bit_cast(uint32x4_t, __p0 < __p1);
7793	7814	return __ret;
7794	7815	}
7795	7816	#else
7796	7817	__ai __attribute__((target("neon"))) uint32x4_t vcltq_f32(float32x4_t __p0, float32x4_t __p1) {
7797	7818	uint32x4_t __ret;
7798		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
7799		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
7800		__ret = (uint32x4_t)(__rev0 < __rev1);
7801		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	7819	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	7820	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	7821	__ret = __builtin_bit_cast(uint32x4_t, __rev0 < __rev1);
	7822	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
7802	7823	return __ret;
7803	7824	}
7804	7825	#endif
...	...	@@ -7806,16 +7827,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vcltq_f32(float32x4_t __p0, floa
7806	7827	#ifdef __LITTLE_ENDIAN__
7807	7828	__ai __attribute__((target("neon"))) uint32x4_t vcltq_s32(int32x4_t __p0, int32x4_t __p1) {
7808	7829	uint32x4_t __ret;
7809		__ret = (uint32x4_t)(__p0 < __p1);
	7830	__ret = __builtin_bit_cast(uint32x4_t, __p0 < __p1);
7810	7831	return __ret;
7811	7832	}
7812	7833	#else
7813	7834	__ai __attribute__((target("neon"))) uint32x4_t vcltq_s32(int32x4_t __p0, int32x4_t __p1) {
7814	7835	uint32x4_t __ret;
7815		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
7816		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
7817		__ret = (uint32x4_t)(__rev0 < __rev1);
7818		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	7836	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	7837	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	7838	__ret = __builtin_bit_cast(uint32x4_t, __rev0 < __rev1);
	7839	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
7819	7840	return __ret;
7820	7841	}
7821	7842	#endif
...	...	@@ -7823,16 +7844,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vcltq_s32(int32x4_t __p0, int32x
7823	7844	#ifdef __LITTLE_ENDIAN__
7824	7845	__ai __attribute__((target("neon"))) uint16x8_t vcltq_s16(int16x8_t __p0, int16x8_t __p1) {
7825	7846	uint16x8_t __ret;
7826		__ret = (uint16x8_t)(__p0 < __p1);
	7847	__ret = __builtin_bit_cast(uint16x8_t, __p0 < __p1);
7827	7848	return __ret;
7828	7849	}
7829	7850	#else
7830	7851	__ai __attribute__((target("neon"))) uint16x8_t vcltq_s16(int16x8_t __p0, int16x8_t __p1) {
7831	7852	uint16x8_t __ret;
7832		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
7833		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
7834		__ret = (uint16x8_t)(__rev0 < __rev1);
7835		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	7853	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	7854	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	7855	__ret = __builtin_bit_cast(uint16x8_t, __rev0 < __rev1);
	7856	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
7836	7857	return __ret;
7837	7858	}
7838	7859	#endif
...	...	@@ -7840,16 +7861,16 @@ __ai __attribute__((target("neon"))) uint16x8_t vcltq_s16(int16x8_t __p0, int16x
7840	7861	#ifdef __LITTLE_ENDIAN__
7841	7862	__ai __attribute__((target("neon"))) uint8x8_t vclt_u8(uint8x8_t __p0, uint8x8_t __p1) {
7842	7863	uint8x8_t __ret;
7843		__ret = (uint8x8_t)(__p0 < __p1);
	7864	__ret = __builtin_bit_cast(uint8x8_t, __p0 < __p1);
7844	7865	return __ret;
7845	7866	}
7846	7867	#else
7847	7868	__ai __attribute__((target("neon"))) uint8x8_t vclt_u8(uint8x8_t __p0, uint8x8_t __p1) {
7848	7869	uint8x8_t __ret;
7849		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
7850		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
7851		__ret = (uint8x8_t)(__rev0 < __rev1);
7852		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	7870	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	7871	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	7872	__ret = __builtin_bit_cast(uint8x8_t, __rev0 < __rev1);
	7873	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
7853	7874	return __ret;
7854	7875	}
7855	7876	#endif
...	...	@@ -7857,16 +7878,16 @@ __ai __attribute__((target("neon"))) uint8x8_t vclt_u8(uint8x8_t __p0, uint8x8_t
7857	7878	#ifdef __LITTLE_ENDIAN__
7858	7879	__ai __attribute__((target("neon"))) uint32x2_t vclt_u32(uint32x2_t __p0, uint32x2_t __p1) {
7859	7880	uint32x2_t __ret;
7860		__ret = (uint32x2_t)(__p0 < __p1);
	7881	__ret = __builtin_bit_cast(uint32x2_t, __p0 < __p1);
7861	7882	return __ret;
7862	7883	}
7863	7884	#else
7864	7885	__ai __attribute__((target("neon"))) uint32x2_t vclt_u32(uint32x2_t __p0, uint32x2_t __p1) {
7865	7886	uint32x2_t __ret;
7866		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
7867		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
7868		__ret = (uint32x2_t)(__rev0 < __rev1);
7869		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	7887	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	7888	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	7889	__ret = __builtin_bit_cast(uint32x2_t, __rev0 < __rev1);
	7890	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
7870	7891	return __ret;
7871	7892	}
7872	7893	#endif
...	...	@@ -7874,16 +7895,16 @@ __ai __attribute__((target("neon"))) uint32x2_t vclt_u32(uint32x2_t __p0, uint32
7874	7895	#ifdef __LITTLE_ENDIAN__
7875	7896	__ai __attribute__((target("neon"))) uint16x4_t vclt_u16(uint16x4_t __p0, uint16x4_t __p1) {
7876	7897	uint16x4_t __ret;
7877		__ret = (uint16x4_t)(__p0 < __p1);
	7898	__ret = __builtin_bit_cast(uint16x4_t, __p0 < __p1);
7878	7899	return __ret;
7879	7900	}
7880	7901	#else
7881	7902	__ai __attribute__((target("neon"))) uint16x4_t vclt_u16(uint16x4_t __p0, uint16x4_t __p1) {
7882	7903	uint16x4_t __ret;
7883		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
7884		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
7885		__ret = (uint16x4_t)(__rev0 < __rev1);
7886		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	7904	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	7905	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	7906	__ret = __builtin_bit_cast(uint16x4_t, __rev0 < __rev1);
	7907	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
7887	7908	return __ret;
7888	7909	}
7889	7910	#endif
...	...	@@ -7891,16 +7912,16 @@ __ai __attribute__((target("neon"))) uint16x4_t vclt_u16(uint16x4_t __p0, uint16
7891	7912	#ifdef __LITTLE_ENDIAN__
7892	7913	__ai __attribute__((target("neon"))) uint8x8_t vclt_s8(int8x8_t __p0, int8x8_t __p1) {
7893	7914	uint8x8_t __ret;
7894		__ret = (uint8x8_t)(__p0 < __p1);
	7915	__ret = __builtin_bit_cast(uint8x8_t, __p0 < __p1);
7895	7916	return __ret;
7896	7917	}
7897	7918	#else
7898	7919	__ai __attribute__((target("neon"))) uint8x8_t vclt_s8(int8x8_t __p0, int8x8_t __p1) {
7899	7920	uint8x8_t __ret;
7900		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
7901		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
7902		__ret = (uint8x8_t)(__rev0 < __rev1);
7903		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	7921	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	7922	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	7923	__ret = __builtin_bit_cast(uint8x8_t, __rev0 < __rev1);
	7924	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
7904	7925	return __ret;
7905	7926	}
7906	7927	#endif
...	...	@@ -7908,16 +7929,16 @@ __ai __attribute__((target("neon"))) uint8x8_t vclt_s8(int8x8_t __p0, int8x8_t _
7908	7929	#ifdef __LITTLE_ENDIAN__
7909	7930	__ai __attribute__((target("neon"))) uint32x2_t vclt_f32(float32x2_t __p0, float32x2_t __p1) {
7910	7931	uint32x2_t __ret;
7911		__ret = (uint32x2_t)(__p0 < __p1);
	7932	__ret = __builtin_bit_cast(uint32x2_t, __p0 < __p1);
7912	7933	return __ret;
7913	7934	}
7914	7935	#else
7915	7936	__ai __attribute__((target("neon"))) uint32x2_t vclt_f32(float32x2_t __p0, float32x2_t __p1) {
7916	7937	uint32x2_t __ret;
7917		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
7918		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
7919		__ret = (uint32x2_t)(__rev0 < __rev1);
7920		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	7938	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	7939	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	7940	__ret = __builtin_bit_cast(uint32x2_t, __rev0 < __rev1);
	7941	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
7921	7942	return __ret;
7922	7943	}
7923	7944	#endif
...	...	@@ -7925,16 +7946,16 @@ __ai __attribute__((target("neon"))) uint32x2_t vclt_f32(float32x2_t __p0, float
7925	7946	#ifdef __LITTLE_ENDIAN__
7926	7947	__ai __attribute__((target("neon"))) uint32x2_t vclt_s32(int32x2_t __p0, int32x2_t __p1) {
7927	7948	uint32x2_t __ret;
7928		__ret = (uint32x2_t)(__p0 < __p1);
	7949	__ret = __builtin_bit_cast(uint32x2_t, __p0 < __p1);
7929	7950	return __ret;
7930	7951	}
7931	7952	#else
7932	7953	__ai __attribute__((target("neon"))) uint32x2_t vclt_s32(int32x2_t __p0, int32x2_t __p1) {
7933	7954	uint32x2_t __ret;
7934		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
7935		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
7936		__ret = (uint32x2_t)(__rev0 < __rev1);
7937		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	7955	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	7956	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	7957	__ret = __builtin_bit_cast(uint32x2_t, __rev0 < __rev1);
	7958	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
7938	7959	return __ret;
7939	7960	}
7940	7961	#endif
...	...	@@ -7942,16 +7963,16 @@ __ai __attribute__((target("neon"))) uint32x2_t vclt_s32(int32x2_t __p0, int32x2
7942	7963	#ifdef __LITTLE_ENDIAN__
7943	7964	__ai __attribute__((target("neon"))) uint16x4_t vclt_s16(int16x4_t __p0, int16x4_t __p1) {
7944	7965	uint16x4_t __ret;
7945		__ret = (uint16x4_t)(__p0 < __p1);
	7966	__ret = __builtin_bit_cast(uint16x4_t, __p0 < __p1);
7946	7967	return __ret;
7947	7968	}
7948	7969	#else
7949	7970	__ai __attribute__((target("neon"))) uint16x4_t vclt_s16(int16x4_t __p0, int16x4_t __p1) {
7950	7971	uint16x4_t __ret;
7951		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
7952		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
7953		__ret = (uint16x4_t)(__rev0 < __rev1);
7954		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	7972	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	7973	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	7974	__ret = __builtin_bit_cast(uint16x4_t, __rev0 < __rev1);
	7975	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
7955	7976	return __ret;
7956	7977	}
7957	7978	#endif
...	...	@@ -7959,15 +7980,15 @@ __ai __attribute__((target("neon"))) uint16x4_t vclt_s16(int16x4_t __p0, int16x4
7959	7980	#ifdef __LITTLE_ENDIAN__
7960	7981	__ai __attribute__((target("neon"))) uint8x16_t vclzq_u8(uint8x16_t __p0) {
7961	7982	uint8x16_t __ret;
7962		__ret = (uint8x16_t) __builtin_neon_vclzq_v((int8x16_t)__p0, 48);
	7983	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vclzq_v(__builtin_bit_cast(int8x16_t, __p0), 48));
7963	7984	return __ret;
7964	7985	}
7965	7986	#else
7966	7987	__ai __attribute__((target("neon"))) uint8x16_t vclzq_u8(uint8x16_t __p0) {
7967	7988	uint8x16_t __ret;
7968		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
7969		__ret = (uint8x16_t) __builtin_neon_vclzq_v((int8x16_t)__rev0, 48);
7970		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	7989	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	7990	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vclzq_v(__builtin_bit_cast(int8x16_t, __rev0), 48));
	7991	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
7971	7992	return __ret;
7972	7993	}
7973	7994	#endif
...	...	@@ -7975,15 +7996,15 @@ __ai __attribute__((target("neon"))) uint8x16_t vclzq_u8(uint8x16_t __p0) {
7975	7996	#ifdef __LITTLE_ENDIAN__
7976	7997	__ai __attribute__((target("neon"))) uint32x4_t vclzq_u32(uint32x4_t __p0) {
7977	7998	uint32x4_t __ret;
7978		__ret = (uint32x4_t) __builtin_neon_vclzq_v((int8x16_t)__p0, 50);
	7999	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vclzq_v(__builtin_bit_cast(int8x16_t, __p0), 50));
7979	8000	return __ret;
7980	8001	}
7981	8002	#else
7982	8003	__ai __attribute__((target("neon"))) uint32x4_t vclzq_u32(uint32x4_t __p0) {
7983	8004	uint32x4_t __ret;
7984		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
7985		__ret = (uint32x4_t) __builtin_neon_vclzq_v((int8x16_t)__rev0, 50);
7986		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	8005	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	8006	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vclzq_v(__builtin_bit_cast(int8x16_t, __rev0), 50));
	8007	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
7987	8008	return __ret;
7988	8009	}
7989	8010	#endif
...	...	@@ -7991,15 +8012,15 @@ __ai __attribute__((target("neon"))) uint32x4_t vclzq_u32(uint32x4_t __p0) {
7991	8012	#ifdef __LITTLE_ENDIAN__
7992	8013	__ai __attribute__((target("neon"))) uint16x8_t vclzq_u16(uint16x8_t __p0) {
7993	8014	uint16x8_t __ret;
7994		__ret = (uint16x8_t) __builtin_neon_vclzq_v((int8x16_t)__p0, 49);
	8015	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vclzq_v(__builtin_bit_cast(int8x16_t, __p0), 49));
7995	8016	return __ret;
7996	8017	}
7997	8018	#else
7998	8019	__ai __attribute__((target("neon"))) uint16x8_t vclzq_u16(uint16x8_t __p0) {
7999	8020	uint16x8_t __ret;
8000		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
8001		__ret = (uint16x8_t) __builtin_neon_vclzq_v((int8x16_t)__rev0, 49);
8002		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	8021	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	8022	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vclzq_v(__builtin_bit_cast(int8x16_t, __rev0), 49));
	8023	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
8003	8024	return __ret;
8004	8025	}
8005	8026	#endif
...	...	@@ -8007,15 +8028,15 @@ __ai __attribute__((target("neon"))) uint16x8_t vclzq_u16(uint16x8_t __p0) {
8007	8028	#ifdef __LITTLE_ENDIAN__
8008	8029	__ai __attribute__((target("neon"))) int8x16_t vclzq_s8(int8x16_t __p0) {
8009	8030	int8x16_t __ret;
8010		__ret = (int8x16_t) __builtin_neon_vclzq_v((int8x16_t)__p0, 32);
	8031	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vclzq_v(__builtin_bit_cast(int8x16_t, __p0), 32));
8011	8032	return __ret;
8012	8033	}
8013	8034	#else
8014	8035	__ai __attribute__((target("neon"))) int8x16_t vclzq_s8(int8x16_t __p0) {
8015	8036	int8x16_t __ret;
8016		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
8017		__ret = (int8x16_t) __builtin_neon_vclzq_v((int8x16_t)__rev0, 32);
8018		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	8037	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	8038	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vclzq_v(__builtin_bit_cast(int8x16_t, __rev0), 32));
	8039	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
8019	8040	return __ret;
8020	8041	}
8021	8042	#endif
...	...	@@ -8023,15 +8044,15 @@ __ai __attribute__((target("neon"))) int8x16_t vclzq_s8(int8x16_t __p0) {
8023	8044	#ifdef __LITTLE_ENDIAN__
8024	8045	__ai __attribute__((target("neon"))) int32x4_t vclzq_s32(int32x4_t __p0) {
8025	8046	int32x4_t __ret;
8026		__ret = (int32x4_t) __builtin_neon_vclzq_v((int8x16_t)__p0, 34);
	8047	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vclzq_v(__builtin_bit_cast(int8x16_t, __p0), 34));
8027	8048	return __ret;
8028	8049	}
8029	8050	#else
8030	8051	__ai __attribute__((target("neon"))) int32x4_t vclzq_s32(int32x4_t __p0) {
8031	8052	int32x4_t __ret;
8032		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
8033		__ret = (int32x4_t) __builtin_neon_vclzq_v((int8x16_t)__rev0, 34);
8034		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	8053	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	8054	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vclzq_v(__builtin_bit_cast(int8x16_t, __rev0), 34));
	8055	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
8035	8056	return __ret;
8036	8057	}
8037	8058	#endif
...	...	@@ -8039,15 +8060,15 @@ __ai __attribute__((target("neon"))) int32x4_t vclzq_s32(int32x4_t __p0) {
8039	8060	#ifdef __LITTLE_ENDIAN__
8040	8061	__ai __attribute__((target("neon"))) int16x8_t vclzq_s16(int16x8_t __p0) {
8041	8062	int16x8_t __ret;
8042		__ret = (int16x8_t) __builtin_neon_vclzq_v((int8x16_t)__p0, 33);
	8063	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vclzq_v(__builtin_bit_cast(int8x16_t, __p0), 33));
8043	8064	return __ret;
8044	8065	}
8045	8066	#else
8046	8067	__ai __attribute__((target("neon"))) int16x8_t vclzq_s16(int16x8_t __p0) {
8047	8068	int16x8_t __ret;
8048		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
8049		__ret = (int16x8_t) __builtin_neon_vclzq_v((int8x16_t)__rev0, 33);
8050		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	8069	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	8070	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vclzq_v(__builtin_bit_cast(int8x16_t, __rev0), 33));
	8071	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
8051	8072	return __ret;
8052	8073	}
8053	8074	#endif
...	...	@@ -8055,15 +8076,15 @@ __ai __attribute__((target("neon"))) int16x8_t vclzq_s16(int16x8_t __p0) {
8055	8076	#ifdef __LITTLE_ENDIAN__
8056	8077	__ai __attribute__((target("neon"))) uint8x8_t vclz_u8(uint8x8_t __p0) {
8057	8078	uint8x8_t __ret;
8058		__ret = (uint8x8_t) __builtin_neon_vclz_v((int8x8_t)__p0, 16);
	8079	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vclz_v(__builtin_bit_cast(int8x8_t, __p0), 16));
8059	8080	return __ret;
8060	8081	}
8061	8082	#else
8062	8083	__ai __attribute__((target("neon"))) uint8x8_t vclz_u8(uint8x8_t __p0) {
8063	8084	uint8x8_t __ret;
8064		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
8065		__ret = (uint8x8_t) __builtin_neon_vclz_v((int8x8_t)__rev0, 16);
8066		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	8085	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	8086	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vclz_v(__builtin_bit_cast(int8x8_t, __rev0), 16));
	8087	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
8067	8088	return __ret;
8068	8089	}
8069	8090	#endif
...	...	@@ -8071,15 +8092,15 @@ __ai __attribute__((target("neon"))) uint8x8_t vclz_u8(uint8x8_t __p0) {
8071	8092	#ifdef __LITTLE_ENDIAN__
8072	8093	__ai __attribute__((target("neon"))) uint32x2_t vclz_u32(uint32x2_t __p0) {
8073	8094	uint32x2_t __ret;
8074		__ret = (uint32x2_t) __builtin_neon_vclz_v((int8x8_t)__p0, 18);
	8095	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vclz_v(__builtin_bit_cast(int8x8_t, __p0), 18));
8075	8096	return __ret;
8076	8097	}
8077	8098	#else
8078	8099	__ai __attribute__((target("neon"))) uint32x2_t vclz_u32(uint32x2_t __p0) {
8079	8100	uint32x2_t __ret;
8080		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
8081		__ret = (uint32x2_t) __builtin_neon_vclz_v((int8x8_t)__rev0, 18);
8082		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	8101	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	8102	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vclz_v(__builtin_bit_cast(int8x8_t, __rev0), 18));
	8103	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
8083	8104	return __ret;
8084	8105	}
8085	8106	#endif
...	...	@@ -8087,15 +8108,15 @@ __ai __attribute__((target("neon"))) uint32x2_t vclz_u32(uint32x2_t __p0) {
8087	8108	#ifdef __LITTLE_ENDIAN__
8088	8109	__ai __attribute__((target("neon"))) uint16x4_t vclz_u16(uint16x4_t __p0) {
8089	8110	uint16x4_t __ret;
8090		__ret = (uint16x4_t) __builtin_neon_vclz_v((int8x8_t)__p0, 17);
	8111	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vclz_v(__builtin_bit_cast(int8x8_t, __p0), 17));
8091	8112	return __ret;
8092	8113	}
8093	8114	#else
8094	8115	__ai __attribute__((target("neon"))) uint16x4_t vclz_u16(uint16x4_t __p0) {
8095	8116	uint16x4_t __ret;
8096		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
8097		__ret = (uint16x4_t) __builtin_neon_vclz_v((int8x8_t)__rev0, 17);
8098		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	8117	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	8118	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vclz_v(__builtin_bit_cast(int8x8_t, __rev0), 17));
	8119	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
8099	8120	return __ret;
8100	8121	}
8101	8122	#endif
...	...	@@ -8103,15 +8124,15 @@ __ai __attribute__((target("neon"))) uint16x4_t vclz_u16(uint16x4_t __p0) {
8103	8124	#ifdef __LITTLE_ENDIAN__
8104	8125	__ai __attribute__((target("neon"))) int8x8_t vclz_s8(int8x8_t __p0) {
8105	8126	int8x8_t __ret;
8106		__ret = (int8x8_t) __builtin_neon_vclz_v((int8x8_t)__p0, 0);
	8127	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vclz_v(__builtin_bit_cast(int8x8_t, __p0), 0));
8107	8128	return __ret;
8108	8129	}
8109	8130	#else
8110	8131	__ai __attribute__((target("neon"))) int8x8_t vclz_s8(int8x8_t __p0) {
8111	8132	int8x8_t __ret;
8112		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
8113		__ret = (int8x8_t) __builtin_neon_vclz_v((int8x8_t)__rev0, 0);
8114		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	8133	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	8134	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vclz_v(__builtin_bit_cast(int8x8_t, __rev0), 0));
	8135	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
8115	8136	return __ret;
8116	8137	}
8117	8138	#endif
...	...	@@ -8119,15 +8140,15 @@ __ai __attribute__((target("neon"))) int8x8_t vclz_s8(int8x8_t __p0) {
8119	8140	#ifdef __LITTLE_ENDIAN__
8120	8141	__ai __attribute__((target("neon"))) int32x2_t vclz_s32(int32x2_t __p0) {
8121	8142	int32x2_t __ret;
8122		__ret = (int32x2_t) __builtin_neon_vclz_v((int8x8_t)__p0, 2);
	8143	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vclz_v(__builtin_bit_cast(int8x8_t, __p0), 2));
8123	8144	return __ret;
8124	8145	}
8125	8146	#else
8126	8147	__ai __attribute__((target("neon"))) int32x2_t vclz_s32(int32x2_t __p0) {
8127	8148	int32x2_t __ret;
8128		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
8129		__ret = (int32x2_t) __builtin_neon_vclz_v((int8x8_t)__rev0, 2);
8130		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	8149	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	8150	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vclz_v(__builtin_bit_cast(int8x8_t, __rev0), 2));
	8151	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
8131	8152	return __ret;
8132	8153	}
8133	8154	#endif
...	...	@@ -8135,15 +8156,15 @@ __ai __attribute__((target("neon"))) int32x2_t vclz_s32(int32x2_t __p0) {
8135	8156	#ifdef __LITTLE_ENDIAN__
8136	8157	__ai __attribute__((target("neon"))) int16x4_t vclz_s16(int16x4_t __p0) {
8137	8158	int16x4_t __ret;
8138		__ret = (int16x4_t) __builtin_neon_vclz_v((int8x8_t)__p0, 1);
	8159	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vclz_v(__builtin_bit_cast(int8x8_t, __p0), 1));
8139	8160	return __ret;
8140	8161	}
8141	8162	#else
8142	8163	__ai __attribute__((target("neon"))) int16x4_t vclz_s16(int16x4_t __p0) {
8143	8164	int16x4_t __ret;
8144		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
8145		__ret = (int16x4_t) __builtin_neon_vclz_v((int8x8_t)__rev0, 1);
8146		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	8165	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	8166	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vclz_v(__builtin_bit_cast(int8x8_t, __rev0), 1));
	8167	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
8147	8168	return __ret;
8148	8169	}
8149	8170	#endif
...	...	@@ -8151,15 +8172,15 @@ __ai __attribute__((target("neon"))) int16x4_t vclz_s16(int16x4_t __p0) {
8151	8172	#ifdef __LITTLE_ENDIAN__
8152	8173	__ai __attribute__((target("neon"))) poly8x8_t vcnt_p8(poly8x8_t __p0) {
8153	8174	poly8x8_t __ret;
8154		__ret = (poly8x8_t) __builtin_neon_vcnt_v((int8x8_t)__p0, 4);
	8175	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vcnt_v(__builtin_bit_cast(int8x8_t, __p0), 4));
8155	8176	return __ret;
8156	8177	}
8157	8178	#else
8158	8179	__ai __attribute__((target("neon"))) poly8x8_t vcnt_p8(poly8x8_t __p0) {
8159	8180	poly8x8_t __ret;
8160		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
8161		__ret = (poly8x8_t) __builtin_neon_vcnt_v((int8x8_t)__rev0, 4);
8162		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	8181	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	8182	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vcnt_v(__builtin_bit_cast(int8x8_t, __rev0), 4));
	8183	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
8163	8184	return __ret;
8164	8185	}
8165	8186	#endif
...	...	@@ -8167,15 +8188,15 @@ __ai __attribute__((target("neon"))) poly8x8_t vcnt_p8(poly8x8_t __p0) {
8167	8188	#ifdef __LITTLE_ENDIAN__
8168	8189	__ai __attribute__((target("neon"))) poly8x16_t vcntq_p8(poly8x16_t __p0) {
8169	8190	poly8x16_t __ret;
8170		__ret = (poly8x16_t) __builtin_neon_vcntq_v((int8x16_t)__p0, 36);
	8191	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vcntq_v(__builtin_bit_cast(int8x16_t, __p0), 36));
8171	8192	return __ret;
8172	8193	}
8173	8194	#else
8174	8195	__ai __attribute__((target("neon"))) poly8x16_t vcntq_p8(poly8x16_t __p0) {
8175	8196	poly8x16_t __ret;
8176		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
8177		__ret = (poly8x16_t) __builtin_neon_vcntq_v((int8x16_t)__rev0, 36);
8178		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	8197	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	8198	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vcntq_v(__builtin_bit_cast(int8x16_t, __rev0), 36));
	8199	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
8179	8200	return __ret;
8180	8201	}
8181	8202	#endif
...	...	@@ -8183,15 +8204,15 @@ __ai __attribute__((target("neon"))) poly8x16_t vcntq_p8(poly8x16_t __p0) {
8183	8204	#ifdef __LITTLE_ENDIAN__
8184	8205	__ai __attribute__((target("neon"))) uint8x16_t vcntq_u8(uint8x16_t __p0) {
8185	8206	uint8x16_t __ret;
8186		__ret = (uint8x16_t) __builtin_neon_vcntq_v((int8x16_t)__p0, 48);
	8207	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vcntq_v(__builtin_bit_cast(int8x16_t, __p0), 48));
8187	8208	return __ret;
8188	8209	}
8189	8210	#else
8190	8211	__ai __attribute__((target("neon"))) uint8x16_t vcntq_u8(uint8x16_t __p0) {
8191	8212	uint8x16_t __ret;
8192		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
8193		__ret = (uint8x16_t) __builtin_neon_vcntq_v((int8x16_t)__rev0, 48);
8194		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	8213	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	8214	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vcntq_v(__builtin_bit_cast(int8x16_t, __rev0), 48));
	8215	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
8195	8216	return __ret;
8196	8217	}
8197	8218	#endif
...	...	@@ -8199,15 +8220,15 @@ __ai __attribute__((target("neon"))) uint8x16_t vcntq_u8(uint8x16_t __p0) {
8199	8220	#ifdef __LITTLE_ENDIAN__
8200	8221	__ai __attribute__((target("neon"))) int8x16_t vcntq_s8(int8x16_t __p0) {
8201	8222	int8x16_t __ret;
8202		__ret = (int8x16_t) __builtin_neon_vcntq_v((int8x16_t)__p0, 32);
	8223	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vcntq_v(__builtin_bit_cast(int8x16_t, __p0), 32));
8203	8224	return __ret;
8204	8225	}
8205	8226	#else
8206	8227	__ai __attribute__((target("neon"))) int8x16_t vcntq_s8(int8x16_t __p0) {
8207	8228	int8x16_t __ret;
8208		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
8209		__ret = (int8x16_t) __builtin_neon_vcntq_v((int8x16_t)__rev0, 32);
8210		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	8229	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	8230	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vcntq_v(__builtin_bit_cast(int8x16_t, __rev0), 32));
	8231	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
8211	8232	return __ret;
8212	8233	}
8213	8234	#endif
...	...	@@ -8215,15 +8236,15 @@ __ai __attribute__((target("neon"))) int8x16_t vcntq_s8(int8x16_t __p0) {
8215	8236	#ifdef __LITTLE_ENDIAN__
8216	8237	__ai __attribute__((target("neon"))) uint8x8_t vcnt_u8(uint8x8_t __p0) {
8217	8238	uint8x8_t __ret;
8218		__ret = (uint8x8_t) __builtin_neon_vcnt_v((int8x8_t)__p0, 16);
	8239	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vcnt_v(__builtin_bit_cast(int8x8_t, __p0), 16));
8219	8240	return __ret;
8220	8241	}
8221	8242	#else
8222	8243	__ai __attribute__((target("neon"))) uint8x8_t vcnt_u8(uint8x8_t __p0) {
8223	8244	uint8x8_t __ret;
8224		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
8225		__ret = (uint8x8_t) __builtin_neon_vcnt_v((int8x8_t)__rev0, 16);
8226		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	8245	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	8246	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vcnt_v(__builtin_bit_cast(int8x8_t, __rev0), 16));
	8247	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
8227	8248	return __ret;
8228	8249	}
8229	8250	#endif
...	...	@@ -8231,15 +8252,15 @@ __ai __attribute__((target("neon"))) uint8x8_t vcnt_u8(uint8x8_t __p0) {
8231	8252	#ifdef __LITTLE_ENDIAN__
8232	8253	__ai __attribute__((target("neon"))) int8x8_t vcnt_s8(int8x8_t __p0) {
8233	8254	int8x8_t __ret;
8234		__ret = (int8x8_t) __builtin_neon_vcnt_v((int8x8_t)__p0, 0);
	8255	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vcnt_v(__builtin_bit_cast(int8x8_t, __p0), 0));
8235	8256	return __ret;
8236	8257	}
8237	8258	#else
8238	8259	__ai __attribute__((target("neon"))) int8x8_t vcnt_s8(int8x8_t __p0) {
8239	8260	int8x8_t __ret;
8240		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
8241		__ret = (int8x8_t) __builtin_neon_vcnt_v((int8x8_t)__rev0, 0);
8242		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	8261	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	8262	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vcnt_v(__builtin_bit_cast(int8x8_t, __rev0), 0));
	8263	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
8243	8264	return __ret;
8244	8265	}
8245	8266	#endif
...	...	@@ -8253,10 +8274,10 @@ __ai __attribute__((target("neon"))) poly8x16_t vcombine_p8(poly8x8_t __p0, poly
8253	8274	#else
8254	8275	__ai __attribute__((target("neon"))) poly8x16_t vcombine_p8(poly8x8_t __p0, poly8x8_t __p1) {
8255	8276	poly8x16_t __ret;
8256		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
8257		poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	8277	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	8278	poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
8258	8279	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15);
8259		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	8280	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
8260	8281	return __ret;
8261	8282	}
8262	8283	#endif
...	...	@@ -8270,10 +8291,10 @@ __ai __attribute__((target("neon"))) poly16x8_t vcombine_p16(poly16x4_t __p0, po
8270	8291	#else
8271	8292	__ai __attribute__((target("neon"))) poly16x8_t vcombine_p16(poly16x4_t __p0, poly16x4_t __p1) {
8272	8293	poly16x8_t __ret;
8273		poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
8274		poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	8294	poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	8295	poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
8275	8296	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 1, 2, 3, 4, 5, 6, 7);
8276		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	8297	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
8277	8298	return __ret;
8278	8299	}
8279	8300	#endif
...	...	@@ -8287,10 +8308,10 @@ __ai __attribute__((target("neon"))) uint8x16_t vcombine_u8(uint8x8_t __p0, uint
8287	8308	#else
8288	8309	__ai __attribute__((target("neon"))) uint8x16_t vcombine_u8(uint8x8_t __p0, uint8x8_t __p1) {
8289	8310	uint8x16_t __ret;
8290		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
8291		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	8311	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	8312	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
8292	8313	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15);
8293		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	8314	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
8294	8315	return __ret;
8295	8316	}
8296	8317	__ai __attribute__((target("neon"))) uint8x16_t __noswap_vcombine_u8(uint8x8_t __p0, uint8x8_t __p1) {
...	...	@@ -8309,10 +8330,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vcombine_u32(uint32x2_t __p0, ui
8309	8330	#else
8310	8331	__ai __attribute__((target("neon"))) uint32x4_t vcombine_u32(uint32x2_t __p0, uint32x2_t __p1) {
8311	8332	uint32x4_t __ret;
8312		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
8313		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	8333	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	8334	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
8314	8335	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 1, 2, 3);
8315		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	8336	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
8316	8337	return __ret;
8317	8338	}
8318	8339	__ai __attribute__((target("neon"))) uint32x4_t __noswap_vcombine_u32(uint32x2_t __p0, uint32x2_t __p1) {
...	...	@@ -8332,7 +8353,7 @@ __ai __attribute__((target("neon"))) uint64x2_t vcombine_u64(uint64x1_t __p0, ui
8332	8353	__ai __attribute__((target("neon"))) uint64x2_t vcombine_u64(uint64x1_t __p0, uint64x1_t __p1) {
8333	8354	uint64x2_t __ret;
8334	8355	__ret = __builtin_shufflevector(__p0, __p1, 0, 1);
8335		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	8356	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
8336	8357	return __ret;
8337	8358	}
8338	8359	#endif
...	...	@@ -8346,10 +8367,10 @@ __ai __attribute__((target("neon"))) uint16x8_t vcombine_u16(uint16x4_t __p0, ui
8346	8367	#else
8347	8368	__ai __attribute__((target("neon"))) uint16x8_t vcombine_u16(uint16x4_t __p0, uint16x4_t __p1) {
8348	8369	uint16x8_t __ret;
8349		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
8350		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	8370	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	8371	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
8351	8372	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 1, 2, 3, 4, 5, 6, 7);
8352		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	8373	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
8353	8374	return __ret;
8354	8375	}
8355	8376	__ai __attribute__((target("neon"))) uint16x8_t __noswap_vcombine_u16(uint16x4_t __p0, uint16x4_t __p1) {
...	...	@@ -8368,10 +8389,10 @@ __ai __attribute__((target("neon"))) int8x16_t vcombine_s8(int8x8_t __p0, int8x8
8368	8389	#else
8369	8390	__ai __attribute__((target("neon"))) int8x16_t vcombine_s8(int8x8_t __p0, int8x8_t __p1) {
8370	8391	int8x16_t __ret;
8371		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
8372		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	8392	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	8393	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
8373	8394	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15);
8374		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	8395	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
8375	8396	return __ret;
8376	8397	}
8377	8398	__ai __attribute__((target("neon"))) int8x16_t __noswap_vcombine_s8(int8x8_t __p0, int8x8_t __p1) {
...	...	@@ -8390,10 +8411,10 @@ __ai __attribute__((target("neon"))) float32x4_t vcombine_f32(float32x2_t __p0,
8390	8411	#else
8391	8412	__ai __attribute__((target("neon"))) float32x4_t vcombine_f32(float32x2_t __p0, float32x2_t __p1) {
8392	8413	float32x4_t __ret;
8393		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
8394		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	8414	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	8415	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
8395	8416	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 1, 2, 3);
8396		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	8417	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
8397	8418	return __ret;
8398	8419	}
8399	8420	__ai __attribute__((target("neon"))) float32x4_t __noswap_vcombine_f32(float32x2_t __p0, float32x2_t __p1) {
...	...	@@ -8412,10 +8433,10 @@ __ai __attribute__((target("neon"))) float16x8_t vcombine_f16(float16x4_t __p0,
8412	8433	#else
8413	8434	__ai __attribute__((target("neon"))) float16x8_t vcombine_f16(float16x4_t __p0, float16x4_t __p1) {
8414	8435	float16x8_t __ret;
8415		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
8416		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	8436	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	8437	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
8417	8438	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 1, 2, 3, 4, 5, 6, 7);
8418		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	8439	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
8419	8440	return __ret;
8420	8441	}
8421	8442	__ai __attribute__((target("neon"))) float16x8_t __noswap_vcombine_f16(float16x4_t __p0, float16x4_t __p1) {
...	...	@@ -8434,10 +8455,10 @@ __ai __attribute__((target("neon"))) int32x4_t vcombine_s32(int32x2_t __p0, int3
8434	8455	#else
8435	8456	__ai __attribute__((target("neon"))) int32x4_t vcombine_s32(int32x2_t __p0, int32x2_t __p1) {
8436	8457	int32x4_t __ret;
8437		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
8438		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	8458	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	8459	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
8439	8460	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 1, 2, 3);
8440		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	8461	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
8441	8462	return __ret;
8442	8463	}
8443	8464	__ai __attribute__((target("neon"))) int32x4_t __noswap_vcombine_s32(int32x2_t __p0, int32x2_t __p1) {
...	...	@@ -8457,7 +8478,7 @@ __ai __attribute__((target("neon"))) int64x2_t vcombine_s64(int64x1_t __p0, int6
8457	8478	__ai __attribute__((target("neon"))) int64x2_t vcombine_s64(int64x1_t __p0, int64x1_t __p1) {
8458	8479	int64x2_t __ret;
8459	8480	__ret = __builtin_shufflevector(__p0, __p1, 0, 1);
8460		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	8481	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
8461	8482	return __ret;
8462	8483	}
8463	8484	#endif
...	...	@@ -8471,10 +8492,10 @@ __ai __attribute__((target("neon"))) int16x8_t vcombine_s16(int16x4_t __p0, int1
8471	8492	#else
8472	8493	__ai __attribute__((target("neon"))) int16x8_t vcombine_s16(int16x4_t __p0, int16x4_t __p1) {
8473	8494	int16x8_t __ret;
8474		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
8475		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	8495	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	8496	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
8476	8497	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 1, 2, 3, 4, 5, 6, 7);
8477		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	8498	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
8478	8499	return __ret;
8479	8500	}
8480	8501	__ai __attribute__((target("neon"))) int16x8_t __noswap_vcombine_s16(int16x4_t __p0, int16x4_t __p1) {
...	...	@@ -8487,87 +8508,87 @@ __ai __attribute__((target("neon"))) int16x8_t __noswap_vcombine_s16(int16x4_t _
8487	8508	#define vcreate_p8(__p0) __extension__ ({ \
8488	8509	poly8x8_t __ret; \
8489	8510	uint64_t __promote = __p0; \
8490		__ret = (poly8x8_t)(__promote); \
	8511	__ret = __builtin_bit_cast(poly8x8_t, __promote); \
8491	8512	__ret; \
8492	8513	})
8493	8514	#define vcreate_p16(__p0) __extension__ ({ \
8494	8515	poly16x4_t __ret; \
8495	8516	uint64_t __promote = __p0; \
8496		__ret = (poly16x4_t)(__promote); \
	8517	__ret = __builtin_bit_cast(poly16x4_t, __promote); \
8497	8518	__ret; \
8498	8519	})
8499	8520	#define vcreate_u8(__p0) __extension__ ({ \
8500	8521	uint8x8_t __ret; \
8501	8522	uint64_t __promote = __p0; \
8502		__ret = (uint8x8_t)(__promote); \
	8523	__ret = __builtin_bit_cast(uint8x8_t, __promote); \
8503	8524	__ret; \
8504	8525	})
8505	8526	#define vcreate_u32(__p0) __extension__ ({ \
8506	8527	uint32x2_t __ret; \
8507	8528	uint64_t __promote = __p0; \
8508		__ret = (uint32x2_t)(__promote); \
	8529	__ret = __builtin_bit_cast(uint32x2_t, __promote); \
8509	8530	__ret; \
8510	8531	})
8511	8532	#define vcreate_u64(__p0) __extension__ ({ \
8512	8533	uint64x1_t __ret; \
8513	8534	uint64_t __promote = __p0; \
8514		__ret = (uint64x1_t)(__promote); \
	8535	__ret = __builtin_bit_cast(uint64x1_t, __promote); \
8515	8536	__ret; \
8516	8537	})
8517	8538	#define vcreate_u16(__p0) __extension__ ({ \
8518	8539	uint16x4_t __ret; \
8519	8540	uint64_t __promote = __p0; \
8520		__ret = (uint16x4_t)(__promote); \
	8541	__ret = __builtin_bit_cast(uint16x4_t, __promote); \
8521	8542	__ret; \
8522	8543	})
8523	8544	#define vcreate_s8(__p0) __extension__ ({ \
8524	8545	int8x8_t __ret; \
8525	8546	uint64_t __promote = __p0; \
8526		__ret = (int8x8_t)(__promote); \
	8547	__ret = __builtin_bit_cast(int8x8_t, __promote); \
8527	8548	__ret; \
8528	8549	})
8529	8550	#define vcreate_f32(__p0) __extension__ ({ \
8530	8551	float32x2_t __ret; \
8531	8552	uint64_t __promote = __p0; \
8532		__ret = (float32x2_t)(__promote); \
	8553	__ret = __builtin_bit_cast(float32x2_t, __promote); \
8533	8554	__ret; \
8534	8555	})
8535	8556	#define vcreate_f16(__p0) __extension__ ({ \
8536	8557	float16x4_t __ret; \
8537	8558	uint64_t __promote = __p0; \
8538		__ret = (float16x4_t)(__promote); \
	8559	__ret = __builtin_bit_cast(float16x4_t, __promote); \
8539	8560	__ret; \
8540	8561	})
8541	8562	#define vcreate_s32(__p0) __extension__ ({ \
8542	8563	int32x2_t __ret; \
8543	8564	uint64_t __promote = __p0; \
8544		__ret = (int32x2_t)(__promote); \
	8565	__ret = __builtin_bit_cast(int32x2_t, __promote); \
8545	8566	__ret; \
8546	8567	})
8547	8568	#define vcreate_s64(__p0) __extension__ ({ \
8548	8569	int64x1_t __ret; \
8549	8570	uint64_t __promote = __p0; \
8550		__ret = (int64x1_t)(__promote); \
	8571	__ret = __builtin_bit_cast(int64x1_t, __promote); \
8551	8572	__ret; \
8552	8573	})
8553	8574	#define vcreate_s16(__p0) __extension__ ({ \
8554	8575	int16x4_t __ret; \
8555	8576	uint64_t __promote = __p0; \
8556		__ret = (int16x4_t)(__promote); \
	8577	__ret = __builtin_bit_cast(int16x4_t, __promote); \
8557	8578	__ret; \
8558	8579	})
8559	8580	#ifdef __LITTLE_ENDIAN__
8560	8581	__ai __attribute__((target("neon"))) float32x4_t vcvtq_f32_u32(uint32x4_t __p0) {
8561	8582	float32x4_t __ret;
8562		__ret = (float32x4_t) __builtin_neon_vcvtq_f32_v((int8x16_t)__p0, 50);
	8583	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vcvtq_f32_v(__builtin_bit_cast(int8x16_t, __p0), 50));
8563	8584	return __ret;
8564	8585	}
8565	8586	#else
8566	8587	__ai __attribute__((target("neon"))) float32x4_t vcvtq_f32_u32(uint32x4_t __p0) {
8567	8588	float32x4_t __ret;
8568		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
8569		__ret = (float32x4_t) __builtin_neon_vcvtq_f32_v((int8x16_t)__rev0, 50);
8570		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	8589	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	8590	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vcvtq_f32_v(__builtin_bit_cast(int8x16_t, __rev0), 50));
	8591	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
8571	8592	return __ret;
8572	8593	}
8573	8594	#endif
...	...	@@ -8575,15 +8596,15 @@ __ai __attribute__((target("neon"))) float32x4_t vcvtq_f32_u32(uint32x4_t __p0)
8575	8596	#ifdef __LITTLE_ENDIAN__
8576	8597	__ai __attribute__((target("neon"))) float32x4_t vcvtq_f32_s32(int32x4_t __p0) {
8577	8598	float32x4_t __ret;
8578		__ret = (float32x4_t) __builtin_neon_vcvtq_f32_v((int8x16_t)__p0, 34);
	8599	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vcvtq_f32_v(__builtin_bit_cast(int8x16_t, __p0), 34));
8579	8600	return __ret;
8580	8601	}
8581	8602	#else
8582	8603	__ai __attribute__((target("neon"))) float32x4_t vcvtq_f32_s32(int32x4_t __p0) {
8583	8604	float32x4_t __ret;
8584		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
8585		__ret = (float32x4_t) __builtin_neon_vcvtq_f32_v((int8x16_t)__rev0, 34);
8586		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	8605	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	8606	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vcvtq_f32_v(__builtin_bit_cast(int8x16_t, __rev0), 34));
	8607	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
8587	8608	return __ret;
8588	8609	}
8589	8610	#endif
...	...	@@ -8591,15 +8612,15 @@ __ai __attribute__((target("neon"))) float32x4_t vcvtq_f32_s32(int32x4_t __p0) {
8591	8612	#ifdef __LITTLE_ENDIAN__
8592	8613	__ai __attribute__((target("neon"))) float32x2_t vcvt_f32_u32(uint32x2_t __p0) {
8593	8614	float32x2_t __ret;
8594		__ret = (float32x2_t) __builtin_neon_vcvt_f32_v((int8x8_t)__p0, 18);
	8615	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vcvt_f32_v(__builtin_bit_cast(int8x8_t, __p0), 18));
8595	8616	return __ret;
8596	8617	}
8597	8618	#else
8598	8619	__ai __attribute__((target("neon"))) float32x2_t vcvt_f32_u32(uint32x2_t __p0) {
8599	8620	float32x2_t __ret;
8600		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
8601		__ret = (float32x2_t) __builtin_neon_vcvt_f32_v((int8x8_t)__rev0, 18);
8602		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	8621	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	8622	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vcvt_f32_v(__builtin_bit_cast(int8x8_t, __rev0), 18));
	8623	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
8603	8624	return __ret;
8604	8625	}
8605	8626	#endif
...	...	@@ -8607,15 +8628,15 @@ __ai __attribute__((target("neon"))) float32x2_t vcvt_f32_u32(uint32x2_t __p0) {
8607	8628	#ifdef __LITTLE_ENDIAN__
8608	8629	__ai __attribute__((target("neon"))) float32x2_t vcvt_f32_s32(int32x2_t __p0) {
8609	8630	float32x2_t __ret;
8610		__ret = (float32x2_t) __builtin_neon_vcvt_f32_v((int8x8_t)__p0, 2);
	8631	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vcvt_f32_v(__builtin_bit_cast(int8x8_t, __p0), 2));
8611	8632	return __ret;
8612	8633	}
8613	8634	#else
8614	8635	__ai __attribute__((target("neon"))) float32x2_t vcvt_f32_s32(int32x2_t __p0) {
8615	8636	float32x2_t __ret;
8616		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
8617		__ret = (float32x2_t) __builtin_neon_vcvt_f32_v((int8x8_t)__rev0, 2);
8618		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	8637	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	8638	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vcvt_f32_v(__builtin_bit_cast(int8x8_t, __rev0), 2));
	8639	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
8619	8640	return __ret;
8620	8641	}
8621	8642	#endif
...	...	@@ -8624,16 +8645,16 @@ __ai __attribute__((target("neon"))) float32x2_t vcvt_f32_s32(int32x2_t __p0) {
8624	8645	#define vcvtq_n_f32_u32(__p0, __p1) __extension__ ({ \
8625	8646	float32x4_t __ret; \
8626	8647	uint32x4_t __s0 = __p0; \
8627		__ret = (float32x4_t) __builtin_neon_vcvtq_n_f32_v((int8x16_t)__s0, __p1, 50); \
	8648	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vcvtq_n_f32_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 50)); \
8628	8649	__ret; \
8629	8650	})
8630	8651	#else
8631	8652	#define vcvtq_n_f32_u32(__p0, __p1) __extension__ ({ \
8632	8653	float32x4_t __ret; \
8633	8654	uint32x4_t __s0 = __p0; \
8634		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
8635		__ret = (float32x4_t) __builtin_neon_vcvtq_n_f32_v((int8x16_t)__rev0, __p1, 50); \
8636		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	8655	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	8656	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vcvtq_n_f32_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 50)); \
	8657	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
8637	8658	__ret; \
8638	8659	})
8639	8660	#endif
...	...	@@ -8642,16 +8663,16 @@ __ai __attribute__((target("neon"))) float32x2_t vcvt_f32_s32(int32x2_t __p0) {
8642	8663	#define vcvtq_n_f32_s32(__p0, __p1) __extension__ ({ \
8643	8664	float32x4_t __ret; \
8644	8665	int32x4_t __s0 = __p0; \
8645		__ret = (float32x4_t) __builtin_neon_vcvtq_n_f32_v((int8x16_t)__s0, __p1, 34); \
	8666	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vcvtq_n_f32_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 34)); \
8646	8667	__ret; \
8647	8668	})
8648	8669	#else
8649	8670	#define vcvtq_n_f32_s32(__p0, __p1) __extension__ ({ \
8650	8671	float32x4_t __ret; \
8651	8672	int32x4_t __s0 = __p0; \
8652		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
8653		__ret = (float32x4_t) __builtin_neon_vcvtq_n_f32_v((int8x16_t)__rev0, __p1, 34); \
8654		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	8673	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	8674	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vcvtq_n_f32_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 34)); \
	8675	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
8655	8676	__ret; \
8656	8677	})
8657	8678	#endif
...	...	@@ -8660,16 +8681,16 @@ __ai __attribute__((target("neon"))) float32x2_t vcvt_f32_s32(int32x2_t __p0) {
8660	8681	#define vcvt_n_f32_u32(__p0, __p1) __extension__ ({ \
8661	8682	float32x2_t __ret; \
8662	8683	uint32x2_t __s0 = __p0; \
8663		__ret = (float32x2_t) __builtin_neon_vcvt_n_f32_v((int8x8_t)__s0, __p1, 18); \
	8684	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vcvt_n_f32_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 18)); \
8664	8685	__ret; \
8665	8686	})
8666	8687	#else
8667	8688	#define vcvt_n_f32_u32(__p0, __p1) __extension__ ({ \
8668	8689	float32x2_t __ret; \
8669	8690	uint32x2_t __s0 = __p0; \
8670		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
8671		__ret = (float32x2_t) __builtin_neon_vcvt_n_f32_v((int8x8_t)__rev0, __p1, 18); \
8672		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	8691	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	8692	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vcvt_n_f32_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 18)); \
	8693	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
8673	8694	__ret; \
8674	8695	})
8675	8696	#endif
...	...	@@ -8678,16 +8699,16 @@ __ai __attribute__((target("neon"))) float32x2_t vcvt_f32_s32(int32x2_t __p0) {
8678	8699	#define vcvt_n_f32_s32(__p0, __p1) __extension__ ({ \
8679	8700	float32x2_t __ret; \
8680	8701	int32x2_t __s0 = __p0; \
8681		__ret = (float32x2_t) __builtin_neon_vcvt_n_f32_v((int8x8_t)__s0, __p1, 2); \
	8702	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vcvt_n_f32_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 2)); \
8682	8703	__ret; \
8683	8704	})
8684	8705	#else
8685	8706	#define vcvt_n_f32_s32(__p0, __p1) __extension__ ({ \
8686	8707	float32x2_t __ret; \
8687	8708	int32x2_t __s0 = __p0; \
8688		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
8689		__ret = (float32x2_t) __builtin_neon_vcvt_n_f32_v((int8x8_t)__rev0, __p1, 2); \
8690		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	8709	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	8710	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vcvt_n_f32_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 2)); \
	8711	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
8691	8712	__ret; \
8692	8713	})
8693	8714	#endif
...	...	@@ -8696,16 +8717,16 @@ __ai __attribute__((target("neon"))) float32x2_t vcvt_f32_s32(int32x2_t __p0) {
8696	8717	#define vcvtq_n_s32_f32(__p0, __p1) __extension__ ({ \
8697	8718	int32x4_t __ret; \
8698	8719	float32x4_t __s0 = __p0; \
8699		__ret = (int32x4_t) __builtin_neon_vcvtq_n_s32_v((int8x16_t)__s0, __p1, 34); \
	8720	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vcvtq_n_s32_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 34)); \
8700	8721	__ret; \
8701	8722	})
8702	8723	#else
8703	8724	#define vcvtq_n_s32_f32(__p0, __p1) __extension__ ({ \
8704	8725	int32x4_t __ret; \
8705	8726	float32x4_t __s0 = __p0; \
8706		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
8707		__ret = (int32x4_t) __builtin_neon_vcvtq_n_s32_v((int8x16_t)__rev0, __p1, 34); \
8708		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	8727	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	8728	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vcvtq_n_s32_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 34)); \
	8729	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
8709	8730	__ret; \
8710	8731	})
8711	8732	#endif
...	...	@@ -8714,16 +8735,16 @@ __ai __attribute__((target("neon"))) float32x2_t vcvt_f32_s32(int32x2_t __p0) {
8714	8735	#define vcvt_n_s32_f32(__p0, __p1) __extension__ ({ \
8715	8736	int32x2_t __ret; \
8716	8737	float32x2_t __s0 = __p0; \
8717		__ret = (int32x2_t) __builtin_neon_vcvt_n_s32_v((int8x8_t)__s0, __p1, 2); \
	8738	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vcvt_n_s32_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 2)); \
8718	8739	__ret; \
8719	8740	})
8720	8741	#else
8721	8742	#define vcvt_n_s32_f32(__p0, __p1) __extension__ ({ \
8722	8743	int32x2_t __ret; \
8723	8744	float32x2_t __s0 = __p0; \
8724		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
8725		__ret = (int32x2_t) __builtin_neon_vcvt_n_s32_v((int8x8_t)__rev0, __p1, 2); \
8726		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	8745	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	8746	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vcvt_n_s32_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 2)); \
	8747	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
8727	8748	__ret; \
8728	8749	})
8729	8750	#endif
...	...	@@ -8732,16 +8753,16 @@ __ai __attribute__((target("neon"))) float32x2_t vcvt_f32_s32(int32x2_t __p0) {
8732	8753	#define vcvtq_n_u32_f32(__p0, __p1) __extension__ ({ \
8733	8754	uint32x4_t __ret; \
8734	8755	float32x4_t __s0 = __p0; \
8735		__ret = (uint32x4_t) __builtin_neon_vcvtq_n_u32_v((int8x16_t)__s0, __p1, 50); \
	8756	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcvtq_n_u32_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 50)); \
8736	8757	__ret; \
8737	8758	})
8738	8759	#else
8739	8760	#define vcvtq_n_u32_f32(__p0, __p1) __extension__ ({ \
8740	8761	uint32x4_t __ret; \
8741	8762	float32x4_t __s0 = __p0; \
8742		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
8743		__ret = (uint32x4_t) __builtin_neon_vcvtq_n_u32_v((int8x16_t)__rev0, __p1, 50); \
8744		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	8763	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	8764	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcvtq_n_u32_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 50)); \
	8765	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
8745	8766	__ret; \
8746	8767	})
8747	8768	#endif
...	...	@@ -8750,16 +8771,16 @@ __ai __attribute__((target("neon"))) float32x2_t vcvt_f32_s32(int32x2_t __p0) {
8750	8771	#define vcvt_n_u32_f32(__p0, __p1) __extension__ ({ \
8751	8772	uint32x2_t __ret; \
8752	8773	float32x2_t __s0 = __p0; \
8753		__ret = (uint32x2_t) __builtin_neon_vcvt_n_u32_v((int8x8_t)__s0, __p1, 18); \
	8774	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcvt_n_u32_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 18)); \
8754	8775	__ret; \
8755	8776	})
8756	8777	#else
8757	8778	#define vcvt_n_u32_f32(__p0, __p1) __extension__ ({ \
8758	8779	uint32x2_t __ret; \
8759	8780	float32x2_t __s0 = __p0; \
8760		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
8761		__ret = (uint32x2_t) __builtin_neon_vcvt_n_u32_v((int8x8_t)__rev0, __p1, 18); \
8762		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	8781	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	8782	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcvt_n_u32_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 18)); \
	8783	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
8763	8784	__ret; \
8764	8785	})
8765	8786	#endif
...	...	@@ -8767,15 +8788,15 @@ __ai __attribute__((target("neon"))) float32x2_t vcvt_f32_s32(int32x2_t __p0) {
8767	8788	#ifdef __LITTLE_ENDIAN__
8768	8789	__ai __attribute__((target("neon"))) int32x4_t vcvtq_s32_f32(float32x4_t __p0) {
8769	8790	int32x4_t __ret;
8770		__ret = (int32x4_t) __builtin_neon_vcvtq_s32_v((int8x16_t)__p0, 34);
	8791	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vcvtq_s32_v(__builtin_bit_cast(int8x16_t, __p0), 34));
8771	8792	return __ret;
8772	8793	}
8773	8794	#else
8774	8795	__ai __attribute__((target("neon"))) int32x4_t vcvtq_s32_f32(float32x4_t __p0) {
8775	8796	int32x4_t __ret;
8776		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
8777		__ret = (int32x4_t) __builtin_neon_vcvtq_s32_v((int8x16_t)__rev0, 34);
8778		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	8797	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	8798	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vcvtq_s32_v(__builtin_bit_cast(int8x16_t, __rev0), 34));
	8799	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
8779	8800	return __ret;
8780	8801	}
8781	8802	#endif
...	...	@@ -8783,15 +8804,15 @@ __ai __attribute__((target("neon"))) int32x4_t vcvtq_s32_f32(float32x4_t __p0) {
8783	8804	#ifdef __LITTLE_ENDIAN__
8784	8805	__ai __attribute__((target("neon"))) int32x2_t vcvt_s32_f32(float32x2_t __p0) {
8785	8806	int32x2_t __ret;
8786		__ret = (int32x2_t) __builtin_neon_vcvt_s32_v((int8x8_t)__p0, 2);
	8807	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vcvt_s32_v(__builtin_bit_cast(int8x8_t, __p0), 2));
8787	8808	return __ret;
8788	8809	}
8789	8810	#else
8790	8811	__ai __attribute__((target("neon"))) int32x2_t vcvt_s32_f32(float32x2_t __p0) {
8791	8812	int32x2_t __ret;
8792		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
8793		__ret = (int32x2_t) __builtin_neon_vcvt_s32_v((int8x8_t)__rev0, 2);
8794		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	8813	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	8814	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vcvt_s32_v(__builtin_bit_cast(int8x8_t, __rev0), 2));
	8815	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
8795	8816	return __ret;
8796	8817	}
8797	8818	#endif
...	...	@@ -8799,15 +8820,15 @@ __ai __attribute__((target("neon"))) int32x2_t vcvt_s32_f32(float32x2_t __p0) {
8799	8820	#ifdef __LITTLE_ENDIAN__
8800	8821	__ai __attribute__((target("neon"))) uint32x4_t vcvtq_u32_f32(float32x4_t __p0) {
8801	8822	uint32x4_t __ret;
8802		__ret = (uint32x4_t) __builtin_neon_vcvtq_u32_v((int8x16_t)__p0, 50);
	8823	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcvtq_u32_v(__builtin_bit_cast(int8x16_t, __p0), 50));
8803	8824	return __ret;
8804	8825	}
8805	8826	#else
8806	8827	__ai __attribute__((target("neon"))) uint32x4_t vcvtq_u32_f32(float32x4_t __p0) {
8807	8828	uint32x4_t __ret;
8808		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
8809		__ret = (uint32x4_t) __builtin_neon_vcvtq_u32_v((int8x16_t)__rev0, 50);
8810		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	8829	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	8830	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcvtq_u32_v(__builtin_bit_cast(int8x16_t, __rev0), 50));
	8831	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
8811	8832	return __ret;
8812	8833	}
8813	8834	#endif
...	...	@@ -8815,15 +8836,15 @@ __ai __attribute__((target("neon"))) uint32x4_t vcvtq_u32_f32(float32x4_t __p0)
8815	8836	#ifdef __LITTLE_ENDIAN__
8816	8837	__ai __attribute__((target("neon"))) uint32x2_t vcvt_u32_f32(float32x2_t __p0) {
8817	8838	uint32x2_t __ret;
8818		__ret = (uint32x2_t) __builtin_neon_vcvt_u32_v((int8x8_t)__p0, 18);
	8839	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcvt_u32_v(__builtin_bit_cast(int8x8_t, __p0), 18));
8819	8840	return __ret;
8820	8841	}
8821	8842	#else
8822	8843	__ai __attribute__((target("neon"))) uint32x2_t vcvt_u32_f32(float32x2_t __p0) {
8823	8844	uint32x2_t __ret;
8824		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
8825		__ret = (uint32x2_t) __builtin_neon_vcvt_u32_v((int8x8_t)__rev0, 18);
8826		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	8845	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	8846	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcvt_u32_v(__builtin_bit_cast(int8x8_t, __rev0), 18));
	8847	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
8827	8848	return __ret;
8828	8849	}
8829	8850	#endif
...	...	@@ -8839,9 +8860,9 @@ __ai __attribute__((target("neon"))) uint32x2_t vcvt_u32_f32(float32x2_t __p0) {
8839	8860	#define vdup_lane_p8(__p0_9, __p1_9) __extension__ ({ \
8840	8861	poly8x8_t __ret_9; \
8841	8862	poly8x8_t __s0_9 = __p0_9; \
8842		poly8x8_t __rev0_9; __rev0_9 = __builtin_shufflevector(__s0_9, __s0_9, 7, 6, 5, 4, 3, 2, 1, 0); \
	8863	poly8x8_t __rev0_9; __rev0_9 = __builtin_shufflevector(__s0_9, __s0_9, __lane_reverse_64_8); \
8843	8864	__ret_9 = __noswap_splat_lane_p8(__rev0_9, __p1_9); \
8844		__ret_9 = __builtin_shufflevector(__ret_9, __ret_9, 7, 6, 5, 4, 3, 2, 1, 0); \
	8865	__ret_9 = __builtin_shufflevector(__ret_9, __ret_9, __lane_reverse_64_8); \
8845	8866	__ret_9; \
8846	8867	})
8847	8868	#endif
...	...	@@ -8857,9 +8878,9 @@ __ai __attribute__((target("neon"))) uint32x2_t vcvt_u32_f32(float32x2_t __p0) {
8857	8878	#define vdup_lane_p16(__p0_11, __p1_11) __extension__ ({ \
8858	8879	poly16x4_t __ret_11; \
8859	8880	poly16x4_t __s0_11 = __p0_11; \
8860		poly16x4_t __rev0_11; __rev0_11 = __builtin_shufflevector(__s0_11, __s0_11, 3, 2, 1, 0); \
	8881	poly16x4_t __rev0_11; __rev0_11 = __builtin_shufflevector(__s0_11, __s0_11, __lane_reverse_64_16); \
8861	8882	__ret_11 = __noswap_splat_lane_p16(__rev0_11, __p1_11); \
8862		__ret_11 = __builtin_shufflevector(__ret_11, __ret_11, 3, 2, 1, 0); \
	8883	__ret_11 = __builtin_shufflevector(__ret_11, __ret_11, __lane_reverse_64_16); \
8863	8884	__ret_11; \
8864	8885	})
8865	8886	#endif
...	...	@@ -8875,9 +8896,9 @@ __ai __attribute__((target("neon"))) uint32x2_t vcvt_u32_f32(float32x2_t __p0) {
8875	8896	#define vdupq_lane_p8(__p0_13, __p1_13) __extension__ ({ \
8876	8897	poly8x16_t __ret_13; \
8877	8898	poly8x8_t __s0_13 = __p0_13; \
8878		poly8x8_t __rev0_13; __rev0_13 = __builtin_shufflevector(__s0_13, __s0_13, 7, 6, 5, 4, 3, 2, 1, 0); \
	8899	poly8x8_t __rev0_13; __rev0_13 = __builtin_shufflevector(__s0_13, __s0_13, __lane_reverse_64_8); \
8879	8900	__ret_13 = __noswap_splatq_lane_p8(__rev0_13, __p1_13); \
8880		__ret_13 = __builtin_shufflevector(__ret_13, __ret_13, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	8901	__ret_13 = __builtin_shufflevector(__ret_13, __ret_13, __lane_reverse_128_8); \
8881	8902	__ret_13; \
8882	8903	})
8883	8904	#endif
...	...	@@ -8893,9 +8914,9 @@ __ai __attribute__((target("neon"))) uint32x2_t vcvt_u32_f32(float32x2_t __p0) {
8893	8914	#define vdupq_lane_p16(__p0_15, __p1_15) __extension__ ({ \
8894	8915	poly16x8_t __ret_15; \
8895	8916	poly16x4_t __s0_15 = __p0_15; \
8896		poly16x4_t __rev0_15; __rev0_15 = __builtin_shufflevector(__s0_15, __s0_15, 3, 2, 1, 0); \
	8917	poly16x4_t __rev0_15; __rev0_15 = __builtin_shufflevector(__s0_15, __s0_15, __lane_reverse_64_16); \
8897	8918	__ret_15 = __noswap_splatq_lane_p16(__rev0_15, __p1_15); \
8898		__ret_15 = __builtin_shufflevector(__ret_15, __ret_15, 7, 6, 5, 4, 3, 2, 1, 0); \
	8919	__ret_15 = __builtin_shufflevector(__ret_15, __ret_15, __lane_reverse_128_16); \
8899	8920	__ret_15; \
8900	8921	})
8901	8922	#endif
...	...	@@ -8911,9 +8932,9 @@ __ai __attribute__((target("neon"))) uint32x2_t vcvt_u32_f32(float32x2_t __p0) {
8911	8932	#define vdupq_lane_u8(__p0_17, __p1_17) __extension__ ({ \
8912	8933	uint8x16_t __ret_17; \
8913	8934	uint8x8_t __s0_17 = __p0_17; \
8914		uint8x8_t __rev0_17; __rev0_17 = __builtin_shufflevector(__s0_17, __s0_17, 7, 6, 5, 4, 3, 2, 1, 0); \
	8935	uint8x8_t __rev0_17; __rev0_17 = __builtin_shufflevector(__s0_17, __s0_17, __lane_reverse_64_8); \
8915	8936	__ret_17 = __noswap_splatq_lane_u8(__rev0_17, __p1_17); \
8916		__ret_17 = __builtin_shufflevector(__ret_17, __ret_17, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	8937	__ret_17 = __builtin_shufflevector(__ret_17, __ret_17, __lane_reverse_128_8); \
8917	8938	__ret_17; \
8918	8939	})
8919	8940	#endif
...	...	@@ -8929,9 +8950,9 @@ __ai __attribute__((target("neon"))) uint32x2_t vcvt_u32_f32(float32x2_t __p0) {
8929	8950	#define vdupq_lane_u32(__p0_19, __p1_19) __extension__ ({ \
8930	8951	uint32x4_t __ret_19; \
8931	8952	uint32x2_t __s0_19 = __p0_19; \
8932		uint32x2_t __rev0_19; __rev0_19 = __builtin_shufflevector(__s0_19, __s0_19, 1, 0); \
	8953	uint32x2_t __rev0_19; __rev0_19 = __builtin_shufflevector(__s0_19, __s0_19, __lane_reverse_64_32); \
8933	8954	__ret_19 = __noswap_splatq_lane_u32(__rev0_19, __p1_19); \
8934		__ret_19 = __builtin_shufflevector(__ret_19, __ret_19, 3, 2, 1, 0); \
	8955	__ret_19 = __builtin_shufflevector(__ret_19, __ret_19, __lane_reverse_128_32); \
8935	8956	__ret_19; \
8936	8957	})
8937	8958	#endif
...	...	@@ -8948,7 +8969,7 @@ __ai __attribute__((target("neon"))) uint32x2_t vcvt_u32_f32(float32x2_t __p0) {
8948	8969	uint64x2_t __ret_21; \
8949	8970	uint64x1_t __s0_21 = __p0_21; \
8950	8971	__ret_21 = __noswap_splatq_lane_u64(__s0_21, __p1_21); \
8951		__ret_21 = __builtin_shufflevector(__ret_21, __ret_21, 1, 0); \
	8972	__ret_21 = __builtin_shufflevector(__ret_21, __ret_21, __lane_reverse_128_64); \
8952	8973	__ret_21; \
8953	8974	})
8954	8975	#endif
...	...	@@ -8964,9 +8985,9 @@ __ai __attribute__((target("neon"))) uint32x2_t vcvt_u32_f32(float32x2_t __p0) {
8964	8985	#define vdupq_lane_u16(__p0_23, __p1_23) __extension__ ({ \
8965	8986	uint16x8_t __ret_23; \
8966	8987	uint16x4_t __s0_23 = __p0_23; \
8967		uint16x4_t __rev0_23; __rev0_23 = __builtin_shufflevector(__s0_23, __s0_23, 3, 2, 1, 0); \
	8988	uint16x4_t __rev0_23; __rev0_23 = __builtin_shufflevector(__s0_23, __s0_23, __lane_reverse_64_16); \
8968	8989	__ret_23 = __noswap_splatq_lane_u16(__rev0_23, __p1_23); \
8969		__ret_23 = __builtin_shufflevector(__ret_23, __ret_23, 7, 6, 5, 4, 3, 2, 1, 0); \
	8990	__ret_23 = __builtin_shufflevector(__ret_23, __ret_23, __lane_reverse_128_16); \
8970	8991	__ret_23; \
8971	8992	})
8972	8993	#endif
...	...	@@ -8982,9 +9003,9 @@ __ai __attribute__((target("neon"))) uint32x2_t vcvt_u32_f32(float32x2_t __p0) {
8982	9003	#define vdupq_lane_s8(__p0_25, __p1_25) __extension__ ({ \
8983	9004	int8x16_t __ret_25; \
8984	9005	int8x8_t __s0_25 = __p0_25; \
8985		int8x8_t __rev0_25; __rev0_25 = __builtin_shufflevector(__s0_25, __s0_25, 7, 6, 5, 4, 3, 2, 1, 0); \
	9006	int8x8_t __rev0_25; __rev0_25 = __builtin_shufflevector(__s0_25, __s0_25, __lane_reverse_64_8); \
8986	9007	__ret_25 = __noswap_splatq_lane_s8(__rev0_25, __p1_25); \
8987		__ret_25 = __builtin_shufflevector(__ret_25, __ret_25, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	9008	__ret_25 = __builtin_shufflevector(__ret_25, __ret_25, __lane_reverse_128_8); \
8988	9009	__ret_25; \
8989	9010	})
8990	9011	#endif
...	...	@@ -9000,9 +9021,9 @@ __ai __attribute__((target("neon"))) uint32x2_t vcvt_u32_f32(float32x2_t __p0) {
9000	9021	#define vdupq_lane_f32(__p0_27, __p1_27) __extension__ ({ \
9001	9022	float32x4_t __ret_27; \
9002	9023	float32x2_t __s0_27 = __p0_27; \
9003		float32x2_t __rev0_27; __rev0_27 = __builtin_shufflevector(__s0_27, __s0_27, 1, 0); \
	9024	float32x2_t __rev0_27; __rev0_27 = __builtin_shufflevector(__s0_27, __s0_27, __lane_reverse_64_32); \
9004	9025	__ret_27 = __noswap_splatq_lane_f32(__rev0_27, __p1_27); \
9005		__ret_27 = __builtin_shufflevector(__ret_27, __ret_27, 3, 2, 1, 0); \
	9026	__ret_27 = __builtin_shufflevector(__ret_27, __ret_27, __lane_reverse_128_32); \
9006	9027	__ret_27; \
9007	9028	})
9008	9029	#endif
...	...	@@ -9018,9 +9039,9 @@ __ai __attribute__((target("neon"))) uint32x2_t vcvt_u32_f32(float32x2_t __p0) {
9018	9039	#define vdupq_lane_f16(__p0_29, __p1_29) __extension__ ({ \
9019	9040	float16x8_t __ret_29; \
9020	9041	float16x4_t __s0_29 = __p0_29; \
9021		float16x4_t __rev0_29; __rev0_29 = __builtin_shufflevector(__s0_29, __s0_29, 3, 2, 1, 0); \
	9042	float16x4_t __rev0_29; __rev0_29 = __builtin_shufflevector(__s0_29, __s0_29, __lane_reverse_64_16); \
9022	9043	__ret_29 = __noswap_splatq_lane_f16(__rev0_29, __p1_29); \
9023		__ret_29 = __builtin_shufflevector(__ret_29, __ret_29, 7, 6, 5, 4, 3, 2, 1, 0); \
	9044	__ret_29 = __builtin_shufflevector(__ret_29, __ret_29, __lane_reverse_128_16); \
9024	9045	__ret_29; \
9025	9046	})
9026	9047	#endif
...	...	@@ -9036,9 +9057,9 @@ __ai __attribute__((target("neon"))) uint32x2_t vcvt_u32_f32(float32x2_t __p0) {
9036	9057	#define vdupq_lane_s32(__p0_31, __p1_31) __extension__ ({ \
9037	9058	int32x4_t __ret_31; \
9038	9059	int32x2_t __s0_31 = __p0_31; \
9039		int32x2_t __rev0_31; __rev0_31 = __builtin_shufflevector(__s0_31, __s0_31, 1, 0); \
	9060	int32x2_t __rev0_31; __rev0_31 = __builtin_shufflevector(__s0_31, __s0_31, __lane_reverse_64_32); \
9040	9061	__ret_31 = __noswap_splatq_lane_s32(__rev0_31, __p1_31); \
9041		__ret_31 = __builtin_shufflevector(__ret_31, __ret_31, 3, 2, 1, 0); \
	9062	__ret_31 = __builtin_shufflevector(__ret_31, __ret_31, __lane_reverse_128_32); \
9042	9063	__ret_31; \
9043	9064	})
9044	9065	#endif
...	...	@@ -9055,7 +9076,7 @@ __ai __attribute__((target("neon"))) uint32x2_t vcvt_u32_f32(float32x2_t __p0) {
9055	9076	int64x2_t __ret_33; \
9056	9077	int64x1_t __s0_33 = __p0_33; \
9057	9078	__ret_33 = __noswap_splatq_lane_s64(__s0_33, __p1_33); \
9058		__ret_33 = __builtin_shufflevector(__ret_33, __ret_33, 1, 0); \
	9079	__ret_33 = __builtin_shufflevector(__ret_33, __ret_33, __lane_reverse_128_64); \
9059	9080	__ret_33; \
9060	9081	})
9061	9082	#endif
...	...	@@ -9071,9 +9092,9 @@ __ai __attribute__((target("neon"))) uint32x2_t vcvt_u32_f32(float32x2_t __p0) {
9071	9092	#define vdupq_lane_s16(__p0_35, __p1_35) __extension__ ({ \
9072	9093	int16x8_t __ret_35; \
9073	9094	int16x4_t __s0_35 = __p0_35; \
9074		int16x4_t __rev0_35; __rev0_35 = __builtin_shufflevector(__s0_35, __s0_35, 3, 2, 1, 0); \
	9095	int16x4_t __rev0_35; __rev0_35 = __builtin_shufflevector(__s0_35, __s0_35, __lane_reverse_64_16); \
9075	9096	__ret_35 = __noswap_splatq_lane_s16(__rev0_35, __p1_35); \
9076		__ret_35 = __builtin_shufflevector(__ret_35, __ret_35, 7, 6, 5, 4, 3, 2, 1, 0); \
	9097	__ret_35 = __builtin_shufflevector(__ret_35, __ret_35, __lane_reverse_128_16); \
9077	9098	__ret_35; \
9078	9099	})
9079	9100	#endif
...	...	@@ -9089,9 +9110,9 @@ __ai __attribute__((target("neon"))) uint32x2_t vcvt_u32_f32(float32x2_t __p0) {
9089	9110	#define vdup_lane_u8(__p0_37, __p1_37) __extension__ ({ \
9090	9111	uint8x8_t __ret_37; \
9091	9112	uint8x8_t __s0_37 = __p0_37; \
9092		uint8x8_t __rev0_37; __rev0_37 = __builtin_shufflevector(__s0_37, __s0_37, 7, 6, 5, 4, 3, 2, 1, 0); \
	9113	uint8x8_t __rev0_37; __rev0_37 = __builtin_shufflevector(__s0_37, __s0_37, __lane_reverse_64_8); \
9093	9114	__ret_37 = __noswap_splat_lane_u8(__rev0_37, __p1_37); \
9094		__ret_37 = __builtin_shufflevector(__ret_37, __ret_37, 7, 6, 5, 4, 3, 2, 1, 0); \
	9115	__ret_37 = __builtin_shufflevector(__ret_37, __ret_37, __lane_reverse_64_8); \
9095	9116	__ret_37; \
9096	9117	})
9097	9118	#endif
...	...	@@ -9107,9 +9128,9 @@ __ai __attribute__((target("neon"))) uint32x2_t vcvt_u32_f32(float32x2_t __p0) {
9107	9128	#define vdup_lane_u32(__p0_39, __p1_39) __extension__ ({ \
9108	9129	uint32x2_t __ret_39; \
9109	9130	uint32x2_t __s0_39 = __p0_39; \
9110		uint32x2_t __rev0_39; __rev0_39 = __builtin_shufflevector(__s0_39, __s0_39, 1, 0); \
	9131	uint32x2_t __rev0_39; __rev0_39 = __builtin_shufflevector(__s0_39, __s0_39, __lane_reverse_64_32); \
9111	9132	__ret_39 = __noswap_splat_lane_u32(__rev0_39, __p1_39); \
9112		__ret_39 = __builtin_shufflevector(__ret_39, __ret_39, 1, 0); \
	9133	__ret_39 = __builtin_shufflevector(__ret_39, __ret_39, __lane_reverse_64_32); \
9113	9134	__ret_39; \
9114	9135	})
9115	9136	#endif
...	...	@@ -9131,9 +9152,9 @@ __ai __attribute__((target("neon"))) uint32x2_t vcvt_u32_f32(float32x2_t __p0) {
9131	9152	#define vdup_lane_u16(__p0_42, __p1_42) __extension__ ({ \
9132	9153	uint16x4_t __ret_42; \
9133	9154	uint16x4_t __s0_42 = __p0_42; \
9134		uint16x4_t __rev0_42; __rev0_42 = __builtin_shufflevector(__s0_42, __s0_42, 3, 2, 1, 0); \
	9155	uint16x4_t __rev0_42; __rev0_42 = __builtin_shufflevector(__s0_42, __s0_42, __lane_reverse_64_16); \
9135	9156	__ret_42 = __noswap_splat_lane_u16(__rev0_42, __p1_42); \
9136		__ret_42 = __builtin_shufflevector(__ret_42, __ret_42, 3, 2, 1, 0); \
	9157	__ret_42 = __builtin_shufflevector(__ret_42, __ret_42, __lane_reverse_64_16); \
9137	9158	__ret_42; \
9138	9159	})
9139	9160	#endif
...	...	@@ -9149,9 +9170,9 @@ __ai __attribute__((target("neon"))) uint32x2_t vcvt_u32_f32(float32x2_t __p0) {
9149	9170	#define vdup_lane_s8(__p0_44, __p1_44) __extension__ ({ \
9150	9171	int8x8_t __ret_44; \
9151	9172	int8x8_t __s0_44 = __p0_44; \
9152		int8x8_t __rev0_44; __rev0_44 = __builtin_shufflevector(__s0_44, __s0_44, 7, 6, 5, 4, 3, 2, 1, 0); \
	9173	int8x8_t __rev0_44; __rev0_44 = __builtin_shufflevector(__s0_44, __s0_44, __lane_reverse_64_8); \
9153	9174	__ret_44 = __noswap_splat_lane_s8(__rev0_44, __p1_44); \
9154		__ret_44 = __builtin_shufflevector(__ret_44, __ret_44, 7, 6, 5, 4, 3, 2, 1, 0); \
	9175	__ret_44 = __builtin_shufflevector(__ret_44, __ret_44, __lane_reverse_64_8); \
9155	9176	__ret_44; \
9156	9177	})
9157	9178	#endif
...	...	@@ -9167,9 +9188,9 @@ __ai __attribute__((target("neon"))) uint32x2_t vcvt_u32_f32(float32x2_t __p0) {
9167	9188	#define vdup_lane_f32(__p0_46, __p1_46) __extension__ ({ \
9168	9189	float32x2_t __ret_46; \
9169	9190	float32x2_t __s0_46 = __p0_46; \
9170		float32x2_t __rev0_46; __rev0_46 = __builtin_shufflevector(__s0_46, __s0_46, 1, 0); \
	9191	float32x2_t __rev0_46; __rev0_46 = __builtin_shufflevector(__s0_46, __s0_46, __lane_reverse_64_32); \
9171	9192	__ret_46 = __noswap_splat_lane_f32(__rev0_46, __p1_46); \
9172		__ret_46 = __builtin_shufflevector(__ret_46, __ret_46, 1, 0); \
	9193	__ret_46 = __builtin_shufflevector(__ret_46, __ret_46, __lane_reverse_64_32); \
9173	9194	__ret_46; \
9174	9195	})
9175	9196	#endif
...	...	@@ -9185,9 +9206,9 @@ __ai __attribute__((target("neon"))) uint32x2_t vcvt_u32_f32(float32x2_t __p0) {
9185	9206	#define vdup_lane_f16(__p0_48, __p1_48) __extension__ ({ \
9186	9207	float16x4_t __ret_48; \
9187	9208	float16x4_t __s0_48 = __p0_48; \
9188		float16x4_t __rev0_48; __rev0_48 = __builtin_shufflevector(__s0_48, __s0_48, 3, 2, 1, 0); \
	9209	float16x4_t __rev0_48; __rev0_48 = __builtin_shufflevector(__s0_48, __s0_48, __lane_reverse_64_16); \
9189	9210	__ret_48 = __noswap_splat_lane_f16(__rev0_48, __p1_48); \
9190		__ret_48 = __builtin_shufflevector(__ret_48, __ret_48, 3, 2, 1, 0); \
	9211	__ret_48 = __builtin_shufflevector(__ret_48, __ret_48, __lane_reverse_64_16); \
9191	9212	__ret_48; \
9192	9213	})
9193	9214	#endif
...	...	@@ -9203,9 +9224,9 @@ __ai __attribute__((target("neon"))) uint32x2_t vcvt_u32_f32(float32x2_t __p0) {
9203	9224	#define vdup_lane_s32(__p0_50, __p1_50) __extension__ ({ \
9204	9225	int32x2_t __ret_50; \
9205	9226	int32x2_t __s0_50 = __p0_50; \
9206		int32x2_t __rev0_50; __rev0_50 = __builtin_shufflevector(__s0_50, __s0_50, 1, 0); \
	9227	int32x2_t __rev0_50; __rev0_50 = __builtin_shufflevector(__s0_50, __s0_50, __lane_reverse_64_32); \
9207	9228	__ret_50 = __noswap_splat_lane_s32(__rev0_50, __p1_50); \
9208		__ret_50 = __builtin_shufflevector(__ret_50, __ret_50, 1, 0); \
	9229	__ret_50 = __builtin_shufflevector(__ret_50, __ret_50, __lane_reverse_64_32); \
9209	9230	__ret_50; \
9210	9231	})
9211	9232	#endif
...	...	@@ -9227,9 +9248,9 @@ __ai __attribute__((target("neon"))) uint32x2_t vcvt_u32_f32(float32x2_t __p0) {
9227	9248	#define vdup_lane_s16(__p0_53, __p1_53) __extension__ ({ \
9228	9249	int16x4_t __ret_53; \
9229	9250	int16x4_t __s0_53 = __p0_53; \
9230		int16x4_t __rev0_53; __rev0_53 = __builtin_shufflevector(__s0_53, __s0_53, 3, 2, 1, 0); \
	9251	int16x4_t __rev0_53; __rev0_53 = __builtin_shufflevector(__s0_53, __s0_53, __lane_reverse_64_16); \
9231	9252	__ret_53 = __noswap_splat_lane_s16(__rev0_53, __p1_53); \
9232		__ret_53 = __builtin_shufflevector(__ret_53, __ret_53, 3, 2, 1, 0); \
	9253	__ret_53 = __builtin_shufflevector(__ret_53, __ret_53, __lane_reverse_64_16); \
9233	9254	__ret_53; \
9234	9255	})
9235	9256	#endif
...	...	@@ -9244,7 +9265,7 @@ __ai __attribute__((target("neon"))) poly8x8_t vdup_n_p8(poly8_t __p0) {
9244	9265	__ai __attribute__((target("neon"))) poly8x8_t vdup_n_p8(poly8_t __p0) {
9245	9266	poly8x8_t __ret;
9246	9267	__ret = (poly8x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
9247		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	9268	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
9248	9269	return __ret;
9249	9270	}
9250	9271	#endif
...	...	@@ -9259,7 +9280,7 @@ __ai __attribute__((target("neon"))) poly16x4_t vdup_n_p16(poly16_t __p0) {
9259	9280	__ai __attribute__((target("neon"))) poly16x4_t vdup_n_p16(poly16_t __p0) {
9260	9281	poly16x4_t __ret;
9261	9282	__ret = (poly16x4_t) {__p0, __p0, __p0, __p0};
9262		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	9283	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
9263	9284	return __ret;
9264	9285	}
9265	9286	#endif
...	...	@@ -9274,7 +9295,7 @@ __ai __attribute__((target("neon"))) poly8x16_t vdupq_n_p8(poly8_t __p0) {
9274	9295	__ai __attribute__((target("neon"))) poly8x16_t vdupq_n_p8(poly8_t __p0) {
9275	9296	poly8x16_t __ret;
9276	9297	__ret = (poly8x16_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
9277		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	9298	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
9278	9299	return __ret;
9279	9300	}
9280	9301	#endif
...	...	@@ -9289,7 +9310,7 @@ __ai __attribute__((target("neon"))) poly16x8_t vdupq_n_p16(poly16_t __p0) {
9289	9310	__ai __attribute__((target("neon"))) poly16x8_t vdupq_n_p16(poly16_t __p0) {
9290	9311	poly16x8_t __ret;
9291	9312	__ret = (poly16x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
9292		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	9313	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
9293	9314	return __ret;
9294	9315	}
9295	9316	#endif
...	...	@@ -9304,7 +9325,7 @@ __ai __attribute__((target("neon"))) uint8x16_t vdupq_n_u8(uint8_t __p0) {
9304	9325	__ai __attribute__((target("neon"))) uint8x16_t vdupq_n_u8(uint8_t __p0) {
9305	9326	uint8x16_t __ret;
9306	9327	__ret = (uint8x16_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
9307		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	9328	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
9308	9329	return __ret;
9309	9330	}
9310	9331	#endif
...	...	@@ -9319,7 +9340,7 @@ __ai __attribute__((target("neon"))) uint32x4_t vdupq_n_u32(uint32_t __p0) {
9319	9340	__ai __attribute__((target("neon"))) uint32x4_t vdupq_n_u32(uint32_t __p0) {
9320	9341	uint32x4_t __ret;
9321	9342	__ret = (uint32x4_t) {__p0, __p0, __p0, __p0};
9322		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	9343	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
9323	9344	return __ret;
9324	9345	}
9325	9346	#endif
...	...	@@ -9334,7 +9355,7 @@ __ai __attribute__((target("neon"))) uint64x2_t vdupq_n_u64(uint64_t __p0) {
9334	9355	__ai __attribute__((target("neon"))) uint64x2_t vdupq_n_u64(uint64_t __p0) {
9335	9356	uint64x2_t __ret;
9336	9357	__ret = (uint64x2_t) {__p0, __p0};
9337		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	9358	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
9338	9359	return __ret;
9339	9360	}
9340	9361	#endif
...	...	@@ -9349,7 +9370,7 @@ __ai __attribute__((target("neon"))) uint16x8_t vdupq_n_u16(uint16_t __p0) {
9349	9370	__ai __attribute__((target("neon"))) uint16x8_t vdupq_n_u16(uint16_t __p0) {
9350	9371	uint16x8_t __ret;
9351	9372	__ret = (uint16x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
9352		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	9373	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
9353	9374	return __ret;
9354	9375	}
9355	9376	#endif
...	...	@@ -9364,7 +9385,7 @@ __ai __attribute__((target("neon"))) int8x16_t vdupq_n_s8(int8_t __p0) {
9364	9385	__ai __attribute__((target("neon"))) int8x16_t vdupq_n_s8(int8_t __p0) {
9365	9386	int8x16_t __ret;
9366	9387	__ret = (int8x16_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
9367		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	9388	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
9368	9389	return __ret;
9369	9390	}
9370	9391	#endif
...	...	@@ -9379,7 +9400,7 @@ __ai __attribute__((target("neon"))) float32x4_t vdupq_n_f32(float32_t __p0) {
9379	9400	__ai __attribute__((target("neon"))) float32x4_t vdupq_n_f32(float32_t __p0) {
9380	9401	float32x4_t __ret;
9381	9402	__ret = (float32x4_t) {__p0, __p0, __p0, __p0};
9382		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	9403	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
9383	9404	return __ret;
9384	9405	}
9385	9406	#endif
...	...	@@ -9396,7 +9417,7 @@ __ai __attribute__((target("neon"))) float32x4_t vdupq_n_f32(float32_t __p0) {
9396	9417	float16x8_t __ret; \
9397	9418	float16_t __s0 = __p0; \
9398	9419	__ret = (float16x8_t) {__s0, __s0, __s0, __s0, __s0, __s0, __s0, __s0}; \
9399		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	9420	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
9400	9421	__ret; \
9401	9422	})
9402	9423	#endif
...	...	@@ -9411,7 +9432,7 @@ __ai __attribute__((target("neon"))) int32x4_t vdupq_n_s32(int32_t __p0) {
9411	9432	__ai __attribute__((target("neon"))) int32x4_t vdupq_n_s32(int32_t __p0) {
9412	9433	int32x4_t __ret;
9413	9434	__ret = (int32x4_t) {__p0, __p0, __p0, __p0};
9414		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	9435	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
9415	9436	return __ret;
9416	9437	}
9417	9438	#endif
...	...	@@ -9426,7 +9447,7 @@ __ai __attribute__((target("neon"))) int64x2_t vdupq_n_s64(int64_t __p0) {
9426	9447	__ai __attribute__((target("neon"))) int64x2_t vdupq_n_s64(int64_t __p0) {
9427	9448	int64x2_t __ret;
9428	9449	__ret = (int64x2_t) {__p0, __p0};
9429		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	9450	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
9430	9451	return __ret;
9431	9452	}
9432	9453	#endif
...	...	@@ -9441,7 +9462,7 @@ __ai __attribute__((target("neon"))) int16x8_t vdupq_n_s16(int16_t __p0) {
9441	9462	__ai __attribute__((target("neon"))) int16x8_t vdupq_n_s16(int16_t __p0) {
9442	9463	int16x8_t __ret;
9443	9464	__ret = (int16x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
9444		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	9465	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
9445	9466	return __ret;
9446	9467	}
9447	9468	#endif
...	...	@@ -9456,7 +9477,7 @@ __ai __attribute__((target("neon"))) uint8x8_t vdup_n_u8(uint8_t __p0) {
9456	9477	__ai __attribute__((target("neon"))) uint8x8_t vdup_n_u8(uint8_t __p0) {
9457	9478	uint8x8_t __ret;
9458	9479	__ret = (uint8x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
9459		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	9480	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
9460	9481	return __ret;
9461	9482	}
9462	9483	#endif
...	...	@@ -9471,7 +9492,7 @@ __ai __attribute__((target("neon"))) uint32x2_t vdup_n_u32(uint32_t __p0) {
9471	9492	__ai __attribute__((target("neon"))) uint32x2_t vdup_n_u32(uint32_t __p0) {
9472	9493	uint32x2_t __ret;
9473	9494	__ret = (uint32x2_t) {__p0, __p0};
9474		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	9495	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
9475	9496	return __ret;
9476	9497	}
9477	9498	#endif
...	...	@@ -9491,7 +9512,7 @@ __ai __attribute__((target("neon"))) uint16x4_t vdup_n_u16(uint16_t __p0) {
9491	9512	__ai __attribute__((target("neon"))) uint16x4_t vdup_n_u16(uint16_t __p0) {
9492	9513	uint16x4_t __ret;
9493	9514	__ret = (uint16x4_t) {__p0, __p0, __p0, __p0};
9494		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	9515	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
9495	9516	return __ret;
9496	9517	}
9497	9518	#endif
...	...	@@ -9506,7 +9527,7 @@ __ai __attribute__((target("neon"))) int8x8_t vdup_n_s8(int8_t __p0) {
9506	9527	__ai __attribute__((target("neon"))) int8x8_t vdup_n_s8(int8_t __p0) {
9507	9528	int8x8_t __ret;
9508	9529	__ret = (int8x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
9509		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	9530	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
9510	9531	return __ret;
9511	9532	}
9512	9533	#endif
...	...	@@ -9521,7 +9542,7 @@ __ai __attribute__((target("neon"))) float32x2_t vdup_n_f32(float32_t __p0) {
9521	9542	__ai __attribute__((target("neon"))) float32x2_t vdup_n_f32(float32_t __p0) {
9522	9543	float32x2_t __ret;
9523	9544	__ret = (float32x2_t) {__p0, __p0};
9524		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	9545	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
9525	9546	return __ret;
9526	9547	}
9527	9548	#endif
...	...	@@ -9538,7 +9559,7 @@ __ai __attribute__((target("neon"))) float32x2_t vdup_n_f32(float32_t __p0) {
9538	9559	float16x4_t __ret; \
9539	9560	float16_t __s0 = __p0; \
9540	9561	__ret = (float16x4_t) {__s0, __s0, __s0, __s0}; \
9541		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	9562	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
9542	9563	__ret; \
9543	9564	})
9544	9565	#endif
...	...	@@ -9553,7 +9574,7 @@ __ai __attribute__((target("neon"))) int32x2_t vdup_n_s32(int32_t __p0) {
9553	9574	__ai __attribute__((target("neon"))) int32x2_t vdup_n_s32(int32_t __p0) {
9554	9575	int32x2_t __ret;
9555	9576	__ret = (int32x2_t) {__p0, __p0};
9556		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	9577	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
9557	9578	return __ret;
9558	9579	}
9559	9580	#endif
...	...	@@ -9573,7 +9594,7 @@ __ai __attribute__((target("neon"))) int16x4_t vdup_n_s16(int16_t __p0) {
9573	9594	__ai __attribute__((target("neon"))) int16x4_t vdup_n_s16(int16_t __p0) {
9574	9595	int16x4_t __ret;
9575	9596	__ret = (int16x4_t) {__p0, __p0, __p0, __p0};
9576		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	9597	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
9577	9598	return __ret;
9578	9599	}
9579	9600	#endif
...	...	@@ -9587,10 +9608,10 @@ __ai __attribute__((target("neon"))) uint8x16_t veorq_u8(uint8x16_t __p0, uint8x
9587	9608	#else
9588	9609	__ai __attribute__((target("neon"))) uint8x16_t veorq_u8(uint8x16_t __p0, uint8x16_t __p1) {
9589	9610	uint8x16_t __ret;
9590		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
9591		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	9611	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	9612	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
9592	9613	__ret = __rev0 ^ __rev1;
9593		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	9614	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
9594	9615	return __ret;
9595	9616	}
9596	9617	#endif
...	...	@@ -9604,10 +9625,10 @@ __ai __attribute__((target("neon"))) uint32x4_t veorq_u32(uint32x4_t __p0, uint3
9604	9625	#else
9605	9626	__ai __attribute__((target("neon"))) uint32x4_t veorq_u32(uint32x4_t __p0, uint32x4_t __p1) {
9606	9627	uint32x4_t __ret;
9607		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
9608		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	9628	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	9629	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
9609	9630	__ret = __rev0 ^ __rev1;
9610		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	9631	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
9611	9632	return __ret;
9612	9633	}
9613	9634	#endif
...	...	@@ -9621,10 +9642,10 @@ __ai __attribute__((target("neon"))) uint64x2_t veorq_u64(uint64x2_t __p0, uint6
9621	9642	#else
9622	9643	__ai __attribute__((target("neon"))) uint64x2_t veorq_u64(uint64x2_t __p0, uint64x2_t __p1) {
9623	9644	uint64x2_t __ret;
9624		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
9625		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	9645	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	9646	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
9626	9647	__ret = __rev0 ^ __rev1;
9627		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	9648	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
9628	9649	return __ret;
9629	9650	}
9630	9651	#endif
...	...	@@ -9638,10 +9659,10 @@ __ai __attribute__((target("neon"))) uint16x8_t veorq_u16(uint16x8_t __p0, uint1
9638	9659	#else
9639	9660	__ai __attribute__((target("neon"))) uint16x8_t veorq_u16(uint16x8_t __p0, uint16x8_t __p1) {
9640	9661	uint16x8_t __ret;
9641		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
9642		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	9662	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	9663	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
9643	9664	__ret = __rev0 ^ __rev1;
9644		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	9665	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
9645	9666	return __ret;
9646	9667	}
9647	9668	#endif
...	...	@@ -9655,10 +9676,10 @@ __ai __attribute__((target("neon"))) int8x16_t veorq_s8(int8x16_t __p0, int8x16_
9655	9676	#else
9656	9677	__ai __attribute__((target("neon"))) int8x16_t veorq_s8(int8x16_t __p0, int8x16_t __p1) {
9657	9678	int8x16_t __ret;
9658		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
9659		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	9679	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	9680	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
9660	9681	__ret = __rev0 ^ __rev1;
9661		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	9682	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
9662	9683	return __ret;
9663	9684	}
9664	9685	#endif
...	...	@@ -9672,10 +9693,10 @@ __ai __attribute__((target("neon"))) int32x4_t veorq_s32(int32x4_t __p0, int32x4
9672	9693	#else
9673	9694	__ai __attribute__((target("neon"))) int32x4_t veorq_s32(int32x4_t __p0, int32x4_t __p1) {
9674	9695	int32x4_t __ret;
9675		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
9676		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	9696	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	9697	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
9677	9698	__ret = __rev0 ^ __rev1;
9678		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	9699	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
9679	9700	return __ret;
9680	9701	}
9681	9702	#endif
...	...	@@ -9689,10 +9710,10 @@ __ai __attribute__((target("neon"))) int64x2_t veorq_s64(int64x2_t __p0, int64x2
9689	9710	#else
9690	9711	__ai __attribute__((target("neon"))) int64x2_t veorq_s64(int64x2_t __p0, int64x2_t __p1) {
9691	9712	int64x2_t __ret;
9692		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
9693		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	9713	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	9714	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
9694	9715	__ret = __rev0 ^ __rev1;
9695		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	9716	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
9696	9717	return __ret;
9697	9718	}
9698	9719	#endif
...	...	@@ -9706,10 +9727,10 @@ __ai __attribute__((target("neon"))) int16x8_t veorq_s16(int16x8_t __p0, int16x8
9706	9727	#else
9707	9728	__ai __attribute__((target("neon"))) int16x8_t veorq_s16(int16x8_t __p0, int16x8_t __p1) {
9708	9729	int16x8_t __ret;
9709		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
9710		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	9730	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	9731	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
9711	9732	__ret = __rev0 ^ __rev1;
9712		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	9733	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
9713	9734	return __ret;
9714	9735	}
9715	9736	#endif
...	...	@@ -9723,10 +9744,10 @@ __ai __attribute__((target("neon"))) uint8x8_t veor_u8(uint8x8_t __p0, uint8x8_t
9723	9744	#else
9724	9745	__ai __attribute__((target("neon"))) uint8x8_t veor_u8(uint8x8_t __p0, uint8x8_t __p1) {
9725	9746	uint8x8_t __ret;
9726		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
9727		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	9747	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	9748	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
9728	9749	__ret = __rev0 ^ __rev1;
9729		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	9750	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
9730	9751	return __ret;
9731	9752	}
9732	9753	#endif
...	...	@@ -9740,10 +9761,10 @@ __ai __attribute__((target("neon"))) uint32x2_t veor_u32(uint32x2_t __p0, uint32
9740	9761	#else
9741	9762	__ai __attribute__((target("neon"))) uint32x2_t veor_u32(uint32x2_t __p0, uint32x2_t __p1) {
9742	9763	uint32x2_t __ret;
9743		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
9744		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	9764	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	9765	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
9745	9766	__ret = __rev0 ^ __rev1;
9746		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	9767	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
9747	9768	return __ret;
9748	9769	}
9749	9770	#endif
...	...	@@ -9762,10 +9783,10 @@ __ai __attribute__((target("neon"))) uint16x4_t veor_u16(uint16x4_t __p0, uint16
9762	9783	#else
9763	9784	__ai __attribute__((target("neon"))) uint16x4_t veor_u16(uint16x4_t __p0, uint16x4_t __p1) {
9764	9785	uint16x4_t __ret;
9765		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
9766		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	9786	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	9787	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
9767	9788	__ret = __rev0 ^ __rev1;
9768		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	9789	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
9769	9790	return __ret;
9770	9791	}
9771	9792	#endif
...	...	@@ -9779,10 +9800,10 @@ __ai __attribute__((target("neon"))) int8x8_t veor_s8(int8x8_t __p0, int8x8_t __
9779	9800	#else
9780	9801	__ai __attribute__((target("neon"))) int8x8_t veor_s8(int8x8_t __p0, int8x8_t __p1) {
9781	9802	int8x8_t __ret;
9782		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
9783		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	9803	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	9804	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
9784	9805	__ret = __rev0 ^ __rev1;
9785		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	9806	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
9786	9807	return __ret;
9787	9808	}
9788	9809	#endif
...	...	@@ -9796,10 +9817,10 @@ __ai __attribute__((target("neon"))) int32x2_t veor_s32(int32x2_t __p0, int32x2_
9796	9817	#else
9797	9818	__ai __attribute__((target("neon"))) int32x2_t veor_s32(int32x2_t __p0, int32x2_t __p1) {
9798	9819	int32x2_t __ret;
9799		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
9800		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	9820	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	9821	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
9801	9822	__ret = __rev0 ^ __rev1;
9802		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	9823	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
9803	9824	return __ret;
9804	9825	}
9805	9826	#endif
...	...	@@ -9818,10 +9839,10 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
9818	9839	#else
9819	9840	__ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_t __p1) {
9820	9841	int16x4_t __ret;
9821		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
9822		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	9842	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	9843	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
9823	9844	__ret = __rev0 ^ __rev1;
9824		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	9845	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
9825	9846	return __ret;
9826	9847	}
9827	9848	#endif
...	...	@@ -9831,7 +9852,7 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
9831	9852	poly8x8_t __ret; \
9832	9853	poly8x8_t __s0 = __p0; \
9833	9854	poly8x8_t __s1 = __p1; \
9834		__ret = (poly8x8_t) __builtin_neon_vext_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 4); \
	9855	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vext_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 4)); \
9835	9856	__ret; \
9836	9857	})
9837	9858	#else
...	...	@@ -9839,10 +9860,10 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
9839	9860	poly8x8_t __ret; \
9840	9861	poly8x8_t __s0 = __p0; \
9841	9862	poly8x8_t __s1 = __p1; \
9842		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
9843		poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
9844		__ret = (poly8x8_t) __builtin_neon_vext_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 4); \
9845		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	9863	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	9864	poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	9865	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vext_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 4)); \
	9866	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
9846	9867	__ret; \
9847	9868	})
9848	9869	#endif
...	...	@@ -9852,7 +9873,7 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
9852	9873	poly16x4_t __ret; \
9853	9874	poly16x4_t __s0 = __p0; \
9854	9875	poly16x4_t __s1 = __p1; \
9855		__ret = (poly16x4_t) __builtin_neon_vext_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 5); \
	9876	__ret = __builtin_bit_cast(poly16x4_t, __builtin_neon_vext_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 5)); \
9856	9877	__ret; \
9857	9878	})
9858	9879	#else
...	...	@@ -9860,10 +9881,10 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
9860	9881	poly16x4_t __ret; \
9861	9882	poly16x4_t __s0 = __p0; \
9862	9883	poly16x4_t __s1 = __p1; \
9863		poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
9864		poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
9865		__ret = (poly16x4_t) __builtin_neon_vext_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 5); \
9866		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	9884	poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	9885	poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	9886	__ret = __builtin_bit_cast(poly16x4_t, __builtin_neon_vext_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 5)); \
	9887	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
9867	9888	__ret; \
9868	9889	})
9869	9890	#endif
...	...	@@ -9873,7 +9894,7 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
9873	9894	poly8x16_t __ret; \
9874	9895	poly8x16_t __s0 = __p0; \
9875	9896	poly8x16_t __s1 = __p1; \
9876		__ret = (poly8x16_t) __builtin_neon_vextq_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 36); \
	9897	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vextq_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 36)); \
9877	9898	__ret; \
9878	9899	})
9879	9900	#else
...	...	@@ -9881,10 +9902,10 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
9881	9902	poly8x16_t __ret; \
9882	9903	poly8x16_t __s0 = __p0; \
9883	9904	poly8x16_t __s1 = __p1; \
9884		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
9885		poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
9886		__ret = (poly8x16_t) __builtin_neon_vextq_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 36); \
9887		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	9905	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	9906	poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	9907	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vextq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 36)); \
	9908	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
9888	9909	__ret; \
9889	9910	})
9890	9911	#endif
...	...	@@ -9894,7 +9915,7 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
9894	9915	poly16x8_t __ret; \
9895	9916	poly16x8_t __s0 = __p0; \
9896	9917	poly16x8_t __s1 = __p1; \
9897		__ret = (poly16x8_t) __builtin_neon_vextq_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 37); \
	9918	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vextq_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 37)); \
9898	9919	__ret; \
9899	9920	})
9900	9921	#else
...	...	@@ -9902,10 +9923,10 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
9902	9923	poly16x8_t __ret; \
9903	9924	poly16x8_t __s0 = __p0; \
9904	9925	poly16x8_t __s1 = __p1; \
9905		poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
9906		poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
9907		__ret = (poly16x8_t) __builtin_neon_vextq_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 37); \
9908		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	9926	poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	9927	poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	9928	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vextq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 37)); \
	9929	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
9909	9930	__ret; \
9910	9931	})
9911	9932	#endif
...	...	@@ -9915,7 +9936,7 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
9915	9936	uint8x16_t __ret; \
9916	9937	uint8x16_t __s0 = __p0; \
9917	9938	uint8x16_t __s1 = __p1; \
9918		__ret = (uint8x16_t) __builtin_neon_vextq_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 48); \
	9939	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vextq_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 48)); \
9919	9940	__ret; \
9920	9941	})
9921	9942	#else
...	...	@@ -9923,10 +9944,10 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
9923	9944	uint8x16_t __ret; \
9924	9945	uint8x16_t __s0 = __p0; \
9925	9946	uint8x16_t __s1 = __p1; \
9926		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
9927		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
9928		__ret = (uint8x16_t) __builtin_neon_vextq_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 48); \
9929		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	9947	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	9948	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	9949	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vextq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 48)); \
	9950	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
9930	9951	__ret; \
9931	9952	})
9932	9953	#endif
...	...	@@ -9936,7 +9957,7 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
9936	9957	uint32x4_t __ret; \
9937	9958	uint32x4_t __s0 = __p0; \
9938	9959	uint32x4_t __s1 = __p1; \
9939		__ret = (uint32x4_t) __builtin_neon_vextq_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 50); \
	9960	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vextq_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 50)); \
9940	9961	__ret; \
9941	9962	})
9942	9963	#else
...	...	@@ -9944,10 +9965,10 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
9944	9965	uint32x4_t __ret; \
9945	9966	uint32x4_t __s0 = __p0; \
9946	9967	uint32x4_t __s1 = __p1; \
9947		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
9948		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
9949		__ret = (uint32x4_t) __builtin_neon_vextq_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 50); \
9950		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	9968	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	9969	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	9970	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vextq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 50)); \
	9971	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
9951	9972	__ret; \
9952	9973	})
9953	9974	#endif
...	...	@@ -9957,7 +9978,7 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
9957	9978	uint64x2_t __ret; \
9958	9979	uint64x2_t __s0 = __p0; \
9959	9980	uint64x2_t __s1 = __p1; \
9960		__ret = (uint64x2_t) __builtin_neon_vextq_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 51); \
	9981	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vextq_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 51)); \
9961	9982	__ret; \
9962	9983	})
9963	9984	#else
...	...	@@ -9965,10 +9986,10 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
9965	9986	uint64x2_t __ret; \
9966	9987	uint64x2_t __s0 = __p0; \
9967	9988	uint64x2_t __s1 = __p1; \
9968		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
9969		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
9970		__ret = (uint64x2_t) __builtin_neon_vextq_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 51); \
9971		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	9989	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	9990	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	9991	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vextq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 51)); \
	9992	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
9972	9993	__ret; \
9973	9994	})
9974	9995	#endif
...	...	@@ -9978,7 +9999,7 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
9978	9999	uint16x8_t __ret; \
9979	10000	uint16x8_t __s0 = __p0; \
9980	10001	uint16x8_t __s1 = __p1; \
9981		__ret = (uint16x8_t) __builtin_neon_vextq_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 49); \
	10002	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vextq_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 49)); \
9982	10003	__ret; \
9983	10004	})
9984	10005	#else
...	...	@@ -9986,10 +10007,10 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
9986	10007	uint16x8_t __ret; \
9987	10008	uint16x8_t __s0 = __p0; \
9988	10009	uint16x8_t __s1 = __p1; \
9989		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
9990		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
9991		__ret = (uint16x8_t) __builtin_neon_vextq_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 49); \
9992		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	10010	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	10011	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	10012	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vextq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 49)); \
	10013	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
9993	10014	__ret; \
9994	10015	})
9995	10016	#endif
...	...	@@ -9999,7 +10020,7 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
9999	10020	int8x16_t __ret; \
10000	10021	int8x16_t __s0 = __p0; \
10001	10022	int8x16_t __s1 = __p1; \
10002		__ret = (int8x16_t) __builtin_neon_vextq_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 32); \
	10023	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vextq_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 32)); \
10003	10024	__ret; \
10004	10025	})
10005	10026	#else
...	...	@@ -10007,10 +10028,10 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
10007	10028	int8x16_t __ret; \
10008	10029	int8x16_t __s0 = __p0; \
10009	10030	int8x16_t __s1 = __p1; \
10010		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
10011		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
10012		__ret = (int8x16_t) __builtin_neon_vextq_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 32); \
10013		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	10031	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	10032	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	10033	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vextq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 32)); \
	10034	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
10014	10035	__ret; \
10015	10036	})
10016	10037	#endif
...	...	@@ -10020,7 +10041,7 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
10020	10041	float32x4_t __ret; \
10021	10042	float32x4_t __s0 = __p0; \
10022	10043	float32x4_t __s1 = __p1; \
10023		__ret = (float32x4_t) __builtin_neon_vextq_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 41); \
	10044	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vextq_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 41)); \
10024	10045	__ret; \
10025	10046	})
10026	10047	#else
...	...	@@ -10028,10 +10049,10 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
10028	10049	float32x4_t __ret; \
10029	10050	float32x4_t __s0 = __p0; \
10030	10051	float32x4_t __s1 = __p1; \
10031		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
10032		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
10033		__ret = (float32x4_t) __builtin_neon_vextq_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 41); \
10034		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	10052	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	10053	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	10054	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vextq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 41)); \
	10055	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
10035	10056	__ret; \
10036	10057	})
10037	10058	#endif
...	...	@@ -10041,7 +10062,7 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
10041	10062	int32x4_t __ret; \
10042	10063	int32x4_t __s0 = __p0; \
10043	10064	int32x4_t __s1 = __p1; \
10044		__ret = (int32x4_t) __builtin_neon_vextq_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 34); \
	10065	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vextq_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 34)); \
10045	10066	__ret; \
10046	10067	})
10047	10068	#else
...	...	@@ -10049,10 +10070,10 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
10049	10070	int32x4_t __ret; \
10050	10071	int32x4_t __s0 = __p0; \
10051	10072	int32x4_t __s1 = __p1; \
10052		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
10053		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
10054		__ret = (int32x4_t) __builtin_neon_vextq_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 34); \
10055		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	10073	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	10074	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	10075	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vextq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 34)); \
	10076	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
10056	10077	__ret; \
10057	10078	})
10058	10079	#endif
...	...	@@ -10062,7 +10083,7 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
10062	10083	int64x2_t __ret; \
10063	10084	int64x2_t __s0 = __p0; \
10064	10085	int64x2_t __s1 = __p1; \
10065		__ret = (int64x2_t) __builtin_neon_vextq_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 35); \
	10086	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vextq_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 35)); \
10066	10087	__ret; \
10067	10088	})
10068	10089	#else
...	...	@@ -10070,10 +10091,10 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
10070	10091	int64x2_t __ret; \
10071	10092	int64x2_t __s0 = __p0; \
10072	10093	int64x2_t __s1 = __p1; \
10073		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
10074		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
10075		__ret = (int64x2_t) __builtin_neon_vextq_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 35); \
10076		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	10094	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	10095	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	10096	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vextq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 35)); \
	10097	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
10077	10098	__ret; \
10078	10099	})
10079	10100	#endif
...	...	@@ -10083,7 +10104,7 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
10083	10104	int16x8_t __ret; \
10084	10105	int16x8_t __s0 = __p0; \
10085	10106	int16x8_t __s1 = __p1; \
10086		__ret = (int16x8_t) __builtin_neon_vextq_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 33); \
	10107	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vextq_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 33)); \
10087	10108	__ret; \
10088	10109	})
10089	10110	#else
...	...	@@ -10091,10 +10112,10 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
10091	10112	int16x8_t __ret; \
10092	10113	int16x8_t __s0 = __p0; \
10093	10114	int16x8_t __s1 = __p1; \
10094		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
10095		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
10096		__ret = (int16x8_t) __builtin_neon_vextq_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 33); \
10097		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	10115	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	10116	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	10117	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vextq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 33)); \
	10118	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
10098	10119	__ret; \
10099	10120	})
10100	10121	#endif
...	...	@@ -10104,7 +10125,7 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
10104	10125	uint8x8_t __ret; \
10105	10126	uint8x8_t __s0 = __p0; \
10106	10127	uint8x8_t __s1 = __p1; \
10107		__ret = (uint8x8_t) __builtin_neon_vext_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 16); \
	10128	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vext_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 16)); \
10108	10129	__ret; \
10109	10130	})
10110	10131	#else
...	...	@@ -10112,10 +10133,10 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
10112	10133	uint8x8_t __ret; \
10113	10134	uint8x8_t __s0 = __p0; \
10114	10135	uint8x8_t __s1 = __p1; \
10115		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
10116		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
10117		__ret = (uint8x8_t) __builtin_neon_vext_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 16); \
10118		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	10136	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	10137	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	10138	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vext_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 16)); \
	10139	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
10119	10140	__ret; \
10120	10141	})
10121	10142	#endif
...	...	@@ -10125,7 +10146,7 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
10125	10146	uint32x2_t __ret; \
10126	10147	uint32x2_t __s0 = __p0; \
10127	10148	uint32x2_t __s1 = __p1; \
10128		__ret = (uint32x2_t) __builtin_neon_vext_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 18); \
	10149	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vext_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 18)); \
10129	10150	__ret; \
10130	10151	})
10131	10152	#else
...	...	@@ -10133,10 +10154,10 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
10133	10154	uint32x2_t __ret; \
10134	10155	uint32x2_t __s0 = __p0; \
10135	10156	uint32x2_t __s1 = __p1; \
10136		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
10137		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
10138		__ret = (uint32x2_t) __builtin_neon_vext_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 18); \
10139		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	10157	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	10158	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_32); \
	10159	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vext_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 18)); \
	10160	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
10140	10161	__ret; \
10141	10162	})
10142	10163	#endif
...	...	@@ -10145,7 +10166,7 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
10145	10166	uint64x1_t __ret; \
10146	10167	uint64x1_t __s0 = __p0; \
10147	10168	uint64x1_t __s1 = __p1; \
10148		__ret = (uint64x1_t) __builtin_neon_vext_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 19); \
	10169	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vext_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 19)); \
10149	10170	__ret; \
10150	10171	})
10151	10172	#ifdef __LITTLE_ENDIAN__
...	...	@@ -10153,7 +10174,7 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
10153	10174	uint16x4_t __ret; \
10154	10175	uint16x4_t __s0 = __p0; \
10155	10176	uint16x4_t __s1 = __p1; \
10156		__ret = (uint16x4_t) __builtin_neon_vext_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 17); \
	10177	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vext_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 17)); \
10157	10178	__ret; \
10158	10179	})
10159	10180	#else
...	...	@@ -10161,10 +10182,10 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
10161	10182	uint16x4_t __ret; \
10162	10183	uint16x4_t __s0 = __p0; \
10163	10184	uint16x4_t __s1 = __p1; \
10164		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
10165		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
10166		__ret = (uint16x4_t) __builtin_neon_vext_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 17); \
10167		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	10185	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	10186	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	10187	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vext_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 17)); \
	10188	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
10168	10189	__ret; \
10169	10190	})
10170	10191	#endif
...	...	@@ -10174,7 +10195,7 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
10174	10195	int8x8_t __ret; \
10175	10196	int8x8_t __s0 = __p0; \
10176	10197	int8x8_t __s1 = __p1; \
10177		__ret = (int8x8_t) __builtin_neon_vext_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 0); \
	10198	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vext_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 0)); \
10178	10199	__ret; \
10179	10200	})
10180	10201	#else
...	...	@@ -10182,10 +10203,10 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
10182	10203	int8x8_t __ret; \
10183	10204	int8x8_t __s0 = __p0; \
10184	10205	int8x8_t __s1 = __p1; \
10185		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
10186		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
10187		__ret = (int8x8_t) __builtin_neon_vext_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 0); \
10188		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	10206	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	10207	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	10208	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vext_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 0)); \
	10209	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
10189	10210	__ret; \
10190	10211	})
10191	10212	#endif
...	...	@@ -10195,7 +10216,7 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
10195	10216	float32x2_t __ret; \
10196	10217	float32x2_t __s0 = __p0; \
10197	10218	float32x2_t __s1 = __p1; \
10198		__ret = (float32x2_t) __builtin_neon_vext_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 9); \
	10219	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vext_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 9)); \
10199	10220	__ret; \
10200	10221	})
10201	10222	#else
...	...	@@ -10203,10 +10224,10 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
10203	10224	float32x2_t __ret; \
10204	10225	float32x2_t __s0 = __p0; \
10205	10226	float32x2_t __s1 = __p1; \
10206		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
10207		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
10208		__ret = (float32x2_t) __builtin_neon_vext_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 9); \
10209		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	10227	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	10228	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_32); \
	10229	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vext_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 9)); \
	10230	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
10210	10231	__ret; \
10211	10232	})
10212	10233	#endif
...	...	@@ -10216,7 +10237,7 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
10216	10237	int32x2_t __ret; \
10217	10238	int32x2_t __s0 = __p0; \
10218	10239	int32x2_t __s1 = __p1; \
10219		__ret = (int32x2_t) __builtin_neon_vext_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 2); \
	10240	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vext_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 2)); \
10220	10241	__ret; \
10221	10242	})
10222	10243	#else
...	...	@@ -10224,10 +10245,10 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
10224	10245	int32x2_t __ret; \
10225	10246	int32x2_t __s0 = __p0; \
10226	10247	int32x2_t __s1 = __p1; \
10227		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
10228		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
10229		__ret = (int32x2_t) __builtin_neon_vext_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 2); \
10230		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	10248	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	10249	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_32); \
	10250	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vext_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 2)); \
	10251	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
10231	10252	__ret; \
10232	10253	})
10233	10254	#endif
...	...	@@ -10236,7 +10257,7 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
10236	10257	int64x1_t __ret; \
10237	10258	int64x1_t __s0 = __p0; \
10238	10259	int64x1_t __s1 = __p1; \
10239		__ret = (int64x1_t) __builtin_neon_vext_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 3); \
	10260	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vext_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 3)); \
10240	10261	__ret; \
10241	10262	})
10242	10263	#ifdef __LITTLE_ENDIAN__
...	...	@@ -10244,7 +10265,7 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
10244	10265	int16x4_t __ret; \
10245	10266	int16x4_t __s0 = __p0; \
10246	10267	int16x4_t __s1 = __p1; \
10247		__ret = (int16x4_t) __builtin_neon_vext_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 1); \
	10268	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vext_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 1)); \
10248	10269	__ret; \
10249	10270	})
10250	10271	#else
...	...	@@ -10252,10 +10273,10 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
10252	10273	int16x4_t __ret; \
10253	10274	int16x4_t __s0 = __p0; \
10254	10275	int16x4_t __s1 = __p1; \
10255		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
10256		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
10257		__ret = (int16x4_t) __builtin_neon_vext_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 1); \
10258		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	10276	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	10277	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	10278	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vext_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 1)); \
	10279	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
10259	10280	__ret; \
10260	10281	})
10261	10282	#endif
...	...	@@ -10265,7 +10286,7 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
10265	10286	float16x8_t __ret; \
10266	10287	float16x8_t __s0 = __p0; \
10267	10288	float16x8_t __s1 = __p1; \
10268		__ret = (float16x8_t) __builtin_neon_vextq_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 40); \
	10289	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vextq_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 40)); \
10269	10290	__ret; \
10270	10291	})
10271	10292	#else
...	...	@@ -10273,10 +10294,10 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
10273	10294	float16x8_t __ret; \
10274	10295	float16x8_t __s0 = __p0; \
10275	10296	float16x8_t __s1 = __p1; \
10276		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
10277		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
10278		__ret = (float16x8_t) __builtin_neon_vextq_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 40); \
10279		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	10297	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	10298	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	10299	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vextq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 40)); \
	10300	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
10280	10301	__ret; \
10281	10302	})
10282	10303	#endif
...	...	@@ -10286,7 +10307,7 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
10286	10307	float16x4_t __ret; \
10287	10308	float16x4_t __s0 = __p0; \
10288	10309	float16x4_t __s1 = __p1; \
10289		__ret = (float16x4_t) __builtin_neon_vext_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 8); \
	10310	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vext_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 8)); \
10290	10311	__ret; \
10291	10312	})
10292	10313	#else
...	...	@@ -10294,10 +10315,10 @@ __ai __attribute__((target("neon"))) int16x4_t veor_s16(int16x4_t __p0, int16x4_
10294	10315	float16x4_t __ret; \
10295	10316	float16x4_t __s0 = __p0; \
10296	10317	float16x4_t __s1 = __p1; \
10297		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
10298		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
10299		__ret = (float16x4_t) __builtin_neon_vext_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 8); \
10300		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	10318	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	10319	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	10320	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vext_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 8)); \
	10321	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
10301	10322	__ret; \
10302	10323	})
10303	10324	#endif
...	...	@@ -10311,9 +10332,9 @@ __ai __attribute__((target("neon"))) poly8x8_t vget_high_p8(poly8x16_t __p0) {
10311	10332	#else
10312	10333	__ai __attribute__((target("neon"))) poly8x8_t vget_high_p8(poly8x16_t __p0) {
10313	10334	poly8x8_t __ret;
10314		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	10335	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
10315	10336	__ret = __builtin_shufflevector(__rev0, __rev0, 8, 9, 10, 11, 12, 13, 14, 15);
10316		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	10337	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
10317	10338	return __ret;
10318	10339	}
10319	10340	__ai __attribute__((target("neon"))) poly8x8_t __noswap_vget_high_p8(poly8x16_t __p0) {
...	...	@@ -10332,9 +10353,9 @@ __ai __attribute__((target("neon"))) poly16x4_t vget_high_p16(poly16x8_t __p0) {
10332	10353	#else
10333	10354	__ai __attribute__((target("neon"))) poly16x4_t vget_high_p16(poly16x8_t __p0) {
10334	10355	poly16x4_t __ret;
10335		poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	10356	poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
10336	10357	__ret = __builtin_shufflevector(__rev0, __rev0, 4, 5, 6, 7);
10337		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	10358	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
10338	10359	return __ret;
10339	10360	}
10340	10361	#endif
...	...	@@ -10348,9 +10369,9 @@ __ai __attribute__((target("neon"))) uint8x8_t vget_high_u8(uint8x16_t __p0) {
10348	10369	#else
10349	10370	__ai __attribute__((target("neon"))) uint8x8_t vget_high_u8(uint8x16_t __p0) {
10350	10371	uint8x8_t __ret;
10351		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	10372	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
10352	10373	__ret = __builtin_shufflevector(__rev0, __rev0, 8, 9, 10, 11, 12, 13, 14, 15);
10353		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	10374	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
10354	10375	return __ret;
10355	10376	}
10356	10377	__ai __attribute__((target("neon"))) uint8x8_t __noswap_vget_high_u8(uint8x16_t __p0) {
...	...	@@ -10369,9 +10390,9 @@ __ai __attribute__((target("neon"))) uint32x2_t vget_high_u32(uint32x4_t __p0) {
10369	10390	#else
10370	10391	__ai __attribute__((target("neon"))) uint32x2_t vget_high_u32(uint32x4_t __p0) {
10371	10392	uint32x2_t __ret;
10372		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	10393	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
10373	10394	__ret = __builtin_shufflevector(__rev0, __rev0, 2, 3);
10374		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	10395	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
10375	10396	return __ret;
10376	10397	}
10377	10398	__ai __attribute__((target("neon"))) uint32x2_t __noswap_vget_high_u32(uint32x4_t __p0) {
...	...	@@ -10390,7 +10411,7 @@ __ai __attribute__((target("neon"))) uint64x1_t vget_high_u64(uint64x2_t __p0) {
10390	10411	#else
10391	10412	__ai __attribute__((target("neon"))) uint64x1_t vget_high_u64(uint64x2_t __p0) {
10392	10413	uint64x1_t __ret;
10393		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
	10414	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
10394	10415	__ret = __builtin_shufflevector(__rev0, __rev0, 1);
10395	10416	return __ret;
10396	10417	}
...	...	@@ -10405,9 +10426,9 @@ __ai __attribute__((target("neon"))) uint16x4_t vget_high_u16(uint16x8_t __p0) {
10405	10426	#else
10406	10427	__ai __attribute__((target("neon"))) uint16x4_t vget_high_u16(uint16x8_t __p0) {
10407	10428	uint16x4_t __ret;
10408		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	10429	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
10409	10430	__ret = __builtin_shufflevector(__rev0, __rev0, 4, 5, 6, 7);
10410		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	10431	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
10411	10432	return __ret;
10412	10433	}
10413	10434	__ai __attribute__((target("neon"))) uint16x4_t __noswap_vget_high_u16(uint16x8_t __p0) {
...	...	@@ -10426,9 +10447,9 @@ __ai __attribute__((target("neon"))) int8x8_t vget_high_s8(int8x16_t __p0) {
10426	10447	#else
10427	10448	__ai __attribute__((target("neon"))) int8x8_t vget_high_s8(int8x16_t __p0) {
10428	10449	int8x8_t __ret;
10429		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	10450	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
10430	10451	__ret = __builtin_shufflevector(__rev0, __rev0, 8, 9, 10, 11, 12, 13, 14, 15);
10431		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	10452	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
10432	10453	return __ret;
10433	10454	}
10434	10455	__ai __attribute__((target("neon"))) int8x8_t __noswap_vget_high_s8(int8x16_t __p0) {
...	...	@@ -10447,9 +10468,9 @@ __ai __attribute__((target("neon"))) float32x2_t vget_high_f32(float32x4_t __p0)
10447	10468	#else
10448	10469	__ai __attribute__((target("neon"))) float32x2_t vget_high_f32(float32x4_t __p0) {
10449	10470	float32x2_t __ret;
10450		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	10471	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
10451	10472	__ret = __builtin_shufflevector(__rev0, __rev0, 2, 3);
10452		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	10473	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
10453	10474	return __ret;
10454	10475	}
10455	10476	__ai __attribute__((target("neon"))) float32x2_t __noswap_vget_high_f32(float32x4_t __p0) {
...	...	@@ -10468,9 +10489,9 @@ __ai __attribute__((target("neon"))) float16x4_t vget_high_f16(float16x8_t __p0)
10468	10489	#else
10469	10490	__ai __attribute__((target("neon"))) float16x4_t vget_high_f16(float16x8_t __p0) {
10470	10491	float16x4_t __ret;
10471		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	10492	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
10472	10493	__ret = __builtin_shufflevector(__rev0, __rev0, 4, 5, 6, 7);
10473		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	10494	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
10474	10495	return __ret;
10475	10496	}
10476	10497	__ai __attribute__((target("neon"))) float16x4_t __noswap_vget_high_f16(float16x8_t __p0) {
...	...	@@ -10489,9 +10510,9 @@ __ai __attribute__((target("neon"))) int32x2_t vget_high_s32(int32x4_t __p0) {
10489	10510	#else
10490	10511	__ai __attribute__((target("neon"))) int32x2_t vget_high_s32(int32x4_t __p0) {
10491	10512	int32x2_t __ret;
10492		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	10513	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
10493	10514	__ret = __builtin_shufflevector(__rev0, __rev0, 2, 3);
10494		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	10515	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
10495	10516	return __ret;
10496	10517	}
10497	10518	__ai __attribute__((target("neon"))) int32x2_t __noswap_vget_high_s32(int32x4_t __p0) {
...	...	@@ -10510,7 +10531,7 @@ __ai __attribute__((target("neon"))) int64x1_t vget_high_s64(int64x2_t __p0) {
10510	10531	#else
10511	10532	__ai __attribute__((target("neon"))) int64x1_t vget_high_s64(int64x2_t __p0) {
10512	10533	int64x1_t __ret;
10513		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
	10534	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
10514	10535	__ret = __builtin_shufflevector(__rev0, __rev0, 1);
10515	10536	return __ret;
10516	10537	}
...	...	@@ -10525,9 +10546,9 @@ __ai __attribute__((target("neon"))) int16x4_t vget_high_s16(int16x8_t __p0) {
10525	10546	#else
10526	10547	__ai __attribute__((target("neon"))) int16x4_t vget_high_s16(int16x8_t __p0) {
10527	10548	int16x4_t __ret;
10528		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	10549	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
10529	10550	__ret = __builtin_shufflevector(__rev0, __rev0, 4, 5, 6, 7);
10530		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	10551	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
10531	10552	return __ret;
10532	10553	}
10533	10554	__ai __attribute__((target("neon"))) int16x4_t __noswap_vget_high_s16(int16x8_t __p0) {
...	...	@@ -10541,21 +10562,21 @@ __ai __attribute__((target("neon"))) int16x4_t __noswap_vget_high_s16(int16x8_t
10541	10562	#define vget_lane_p8(__p0, __p1) __extension__ ({ \
10542	10563	poly8_t __ret; \
10543	10564	poly8x8_t __s0 = __p0; \
10544		__ret = (poly8_t) __builtin_neon_vget_lane_i8((poly8x8_t)__s0, __p1); \
	10565	__ret = __builtin_bit_cast(poly8_t, __builtin_neon_vget_lane_i8(__s0, __p1)); \
10545	10566	__ret; \
10546	10567	})
10547	10568	#else
10548	10569	#define vget_lane_p8(__p0, __p1) __extension__ ({ \
10549	10570	poly8_t __ret; \
10550	10571	poly8x8_t __s0 = __p0; \
10551		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
10552		__ret = (poly8_t) __builtin_neon_vget_lane_i8((poly8x8_t)__rev0, __p1); \
	10572	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	10573	__ret = __builtin_bit_cast(poly8_t, __builtin_neon_vget_lane_i8(__rev0, __p1)); \
10553	10574	__ret; \
10554	10575	})
10555	10576	#define __noswap_vget_lane_p8(__p0, __p1) __extension__ ({ \
10556	10577	poly8_t __ret; \
10557	10578	poly8x8_t __s0 = __p0; \
10558		__ret = (poly8_t) __builtin_neon_vget_lane_i8((poly8x8_t)__s0, __p1); \
	10579	__ret = __builtin_bit_cast(poly8_t, __builtin_neon_vget_lane_i8(__s0, __p1)); \
10559	10580	__ret; \
10560	10581	})
10561	10582	#endif
...	...	@@ -10564,21 +10585,21 @@ __ai __attribute__((target("neon"))) int16x4_t __noswap_vget_high_s16(int16x8_t
10564	10585	#define vget_lane_p16(__p0, __p1) __extension__ ({ \
10565	10586	poly16_t __ret; \
10566	10587	poly16x4_t __s0 = __p0; \
10567		__ret = (poly16_t) __builtin_neon_vget_lane_i16((poly16x4_t)__s0, __p1); \
	10588	__ret = __builtin_bit_cast(poly16_t, __builtin_neon_vget_lane_i16(__s0, __p1)); \
10568	10589	__ret; \
10569	10590	})
10570	10591	#else
10571	10592	#define vget_lane_p16(__p0, __p1) __extension__ ({ \
10572	10593	poly16_t __ret; \
10573	10594	poly16x4_t __s0 = __p0; \
10574		poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
10575		__ret = (poly16_t) __builtin_neon_vget_lane_i16((poly16x4_t)__rev0, __p1); \
	10595	poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	10596	__ret = __builtin_bit_cast(poly16_t, __builtin_neon_vget_lane_i16(__rev0, __p1)); \
10576	10597	__ret; \
10577	10598	})
10578	10599	#define __noswap_vget_lane_p16(__p0, __p1) __extension__ ({ \
10579	10600	poly16_t __ret; \
10580	10601	poly16x4_t __s0 = __p0; \
10581		__ret = (poly16_t) __builtin_neon_vget_lane_i16((poly16x4_t)__s0, __p1); \
	10602	__ret = __builtin_bit_cast(poly16_t, __builtin_neon_vget_lane_i16(__s0, __p1)); \
10582	10603	__ret; \
10583	10604	})
10584	10605	#endif
...	...	@@ -10587,21 +10608,21 @@ __ai __attribute__((target("neon"))) int16x4_t __noswap_vget_high_s16(int16x8_t
10587	10608	#define vgetq_lane_p8(__p0, __p1) __extension__ ({ \
10588	10609	poly8_t __ret; \
10589	10610	poly8x16_t __s0 = __p0; \
10590		__ret = (poly8_t) __builtin_neon_vgetq_lane_i8((poly8x16_t)__s0, __p1); \
	10611	__ret = __builtin_bit_cast(poly8_t, __builtin_neon_vgetq_lane_i8(__s0, __p1)); \
10591	10612	__ret; \
10592	10613	})
10593	10614	#else
10594	10615	#define vgetq_lane_p8(__p0, __p1) __extension__ ({ \
10595	10616	poly8_t __ret; \
10596	10617	poly8x16_t __s0 = __p0; \
10597		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
10598		__ret = (poly8_t) __builtin_neon_vgetq_lane_i8((poly8x16_t)__rev0, __p1); \
	10618	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	10619	__ret = __builtin_bit_cast(poly8_t, __builtin_neon_vgetq_lane_i8(__rev0, __p1)); \
10599	10620	__ret; \
10600	10621	})
10601	10622	#define __noswap_vgetq_lane_p8(__p0, __p1) __extension__ ({ \
10602	10623	poly8_t __ret; \
10603	10624	poly8x16_t __s0 = __p0; \
10604		__ret = (poly8_t) __builtin_neon_vgetq_lane_i8((poly8x16_t)__s0, __p1); \
	10625	__ret = __builtin_bit_cast(poly8_t, __builtin_neon_vgetq_lane_i8(__s0, __p1)); \
10605	10626	__ret; \
10606	10627	})
10607	10628	#endif
...	...	@@ -10610,21 +10631,21 @@ __ai __attribute__((target("neon"))) int16x4_t __noswap_vget_high_s16(int16x8_t
10610	10631	#define vgetq_lane_p16(__p0, __p1) __extension__ ({ \
10611	10632	poly16_t __ret; \
10612	10633	poly16x8_t __s0 = __p0; \
10613		__ret = (poly16_t) __builtin_neon_vgetq_lane_i16((poly16x8_t)__s0, __p1); \
	10634	__ret = __builtin_bit_cast(poly16_t, __builtin_neon_vgetq_lane_i16(__s0, __p1)); \
10614	10635	__ret; \
10615	10636	})
10616	10637	#else
10617	10638	#define vgetq_lane_p16(__p0, __p1) __extension__ ({ \
10618	10639	poly16_t __ret; \
10619	10640	poly16x8_t __s0 = __p0; \
10620		poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
10621		__ret = (poly16_t) __builtin_neon_vgetq_lane_i16((poly16x8_t)__rev0, __p1); \
	10641	poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	10642	__ret = __builtin_bit_cast(poly16_t, __builtin_neon_vgetq_lane_i16(__rev0, __p1)); \
10622	10643	__ret; \
10623	10644	})
10624	10645	#define __noswap_vgetq_lane_p16(__p0, __p1) __extension__ ({ \
10625	10646	poly16_t __ret; \
10626	10647	poly16x8_t __s0 = __p0; \
10627		__ret = (poly16_t) __builtin_neon_vgetq_lane_i16((poly16x8_t)__s0, __p1); \
	10648	__ret = __builtin_bit_cast(poly16_t, __builtin_neon_vgetq_lane_i16(__s0, __p1)); \
10628	10649	__ret; \
10629	10650	})
10630	10651	#endif
...	...	@@ -10633,21 +10654,21 @@ __ai __attribute__((target("neon"))) int16x4_t __noswap_vget_high_s16(int16x8_t
10633	10654	#define vgetq_lane_u8(__p0, __p1) __extension__ ({ \
10634	10655	uint8_t __ret; \
10635	10656	uint8x16_t __s0 = __p0; \
10636		__ret = (uint8_t) __builtin_neon_vgetq_lane_i8((int8x16_t)__s0, __p1); \
	10657	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vgetq_lane_i8(__builtin_bit_cast(int8x16_t, __s0), __p1)); \
10637	10658	__ret; \
10638	10659	})
10639	10660	#else
10640	10661	#define vgetq_lane_u8(__p0, __p1) __extension__ ({ \
10641	10662	uint8_t __ret; \
10642	10663	uint8x16_t __s0 = __p0; \
10643		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
10644		__ret = (uint8_t) __builtin_neon_vgetq_lane_i8((int8x16_t)__rev0, __p1); \
	10664	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	10665	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vgetq_lane_i8(__builtin_bit_cast(int8x16_t, __rev0), __p1)); \
10645	10666	__ret; \
10646	10667	})
10647	10668	#define __noswap_vgetq_lane_u8(__p0, __p1) __extension__ ({ \
10648	10669	uint8_t __ret; \
10649	10670	uint8x16_t __s0 = __p0; \
10650		__ret = (uint8_t) __builtin_neon_vgetq_lane_i8((int8x16_t)__s0, __p1); \
	10671	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vgetq_lane_i8(__builtin_bit_cast(int8x16_t, __s0), __p1)); \
10651	10672	__ret; \
10652	10673	})
10653	10674	#endif
...	...	@@ -10656,21 +10677,21 @@ __ai __attribute__((target("neon"))) int16x4_t __noswap_vget_high_s16(int16x8_t
10656	10677	#define vgetq_lane_u32(__p0, __p1) __extension__ ({ \
10657	10678	uint32_t __ret; \
10658	10679	uint32x4_t __s0 = __p0; \
10659		__ret = (uint32_t) __builtin_neon_vgetq_lane_i32((int32x4_t)__s0, __p1); \
	10680	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vgetq_lane_i32(__builtin_bit_cast(int32x4_t, __s0), __p1)); \
10660	10681	__ret; \
10661	10682	})
10662	10683	#else
10663	10684	#define vgetq_lane_u32(__p0, __p1) __extension__ ({ \
10664	10685	uint32_t __ret; \
10665	10686	uint32x4_t __s0 = __p0; \
10666		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
10667		__ret = (uint32_t) __builtin_neon_vgetq_lane_i32((int32x4_t)__rev0, __p1); \
	10687	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	10688	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vgetq_lane_i32(__builtin_bit_cast(int32x4_t, __rev0), __p1)); \
10668	10689	__ret; \
10669	10690	})
10670	10691	#define __noswap_vgetq_lane_u32(__p0, __p1) __extension__ ({ \
10671	10692	uint32_t __ret; \
10672	10693	uint32x4_t __s0 = __p0; \
10673		__ret = (uint32_t) __builtin_neon_vgetq_lane_i32((int32x4_t)__s0, __p1); \
	10694	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vgetq_lane_i32(__builtin_bit_cast(int32x4_t, __s0), __p1)); \
10674	10695	__ret; \
10675	10696	})
10676	10697	#endif
...	...	@@ -10679,21 +10700,21 @@ __ai __attribute__((target("neon"))) int16x4_t __noswap_vget_high_s16(int16x8_t
10679	10700	#define vgetq_lane_u64(__p0, __p1) __extension__ ({ \
10680	10701	uint64_t __ret; \
10681	10702	uint64x2_t __s0 = __p0; \
10682		__ret = (uint64_t) __builtin_neon_vgetq_lane_i64((int64x2_t)__s0, __p1); \
	10703	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vgetq_lane_i64(__builtin_bit_cast(int64x2_t, __s0), __p1)); \
10683	10704	__ret; \
10684	10705	})
10685	10706	#else
10686	10707	#define vgetq_lane_u64(__p0, __p1) __extension__ ({ \
10687	10708	uint64_t __ret; \
10688	10709	uint64x2_t __s0 = __p0; \
10689		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
10690		__ret = (uint64_t) __builtin_neon_vgetq_lane_i64((int64x2_t)__rev0, __p1); \
	10710	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	10711	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vgetq_lane_i64(__builtin_bit_cast(int64x2_t, __rev0), __p1)); \
10691	10712	__ret; \
10692	10713	})
10693	10714	#define __noswap_vgetq_lane_u64(__p0, __p1) __extension__ ({ \
10694	10715	uint64_t __ret; \
10695	10716	uint64x2_t __s0 = __p0; \
10696		__ret = (uint64_t) __builtin_neon_vgetq_lane_i64((int64x2_t)__s0, __p1); \
	10717	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vgetq_lane_i64(__builtin_bit_cast(int64x2_t, __s0), __p1)); \
10697	10718	__ret; \
10698	10719	})
10699	10720	#endif
...	...	@@ -10702,21 +10723,21 @@ __ai __attribute__((target("neon"))) int16x4_t __noswap_vget_high_s16(int16x8_t
10702	10723	#define vgetq_lane_u16(__p0, __p1) __extension__ ({ \
10703	10724	uint16_t __ret; \
10704	10725	uint16x8_t __s0 = __p0; \
10705		__ret = (uint16_t) __builtin_neon_vgetq_lane_i16((int16x8_t)__s0, __p1); \
	10726	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vgetq_lane_i16(__builtin_bit_cast(int16x8_t, __s0), __p1)); \
10706	10727	__ret; \
10707	10728	})
10708	10729	#else
10709	10730	#define vgetq_lane_u16(__p0, __p1) __extension__ ({ \
10710	10731	uint16_t __ret; \
10711	10732	uint16x8_t __s0 = __p0; \
10712		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
10713		__ret = (uint16_t) __builtin_neon_vgetq_lane_i16((int16x8_t)__rev0, __p1); \
	10733	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	10734	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vgetq_lane_i16(__builtin_bit_cast(int16x8_t, __rev0), __p1)); \
10714	10735	__ret; \
10715	10736	})
10716	10737	#define __noswap_vgetq_lane_u16(__p0, __p1) __extension__ ({ \
10717	10738	uint16_t __ret; \
10718	10739	uint16x8_t __s0 = __p0; \
10719		__ret = (uint16_t) __builtin_neon_vgetq_lane_i16((int16x8_t)__s0, __p1); \
	10740	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vgetq_lane_i16(__builtin_bit_cast(int16x8_t, __s0), __p1)); \
10720	10741	__ret; \
10721	10742	})
10722	10743	#endif
...	...	@@ -10725,21 +10746,21 @@ __ai __attribute__((target("neon"))) int16x4_t __noswap_vget_high_s16(int16x8_t
10725	10746	#define vgetq_lane_s8(__p0, __p1) __extension__ ({ \
10726	10747	int8_t __ret; \
10727	10748	int8x16_t __s0 = __p0; \
10728		__ret = (int8_t) __builtin_neon_vgetq_lane_i8((int8x16_t)__s0, __p1); \
	10749	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vgetq_lane_i8(__builtin_bit_cast(int8x16_t, __s0), __p1)); \
10729	10750	__ret; \
10730	10751	})
10731	10752	#else
10732	10753	#define vgetq_lane_s8(__p0, __p1) __extension__ ({ \
10733	10754	int8_t __ret; \
10734	10755	int8x16_t __s0 = __p0; \
10735		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
10736		__ret = (int8_t) __builtin_neon_vgetq_lane_i8((int8x16_t)__rev0, __p1); \
	10756	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	10757	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vgetq_lane_i8(__builtin_bit_cast(int8x16_t, __rev0), __p1)); \
10737	10758	__ret; \
10738	10759	})
10739	10760	#define __noswap_vgetq_lane_s8(__p0, __p1) __extension__ ({ \
10740	10761	int8_t __ret; \
10741	10762	int8x16_t __s0 = __p0; \
10742		__ret = (int8_t) __builtin_neon_vgetq_lane_i8((int8x16_t)__s0, __p1); \
	10763	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vgetq_lane_i8(__builtin_bit_cast(int8x16_t, __s0), __p1)); \
10743	10764	__ret; \
10744	10765	})
10745	10766	#endif
...	...	@@ -10748,21 +10769,21 @@ __ai __attribute__((target("neon"))) int16x4_t __noswap_vget_high_s16(int16x8_t
10748	10769	#define vgetq_lane_f32(__p0, __p1) __extension__ ({ \
10749	10770	float32_t __ret; \
10750	10771	float32x4_t __s0 = __p0; \
10751		__ret = (float32_t) __builtin_neon_vgetq_lane_f32((float32x4_t)__s0, __p1); \
	10772	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vgetq_lane_f32(__s0, __p1)); \
10752	10773	__ret; \
10753	10774	})
10754	10775	#else
10755	10776	#define vgetq_lane_f32(__p0, __p1) __extension__ ({ \
10756	10777	float32_t __ret; \
10757	10778	float32x4_t __s0 = __p0; \
10758		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
10759		__ret = (float32_t) __builtin_neon_vgetq_lane_f32((float32x4_t)__rev0, __p1); \
	10779	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	10780	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vgetq_lane_f32(__rev0, __p1)); \
10760	10781	__ret; \
10761	10782	})
10762	10783	#define __noswap_vgetq_lane_f32(__p0, __p1) __extension__ ({ \
10763	10784	float32_t __ret; \
10764	10785	float32x4_t __s0 = __p0; \
10765		__ret = (float32_t) __builtin_neon_vgetq_lane_f32((float32x4_t)__s0, __p1); \
	10786	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vgetq_lane_f32(__s0, __p1)); \
10766	10787	__ret; \
10767	10788	})
10768	10789	#endif
...	...	@@ -10771,21 +10792,21 @@ __ai __attribute__((target("neon"))) int16x4_t __noswap_vget_high_s16(int16x8_t
10771	10792	#define vgetq_lane_s32(__p0, __p1) __extension__ ({ \
10772	10793	int32_t __ret; \
10773	10794	int32x4_t __s0 = __p0; \
10774		__ret = (int32_t) __builtin_neon_vgetq_lane_i32((int32x4_t)__s0, __p1); \
	10795	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vgetq_lane_i32(__builtin_bit_cast(int32x4_t, __s0), __p1)); \
10775	10796	__ret; \
10776	10797	})
10777	10798	#else
10778	10799	#define vgetq_lane_s32(__p0, __p1) __extension__ ({ \
10779	10800	int32_t __ret; \
10780	10801	int32x4_t __s0 = __p0; \
10781		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
10782		__ret = (int32_t) __builtin_neon_vgetq_lane_i32((int32x4_t)__rev0, __p1); \
	10802	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	10803	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vgetq_lane_i32(__builtin_bit_cast(int32x4_t, __rev0), __p1)); \
10783	10804	__ret; \
10784	10805	})
10785	10806	#define __noswap_vgetq_lane_s32(__p0, __p1) __extension__ ({ \
10786	10807	int32_t __ret; \
10787	10808	int32x4_t __s0 = __p0; \
10788		__ret = (int32_t) __builtin_neon_vgetq_lane_i32((int32x4_t)__s0, __p1); \
	10809	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vgetq_lane_i32(__builtin_bit_cast(int32x4_t, __s0), __p1)); \
10789	10810	__ret; \
10790	10811	})
10791	10812	#endif
...	...	@@ -10794,21 +10815,21 @@ __ai __attribute__((target("neon"))) int16x4_t __noswap_vget_high_s16(int16x8_t
10794	10815	#define vgetq_lane_s64(__p0, __p1) __extension__ ({ \
10795	10816	int64_t __ret; \
10796	10817	int64x2_t __s0 = __p0; \
10797		__ret = (int64_t) __builtin_neon_vgetq_lane_i64((int64x2_t)__s0, __p1); \
	10818	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vgetq_lane_i64(__builtin_bit_cast(int64x2_t, __s0), __p1)); \
10798	10819	__ret; \
10799	10820	})
10800	10821	#else
10801	10822	#define vgetq_lane_s64(__p0, __p1) __extension__ ({ \
10802	10823	int64_t __ret; \
10803	10824	int64x2_t __s0 = __p0; \
10804		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
10805		__ret = (int64_t) __builtin_neon_vgetq_lane_i64((int64x2_t)__rev0, __p1); \
	10825	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	10826	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vgetq_lane_i64(__builtin_bit_cast(int64x2_t, __rev0), __p1)); \
10806	10827	__ret; \
10807	10828	})
10808	10829	#define __noswap_vgetq_lane_s64(__p0, __p1) __extension__ ({ \
10809	10830	int64_t __ret; \
10810	10831	int64x2_t __s0 = __p0; \
10811		__ret = (int64_t) __builtin_neon_vgetq_lane_i64((int64x2_t)__s0, __p1); \
	10832	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vgetq_lane_i64(__builtin_bit_cast(int64x2_t, __s0), __p1)); \
10812	10833	__ret; \
10813	10834	})
10814	10835	#endif
...	...	@@ -10817,21 +10838,21 @@ __ai __attribute__((target("neon"))) int16x4_t __noswap_vget_high_s16(int16x8_t
10817	10838	#define vgetq_lane_s16(__p0, __p1) __extension__ ({ \
10818	10839	int16_t __ret; \
10819	10840	int16x8_t __s0 = __p0; \
10820		__ret = (int16_t) __builtin_neon_vgetq_lane_i16((int16x8_t)__s0, __p1); \
	10841	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vgetq_lane_i16(__builtin_bit_cast(int16x8_t, __s0), __p1)); \
10821	10842	__ret; \
10822	10843	})
10823	10844	#else
10824	10845	#define vgetq_lane_s16(__p0, __p1) __extension__ ({ \
10825	10846	int16_t __ret; \
10826	10847	int16x8_t __s0 = __p0; \
10827		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
10828		__ret = (int16_t) __builtin_neon_vgetq_lane_i16((int16x8_t)__rev0, __p1); \
	10848	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	10849	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vgetq_lane_i16(__builtin_bit_cast(int16x8_t, __rev0), __p1)); \
10829	10850	__ret; \
10830	10851	})
10831	10852	#define __noswap_vgetq_lane_s16(__p0, __p1) __extension__ ({ \
10832	10853	int16_t __ret; \
10833	10854	int16x8_t __s0 = __p0; \
10834		__ret = (int16_t) __builtin_neon_vgetq_lane_i16((int16x8_t)__s0, __p1); \
	10855	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vgetq_lane_i16(__builtin_bit_cast(int16x8_t, __s0), __p1)); \
10835	10856	__ret; \
10836	10857	})
10837	10858	#endif
...	...	@@ -10840,21 +10861,21 @@ __ai __attribute__((target("neon"))) int16x4_t __noswap_vget_high_s16(int16x8_t
10840	10861	#define vget_lane_u8(__p0, __p1) __extension__ ({ \
10841	10862	uint8_t __ret; \
10842	10863	uint8x8_t __s0 = __p0; \
10843		__ret = (uint8_t) __builtin_neon_vget_lane_i8((int8x8_t)__s0, __p1); \
	10864	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vget_lane_i8(__builtin_bit_cast(int8x8_t, __s0), __p1)); \
10844	10865	__ret; \
10845	10866	})
10846	10867	#else
10847	10868	#define vget_lane_u8(__p0, __p1) __extension__ ({ \
10848	10869	uint8_t __ret; \
10849	10870	uint8x8_t __s0 = __p0; \
10850		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
10851		__ret = (uint8_t) __builtin_neon_vget_lane_i8((int8x8_t)__rev0, __p1); \
	10871	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	10872	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vget_lane_i8(__builtin_bit_cast(int8x8_t, __rev0), __p1)); \
10852	10873	__ret; \
10853	10874	})
10854	10875	#define __noswap_vget_lane_u8(__p0, __p1) __extension__ ({ \
10855	10876	uint8_t __ret; \
10856	10877	uint8x8_t __s0 = __p0; \
10857		__ret = (uint8_t) __builtin_neon_vget_lane_i8((int8x8_t)__s0, __p1); \
	10878	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vget_lane_i8(__builtin_bit_cast(int8x8_t, __s0), __p1)); \
10858	10879	__ret; \
10859	10880	})
10860	10881	#endif
...	...	@@ -10863,21 +10884,21 @@ __ai __attribute__((target("neon"))) int16x4_t __noswap_vget_high_s16(int16x8_t
10863	10884	#define vget_lane_u32(__p0, __p1) __extension__ ({ \
10864	10885	uint32_t __ret; \
10865	10886	uint32x2_t __s0 = __p0; \
10866		__ret = (uint32_t) __builtin_neon_vget_lane_i32((int32x2_t)__s0, __p1); \
	10887	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vget_lane_i32(__builtin_bit_cast(int32x2_t, __s0), __p1)); \
10867	10888	__ret; \
10868	10889	})
10869	10890	#else
10870	10891	#define vget_lane_u32(__p0, __p1) __extension__ ({ \
10871	10892	uint32_t __ret; \
10872	10893	uint32x2_t __s0 = __p0; \
10873		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
10874		__ret = (uint32_t) __builtin_neon_vget_lane_i32((int32x2_t)__rev0, __p1); \
	10894	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	10895	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vget_lane_i32(__builtin_bit_cast(int32x2_t, __rev0), __p1)); \
10875	10896	__ret; \
10876	10897	})
10877	10898	#define __noswap_vget_lane_u32(__p0, __p1) __extension__ ({ \
10878	10899	uint32_t __ret; \
10879	10900	uint32x2_t __s0 = __p0; \
10880		__ret = (uint32_t) __builtin_neon_vget_lane_i32((int32x2_t)__s0, __p1); \
	10901	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vget_lane_i32(__builtin_bit_cast(int32x2_t, __s0), __p1)); \
10881	10902	__ret; \
10882	10903	})
10883	10904	#endif
...	...	@@ -10885,28 +10906,28 @@ __ai __attribute__((target("neon"))) int16x4_t __noswap_vget_high_s16(int16x8_t
10885	10906	#define vget_lane_u64(__p0, __p1) __extension__ ({ \
10886	10907	uint64_t __ret; \
10887	10908	uint64x1_t __s0 = __p0; \
10888		__ret = (uint64_t) __builtin_neon_vget_lane_i64((int64x1_t)__s0, __p1); \
	10909	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vget_lane_i64(__builtin_bit_cast(int64x1_t, __s0), __p1)); \
10889	10910	__ret; \
10890	10911	})
10891	10912	#ifdef __LITTLE_ENDIAN__
10892	10913	#define vget_lane_u16(__p0, __p1) __extension__ ({ \
10893	10914	uint16_t __ret; \
10894	10915	uint16x4_t __s0 = __p0; \
10895		__ret = (uint16_t) __builtin_neon_vget_lane_i16((int16x4_t)__s0, __p1); \
	10916	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vget_lane_i16(__builtin_bit_cast(int16x4_t, __s0), __p1)); \
10896	10917	__ret; \
10897	10918	})
10898	10919	#else
10899	10920	#define vget_lane_u16(__p0, __p1) __extension__ ({ \
10900	10921	uint16_t __ret; \
10901	10922	uint16x4_t __s0 = __p0; \
10902		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
10903		__ret = (uint16_t) __builtin_neon_vget_lane_i16((int16x4_t)__rev0, __p1); \
	10923	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	10924	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vget_lane_i16(__builtin_bit_cast(int16x4_t, __rev0), __p1)); \
10904	10925	__ret; \
10905	10926	})
10906	10927	#define __noswap_vget_lane_u16(__p0, __p1) __extension__ ({ \
10907	10928	uint16_t __ret; \
10908	10929	uint16x4_t __s0 = __p0; \
10909		__ret = (uint16_t) __builtin_neon_vget_lane_i16((int16x4_t)__s0, __p1); \
	10930	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vget_lane_i16(__builtin_bit_cast(int16x4_t, __s0), __p1)); \
10910	10931	__ret; \
10911	10932	})
10912	10933	#endif
...	...	@@ -10915,21 +10936,21 @@ __ai __attribute__((target("neon"))) int16x4_t __noswap_vget_high_s16(int16x8_t
10915	10936	#define vget_lane_s8(__p0, __p1) __extension__ ({ \
10916	10937	int8_t __ret; \
10917	10938	int8x8_t __s0 = __p0; \
10918		__ret = (int8_t) __builtin_neon_vget_lane_i8((int8x8_t)__s0, __p1); \
	10939	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vget_lane_i8(__builtin_bit_cast(int8x8_t, __s0), __p1)); \
10919	10940	__ret; \
10920	10941	})
10921	10942	#else
10922	10943	#define vget_lane_s8(__p0, __p1) __extension__ ({ \
10923	10944	int8_t __ret; \
10924	10945	int8x8_t __s0 = __p0; \
10925		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
10926		__ret = (int8_t) __builtin_neon_vget_lane_i8((int8x8_t)__rev0, __p1); \
	10946	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	10947	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vget_lane_i8(__builtin_bit_cast(int8x8_t, __rev0), __p1)); \
10927	10948	__ret; \
10928	10949	})
10929	10950	#define __noswap_vget_lane_s8(__p0, __p1) __extension__ ({ \
10930	10951	int8_t __ret; \
10931	10952	int8x8_t __s0 = __p0; \
10932		__ret = (int8_t) __builtin_neon_vget_lane_i8((int8x8_t)__s0, __p1); \
	10953	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vget_lane_i8(__builtin_bit_cast(int8x8_t, __s0), __p1)); \
10933	10954	__ret; \
10934	10955	})
10935	10956	#endif
...	...	@@ -10938,21 +10959,21 @@ __ai __attribute__((target("neon"))) int16x4_t __noswap_vget_high_s16(int16x8_t
10938	10959	#define vget_lane_f32(__p0, __p1) __extension__ ({ \
10939	10960	float32_t __ret; \
10940	10961	float32x2_t __s0 = __p0; \
10941		__ret = (float32_t) __builtin_neon_vget_lane_f32((float32x2_t)__s0, __p1); \
	10962	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vget_lane_f32(__s0, __p1)); \
10942	10963	__ret; \
10943	10964	})
10944	10965	#else
10945	10966	#define vget_lane_f32(__p0, __p1) __extension__ ({ \
10946	10967	float32_t __ret; \
10947	10968	float32x2_t __s0 = __p0; \
10948		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
10949		__ret = (float32_t) __builtin_neon_vget_lane_f32((float32x2_t)__rev0, __p1); \
	10969	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	10970	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vget_lane_f32(__rev0, __p1)); \
10950	10971	__ret; \
10951	10972	})
10952	10973	#define __noswap_vget_lane_f32(__p0, __p1) __extension__ ({ \
10953	10974	float32_t __ret; \
10954	10975	float32x2_t __s0 = __p0; \
10955		__ret = (float32_t) __builtin_neon_vget_lane_f32((float32x2_t)__s0, __p1); \
	10976	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vget_lane_f32(__s0, __p1)); \
10956	10977	__ret; \
10957	10978	})
10958	10979	#endif
...	...	@@ -10961,21 +10982,21 @@ __ai __attribute__((target("neon"))) int16x4_t __noswap_vget_high_s16(int16x8_t
10961	10982	#define vget_lane_s32(__p0, __p1) __extension__ ({ \
10962	10983	int32_t __ret; \
10963	10984	int32x2_t __s0 = __p0; \
10964		__ret = (int32_t) __builtin_neon_vget_lane_i32((int32x2_t)__s0, __p1); \
	10985	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vget_lane_i32(__builtin_bit_cast(int32x2_t, __s0), __p1)); \
10965	10986	__ret; \
10966	10987	})
10967	10988	#else
10968	10989	#define vget_lane_s32(__p0, __p1) __extension__ ({ \
10969	10990	int32_t __ret; \
10970	10991	int32x2_t __s0 = __p0; \
10971		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
10972		__ret = (int32_t) __builtin_neon_vget_lane_i32((int32x2_t)__rev0, __p1); \
	10992	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	10993	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vget_lane_i32(__builtin_bit_cast(int32x2_t, __rev0), __p1)); \
10973	10994	__ret; \
10974	10995	})
10975	10996	#define __noswap_vget_lane_s32(__p0, __p1) __extension__ ({ \
10976	10997	int32_t __ret; \
10977	10998	int32x2_t __s0 = __p0; \
10978		__ret = (int32_t) __builtin_neon_vget_lane_i32((int32x2_t)__s0, __p1); \
	10999	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vget_lane_i32(__builtin_bit_cast(int32x2_t, __s0), __p1)); \
10979	11000	__ret; \
10980	11001	})
10981	11002	#endif
...	...	@@ -10983,28 +11004,28 @@ __ai __attribute__((target("neon"))) int16x4_t __noswap_vget_high_s16(int16x8_t
10983	11004	#define vget_lane_s64(__p0, __p1) __extension__ ({ \
10984	11005	int64_t __ret; \
10985	11006	int64x1_t __s0 = __p0; \
10986		__ret = (int64_t) __builtin_neon_vget_lane_i64((int64x1_t)__s0, __p1); \
	11007	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vget_lane_i64(__builtin_bit_cast(int64x1_t, __s0), __p1)); \
10987	11008	__ret; \
10988	11009	})
10989	11010	#ifdef __LITTLE_ENDIAN__
10990	11011	#define vget_lane_s16(__p0, __p1) __extension__ ({ \
10991	11012	int16_t __ret; \
10992	11013	int16x4_t __s0 = __p0; \
10993		__ret = (int16_t) __builtin_neon_vget_lane_i16((int16x4_t)__s0, __p1); \
	11014	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vget_lane_i16(__builtin_bit_cast(int16x4_t, __s0), __p1)); \
10994	11015	__ret; \
10995	11016	})
10996	11017	#else
10997	11018	#define vget_lane_s16(__p0, __p1) __extension__ ({ \
10998	11019	int16_t __ret; \
10999	11020	int16x4_t __s0 = __p0; \
11000		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
11001		__ret = (int16_t) __builtin_neon_vget_lane_i16((int16x4_t)__rev0, __p1); \
	11021	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	11022	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vget_lane_i16(__builtin_bit_cast(int16x4_t, __rev0), __p1)); \
11002	11023	__ret; \
11003	11024	})
11004	11025	#define __noswap_vget_lane_s16(__p0, __p1) __extension__ ({ \
11005	11026	int16_t __ret; \
11006	11027	int16x4_t __s0 = __p0; \
11007		__ret = (int16_t) __builtin_neon_vget_lane_i16((int16x4_t)__s0, __p1); \
	11028	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vget_lane_i16(__builtin_bit_cast(int16x4_t, __s0), __p1)); \
11008	11029	__ret; \
11009	11030	})
11010	11031	#endif
...	...	@@ -11018,9 +11039,9 @@ __ai __attribute__((target("neon"))) poly8x8_t vget_low_p8(poly8x16_t __p0) {
11018	11039	#else
11019	11040	__ai __attribute__((target("neon"))) poly8x8_t vget_low_p8(poly8x16_t __p0) {
11020	11041	poly8x8_t __ret;
11021		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	11042	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
11022	11043	__ret = __builtin_shufflevector(__rev0, __rev0, 0, 1, 2, 3, 4, 5, 6, 7);
11023		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	11044	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
11024	11045	return __ret;
11025	11046	}
11026	11047	#endif
...	...	@@ -11034,9 +11055,9 @@ __ai __attribute__((target("neon"))) poly16x4_t vget_low_p16(poly16x8_t __p0) {
11034	11055	#else
11035	11056	__ai __attribute__((target("neon"))) poly16x4_t vget_low_p16(poly16x8_t __p0) {
11036	11057	poly16x4_t __ret;
11037		poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	11058	poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
11038	11059	__ret = __builtin_shufflevector(__rev0, __rev0, 0, 1, 2, 3);
11039		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	11060	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
11040	11061	return __ret;
11041	11062	}
11042	11063	#endif
...	...	@@ -11050,9 +11071,9 @@ __ai __attribute__((target("neon"))) uint8x8_t vget_low_u8(uint8x16_t __p0) {
11050	11071	#else
11051	11072	__ai __attribute__((target("neon"))) uint8x8_t vget_low_u8(uint8x16_t __p0) {
11052	11073	uint8x8_t __ret;
11053		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	11074	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
11054	11075	__ret = __builtin_shufflevector(__rev0, __rev0, 0, 1, 2, 3, 4, 5, 6, 7);
11055		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	11076	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
11056	11077	return __ret;
11057	11078	}
11058	11079	#endif
...	...	@@ -11066,9 +11087,9 @@ __ai __attribute__((target("neon"))) uint32x2_t vget_low_u32(uint32x4_t __p0) {
11066	11087	#else
11067	11088	__ai __attribute__((target("neon"))) uint32x2_t vget_low_u32(uint32x4_t __p0) {
11068	11089	uint32x2_t __ret;
11069		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	11090	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
11070	11091	__ret = __builtin_shufflevector(__rev0, __rev0, 0, 1);
11071		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	11092	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
11072	11093	return __ret;
11073	11094	}
11074	11095	#endif
...	...	@@ -11082,7 +11103,7 @@ __ai __attribute__((target("neon"))) uint64x1_t vget_low_u64(uint64x2_t __p0) {
11082	11103	#else
11083	11104	__ai __attribute__((target("neon"))) uint64x1_t vget_low_u64(uint64x2_t __p0) {
11084	11105	uint64x1_t __ret;
11085		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
	11106	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
11086	11107	__ret = __builtin_shufflevector(__rev0, __rev0, 0);
11087	11108	return __ret;
11088	11109	}
...	...	@@ -11097,9 +11118,9 @@ __ai __attribute__((target("neon"))) uint16x4_t vget_low_u16(uint16x8_t __p0) {
11097	11118	#else
11098	11119	__ai __attribute__((target("neon"))) uint16x4_t vget_low_u16(uint16x8_t __p0) {
11099	11120	uint16x4_t __ret;
11100		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	11121	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
11101	11122	__ret = __builtin_shufflevector(__rev0, __rev0, 0, 1, 2, 3);
11102		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	11123	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
11103	11124	return __ret;
11104	11125	}
11105	11126	#endif
...	...	@@ -11113,9 +11134,9 @@ __ai __attribute__((target("neon"))) int8x8_t vget_low_s8(int8x16_t __p0) {
11113	11134	#else
11114	11135	__ai __attribute__((target("neon"))) int8x8_t vget_low_s8(int8x16_t __p0) {
11115	11136	int8x8_t __ret;
11116		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	11137	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
11117	11138	__ret = __builtin_shufflevector(__rev0, __rev0, 0, 1, 2, 3, 4, 5, 6, 7);
11118		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	11139	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
11119	11140	return __ret;
11120	11141	}
11121	11142	#endif
...	...	@@ -11129,9 +11150,9 @@ __ai __attribute__((target("neon"))) float32x2_t vget_low_f32(float32x4_t __p0)
11129	11150	#else
11130	11151	__ai __attribute__((target("neon"))) float32x2_t vget_low_f32(float32x4_t __p0) {
11131	11152	float32x2_t __ret;
11132		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	11153	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
11133	11154	__ret = __builtin_shufflevector(__rev0, __rev0, 0, 1);
11134		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	11155	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
11135	11156	return __ret;
11136	11157	}
11137	11158	#endif
...	...	@@ -11145,9 +11166,9 @@ __ai __attribute__((target("neon"))) float16x4_t vget_low_f16(float16x8_t __p0)
11145	11166	#else
11146	11167	__ai __attribute__((target("neon"))) float16x4_t vget_low_f16(float16x8_t __p0) {
11147	11168	float16x4_t __ret;
11148		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	11169	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
11149	11170	__ret = __builtin_shufflevector(__rev0, __rev0, 0, 1, 2, 3);
11150		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	11171	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
11151	11172	return __ret;
11152	11173	}
11153	11174	#endif
...	...	@@ -11161,9 +11182,9 @@ __ai __attribute__((target("neon"))) int32x2_t vget_low_s32(int32x4_t __p0) {
11161	11182	#else
11162	11183	__ai __attribute__((target("neon"))) int32x2_t vget_low_s32(int32x4_t __p0) {
11163	11184	int32x2_t __ret;
11164		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	11185	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
11165	11186	__ret = __builtin_shufflevector(__rev0, __rev0, 0, 1);
11166		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	11187	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
11167	11188	return __ret;
11168	11189	}
11169	11190	#endif
...	...	@@ -11177,7 +11198,7 @@ __ai __attribute__((target("neon"))) int64x1_t vget_low_s64(int64x2_t __p0) {
11177	11198	#else
11178	11199	__ai __attribute__((target("neon"))) int64x1_t vget_low_s64(int64x2_t __p0) {
11179	11200	int64x1_t __ret;
11180		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
	11201	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
11181	11202	__ret = __builtin_shufflevector(__rev0, __rev0, 0);
11182	11203	return __ret;
11183	11204	}
...	...	@@ -11192,9 +11213,9 @@ __ai __attribute__((target("neon"))) int16x4_t vget_low_s16(int16x8_t __p0) {
11192	11213	#else
11193	11214	__ai __attribute__((target("neon"))) int16x4_t vget_low_s16(int16x8_t __p0) {
11194	11215	int16x4_t __ret;
11195		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	11216	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
11196	11217	__ret = __builtin_shufflevector(__rev0, __rev0, 0, 1, 2, 3);
11197		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	11218	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
11198	11219	return __ret;
11199	11220	}
11200	11221	#endif
...	...	@@ -11202,16 +11223,16 @@ __ai __attribute__((target("neon"))) int16x4_t vget_low_s16(int16x8_t __p0) {
11202	11223	#ifdef __LITTLE_ENDIAN__
11203	11224	__ai __attribute__((target("neon"))) uint8x16_t vhaddq_u8(uint8x16_t __p0, uint8x16_t __p1) {
11204	11225	uint8x16_t __ret;
11205		__ret = (uint8x16_t) __builtin_neon_vhaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
	11226	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vhaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 48));
11206	11227	return __ret;
11207	11228	}
11208	11229	#else
11209	11230	__ai __attribute__((target("neon"))) uint8x16_t vhaddq_u8(uint8x16_t __p0, uint8x16_t __p1) {
11210	11231	uint8x16_t __ret;
11211		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
11212		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
11213		__ret = (uint8x16_t) __builtin_neon_vhaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
11214		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	11232	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	11233	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	11234	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vhaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 48));
	11235	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
11215	11236	return __ret;
11216	11237	}
11217	11238	#endif
...	...	@@ -11219,16 +11240,16 @@ __ai __attribute__((target("neon"))) uint8x16_t vhaddq_u8(uint8x16_t __p0, uint8
11219	11240	#ifdef __LITTLE_ENDIAN__
11220	11241	__ai __attribute__((target("neon"))) uint32x4_t vhaddq_u32(uint32x4_t __p0, uint32x4_t __p1) {
11221	11242	uint32x4_t __ret;
11222		__ret = (uint32x4_t) __builtin_neon_vhaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
	11243	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vhaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50));
11223	11244	return __ret;
11224	11245	}
11225	11246	#else
11226	11247	__ai __attribute__((target("neon"))) uint32x4_t vhaddq_u32(uint32x4_t __p0, uint32x4_t __p1) {
11227	11248	uint32x4_t __ret;
11228		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
11229		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
11230		__ret = (uint32x4_t) __builtin_neon_vhaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
11231		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	11249	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	11250	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	11251	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vhaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 50));
	11252	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
11232	11253	return __ret;
11233	11254	}
11234	11255	#endif
...	...	@@ -11236,16 +11257,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vhaddq_u32(uint32x4_t __p0, uint
11236	11257	#ifdef __LITTLE_ENDIAN__
11237	11258	__ai __attribute__((target("neon"))) uint16x8_t vhaddq_u16(uint16x8_t __p0, uint16x8_t __p1) {
11238	11259	uint16x8_t __ret;
11239		__ret = (uint16x8_t) __builtin_neon_vhaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
	11260	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vhaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 49));
11240	11261	return __ret;
11241	11262	}
11242	11263	#else
11243	11264	__ai __attribute__((target("neon"))) uint16x8_t vhaddq_u16(uint16x8_t __p0, uint16x8_t __p1) {
11244	11265	uint16x8_t __ret;
11245		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
11246		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
11247		__ret = (uint16x8_t) __builtin_neon_vhaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
11248		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	11266	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	11267	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	11268	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vhaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 49));
	11269	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
11249	11270	return __ret;
11250	11271	}
11251	11272	#endif
...	...	@@ -11253,16 +11274,16 @@ __ai __attribute__((target("neon"))) uint16x8_t vhaddq_u16(uint16x8_t __p0, uint
11253	11274	#ifdef __LITTLE_ENDIAN__
11254	11275	__ai __attribute__((target("neon"))) int8x16_t vhaddq_s8(int8x16_t __p0, int8x16_t __p1) {
11255	11276	int8x16_t __ret;
11256		__ret = (int8x16_t) __builtin_neon_vhaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
	11277	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vhaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 32));
11257	11278	return __ret;
11258	11279	}
11259	11280	#else
11260	11281	__ai __attribute__((target("neon"))) int8x16_t vhaddq_s8(int8x16_t __p0, int8x16_t __p1) {
11261	11282	int8x16_t __ret;
11262		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
11263		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
11264		__ret = (int8x16_t) __builtin_neon_vhaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 32);
11265		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	11283	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	11284	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	11285	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vhaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 32));
	11286	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
11266	11287	return __ret;
11267	11288	}
11268	11289	#endif
...	...	@@ -11270,16 +11291,16 @@ __ai __attribute__((target("neon"))) int8x16_t vhaddq_s8(int8x16_t __p0, int8x16
11270	11291	#ifdef __LITTLE_ENDIAN__
11271	11292	__ai __attribute__((target("neon"))) int32x4_t vhaddq_s32(int32x4_t __p0, int32x4_t __p1) {
11272	11293	int32x4_t __ret;
11273		__ret = (int32x4_t) __builtin_neon_vhaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
	11294	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vhaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 34));
11274	11295	return __ret;
11275	11296	}
11276	11297	#else
11277	11298	__ai __attribute__((target("neon"))) int32x4_t vhaddq_s32(int32x4_t __p0, int32x4_t __p1) {
11278	11299	int32x4_t __ret;
11279		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
11280		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
11281		__ret = (int32x4_t) __builtin_neon_vhaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
11282		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	11300	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	11301	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	11302	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vhaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 34));
	11303	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
11283	11304	return __ret;
11284	11305	}
11285	11306	#endif
...	...	@@ -11287,16 +11308,16 @@ __ai __attribute__((target("neon"))) int32x4_t vhaddq_s32(int32x4_t __p0, int32x
11287	11308	#ifdef __LITTLE_ENDIAN__
11288	11309	__ai __attribute__((target("neon"))) int16x8_t vhaddq_s16(int16x8_t __p0, int16x8_t __p1) {
11289	11310	int16x8_t __ret;
11290		__ret = (int16x8_t) __builtin_neon_vhaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
	11311	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vhaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 33));
11291	11312	return __ret;
11292	11313	}
11293	11314	#else
11294	11315	__ai __attribute__((target("neon"))) int16x8_t vhaddq_s16(int16x8_t __p0, int16x8_t __p1) {
11295	11316	int16x8_t __ret;
11296		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
11297		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
11298		__ret = (int16x8_t) __builtin_neon_vhaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
11299		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	11317	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	11318	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	11319	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vhaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 33));
	11320	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
11300	11321	return __ret;
11301	11322	}
11302	11323	#endif
...	...	@@ -11304,16 +11325,16 @@ __ai __attribute__((target("neon"))) int16x8_t vhaddq_s16(int16x8_t __p0, int16x
11304	11325	#ifdef __LITTLE_ENDIAN__
11305	11326	__ai __attribute__((target("neon"))) uint8x8_t vhadd_u8(uint8x8_t __p0, uint8x8_t __p1) {
11306	11327	uint8x8_t __ret;
11307		__ret = (uint8x8_t) __builtin_neon_vhadd_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
	11328	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vhadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 16));
11308	11329	return __ret;
11309	11330	}
11310	11331	#else
11311	11332	__ai __attribute__((target("neon"))) uint8x8_t vhadd_u8(uint8x8_t __p0, uint8x8_t __p1) {
11312	11333	uint8x8_t __ret;
11313		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
11314		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
11315		__ret = (uint8x8_t) __builtin_neon_vhadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
11316		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	11334	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	11335	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	11336	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vhadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 16));
	11337	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
11317	11338	return __ret;
11318	11339	}
11319	11340	#endif
...	...	@@ -11321,16 +11342,16 @@ __ai __attribute__((target("neon"))) uint8x8_t vhadd_u8(uint8x8_t __p0, uint8x8_
11321	11342	#ifdef __LITTLE_ENDIAN__
11322	11343	__ai __attribute__((target("neon"))) uint32x2_t vhadd_u32(uint32x2_t __p0, uint32x2_t __p1) {
11323	11344	uint32x2_t __ret;
11324		__ret = (uint32x2_t) __builtin_neon_vhadd_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
	11345	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vhadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 18));
11325	11346	return __ret;
11326	11347	}
11327	11348	#else
11328	11349	__ai __attribute__((target("neon"))) uint32x2_t vhadd_u32(uint32x2_t __p0, uint32x2_t __p1) {
11329	11350	uint32x2_t __ret;
11330		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
11331		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
11332		__ret = (uint32x2_t) __builtin_neon_vhadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
11333		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	11351	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	11352	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	11353	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vhadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 18));
	11354	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
11334	11355	return __ret;
11335	11356	}
11336	11357	#endif
...	...	@@ -11338,16 +11359,16 @@ __ai __attribute__((target("neon"))) uint32x2_t vhadd_u32(uint32x2_t __p0, uint3
11338	11359	#ifdef __LITTLE_ENDIAN__
11339	11360	__ai __attribute__((target("neon"))) uint16x4_t vhadd_u16(uint16x4_t __p0, uint16x4_t __p1) {
11340	11361	uint16x4_t __ret;
11341		__ret = (uint16x4_t) __builtin_neon_vhadd_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
	11362	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vhadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 17));
11342	11363	return __ret;
11343	11364	}
11344	11365	#else
11345	11366	__ai __attribute__((target("neon"))) uint16x4_t vhadd_u16(uint16x4_t __p0, uint16x4_t __p1) {
11346	11367	uint16x4_t __ret;
11347		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
11348		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
11349		__ret = (uint16x4_t) __builtin_neon_vhadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
11350		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	11368	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	11369	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	11370	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vhadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 17));
	11371	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
11351	11372	return __ret;
11352	11373	}
11353	11374	#endif
...	...	@@ -11355,16 +11376,16 @@ __ai __attribute__((target("neon"))) uint16x4_t vhadd_u16(uint16x4_t __p0, uint1
11355	11376	#ifdef __LITTLE_ENDIAN__
11356	11377	__ai __attribute__((target("neon"))) int8x8_t vhadd_s8(int8x8_t __p0, int8x8_t __p1) {
11357	11378	int8x8_t __ret;
11358		__ret = (int8x8_t) __builtin_neon_vhadd_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
	11379	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vhadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 0));
11359	11380	return __ret;
11360	11381	}
11361	11382	#else
11362	11383	__ai __attribute__((target("neon"))) int8x8_t vhadd_s8(int8x8_t __p0, int8x8_t __p1) {
11363	11384	int8x8_t __ret;
11364		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
11365		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
11366		__ret = (int8x8_t) __builtin_neon_vhadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
11367		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	11385	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	11386	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	11387	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vhadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 0));
	11388	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
11368	11389	return __ret;
11369	11390	}
11370	11391	#endif
...	...	@@ -11372,16 +11393,16 @@ __ai __attribute__((target("neon"))) int8x8_t vhadd_s8(int8x8_t __p0, int8x8_t _
11372	11393	#ifdef __LITTLE_ENDIAN__
11373	11394	__ai __attribute__((target("neon"))) int32x2_t vhadd_s32(int32x2_t __p0, int32x2_t __p1) {
11374	11395	int32x2_t __ret;
11375		__ret = (int32x2_t) __builtin_neon_vhadd_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
	11396	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vhadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 2));
11376	11397	return __ret;
11377	11398	}
11378	11399	#else
11379	11400	__ai __attribute__((target("neon"))) int32x2_t vhadd_s32(int32x2_t __p0, int32x2_t __p1) {
11380	11401	int32x2_t __ret;
11381		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
11382		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
11383		__ret = (int32x2_t) __builtin_neon_vhadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
11384		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	11402	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	11403	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	11404	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vhadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 2));
	11405	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
11385	11406	return __ret;
11386	11407	}
11387	11408	#endif
...	...	@@ -11389,16 +11410,16 @@ __ai __attribute__((target("neon"))) int32x2_t vhadd_s32(int32x2_t __p0, int32x2
11389	11410	#ifdef __LITTLE_ENDIAN__
11390	11411	__ai __attribute__((target("neon"))) int16x4_t vhadd_s16(int16x4_t __p0, int16x4_t __p1) {
11391	11412	int16x4_t __ret;
11392		__ret = (int16x4_t) __builtin_neon_vhadd_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
	11413	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vhadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 1));
11393	11414	return __ret;
11394	11415	}
11395	11416	#else
11396	11417	__ai __attribute__((target("neon"))) int16x4_t vhadd_s16(int16x4_t __p0, int16x4_t __p1) {
11397	11418	int16x4_t __ret;
11398		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
11399		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
11400		__ret = (int16x4_t) __builtin_neon_vhadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
11401		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	11419	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	11420	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	11421	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vhadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 1));
	11422	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
11402	11423	return __ret;
11403	11424	}
11404	11425	#endif
...	...	@@ -11406,16 +11427,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhadd_s16(int16x4_t __p0, int16x4
11406	11427	#ifdef __LITTLE_ENDIAN__
11407	11428	__ai __attribute__((target("neon"))) uint8x16_t vhsubq_u8(uint8x16_t __p0, uint8x16_t __p1) {
11408	11429	uint8x16_t __ret;
11409		__ret = (uint8x16_t) __builtin_neon_vhsubq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
	11430	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vhsubq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 48));
11410	11431	return __ret;
11411	11432	}
11412	11433	#else
11413	11434	__ai __attribute__((target("neon"))) uint8x16_t vhsubq_u8(uint8x16_t __p0, uint8x16_t __p1) {
11414	11435	uint8x16_t __ret;
11415		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
11416		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
11417		__ret = (uint8x16_t) __builtin_neon_vhsubq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
11418		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	11436	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	11437	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	11438	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vhsubq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 48));
	11439	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
11419	11440	return __ret;
11420	11441	}
11421	11442	#endif
...	...	@@ -11423,16 +11444,16 @@ __ai __attribute__((target("neon"))) uint8x16_t vhsubq_u8(uint8x16_t __p0, uint8
11423	11444	#ifdef __LITTLE_ENDIAN__
11424	11445	__ai __attribute__((target("neon"))) uint32x4_t vhsubq_u32(uint32x4_t __p0, uint32x4_t __p1) {
11425	11446	uint32x4_t __ret;
11426		__ret = (uint32x4_t) __builtin_neon_vhsubq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
	11447	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vhsubq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50));
11427	11448	return __ret;
11428	11449	}
11429	11450	#else
11430	11451	__ai __attribute__((target("neon"))) uint32x4_t vhsubq_u32(uint32x4_t __p0, uint32x4_t __p1) {
11431	11452	uint32x4_t __ret;
11432		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
11433		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
11434		__ret = (uint32x4_t) __builtin_neon_vhsubq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
11435		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	11453	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	11454	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	11455	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vhsubq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 50));
	11456	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
11436	11457	return __ret;
11437	11458	}
11438	11459	#endif
...	...	@@ -11440,16 +11461,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vhsubq_u32(uint32x4_t __p0, uint
11440	11461	#ifdef __LITTLE_ENDIAN__
11441	11462	__ai __attribute__((target("neon"))) uint16x8_t vhsubq_u16(uint16x8_t __p0, uint16x8_t __p1) {
11442	11463	uint16x8_t __ret;
11443		__ret = (uint16x8_t) __builtin_neon_vhsubq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
	11464	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vhsubq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 49));
11444	11465	return __ret;
11445	11466	}
11446	11467	#else
11447	11468	__ai __attribute__((target("neon"))) uint16x8_t vhsubq_u16(uint16x8_t __p0, uint16x8_t __p1) {
11448	11469	uint16x8_t __ret;
11449		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
11450		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
11451		__ret = (uint16x8_t) __builtin_neon_vhsubq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
11452		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	11470	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	11471	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	11472	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vhsubq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 49));
	11473	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
11453	11474	return __ret;
11454	11475	}
11455	11476	#endif
...	...	@@ -11457,16 +11478,16 @@ __ai __attribute__((target("neon"))) uint16x8_t vhsubq_u16(uint16x8_t __p0, uint
11457	11478	#ifdef __LITTLE_ENDIAN__
11458	11479	__ai __attribute__((target("neon"))) int8x16_t vhsubq_s8(int8x16_t __p0, int8x16_t __p1) {
11459	11480	int8x16_t __ret;
11460		__ret = (int8x16_t) __builtin_neon_vhsubq_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
	11481	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vhsubq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 32));
11461	11482	return __ret;
11462	11483	}
11463	11484	#else
11464	11485	__ai __attribute__((target("neon"))) int8x16_t vhsubq_s8(int8x16_t __p0, int8x16_t __p1) {
11465	11486	int8x16_t __ret;
11466		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
11467		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
11468		__ret = (int8x16_t) __builtin_neon_vhsubq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 32);
11469		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	11487	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	11488	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	11489	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vhsubq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 32));
	11490	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
11470	11491	return __ret;
11471	11492	}
11472	11493	#endif
...	...	@@ -11474,16 +11495,16 @@ __ai __attribute__((target("neon"))) int8x16_t vhsubq_s8(int8x16_t __p0, int8x16
11474	11495	#ifdef __LITTLE_ENDIAN__
11475	11496	__ai __attribute__((target("neon"))) int32x4_t vhsubq_s32(int32x4_t __p0, int32x4_t __p1) {
11476	11497	int32x4_t __ret;
11477		__ret = (int32x4_t) __builtin_neon_vhsubq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
	11498	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vhsubq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 34));
11478	11499	return __ret;
11479	11500	}
11480	11501	#else
11481	11502	__ai __attribute__((target("neon"))) int32x4_t vhsubq_s32(int32x4_t __p0, int32x4_t __p1) {
11482	11503	int32x4_t __ret;
11483		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
11484		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
11485		__ret = (int32x4_t) __builtin_neon_vhsubq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
11486		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	11504	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	11505	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	11506	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vhsubq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 34));
	11507	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
11487	11508	return __ret;
11488	11509	}
11489	11510	#endif
...	...	@@ -11491,16 +11512,16 @@ __ai __attribute__((target("neon"))) int32x4_t vhsubq_s32(int32x4_t __p0, int32x
11491	11512	#ifdef __LITTLE_ENDIAN__
11492	11513	__ai __attribute__((target("neon"))) int16x8_t vhsubq_s16(int16x8_t __p0, int16x8_t __p1) {
11493	11514	int16x8_t __ret;
11494		__ret = (int16x8_t) __builtin_neon_vhsubq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
	11515	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vhsubq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 33));
11495	11516	return __ret;
11496	11517	}
11497	11518	#else
11498	11519	__ai __attribute__((target("neon"))) int16x8_t vhsubq_s16(int16x8_t __p0, int16x8_t __p1) {
11499	11520	int16x8_t __ret;
11500		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
11501		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
11502		__ret = (int16x8_t) __builtin_neon_vhsubq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
11503		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	11521	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	11522	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	11523	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vhsubq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 33));
	11524	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
11504	11525	return __ret;
11505	11526	}
11506	11527	#endif
...	...	@@ -11508,16 +11529,16 @@ __ai __attribute__((target("neon"))) int16x8_t vhsubq_s16(int16x8_t __p0, int16x
11508	11529	#ifdef __LITTLE_ENDIAN__
11509	11530	__ai __attribute__((target("neon"))) uint8x8_t vhsub_u8(uint8x8_t __p0, uint8x8_t __p1) {
11510	11531	uint8x8_t __ret;
11511		__ret = (uint8x8_t) __builtin_neon_vhsub_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
	11532	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vhsub_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 16));
11512	11533	return __ret;
11513	11534	}
11514	11535	#else
11515	11536	__ai __attribute__((target("neon"))) uint8x8_t vhsub_u8(uint8x8_t __p0, uint8x8_t __p1) {
11516	11537	uint8x8_t __ret;
11517		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
11518		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
11519		__ret = (uint8x8_t) __builtin_neon_vhsub_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
11520		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	11538	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	11539	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	11540	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vhsub_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 16));
	11541	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
11521	11542	return __ret;
11522	11543	}
11523	11544	#endif
...	...	@@ -11525,16 +11546,16 @@ __ai __attribute__((target("neon"))) uint8x8_t vhsub_u8(uint8x8_t __p0, uint8x8_
11525	11546	#ifdef __LITTLE_ENDIAN__
11526	11547	__ai __attribute__((target("neon"))) uint32x2_t vhsub_u32(uint32x2_t __p0, uint32x2_t __p1) {
11527	11548	uint32x2_t __ret;
11528		__ret = (uint32x2_t) __builtin_neon_vhsub_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
	11549	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vhsub_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 18));
11529	11550	return __ret;
11530	11551	}
11531	11552	#else
11532	11553	__ai __attribute__((target("neon"))) uint32x2_t vhsub_u32(uint32x2_t __p0, uint32x2_t __p1) {
11533	11554	uint32x2_t __ret;
11534		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
11535		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
11536		__ret = (uint32x2_t) __builtin_neon_vhsub_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
11537		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	11555	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	11556	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	11557	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vhsub_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 18));
	11558	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
11538	11559	return __ret;
11539	11560	}
11540	11561	#endif
...	...	@@ -11542,16 +11563,16 @@ __ai __attribute__((target("neon"))) uint32x2_t vhsub_u32(uint32x2_t __p0, uint3
11542	11563	#ifdef __LITTLE_ENDIAN__
11543	11564	__ai __attribute__((target("neon"))) uint16x4_t vhsub_u16(uint16x4_t __p0, uint16x4_t __p1) {
11544	11565	uint16x4_t __ret;
11545		__ret = (uint16x4_t) __builtin_neon_vhsub_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
	11566	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vhsub_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 17));
11546	11567	return __ret;
11547	11568	}
11548	11569	#else
11549	11570	__ai __attribute__((target("neon"))) uint16x4_t vhsub_u16(uint16x4_t __p0, uint16x4_t __p1) {
11550	11571	uint16x4_t __ret;
11551		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
11552		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
11553		__ret = (uint16x4_t) __builtin_neon_vhsub_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
11554		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	11572	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	11573	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	11574	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vhsub_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 17));
	11575	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
11555	11576	return __ret;
11556	11577	}
11557	11578	#endif
...	...	@@ -11559,16 +11580,16 @@ __ai __attribute__((target("neon"))) uint16x4_t vhsub_u16(uint16x4_t __p0, uint1
11559	11580	#ifdef __LITTLE_ENDIAN__
11560	11581	__ai __attribute__((target("neon"))) int8x8_t vhsub_s8(int8x8_t __p0, int8x8_t __p1) {
11561	11582	int8x8_t __ret;
11562		__ret = (int8x8_t) __builtin_neon_vhsub_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
	11583	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vhsub_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 0));
11563	11584	return __ret;
11564	11585	}
11565	11586	#else
11566	11587	__ai __attribute__((target("neon"))) int8x8_t vhsub_s8(int8x8_t __p0, int8x8_t __p1) {
11567	11588	int8x8_t __ret;
11568		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
11569		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
11570		__ret = (int8x8_t) __builtin_neon_vhsub_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
11571		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	11589	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	11590	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	11591	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vhsub_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 0));
	11592	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
11572	11593	return __ret;
11573	11594	}
11574	11595	#endif
...	...	@@ -11576,16 +11597,16 @@ __ai __attribute__((target("neon"))) int8x8_t vhsub_s8(int8x8_t __p0, int8x8_t _
11576	11597	#ifdef __LITTLE_ENDIAN__
11577	11598	__ai __attribute__((target("neon"))) int32x2_t vhsub_s32(int32x2_t __p0, int32x2_t __p1) {
11578	11599	int32x2_t __ret;
11579		__ret = (int32x2_t) __builtin_neon_vhsub_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
	11600	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vhsub_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 2));
11580	11601	return __ret;
11581	11602	}
11582	11603	#else
11583	11604	__ai __attribute__((target("neon"))) int32x2_t vhsub_s32(int32x2_t __p0, int32x2_t __p1) {
11584	11605	int32x2_t __ret;
11585		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
11586		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
11587		__ret = (int32x2_t) __builtin_neon_vhsub_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
11588		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	11606	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	11607	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	11608	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vhsub_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 2));
	11609	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
11589	11610	return __ret;
11590	11611	}
11591	11612	#endif
...	...	@@ -11593,16 +11614,16 @@ __ai __attribute__((target("neon"))) int32x2_t vhsub_s32(int32x2_t __p0, int32x2
11593	11614	#ifdef __LITTLE_ENDIAN__
11594	11615	__ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4_t __p1) {
11595	11616	int16x4_t __ret;
11596		__ret = (int16x4_t) __builtin_neon_vhsub_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
	11617	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vhsub_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 1));
11597	11618	return __ret;
11598	11619	}
11599	11620	#else
11600	11621	__ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4_t __p1) {
11601	11622	int16x4_t __ret;
11602		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
11603		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
11604		__ret = (int16x4_t) __builtin_neon_vhsub_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
11605		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	11623	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	11624	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	11625	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vhsub_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 1));
	11626	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
11606	11627	return __ret;
11607	11628	}
11608	11629	#endif
...	...	@@ -11610,14 +11631,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
11610	11631	#ifdef __LITTLE_ENDIAN__
11611	11632	#define vld1_p8(__p0) __extension__ ({ \
11612	11633	poly8x8_t __ret; \
11613		__ret = (poly8x8_t) __builtin_neon_vld1_v(__p0, 4); \
	11634	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vld1_v(__p0, 4)); \
11614	11635	__ret; \
11615	11636	})
11616	11637	#else
11617	11638	#define vld1_p8(__p0) __extension__ ({ \
11618	11639	poly8x8_t __ret; \
11619		__ret = (poly8x8_t) __builtin_neon_vld1_v(__p0, 4); \
11620		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	11640	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vld1_v(__p0, 4)); \
	11641	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
11621	11642	__ret; \
11622	11643	})
11623	11644	#endif
...	...	@@ -11625,14 +11646,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
11625	11646	#ifdef __LITTLE_ENDIAN__
11626	11647	#define vld1_p16(__p0) __extension__ ({ \
11627	11648	poly16x4_t __ret; \
11628		__ret = (poly16x4_t) __builtin_neon_vld1_v(__p0, 5); \
	11649	__ret = __builtin_bit_cast(poly16x4_t, __builtin_neon_vld1_v(__p0, 5)); \
11629	11650	__ret; \
11630	11651	})
11631	11652	#else
11632	11653	#define vld1_p16(__p0) __extension__ ({ \
11633	11654	poly16x4_t __ret; \
11634		__ret = (poly16x4_t) __builtin_neon_vld1_v(__p0, 5); \
11635		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	11655	__ret = __builtin_bit_cast(poly16x4_t, __builtin_neon_vld1_v(__p0, 5)); \
	11656	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
11636	11657	__ret; \
11637	11658	})
11638	11659	#endif
...	...	@@ -11640,14 +11661,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
11640	11661	#ifdef __LITTLE_ENDIAN__
11641	11662	#define vld1q_p8(__p0) __extension__ ({ \
11642	11663	poly8x16_t __ret; \
11643		__ret = (poly8x16_t) __builtin_neon_vld1q_v(__p0, 36); \
	11664	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vld1q_v(__p0, 36)); \
11644	11665	__ret; \
11645	11666	})
11646	11667	#else
11647	11668	#define vld1q_p8(__p0) __extension__ ({ \
11648	11669	poly8x16_t __ret; \
11649		__ret = (poly8x16_t) __builtin_neon_vld1q_v(__p0, 36); \
11650		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	11670	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vld1q_v(__p0, 36)); \
	11671	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
11651	11672	__ret; \
11652	11673	})
11653	11674	#endif
...	...	@@ -11655,14 +11676,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
11655	11676	#ifdef __LITTLE_ENDIAN__
11656	11677	#define vld1q_p16(__p0) __extension__ ({ \
11657	11678	poly16x8_t __ret; \
11658		__ret = (poly16x8_t) __builtin_neon_vld1q_v(__p0, 37); \
	11679	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vld1q_v(__p0, 37)); \
11659	11680	__ret; \
11660	11681	})
11661	11682	#else
11662	11683	#define vld1q_p16(__p0) __extension__ ({ \
11663	11684	poly16x8_t __ret; \
11664		__ret = (poly16x8_t) __builtin_neon_vld1q_v(__p0, 37); \
11665		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	11685	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vld1q_v(__p0, 37)); \
	11686	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
11666	11687	__ret; \
11667	11688	})
11668	11689	#endif
...	...	@@ -11670,14 +11691,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
11670	11691	#ifdef __LITTLE_ENDIAN__
11671	11692	#define vld1q_u8(__p0) __extension__ ({ \
11672	11693	uint8x16_t __ret; \
11673		__ret = (uint8x16_t) __builtin_neon_vld1q_v(__p0, 48); \
	11694	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vld1q_v(__p0, 48)); \
11674	11695	__ret; \
11675	11696	})
11676	11697	#else
11677	11698	#define vld1q_u8(__p0) __extension__ ({ \
11678	11699	uint8x16_t __ret; \
11679		__ret = (uint8x16_t) __builtin_neon_vld1q_v(__p0, 48); \
11680		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	11700	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vld1q_v(__p0, 48)); \
	11701	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
11681	11702	__ret; \
11682	11703	})
11683	11704	#endif
...	...	@@ -11685,14 +11706,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
11685	11706	#ifdef __LITTLE_ENDIAN__
11686	11707	#define vld1q_u32(__p0) __extension__ ({ \
11687	11708	uint32x4_t __ret; \
11688		__ret = (uint32x4_t) __builtin_neon_vld1q_v(__p0, 50); \
	11709	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vld1q_v(__p0, 50)); \
11689	11710	__ret; \
11690	11711	})
11691	11712	#else
11692	11713	#define vld1q_u32(__p0) __extension__ ({ \
11693	11714	uint32x4_t __ret; \
11694		__ret = (uint32x4_t) __builtin_neon_vld1q_v(__p0, 50); \
11695		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	11715	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vld1q_v(__p0, 50)); \
	11716	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
11696	11717	__ret; \
11697	11718	})
11698	11719	#endif
...	...	@@ -11700,14 +11721,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
11700	11721	#ifdef __LITTLE_ENDIAN__
11701	11722	#define vld1q_u64(__p0) __extension__ ({ \
11702	11723	uint64x2_t __ret; \
11703		__ret = (uint64x2_t) __builtin_neon_vld1q_v(__p0, 51); \
	11724	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vld1q_v(__p0, 51)); \
11704	11725	__ret; \
11705	11726	})
11706	11727	#else
11707	11728	#define vld1q_u64(__p0) __extension__ ({ \
11708	11729	uint64x2_t __ret; \
11709		__ret = (uint64x2_t) __builtin_neon_vld1q_v(__p0, 51); \
11710		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	11730	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vld1q_v(__p0, 51)); \
	11731	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
11711	11732	__ret; \
11712	11733	})
11713	11734	#endif
...	...	@@ -11715,14 +11736,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
11715	11736	#ifdef __LITTLE_ENDIAN__
11716	11737	#define vld1q_u16(__p0) __extension__ ({ \
11717	11738	uint16x8_t __ret; \
11718		__ret = (uint16x8_t) __builtin_neon_vld1q_v(__p0, 49); \
	11739	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vld1q_v(__p0, 49)); \
11719	11740	__ret; \
11720	11741	})
11721	11742	#else
11722	11743	#define vld1q_u16(__p0) __extension__ ({ \
11723	11744	uint16x8_t __ret; \
11724		__ret = (uint16x8_t) __builtin_neon_vld1q_v(__p0, 49); \
11725		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	11745	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vld1q_v(__p0, 49)); \
	11746	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
11726	11747	__ret; \
11727	11748	})
11728	11749	#endif
...	...	@@ -11730,14 +11751,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
11730	11751	#ifdef __LITTLE_ENDIAN__
11731	11752	#define vld1q_s8(__p0) __extension__ ({ \
11732	11753	int8x16_t __ret; \
11733		__ret = (int8x16_t) __builtin_neon_vld1q_v(__p0, 32); \
	11754	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vld1q_v(__p0, 32)); \
11734	11755	__ret; \
11735	11756	})
11736	11757	#else
11737	11758	#define vld1q_s8(__p0) __extension__ ({ \
11738	11759	int8x16_t __ret; \
11739		__ret = (int8x16_t) __builtin_neon_vld1q_v(__p0, 32); \
11740		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	11760	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vld1q_v(__p0, 32)); \
	11761	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
11741	11762	__ret; \
11742	11763	})
11743	11764	#endif
...	...	@@ -11745,14 +11766,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
11745	11766	#ifdef __LITTLE_ENDIAN__
11746	11767	#define vld1q_f32(__p0) __extension__ ({ \
11747	11768	float32x4_t __ret; \
11748		__ret = (float32x4_t) __builtin_neon_vld1q_v(__p0, 41); \
	11769	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vld1q_v(__p0, 41)); \
11749	11770	__ret; \
11750	11771	})
11751	11772	#else
11752	11773	#define vld1q_f32(__p0) __extension__ ({ \
11753	11774	float32x4_t __ret; \
11754		__ret = (float32x4_t) __builtin_neon_vld1q_v(__p0, 41); \
11755		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	11775	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vld1q_v(__p0, 41)); \
	11776	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
11756	11777	__ret; \
11757	11778	})
11758	11779	#endif
...	...	@@ -11760,14 +11781,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
11760	11781	#ifdef __LITTLE_ENDIAN__
11761	11782	#define vld1q_s32(__p0) __extension__ ({ \
11762	11783	int32x4_t __ret; \
11763		__ret = (int32x4_t) __builtin_neon_vld1q_v(__p0, 34); \
	11784	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vld1q_v(__p0, 34)); \
11764	11785	__ret; \
11765	11786	})
11766	11787	#else
11767	11788	#define vld1q_s32(__p0) __extension__ ({ \
11768	11789	int32x4_t __ret; \
11769		__ret = (int32x4_t) __builtin_neon_vld1q_v(__p0, 34); \
11770		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	11790	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vld1q_v(__p0, 34)); \
	11791	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
11771	11792	__ret; \
11772	11793	})
11773	11794	#endif
...	...	@@ -11775,14 +11796,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
11775	11796	#ifdef __LITTLE_ENDIAN__
11776	11797	#define vld1q_s64(__p0) __extension__ ({ \
11777	11798	int64x2_t __ret; \
11778		__ret = (int64x2_t) __builtin_neon_vld1q_v(__p0, 35); \
	11799	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vld1q_v(__p0, 35)); \
11779	11800	__ret; \
11780	11801	})
11781	11802	#else
11782	11803	#define vld1q_s64(__p0) __extension__ ({ \
11783	11804	int64x2_t __ret; \
11784		__ret = (int64x2_t) __builtin_neon_vld1q_v(__p0, 35); \
11785		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	11805	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vld1q_v(__p0, 35)); \
	11806	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
11786	11807	__ret; \
11787	11808	})
11788	11809	#endif
...	...	@@ -11790,14 +11811,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
11790	11811	#ifdef __LITTLE_ENDIAN__
11791	11812	#define vld1q_s16(__p0) __extension__ ({ \
11792	11813	int16x8_t __ret; \
11793		__ret = (int16x8_t) __builtin_neon_vld1q_v(__p0, 33); \
	11814	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vld1q_v(__p0, 33)); \
11794	11815	__ret; \
11795	11816	})
11796	11817	#else
11797	11818	#define vld1q_s16(__p0) __extension__ ({ \
11798	11819	int16x8_t __ret; \
11799		__ret = (int16x8_t) __builtin_neon_vld1q_v(__p0, 33); \
11800		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	11820	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vld1q_v(__p0, 33)); \
	11821	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
11801	11822	__ret; \
11802	11823	})
11803	11824	#endif
...	...	@@ -11805,14 +11826,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
11805	11826	#ifdef __LITTLE_ENDIAN__
11806	11827	#define vld1_u8(__p0) __extension__ ({ \
11807	11828	uint8x8_t __ret; \
11808		__ret = (uint8x8_t) __builtin_neon_vld1_v(__p0, 16); \
	11829	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vld1_v(__p0, 16)); \
11809	11830	__ret; \
11810	11831	})
11811	11832	#else
11812	11833	#define vld1_u8(__p0) __extension__ ({ \
11813	11834	uint8x8_t __ret; \
11814		__ret = (uint8x8_t) __builtin_neon_vld1_v(__p0, 16); \
11815		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	11835	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vld1_v(__p0, 16)); \
	11836	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
11816	11837	__ret; \
11817	11838	})
11818	11839	#endif
...	...	@@ -11820,34 +11841,34 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
11820	11841	#ifdef __LITTLE_ENDIAN__
11821	11842	#define vld1_u32(__p0) __extension__ ({ \
11822	11843	uint32x2_t __ret; \
11823		__ret = (uint32x2_t) __builtin_neon_vld1_v(__p0, 18); \
	11844	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vld1_v(__p0, 18)); \
11824	11845	__ret; \
11825	11846	})
11826	11847	#else
11827	11848	#define vld1_u32(__p0) __extension__ ({ \
11828	11849	uint32x2_t __ret; \
11829		__ret = (uint32x2_t) __builtin_neon_vld1_v(__p0, 18); \
11830		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	11850	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vld1_v(__p0, 18)); \
	11851	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
11831	11852	__ret; \
11832	11853	})
11833	11854	#endif
11834	11855
11835	11856	#define vld1_u64(__p0) __extension__ ({ \
11836	11857	uint64x1_t __ret; \
11837		__ret = (uint64x1_t) __builtin_neon_vld1_v(__p0, 19); \
	11858	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vld1_v(__p0, 19)); \
11838	11859	__ret; \
11839	11860	})
11840	11861	#ifdef __LITTLE_ENDIAN__
11841	11862	#define vld1_u16(__p0) __extension__ ({ \
11842	11863	uint16x4_t __ret; \
11843		__ret = (uint16x4_t) __builtin_neon_vld1_v(__p0, 17); \
	11864	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vld1_v(__p0, 17)); \
11844	11865	__ret; \
11845	11866	})
11846	11867	#else
11847	11868	#define vld1_u16(__p0) __extension__ ({ \
11848	11869	uint16x4_t __ret; \
11849		__ret = (uint16x4_t) __builtin_neon_vld1_v(__p0, 17); \
11850		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	11870	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vld1_v(__p0, 17)); \
	11871	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
11851	11872	__ret; \
11852	11873	})
11853	11874	#endif
...	...	@@ -11855,14 +11876,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
11855	11876	#ifdef __LITTLE_ENDIAN__
11856	11877	#define vld1_s8(__p0) __extension__ ({ \
11857	11878	int8x8_t __ret; \
11858		__ret = (int8x8_t) __builtin_neon_vld1_v(__p0, 0); \
	11879	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vld1_v(__p0, 0)); \
11859	11880	__ret; \
11860	11881	})
11861	11882	#else
11862	11883	#define vld1_s8(__p0) __extension__ ({ \
11863	11884	int8x8_t __ret; \
11864		__ret = (int8x8_t) __builtin_neon_vld1_v(__p0, 0); \
11865		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	11885	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vld1_v(__p0, 0)); \
	11886	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
11866	11887	__ret; \
11867	11888	})
11868	11889	#endif
...	...	@@ -11870,14 +11891,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
11870	11891	#ifdef __LITTLE_ENDIAN__
11871	11892	#define vld1_f32(__p0) __extension__ ({ \
11872	11893	float32x2_t __ret; \
11873		__ret = (float32x2_t) __builtin_neon_vld1_v(__p0, 9); \
	11894	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vld1_v(__p0, 9)); \
11874	11895	__ret; \
11875	11896	})
11876	11897	#else
11877	11898	#define vld1_f32(__p0) __extension__ ({ \
11878	11899	float32x2_t __ret; \
11879		__ret = (float32x2_t) __builtin_neon_vld1_v(__p0, 9); \
11880		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	11900	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vld1_v(__p0, 9)); \
	11901	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
11881	11902	__ret; \
11882	11903	})
11883	11904	#endif
...	...	@@ -11885,34 +11906,34 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
11885	11906	#ifdef __LITTLE_ENDIAN__
11886	11907	#define vld1_s32(__p0) __extension__ ({ \
11887	11908	int32x2_t __ret; \
11888		__ret = (int32x2_t) __builtin_neon_vld1_v(__p0, 2); \
	11909	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vld1_v(__p0, 2)); \
11889	11910	__ret; \
11890	11911	})
11891	11912	#else
11892	11913	#define vld1_s32(__p0) __extension__ ({ \
11893	11914	int32x2_t __ret; \
11894		__ret = (int32x2_t) __builtin_neon_vld1_v(__p0, 2); \
11895		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	11915	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vld1_v(__p0, 2)); \
	11916	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
11896	11917	__ret; \
11897	11918	})
11898	11919	#endif
11899	11920
11900	11921	#define vld1_s64(__p0) __extension__ ({ \
11901	11922	int64x1_t __ret; \
11902		__ret = (int64x1_t) __builtin_neon_vld1_v(__p0, 3); \
	11923	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vld1_v(__p0, 3)); \
11903	11924	__ret; \
11904	11925	})
11905	11926	#ifdef __LITTLE_ENDIAN__
11906	11927	#define vld1_s16(__p0) __extension__ ({ \
11907	11928	int16x4_t __ret; \
11908		__ret = (int16x4_t) __builtin_neon_vld1_v(__p0, 1); \
	11929	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vld1_v(__p0, 1)); \
11909	11930	__ret; \
11910	11931	})
11911	11932	#else
11912	11933	#define vld1_s16(__p0) __extension__ ({ \
11913	11934	int16x4_t __ret; \
11914		__ret = (int16x4_t) __builtin_neon_vld1_v(__p0, 1); \
11915		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	11935	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vld1_v(__p0, 1)); \
	11936	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
11916	11937	__ret; \
11917	11938	})
11918	11939	#endif
...	...	@@ -11920,14 +11941,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
11920	11941	#ifdef __LITTLE_ENDIAN__
11921	11942	#define vld1_dup_p8(__p0) __extension__ ({ \
11922	11943	poly8x8_t __ret; \
11923		__ret = (poly8x8_t) __builtin_neon_vld1_dup_v(__p0, 4); \
	11944	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vld1_dup_v(__p0, 4)); \
11924	11945	__ret; \
11925	11946	})
11926	11947	#else
11927	11948	#define vld1_dup_p8(__p0) __extension__ ({ \
11928	11949	poly8x8_t __ret; \
11929		__ret = (poly8x8_t) __builtin_neon_vld1_dup_v(__p0, 4); \
11930		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	11950	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vld1_dup_v(__p0, 4)); \
	11951	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
11931	11952	__ret; \
11932	11953	})
11933	11954	#endif
...	...	@@ -11935,14 +11956,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
11935	11956	#ifdef __LITTLE_ENDIAN__
11936	11957	#define vld1_dup_p16(__p0) __extension__ ({ \
11937	11958	poly16x4_t __ret; \
11938		__ret = (poly16x4_t) __builtin_neon_vld1_dup_v(__p0, 5); \
	11959	__ret = __builtin_bit_cast(poly16x4_t, __builtin_neon_vld1_dup_v(__p0, 5)); \
11939	11960	__ret; \
11940	11961	})
11941	11962	#else
11942	11963	#define vld1_dup_p16(__p0) __extension__ ({ \
11943	11964	poly16x4_t __ret; \
11944		__ret = (poly16x4_t) __builtin_neon_vld1_dup_v(__p0, 5); \
11945		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	11965	__ret = __builtin_bit_cast(poly16x4_t, __builtin_neon_vld1_dup_v(__p0, 5)); \
	11966	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
11946	11967	__ret; \
11947	11968	})
11948	11969	#endif
...	...	@@ -11950,14 +11971,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
11950	11971	#ifdef __LITTLE_ENDIAN__
11951	11972	#define vld1q_dup_p8(__p0) __extension__ ({ \
11952	11973	poly8x16_t __ret; \
11953		__ret = (poly8x16_t) __builtin_neon_vld1q_dup_v(__p0, 36); \
	11974	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vld1q_dup_v(__p0, 36)); \
11954	11975	__ret; \
11955	11976	})
11956	11977	#else
11957	11978	#define vld1q_dup_p8(__p0) __extension__ ({ \
11958	11979	poly8x16_t __ret; \
11959		__ret = (poly8x16_t) __builtin_neon_vld1q_dup_v(__p0, 36); \
11960		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	11980	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vld1q_dup_v(__p0, 36)); \
	11981	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
11961	11982	__ret; \
11962	11983	})
11963	11984	#endif
...	...	@@ -11965,14 +11986,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
11965	11986	#ifdef __LITTLE_ENDIAN__
11966	11987	#define vld1q_dup_p16(__p0) __extension__ ({ \
11967	11988	poly16x8_t __ret; \
11968		__ret = (poly16x8_t) __builtin_neon_vld1q_dup_v(__p0, 37); \
	11989	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vld1q_dup_v(__p0, 37)); \
11969	11990	__ret; \
11970	11991	})
11971	11992	#else
11972	11993	#define vld1q_dup_p16(__p0) __extension__ ({ \
11973	11994	poly16x8_t __ret; \
11974		__ret = (poly16x8_t) __builtin_neon_vld1q_dup_v(__p0, 37); \
11975		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	11995	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vld1q_dup_v(__p0, 37)); \
	11996	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
11976	11997	__ret; \
11977	11998	})
11978	11999	#endif
...	...	@@ -11980,14 +12001,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
11980	12001	#ifdef __LITTLE_ENDIAN__
11981	12002	#define vld1q_dup_u8(__p0) __extension__ ({ \
11982	12003	uint8x16_t __ret; \
11983		__ret = (uint8x16_t) __builtin_neon_vld1q_dup_v(__p0, 48); \
	12004	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vld1q_dup_v(__p0, 48)); \
11984	12005	__ret; \
11985	12006	})
11986	12007	#else
11987	12008	#define vld1q_dup_u8(__p0) __extension__ ({ \
11988	12009	uint8x16_t __ret; \
11989		__ret = (uint8x16_t) __builtin_neon_vld1q_dup_v(__p0, 48); \
11990		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	12010	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vld1q_dup_v(__p0, 48)); \
	12011	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
11991	12012	__ret; \
11992	12013	})
11993	12014	#endif
...	...	@@ -11995,14 +12016,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
11995	12016	#ifdef __LITTLE_ENDIAN__
11996	12017	#define vld1q_dup_u32(__p0) __extension__ ({ \
11997	12018	uint32x4_t __ret; \
11998		__ret = (uint32x4_t) __builtin_neon_vld1q_dup_v(__p0, 50); \
	12019	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vld1q_dup_v(__p0, 50)); \
11999	12020	__ret; \
12000	12021	})
12001	12022	#else
12002	12023	#define vld1q_dup_u32(__p0) __extension__ ({ \
12003	12024	uint32x4_t __ret; \
12004		__ret = (uint32x4_t) __builtin_neon_vld1q_dup_v(__p0, 50); \
12005		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	12025	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vld1q_dup_v(__p0, 50)); \
	12026	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
12006	12027	__ret; \
12007	12028	})
12008	12029	#endif
...	...	@@ -12010,14 +12031,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12010	12031	#ifdef __LITTLE_ENDIAN__
12011	12032	#define vld1q_dup_u64(__p0) __extension__ ({ \
12012	12033	uint64x2_t __ret; \
12013		__ret = (uint64x2_t) __builtin_neon_vld1q_dup_v(__p0, 51); \
	12034	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vld1q_dup_v(__p0, 51)); \
12014	12035	__ret; \
12015	12036	})
12016	12037	#else
12017	12038	#define vld1q_dup_u64(__p0) __extension__ ({ \
12018	12039	uint64x2_t __ret; \
12019		__ret = (uint64x2_t) __builtin_neon_vld1q_dup_v(__p0, 51); \
12020		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	12040	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vld1q_dup_v(__p0, 51)); \
	12041	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
12021	12042	__ret; \
12022	12043	})
12023	12044	#endif
...	...	@@ -12025,14 +12046,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12025	12046	#ifdef __LITTLE_ENDIAN__
12026	12047	#define vld1q_dup_u16(__p0) __extension__ ({ \
12027	12048	uint16x8_t __ret; \
12028		__ret = (uint16x8_t) __builtin_neon_vld1q_dup_v(__p0, 49); \
	12049	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vld1q_dup_v(__p0, 49)); \
12029	12050	__ret; \
12030	12051	})
12031	12052	#else
12032	12053	#define vld1q_dup_u16(__p0) __extension__ ({ \
12033	12054	uint16x8_t __ret; \
12034		__ret = (uint16x8_t) __builtin_neon_vld1q_dup_v(__p0, 49); \
12035		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	12055	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vld1q_dup_v(__p0, 49)); \
	12056	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
12036	12057	__ret; \
12037	12058	})
12038	12059	#endif
...	...	@@ -12040,14 +12061,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12040	12061	#ifdef __LITTLE_ENDIAN__
12041	12062	#define vld1q_dup_s8(__p0) __extension__ ({ \
12042	12063	int8x16_t __ret; \
12043		__ret = (int8x16_t) __builtin_neon_vld1q_dup_v(__p0, 32); \
	12064	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vld1q_dup_v(__p0, 32)); \
12044	12065	__ret; \
12045	12066	})
12046	12067	#else
12047	12068	#define vld1q_dup_s8(__p0) __extension__ ({ \
12048	12069	int8x16_t __ret; \
12049		__ret = (int8x16_t) __builtin_neon_vld1q_dup_v(__p0, 32); \
12050		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	12070	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vld1q_dup_v(__p0, 32)); \
	12071	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
12051	12072	__ret; \
12052	12073	})
12053	12074	#endif
...	...	@@ -12055,14 +12076,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12055	12076	#ifdef __LITTLE_ENDIAN__
12056	12077	#define vld1q_dup_f32(__p0) __extension__ ({ \
12057	12078	float32x4_t __ret; \
12058		__ret = (float32x4_t) __builtin_neon_vld1q_dup_v(__p0, 41); \
	12079	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vld1q_dup_v(__p0, 41)); \
12059	12080	__ret; \
12060	12081	})
12061	12082	#else
12062	12083	#define vld1q_dup_f32(__p0) __extension__ ({ \
12063	12084	float32x4_t __ret; \
12064		__ret = (float32x4_t) __builtin_neon_vld1q_dup_v(__p0, 41); \
12065		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	12085	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vld1q_dup_v(__p0, 41)); \
	12086	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
12066	12087	__ret; \
12067	12088	})
12068	12089	#endif
...	...	@@ -12070,14 +12091,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12070	12091	#ifdef __LITTLE_ENDIAN__
12071	12092	#define vld1q_dup_s32(__p0) __extension__ ({ \
12072	12093	int32x4_t __ret; \
12073		__ret = (int32x4_t) __builtin_neon_vld1q_dup_v(__p0, 34); \
	12094	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vld1q_dup_v(__p0, 34)); \
12074	12095	__ret; \
12075	12096	})
12076	12097	#else
12077	12098	#define vld1q_dup_s32(__p0) __extension__ ({ \
12078	12099	int32x4_t __ret; \
12079		__ret = (int32x4_t) __builtin_neon_vld1q_dup_v(__p0, 34); \
12080		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	12100	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vld1q_dup_v(__p0, 34)); \
	12101	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
12081	12102	__ret; \
12082	12103	})
12083	12104	#endif
...	...	@@ -12085,14 +12106,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12085	12106	#ifdef __LITTLE_ENDIAN__
12086	12107	#define vld1q_dup_s64(__p0) __extension__ ({ \
12087	12108	int64x2_t __ret; \
12088		__ret = (int64x2_t) __builtin_neon_vld1q_dup_v(__p0, 35); \
	12109	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vld1q_dup_v(__p0, 35)); \
12089	12110	__ret; \
12090	12111	})
12091	12112	#else
12092	12113	#define vld1q_dup_s64(__p0) __extension__ ({ \
12093	12114	int64x2_t __ret; \
12094		__ret = (int64x2_t) __builtin_neon_vld1q_dup_v(__p0, 35); \
12095		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	12115	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vld1q_dup_v(__p0, 35)); \
	12116	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
12096	12117	__ret; \
12097	12118	})
12098	12119	#endif
...	...	@@ -12100,14 +12121,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12100	12121	#ifdef __LITTLE_ENDIAN__
12101	12122	#define vld1q_dup_s16(__p0) __extension__ ({ \
12102	12123	int16x8_t __ret; \
12103		__ret = (int16x8_t) __builtin_neon_vld1q_dup_v(__p0, 33); \
	12124	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vld1q_dup_v(__p0, 33)); \
12104	12125	__ret; \
12105	12126	})
12106	12127	#else
12107	12128	#define vld1q_dup_s16(__p0) __extension__ ({ \
12108	12129	int16x8_t __ret; \
12109		__ret = (int16x8_t) __builtin_neon_vld1q_dup_v(__p0, 33); \
12110		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	12130	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vld1q_dup_v(__p0, 33)); \
	12131	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
12111	12132	__ret; \
12112	12133	})
12113	12134	#endif
...	...	@@ -12115,14 +12136,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12115	12136	#ifdef __LITTLE_ENDIAN__
12116	12137	#define vld1_dup_u8(__p0) __extension__ ({ \
12117	12138	uint8x8_t __ret; \
12118		__ret = (uint8x8_t) __builtin_neon_vld1_dup_v(__p0, 16); \
	12139	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vld1_dup_v(__p0, 16)); \
12119	12140	__ret; \
12120	12141	})
12121	12142	#else
12122	12143	#define vld1_dup_u8(__p0) __extension__ ({ \
12123	12144	uint8x8_t __ret; \
12124		__ret = (uint8x8_t) __builtin_neon_vld1_dup_v(__p0, 16); \
12125		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	12145	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vld1_dup_v(__p0, 16)); \
	12146	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
12126	12147	__ret; \
12127	12148	})
12128	12149	#endif
...	...	@@ -12130,34 +12151,34 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12130	12151	#ifdef __LITTLE_ENDIAN__
12131	12152	#define vld1_dup_u32(__p0) __extension__ ({ \
12132	12153	uint32x2_t __ret; \
12133		__ret = (uint32x2_t) __builtin_neon_vld1_dup_v(__p0, 18); \
	12154	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vld1_dup_v(__p0, 18)); \
12134	12155	__ret; \
12135	12156	})
12136	12157	#else
12137	12158	#define vld1_dup_u32(__p0) __extension__ ({ \
12138	12159	uint32x2_t __ret; \
12139		__ret = (uint32x2_t) __builtin_neon_vld1_dup_v(__p0, 18); \
12140		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	12160	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vld1_dup_v(__p0, 18)); \
	12161	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
12141	12162	__ret; \
12142	12163	})
12143	12164	#endif
12144	12165
12145	12166	#define vld1_dup_u64(__p0) __extension__ ({ \
12146	12167	uint64x1_t __ret; \
12147		__ret = (uint64x1_t) __builtin_neon_vld1_dup_v(__p0, 19); \
	12168	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vld1_dup_v(__p0, 19)); \
12148	12169	__ret; \
12149	12170	})
12150	12171	#ifdef __LITTLE_ENDIAN__
12151	12172	#define vld1_dup_u16(__p0) __extension__ ({ \
12152	12173	uint16x4_t __ret; \
12153		__ret = (uint16x4_t) __builtin_neon_vld1_dup_v(__p0, 17); \
	12174	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vld1_dup_v(__p0, 17)); \
12154	12175	__ret; \
12155	12176	})
12156	12177	#else
12157	12178	#define vld1_dup_u16(__p0) __extension__ ({ \
12158	12179	uint16x4_t __ret; \
12159		__ret = (uint16x4_t) __builtin_neon_vld1_dup_v(__p0, 17); \
12160		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	12180	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vld1_dup_v(__p0, 17)); \
	12181	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
12161	12182	__ret; \
12162	12183	})
12163	12184	#endif
...	...	@@ -12165,14 +12186,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12165	12186	#ifdef __LITTLE_ENDIAN__
12166	12187	#define vld1_dup_s8(__p0) __extension__ ({ \
12167	12188	int8x8_t __ret; \
12168		__ret = (int8x8_t) __builtin_neon_vld1_dup_v(__p0, 0); \
	12189	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vld1_dup_v(__p0, 0)); \
12169	12190	__ret; \
12170	12191	})
12171	12192	#else
12172	12193	#define vld1_dup_s8(__p0) __extension__ ({ \
12173	12194	int8x8_t __ret; \
12174		__ret = (int8x8_t) __builtin_neon_vld1_dup_v(__p0, 0); \
12175		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	12195	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vld1_dup_v(__p0, 0)); \
	12196	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
12176	12197	__ret; \
12177	12198	})
12178	12199	#endif
...	...	@@ -12180,14 +12201,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12180	12201	#ifdef __LITTLE_ENDIAN__
12181	12202	#define vld1_dup_f32(__p0) __extension__ ({ \
12182	12203	float32x2_t __ret; \
12183		__ret = (float32x2_t) __builtin_neon_vld1_dup_v(__p0, 9); \
	12204	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vld1_dup_v(__p0, 9)); \
12184	12205	__ret; \
12185	12206	})
12186	12207	#else
12187	12208	#define vld1_dup_f32(__p0) __extension__ ({ \
12188	12209	float32x2_t __ret; \
12189		__ret = (float32x2_t) __builtin_neon_vld1_dup_v(__p0, 9); \
12190		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	12210	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vld1_dup_v(__p0, 9)); \
	12211	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
12191	12212	__ret; \
12192	12213	})
12193	12214	#endif
...	...	@@ -12195,34 +12216,34 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12195	12216	#ifdef __LITTLE_ENDIAN__
12196	12217	#define vld1_dup_s32(__p0) __extension__ ({ \
12197	12218	int32x2_t __ret; \
12198		__ret = (int32x2_t) __builtin_neon_vld1_dup_v(__p0, 2); \
	12219	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vld1_dup_v(__p0, 2)); \
12199	12220	__ret; \
12200	12221	})
12201	12222	#else
12202	12223	#define vld1_dup_s32(__p0) __extension__ ({ \
12203	12224	int32x2_t __ret; \
12204		__ret = (int32x2_t) __builtin_neon_vld1_dup_v(__p0, 2); \
12205		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	12225	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vld1_dup_v(__p0, 2)); \
	12226	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
12206	12227	__ret; \
12207	12228	})
12208	12229	#endif
12209	12230
12210	12231	#define vld1_dup_s64(__p0) __extension__ ({ \
12211	12232	int64x1_t __ret; \
12212		__ret = (int64x1_t) __builtin_neon_vld1_dup_v(__p0, 3); \
	12233	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vld1_dup_v(__p0, 3)); \
12213	12234	__ret; \
12214	12235	})
12215	12236	#ifdef __LITTLE_ENDIAN__
12216	12237	#define vld1_dup_s16(__p0) __extension__ ({ \
12217	12238	int16x4_t __ret; \
12218		__ret = (int16x4_t) __builtin_neon_vld1_dup_v(__p0, 1); \
	12239	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vld1_dup_v(__p0, 1)); \
12219	12240	__ret; \
12220	12241	})
12221	12242	#else
12222	12243	#define vld1_dup_s16(__p0) __extension__ ({ \
12223	12244	int16x4_t __ret; \
12224		__ret = (int16x4_t) __builtin_neon_vld1_dup_v(__p0, 1); \
12225		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	12245	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vld1_dup_v(__p0, 1)); \
	12246	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
12226	12247	__ret; \
12227	12248	})
12228	12249	#endif
...	...	@@ -12231,16 +12252,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12231	12252	#define vld1_lane_p8(__p0, __p1, __p2) __extension__ ({ \
12232	12253	poly8x8_t __ret; \
12233	12254	poly8x8_t __s1 = __p1; \
12234		__ret = (poly8x8_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__s1, __p2, 4); \
	12255	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vld1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 4)); \
12235	12256	__ret; \
12236	12257	})
12237	12258	#else
12238	12259	#define vld1_lane_p8(__p0, __p1, __p2) __extension__ ({ \
12239	12260	poly8x8_t __ret; \
12240	12261	poly8x8_t __s1 = __p1; \
12241		poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
12242		__ret = (poly8x8_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__rev1, __p2, 4); \
12243		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	12262	poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	12263	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vld1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), __p2, 4)); \
	12264	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
12244	12265	__ret; \
12245	12266	})
12246	12267	#endif
...	...	@@ -12249,16 +12270,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12249	12270	#define vld1_lane_p16(__p0, __p1, __p2) __extension__ ({ \
12250	12271	poly16x4_t __ret; \
12251	12272	poly16x4_t __s1 = __p1; \
12252		__ret = (poly16x4_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__s1, __p2, 5); \
	12273	__ret = __builtin_bit_cast(poly16x4_t, __builtin_neon_vld1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 5)); \
12253	12274	__ret; \
12254	12275	})
12255	12276	#else
12256	12277	#define vld1_lane_p16(__p0, __p1, __p2) __extension__ ({ \
12257	12278	poly16x4_t __ret; \
12258	12279	poly16x4_t __s1 = __p1; \
12259		poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
12260		__ret = (poly16x4_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__rev1, __p2, 5); \
12261		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	12280	poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	12281	__ret = __builtin_bit_cast(poly16x4_t, __builtin_neon_vld1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), __p2, 5)); \
	12282	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
12262	12283	__ret; \
12263	12284	})
12264	12285	#endif
...	...	@@ -12267,16 +12288,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12267	12288	#define vld1q_lane_p8(__p0, __p1, __p2) __extension__ ({ \
12268	12289	poly8x16_t __ret; \
12269	12290	poly8x16_t __s1 = __p1; \
12270		__ret = (poly8x16_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__s1, __p2, 36); \
	12291	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vld1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 36)); \
12271	12292	__ret; \
12272	12293	})
12273	12294	#else
12274	12295	#define vld1q_lane_p8(__p0, __p1, __p2) __extension__ ({ \
12275	12296	poly8x16_t __ret; \
12276	12297	poly8x16_t __s1 = __p1; \
12277		poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
12278		__ret = (poly8x16_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 36); \
12279		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	12298	poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	12299	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vld1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 36)); \
	12300	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
12280	12301	__ret; \
12281	12302	})
12282	12303	#endif
...	...	@@ -12285,16 +12306,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12285	12306	#define vld1q_lane_p16(__p0, __p1, __p2) __extension__ ({ \
12286	12307	poly16x8_t __ret; \
12287	12308	poly16x8_t __s1 = __p1; \
12288		__ret = (poly16x8_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__s1, __p2, 37); \
	12309	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vld1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 37)); \
12289	12310	__ret; \
12290	12311	})
12291	12312	#else
12292	12313	#define vld1q_lane_p16(__p0, __p1, __p2) __extension__ ({ \
12293	12314	poly16x8_t __ret; \
12294	12315	poly16x8_t __s1 = __p1; \
12295		poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
12296		__ret = (poly16x8_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 37); \
12297		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	12316	poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	12317	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vld1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 37)); \
	12318	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
12298	12319	__ret; \
12299	12320	})
12300	12321	#endif
...	...	@@ -12303,16 +12324,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12303	12324	#define vld1q_lane_u8(__p0, __p1, __p2) __extension__ ({ \
12304	12325	uint8x16_t __ret; \
12305	12326	uint8x16_t __s1 = __p1; \
12306		__ret = (uint8x16_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__s1, __p2, 48); \
	12327	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vld1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 48)); \
12307	12328	__ret; \
12308	12329	})
12309	12330	#else
12310	12331	#define vld1q_lane_u8(__p0, __p1, __p2) __extension__ ({ \
12311	12332	uint8x16_t __ret; \
12312	12333	uint8x16_t __s1 = __p1; \
12313		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
12314		__ret = (uint8x16_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 48); \
12315		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	12334	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	12335	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vld1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 48)); \
	12336	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
12316	12337	__ret; \
12317	12338	})
12318	12339	#endif
...	...	@@ -12321,16 +12342,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12321	12342	#define vld1q_lane_u32(__p0, __p1, __p2) __extension__ ({ \
12322	12343	uint32x4_t __ret; \
12323	12344	uint32x4_t __s1 = __p1; \
12324		__ret = (uint32x4_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__s1, __p2, 50); \
	12345	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vld1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 50)); \
12325	12346	__ret; \
12326	12347	})
12327	12348	#else
12328	12349	#define vld1q_lane_u32(__p0, __p1, __p2) __extension__ ({ \
12329	12350	uint32x4_t __ret; \
12330	12351	uint32x4_t __s1 = __p1; \
12331		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
12332		__ret = (uint32x4_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 50); \
12333		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	12352	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	12353	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vld1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 50)); \
	12354	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
12334	12355	__ret; \
12335	12356	})
12336	12357	#endif
...	...	@@ -12339,16 +12360,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12339	12360	#define vld1q_lane_u64(__p0, __p1, __p2) __extension__ ({ \
12340	12361	uint64x2_t __ret; \
12341	12362	uint64x2_t __s1 = __p1; \
12342		__ret = (uint64x2_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__s1, __p2, 51); \
	12363	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vld1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 51)); \
12343	12364	__ret; \
12344	12365	})
12345	12366	#else
12346	12367	#define vld1q_lane_u64(__p0, __p1, __p2) __extension__ ({ \
12347	12368	uint64x2_t __ret; \
12348	12369	uint64x2_t __s1 = __p1; \
12349		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
12350		__ret = (uint64x2_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 51); \
12351		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	12370	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	12371	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vld1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 51)); \
	12372	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
12352	12373	__ret; \
12353	12374	})
12354	12375	#endif
...	...	@@ -12357,16 +12378,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12357	12378	#define vld1q_lane_u16(__p0, __p1, __p2) __extension__ ({ \
12358	12379	uint16x8_t __ret; \
12359	12380	uint16x8_t __s1 = __p1; \
12360		__ret = (uint16x8_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__s1, __p2, 49); \
	12381	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vld1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 49)); \
12361	12382	__ret; \
12362	12383	})
12363	12384	#else
12364	12385	#define vld1q_lane_u16(__p0, __p1, __p2) __extension__ ({ \
12365	12386	uint16x8_t __ret; \
12366	12387	uint16x8_t __s1 = __p1; \
12367		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
12368		__ret = (uint16x8_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 49); \
12369		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	12388	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	12389	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vld1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 49)); \
	12390	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
12370	12391	__ret; \
12371	12392	})
12372	12393	#endif
...	...	@@ -12375,16 +12396,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12375	12396	#define vld1q_lane_s8(__p0, __p1, __p2) __extension__ ({ \
12376	12397	int8x16_t __ret; \
12377	12398	int8x16_t __s1 = __p1; \
12378		__ret = (int8x16_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__s1, __p2, 32); \
	12399	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vld1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 32)); \
12379	12400	__ret; \
12380	12401	})
12381	12402	#else
12382	12403	#define vld1q_lane_s8(__p0, __p1, __p2) __extension__ ({ \
12383	12404	int8x16_t __ret; \
12384	12405	int8x16_t __s1 = __p1; \
12385		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
12386		__ret = (int8x16_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 32); \
12387		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	12406	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	12407	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vld1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 32)); \
	12408	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
12388	12409	__ret; \
12389	12410	})
12390	12411	#endif
...	...	@@ -12393,16 +12414,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12393	12414	#define vld1q_lane_f32(__p0, __p1, __p2) __extension__ ({ \
12394	12415	float32x4_t __ret; \
12395	12416	float32x4_t __s1 = __p1; \
12396		__ret = (float32x4_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__s1, __p2, 41); \
	12417	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vld1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 41)); \
12397	12418	__ret; \
12398	12419	})
12399	12420	#else
12400	12421	#define vld1q_lane_f32(__p0, __p1, __p2) __extension__ ({ \
12401	12422	float32x4_t __ret; \
12402	12423	float32x4_t __s1 = __p1; \
12403		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
12404		__ret = (float32x4_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 41); \
12405		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	12424	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	12425	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vld1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 41)); \
	12426	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
12406	12427	__ret; \
12407	12428	})
12408	12429	#endif
...	...	@@ -12411,16 +12432,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12411	12432	#define vld1q_lane_s32(__p0, __p1, __p2) __extension__ ({ \
12412	12433	int32x4_t __ret; \
12413	12434	int32x4_t __s1 = __p1; \
12414		__ret = (int32x4_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__s1, __p2, 34); \
	12435	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vld1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 34)); \
12415	12436	__ret; \
12416	12437	})
12417	12438	#else
12418	12439	#define vld1q_lane_s32(__p0, __p1, __p2) __extension__ ({ \
12419	12440	int32x4_t __ret; \
12420	12441	int32x4_t __s1 = __p1; \
12421		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
12422		__ret = (int32x4_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 34); \
12423		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	12442	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	12443	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vld1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 34)); \
	12444	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
12424	12445	__ret; \
12425	12446	})
12426	12447	#endif
...	...	@@ -12429,16 +12450,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12429	12450	#define vld1q_lane_s64(__p0, __p1, __p2) __extension__ ({ \
12430	12451	int64x2_t __ret; \
12431	12452	int64x2_t __s1 = __p1; \
12432		__ret = (int64x2_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__s1, __p2, 35); \
	12453	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vld1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 35)); \
12433	12454	__ret; \
12434	12455	})
12435	12456	#else
12436	12457	#define vld1q_lane_s64(__p0, __p1, __p2) __extension__ ({ \
12437	12458	int64x2_t __ret; \
12438	12459	int64x2_t __s1 = __p1; \
12439		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
12440		__ret = (int64x2_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 35); \
12441		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	12460	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	12461	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vld1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 35)); \
	12462	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
12442	12463	__ret; \
12443	12464	})
12444	12465	#endif
...	...	@@ -12447,16 +12468,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12447	12468	#define vld1q_lane_s16(__p0, __p1, __p2) __extension__ ({ \
12448	12469	int16x8_t __ret; \
12449	12470	int16x8_t __s1 = __p1; \
12450		__ret = (int16x8_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__s1, __p2, 33); \
	12471	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vld1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 33)); \
12451	12472	__ret; \
12452	12473	})
12453	12474	#else
12454	12475	#define vld1q_lane_s16(__p0, __p1, __p2) __extension__ ({ \
12455	12476	int16x8_t __ret; \
12456	12477	int16x8_t __s1 = __p1; \
12457		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
12458		__ret = (int16x8_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 33); \
12459		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	12478	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	12479	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vld1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 33)); \
	12480	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
12460	12481	__ret; \
12461	12482	})
12462	12483	#endif
...	...	@@ -12465,16 +12486,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12465	12486	#define vld1_lane_u8(__p0, __p1, __p2) __extension__ ({ \
12466	12487	uint8x8_t __ret; \
12467	12488	uint8x8_t __s1 = __p1; \
12468		__ret = (uint8x8_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__s1, __p2, 16); \
	12489	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vld1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 16)); \
12469	12490	__ret; \
12470	12491	})
12471	12492	#else
12472	12493	#define vld1_lane_u8(__p0, __p1, __p2) __extension__ ({ \
12473	12494	uint8x8_t __ret; \
12474	12495	uint8x8_t __s1 = __p1; \
12475		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
12476		__ret = (uint8x8_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__rev1, __p2, 16); \
12477		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	12496	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	12497	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vld1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), __p2, 16)); \
	12498	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
12478	12499	__ret; \
12479	12500	})
12480	12501	#endif
...	...	@@ -12483,16 +12504,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12483	12504	#define vld1_lane_u32(__p0, __p1, __p2) __extension__ ({ \
12484	12505	uint32x2_t __ret; \
12485	12506	uint32x2_t __s1 = __p1; \
12486		__ret = (uint32x2_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__s1, __p2, 18); \
	12507	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vld1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 18)); \
12487	12508	__ret; \
12488	12509	})
12489	12510	#else
12490	12511	#define vld1_lane_u32(__p0, __p1, __p2) __extension__ ({ \
12491	12512	uint32x2_t __ret; \
12492	12513	uint32x2_t __s1 = __p1; \
12493		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
12494		__ret = (uint32x2_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__rev1, __p2, 18); \
12495		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	12514	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_32); \
	12515	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vld1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), __p2, 18)); \
	12516	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
12496	12517	__ret; \
12497	12518	})
12498	12519	#endif
...	...	@@ -12500,23 +12521,23 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12500	12521	#define vld1_lane_u64(__p0, __p1, __p2) __extension__ ({ \
12501	12522	uint64x1_t __ret; \
12502	12523	uint64x1_t __s1 = __p1; \
12503		__ret = (uint64x1_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__s1, __p2, 19); \
	12524	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vld1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 19)); \
12504	12525	__ret; \
12505	12526	})
12506	12527	#ifdef __LITTLE_ENDIAN__
12507	12528	#define vld1_lane_u16(__p0, __p1, __p2) __extension__ ({ \
12508	12529	uint16x4_t __ret; \
12509	12530	uint16x4_t __s1 = __p1; \
12510		__ret = (uint16x4_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__s1, __p2, 17); \
	12531	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vld1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 17)); \
12511	12532	__ret; \
12512	12533	})
12513	12534	#else
12514	12535	#define vld1_lane_u16(__p0, __p1, __p2) __extension__ ({ \
12515	12536	uint16x4_t __ret; \
12516	12537	uint16x4_t __s1 = __p1; \
12517		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
12518		__ret = (uint16x4_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__rev1, __p2, 17); \
12519		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	12538	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	12539	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vld1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), __p2, 17)); \
	12540	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
12520	12541	__ret; \
12521	12542	})
12522	12543	#endif
...	...	@@ -12525,16 +12546,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12525	12546	#define vld1_lane_s8(__p0, __p1, __p2) __extension__ ({ \
12526	12547	int8x8_t __ret; \
12527	12548	int8x8_t __s1 = __p1; \
12528		__ret = (int8x8_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__s1, __p2, 0); \
	12549	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vld1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 0)); \
12529	12550	__ret; \
12530	12551	})
12531	12552	#else
12532	12553	#define vld1_lane_s8(__p0, __p1, __p2) __extension__ ({ \
12533	12554	int8x8_t __ret; \
12534	12555	int8x8_t __s1 = __p1; \
12535		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
12536		__ret = (int8x8_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__rev1, __p2, 0); \
12537		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	12556	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	12557	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vld1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), __p2, 0)); \
	12558	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
12538	12559	__ret; \
12539	12560	})
12540	12561	#endif
...	...	@@ -12543,16 +12564,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12543	12564	#define vld1_lane_f32(__p0, __p1, __p2) __extension__ ({ \
12544	12565	float32x2_t __ret; \
12545	12566	float32x2_t __s1 = __p1; \
12546		__ret = (float32x2_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__s1, __p2, 9); \
	12567	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vld1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 9)); \
12547	12568	__ret; \
12548	12569	})
12549	12570	#else
12550	12571	#define vld1_lane_f32(__p0, __p1, __p2) __extension__ ({ \
12551	12572	float32x2_t __ret; \
12552	12573	float32x2_t __s1 = __p1; \
12553		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
12554		__ret = (float32x2_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__rev1, __p2, 9); \
12555		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	12574	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_32); \
	12575	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vld1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), __p2, 9)); \
	12576	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
12556	12577	__ret; \
12557	12578	})
12558	12579	#endif
...	...	@@ -12561,16 +12582,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12561	12582	#define vld1_lane_s32(__p0, __p1, __p2) __extension__ ({ \
12562	12583	int32x2_t __ret; \
12563	12584	int32x2_t __s1 = __p1; \
12564		__ret = (int32x2_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__s1, __p2, 2); \
	12585	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vld1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 2)); \
12565	12586	__ret; \
12566	12587	})
12567	12588	#else
12568	12589	#define vld1_lane_s32(__p0, __p1, __p2) __extension__ ({ \
12569	12590	int32x2_t __ret; \
12570	12591	int32x2_t __s1 = __p1; \
12571		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
12572		__ret = (int32x2_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__rev1, __p2, 2); \
12573		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	12592	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_32); \
	12593	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vld1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), __p2, 2)); \
	12594	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
12574	12595	__ret; \
12575	12596	})
12576	12597	#endif
...	...	@@ -12578,23 +12599,23 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12578	12599	#define vld1_lane_s64(__p0, __p1, __p2) __extension__ ({ \
12579	12600	int64x1_t __ret; \
12580	12601	int64x1_t __s1 = __p1; \
12581		__ret = (int64x1_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__s1, __p2, 3); \
	12602	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vld1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 3)); \
12582	12603	__ret; \
12583	12604	})
12584	12605	#ifdef __LITTLE_ENDIAN__
12585	12606	#define vld1_lane_s16(__p0, __p1, __p2) __extension__ ({ \
12586	12607	int16x4_t __ret; \
12587	12608	int16x4_t __s1 = __p1; \
12588		__ret = (int16x4_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__s1, __p2, 1); \
	12609	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vld1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 1)); \
12589	12610	__ret; \
12590	12611	})
12591	12612	#else
12592	12613	#define vld1_lane_s16(__p0, __p1, __p2) __extension__ ({ \
12593	12614	int16x4_t __ret; \
12594	12615	int16x4_t __s1 = __p1; \
12595		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
12596		__ret = (int16x4_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__rev1, __p2, 1); \
12597		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	12616	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	12617	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vld1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), __p2, 1)); \
	12618	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
12598	12619	__ret; \
12599	12620	})
12600	12621	#endif
...	...	@@ -12610,8 +12631,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12610	12631	poly8x8x2_t __ret; \
12611	12632	__builtin_neon_vld1_x2_v(&__ret, __p0, 4); \
12612	12633	\
12613		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
12614		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	12634	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	12635	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
12615	12636	__ret; \
12616	12637	})
12617	12638	#endif
...	...	@@ -12627,8 +12648,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12627	12648	poly16x4x2_t __ret; \
12628	12649	__builtin_neon_vld1_x2_v(&__ret, __p0, 5); \
12629	12650	\
12630		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
12631		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	12651	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	12652	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
12632	12653	__ret; \
12633	12654	})
12634	12655	#endif
...	...	@@ -12644,8 +12665,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12644	12665	poly8x16x2_t __ret; \
12645	12666	__builtin_neon_vld1q_x2_v(&__ret, __p0, 36); \
12646	12667	\
12647		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
12648		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	12668	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	12669	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
12649	12670	__ret; \
12650	12671	})
12651	12672	#endif
...	...	@@ -12661,8 +12682,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12661	12682	poly16x8x2_t __ret; \
12662	12683	__builtin_neon_vld1q_x2_v(&__ret, __p0, 37); \
12663	12684	\
12664		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
12665		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	12685	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	12686	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
12666	12687	__ret; \
12667	12688	})
12668	12689	#endif
...	...	@@ -12678,8 +12699,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12678	12699	uint8x16x2_t __ret; \
12679	12700	__builtin_neon_vld1q_x2_v(&__ret, __p0, 48); \
12680	12701	\
12681		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
12682		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	12702	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	12703	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
12683	12704	__ret; \
12684	12705	})
12685	12706	#endif
...	...	@@ -12695,8 +12716,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12695	12716	uint32x4x2_t __ret; \
12696	12717	__builtin_neon_vld1q_x2_v(&__ret, __p0, 50); \
12697	12718	\
12698		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
12699		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	12719	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	12720	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
12700	12721	__ret; \
12701	12722	})
12702	12723	#endif
...	...	@@ -12712,8 +12733,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12712	12733	uint64x2x2_t __ret; \
12713	12734	__builtin_neon_vld1q_x2_v(&__ret, __p0, 51); \
12714	12735	\
12715		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
12716		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
	12736	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	12737	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
12717	12738	__ret; \
12718	12739	})
12719	12740	#endif
...	...	@@ -12729,8 +12750,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12729	12750	uint16x8x2_t __ret; \
12730	12751	__builtin_neon_vld1q_x2_v(&__ret, __p0, 49); \
12731	12752	\
12732		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
12733		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	12753	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	12754	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
12734	12755	__ret; \
12735	12756	})
12736	12757	#endif
...	...	@@ -12746,8 +12767,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12746	12767	int8x16x2_t __ret; \
12747	12768	__builtin_neon_vld1q_x2_v(&__ret, __p0, 32); \
12748	12769	\
12749		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
12750		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	12770	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	12771	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
12751	12772	__ret; \
12752	12773	})
12753	12774	#endif
...	...	@@ -12763,8 +12784,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12763	12784	float32x4x2_t __ret; \
12764	12785	__builtin_neon_vld1q_x2_v(&__ret, __p0, 41); \
12765	12786	\
12766		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
12767		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	12787	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	12788	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
12768	12789	__ret; \
12769	12790	})
12770	12791	#endif
...	...	@@ -12780,8 +12801,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12780	12801	int32x4x2_t __ret; \
12781	12802	__builtin_neon_vld1q_x2_v(&__ret, __p0, 34); \
12782	12803	\
12783		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
12784		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	12804	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	12805	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
12785	12806	__ret; \
12786	12807	})
12787	12808	#endif
...	...	@@ -12797,8 +12818,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12797	12818	int64x2x2_t __ret; \
12798	12819	__builtin_neon_vld1q_x2_v(&__ret, __p0, 35); \
12799	12820	\
12800		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
12801		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
	12821	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	12822	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
12802	12823	__ret; \
12803	12824	})
12804	12825	#endif
...	...	@@ -12814,8 +12835,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12814	12835	int16x8x2_t __ret; \
12815	12836	__builtin_neon_vld1q_x2_v(&__ret, __p0, 33); \
12816	12837	\
12817		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
12818		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	12838	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	12839	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
12819	12840	__ret; \
12820	12841	})
12821	12842	#endif
...	...	@@ -12831,8 +12852,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12831	12852	uint8x8x2_t __ret; \
12832	12853	__builtin_neon_vld1_x2_v(&__ret, __p0, 16); \
12833	12854	\
12834		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
12835		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	12855	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	12856	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
12836	12857	__ret; \
12837	12858	})
12838	12859	#endif
...	...	@@ -12848,8 +12869,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12848	12869	uint32x2x2_t __ret; \
12849	12870	__builtin_neon_vld1_x2_v(&__ret, __p0, 18); \
12850	12871	\
12851		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
12852		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
	12872	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	12873	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
12853	12874	__ret; \
12854	12875	})
12855	12876	#endif
...	...	@@ -12870,8 +12891,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12870	12891	uint16x4x2_t __ret; \
12871	12892	__builtin_neon_vld1_x2_v(&__ret, __p0, 17); \
12872	12893	\
12873		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
12874		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	12894	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	12895	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
12875	12896	__ret; \
12876	12897	})
12877	12898	#endif
...	...	@@ -12887,8 +12908,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12887	12908	int8x8x2_t __ret; \
12888	12909	__builtin_neon_vld1_x2_v(&__ret, __p0, 0); \
12889	12910	\
12890		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
12891		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	12911	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	12912	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
12892	12913	__ret; \
12893	12914	})
12894	12915	#endif
...	...	@@ -12904,8 +12925,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12904	12925	float32x2x2_t __ret; \
12905	12926	__builtin_neon_vld1_x2_v(&__ret, __p0, 9); \
12906	12927	\
12907		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
12908		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
	12928	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	12929	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
12909	12930	__ret; \
12910	12931	})
12911	12932	#endif
...	...	@@ -12921,8 +12942,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12921	12942	int32x2x2_t __ret; \
12922	12943	__builtin_neon_vld1_x2_v(&__ret, __p0, 2); \
12923	12944	\
12924		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
12925		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
	12945	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	12946	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
12926	12947	__ret; \
12927	12948	})
12928	12949	#endif
...	...	@@ -12943,8 +12964,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12943	12964	int16x4x2_t __ret; \
12944	12965	__builtin_neon_vld1_x2_v(&__ret, __p0, 1); \
12945	12966	\
12946		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
12947		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	12967	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	12968	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
12948	12969	__ret; \
12949	12970	})
12950	12971	#endif
...	...	@@ -12960,9 +12981,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12960	12981	poly8x8x3_t __ret; \
12961	12982	__builtin_neon_vld1_x3_v(&__ret, __p0, 4); \
12962	12983	\
12963		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
12964		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
12965		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	12984	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	12985	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	12986	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
12966	12987	__ret; \
12967	12988	})
12968	12989	#endif
...	...	@@ -12978,9 +12999,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12978	12999	poly16x4x3_t __ret; \
12979	13000	__builtin_neon_vld1_x3_v(&__ret, __p0, 5); \
12980	13001	\
12981		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
12982		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
12983		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	13002	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	13003	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	13004	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
12984	13005	__ret; \
12985	13006	})
12986	13007	#endif
...	...	@@ -12996,9 +13017,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
12996	13017	poly8x16x3_t __ret; \
12997	13018	__builtin_neon_vld1q_x3_v(&__ret, __p0, 36); \
12998	13019	\
12999		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
13000		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
13001		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	13020	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	13021	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	13022	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
13002	13023	__ret; \
13003	13024	})
13004	13025	#endif
...	...	@@ -13014,9 +13035,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13014	13035	poly16x8x3_t __ret; \
13015	13036	__builtin_neon_vld1q_x3_v(&__ret, __p0, 37); \
13016	13037	\
13017		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13018		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
13019		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	13038	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	13039	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	13040	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
13020	13041	__ret; \
13021	13042	})
13022	13043	#endif
...	...	@@ -13032,9 +13053,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13032	13053	uint8x16x3_t __ret; \
13033	13054	__builtin_neon_vld1q_x3_v(&__ret, __p0, 48); \
13034	13055	\
13035		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
13036		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
13037		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	13056	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	13057	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	13058	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
13038	13059	__ret; \
13039	13060	})
13040	13061	#endif
...	...	@@ -13050,9 +13071,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13050	13071	uint32x4x3_t __ret; \
13051	13072	__builtin_neon_vld1q_x3_v(&__ret, __p0, 50); \
13052	13073	\
13053		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
13054		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
13055		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	13074	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	13075	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
	13076	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_32); \
13056	13077	__ret; \
13057	13078	})
13058	13079	#endif
...	...	@@ -13068,9 +13089,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13068	13089	uint64x2x3_t __ret; \
13069	13090	__builtin_neon_vld1q_x3_v(&__ret, __p0, 51); \
13070	13091	\
13071		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
13072		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
13073		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
	13092	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	13093	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	13094	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
13074	13095	__ret; \
13075	13096	})
13076	13097	#endif
...	...	@@ -13086,9 +13107,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13086	13107	uint16x8x3_t __ret; \
13087	13108	__builtin_neon_vld1q_x3_v(&__ret, __p0, 49); \
13088	13109	\
13089		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13090		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
13091		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	13110	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	13111	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	13112	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
13092	13113	__ret; \
13093	13114	})
13094	13115	#endif
...	...	@@ -13104,9 +13125,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13104	13125	int8x16x3_t __ret; \
13105	13126	__builtin_neon_vld1q_x3_v(&__ret, __p0, 32); \
13106	13127	\
13107		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
13108		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
13109		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	13128	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	13129	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	13130	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
13110	13131	__ret; \
13111	13132	})
13112	13133	#endif
...	...	@@ -13122,9 +13143,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13122	13143	float32x4x3_t __ret; \
13123	13144	__builtin_neon_vld1q_x3_v(&__ret, __p0, 41); \
13124	13145	\
13125		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
13126		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
13127		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	13146	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	13147	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
	13148	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_32); \
13128	13149	__ret; \
13129	13150	})
13130	13151	#endif
...	...	@@ -13140,9 +13161,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13140	13161	int32x4x3_t __ret; \
13141	13162	__builtin_neon_vld1q_x3_v(&__ret, __p0, 34); \
13142	13163	\
13143		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
13144		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
13145		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	13164	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	13165	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
	13166	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_32); \
13146	13167	__ret; \
13147	13168	})
13148	13169	#endif
...	...	@@ -13158,9 +13179,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13158	13179	int64x2x3_t __ret; \
13159	13180	__builtin_neon_vld1q_x3_v(&__ret, __p0, 35); \
13160	13181	\
13161		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
13162		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
13163		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
	13182	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	13183	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	13184	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
13164	13185	__ret; \
13165	13186	})
13166	13187	#endif
...	...	@@ -13176,9 +13197,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13176	13197	int16x8x3_t __ret; \
13177	13198	__builtin_neon_vld1q_x3_v(&__ret, __p0, 33); \
13178	13199	\
13179		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13180		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
13181		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	13200	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	13201	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	13202	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
13182	13203	__ret; \
13183	13204	})
13184	13205	#endif
...	...	@@ -13194,9 +13215,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13194	13215	uint8x8x3_t __ret; \
13195	13216	__builtin_neon_vld1_x3_v(&__ret, __p0, 16); \
13196	13217	\
13197		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13198		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
13199		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	13218	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	13219	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	13220	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
13200	13221	__ret; \
13201	13222	})
13202	13223	#endif
...	...	@@ -13212,9 +13233,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13212	13233	uint32x2x3_t __ret; \
13213	13234	__builtin_neon_vld1_x3_v(&__ret, __p0, 18); \
13214	13235	\
13215		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
13216		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
13217		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
	13236	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	13237	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
	13238	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_32); \
13218	13239	__ret; \
13219	13240	})
13220	13241	#endif
...	...	@@ -13235,9 +13256,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13235	13256	uint16x4x3_t __ret; \
13236	13257	__builtin_neon_vld1_x3_v(&__ret, __p0, 17); \
13237	13258	\
13238		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
13239		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
13240		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	13259	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	13260	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	13261	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
13241	13262	__ret; \
13242	13263	})
13243	13264	#endif
...	...	@@ -13253,9 +13274,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13253	13274	int8x8x3_t __ret; \
13254	13275	__builtin_neon_vld1_x3_v(&__ret, __p0, 0); \
13255	13276	\
13256		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13257		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
13258		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	13277	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	13278	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	13279	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
13259	13280	__ret; \
13260	13281	})
13261	13282	#endif
...	...	@@ -13271,9 +13292,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13271	13292	float32x2x3_t __ret; \
13272	13293	__builtin_neon_vld1_x3_v(&__ret, __p0, 9); \
13273	13294	\
13274		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
13275		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
13276		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
	13295	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	13296	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
	13297	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_32); \
13277	13298	__ret; \
13278	13299	})
13279	13300	#endif
...	...	@@ -13289,9 +13310,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13289	13310	int32x2x3_t __ret; \
13290	13311	__builtin_neon_vld1_x3_v(&__ret, __p0, 2); \
13291	13312	\
13292		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
13293		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
13294		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
	13313	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	13314	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
	13315	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_32); \
13295	13316	__ret; \
13296	13317	})
13297	13318	#endif
...	...	@@ -13312,9 +13333,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13312	13333	int16x4x3_t __ret; \
13313	13334	__builtin_neon_vld1_x3_v(&__ret, __p0, 1); \
13314	13335	\
13315		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
13316		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
13317		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	13336	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	13337	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	13338	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
13318	13339	__ret; \
13319	13340	})
13320	13341	#endif
...	...	@@ -13330,10 +13351,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13330	13351	poly8x8x4_t __ret; \
13331	13352	__builtin_neon_vld1_x4_v(&__ret, __p0, 4); \
13332	13353	\
13333		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13334		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
13335		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
13336		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	13354	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	13355	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	13356	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
	13357	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_8); \
13337	13358	__ret; \
13338	13359	})
13339	13360	#endif
...	...	@@ -13349,10 +13370,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13349	13370	poly16x4x4_t __ret; \
13350	13371	__builtin_neon_vld1_x4_v(&__ret, __p0, 5); \
13351	13372	\
13352		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
13353		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
13354		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
13355		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	13373	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	13374	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	13375	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
	13376	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_16); \
13356	13377	__ret; \
13357	13378	})
13358	13379	#endif
...	...	@@ -13368,10 +13389,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13368	13389	poly8x16x4_t __ret; \
13369	13390	__builtin_neon_vld1q_x4_v(&__ret, __p0, 36); \
13370	13391	\
13371		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
13372		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
13373		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
13374		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	13392	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	13393	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	13394	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
	13395	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_8); \
13375	13396	__ret; \
13376	13397	})
13377	13398	#endif
...	...	@@ -13387,10 +13408,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13387	13408	poly16x8x4_t __ret; \
13388	13409	__builtin_neon_vld1q_x4_v(&__ret, __p0, 37); \
13389	13410	\
13390		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13391		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
13392		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
13393		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	13411	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	13412	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	13413	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
	13414	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_16); \
13394	13415	__ret; \
13395	13416	})
13396	13417	#endif
...	...	@@ -13406,10 +13427,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13406	13427	uint8x16x4_t __ret; \
13407	13428	__builtin_neon_vld1q_x4_v(&__ret, __p0, 48); \
13408	13429	\
13409		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
13410		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
13411		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
13412		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	13430	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	13431	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	13432	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
	13433	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_8); \
13413	13434	__ret; \
13414	13435	})
13415	13436	#endif
...	...	@@ -13425,10 +13446,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13425	13446	uint32x4x4_t __ret; \
13426	13447	__builtin_neon_vld1q_x4_v(&__ret, __p0, 50); \
13427	13448	\
13428		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
13429		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
13430		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
13431		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	13449	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	13450	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
	13451	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_32); \
	13452	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_32); \
13432	13453	__ret; \
13433	13454	})
13434	13455	#endif
...	...	@@ -13444,10 +13465,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13444	13465	uint64x2x4_t __ret; \
13445	13466	__builtin_neon_vld1q_x4_v(&__ret, __p0, 51); \
13446	13467	\
13447		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
13448		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
13449		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
13450		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
	13468	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	13469	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	13470	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
	13471	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_64); \
13451	13472	__ret; \
13452	13473	})
13453	13474	#endif
...	...	@@ -13463,10 +13484,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13463	13484	uint16x8x4_t __ret; \
13464	13485	__builtin_neon_vld1q_x4_v(&__ret, __p0, 49); \
13465	13486	\
13466		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13467		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
13468		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
13469		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	13487	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	13488	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	13489	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
	13490	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_16); \
13470	13491	__ret; \
13471	13492	})
13472	13493	#endif
...	...	@@ -13482,10 +13503,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13482	13503	int8x16x4_t __ret; \
13483	13504	__builtin_neon_vld1q_x4_v(&__ret, __p0, 32); \
13484	13505	\
13485		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
13486		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
13487		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
13488		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	13506	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	13507	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	13508	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
	13509	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_8); \
13489	13510	__ret; \
13490	13511	})
13491	13512	#endif
...	...	@@ -13501,10 +13522,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13501	13522	float32x4x4_t __ret; \
13502	13523	__builtin_neon_vld1q_x4_v(&__ret, __p0, 41); \
13503	13524	\
13504		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
13505		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
13506		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
13507		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	13525	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	13526	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
	13527	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_32); \
	13528	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_32); \
13508	13529	__ret; \
13509	13530	})
13510	13531	#endif
...	...	@@ -13520,10 +13541,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13520	13541	int32x4x4_t __ret; \
13521	13542	__builtin_neon_vld1q_x4_v(&__ret, __p0, 34); \
13522	13543	\
13523		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
13524		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
13525		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
13526		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	13544	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	13545	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
	13546	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_32); \
	13547	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_32); \
13527	13548	__ret; \
13528	13549	})
13529	13550	#endif
...	...	@@ -13539,10 +13560,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13539	13560	int64x2x4_t __ret; \
13540	13561	__builtin_neon_vld1q_x4_v(&__ret, __p0, 35); \
13541	13562	\
13542		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
13543		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
13544		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
13545		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
	13563	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	13564	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	13565	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
	13566	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_64); \
13546	13567	__ret; \
13547	13568	})
13548	13569	#endif
...	...	@@ -13558,10 +13579,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13558	13579	int16x8x4_t __ret; \
13559	13580	__builtin_neon_vld1q_x4_v(&__ret, __p0, 33); \
13560	13581	\
13561		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13562		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
13563		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
13564		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	13582	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	13583	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	13584	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
	13585	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_16); \
13565	13586	__ret; \
13566	13587	})
13567	13588	#endif
...	...	@@ -13577,10 +13598,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13577	13598	uint8x8x4_t __ret; \
13578	13599	__builtin_neon_vld1_x4_v(&__ret, __p0, 16); \
13579	13600	\
13580		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13581		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
13582		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
13583		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	13601	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	13602	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	13603	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
	13604	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_8); \
13584	13605	__ret; \
13585	13606	})
13586	13607	#endif
...	...	@@ -13596,10 +13617,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13596	13617	uint32x2x4_t __ret; \
13597	13618	__builtin_neon_vld1_x4_v(&__ret, __p0, 18); \
13598	13619	\
13599		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
13600		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
13601		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
13602		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
	13620	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	13621	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
	13622	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_32); \
	13623	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_32); \
13603	13624	__ret; \
13604	13625	})
13605	13626	#endif
...	...	@@ -13620,10 +13641,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13620	13641	uint16x4x4_t __ret; \
13621	13642	__builtin_neon_vld1_x4_v(&__ret, __p0, 17); \
13622	13643	\
13623		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
13624		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
13625		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
13626		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	13644	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	13645	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	13646	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
	13647	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_16); \
13627	13648	__ret; \
13628	13649	})
13629	13650	#endif
...	...	@@ -13639,10 +13660,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13639	13660	int8x8x4_t __ret; \
13640	13661	__builtin_neon_vld1_x4_v(&__ret, __p0, 0); \
13641	13662	\
13642		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13643		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
13644		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
13645		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	13663	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	13664	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	13665	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
	13666	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_8); \
13646	13667	__ret; \
13647	13668	})
13648	13669	#endif
...	...	@@ -13658,10 +13679,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13658	13679	float32x2x4_t __ret; \
13659	13680	__builtin_neon_vld1_x4_v(&__ret, __p0, 9); \
13660	13681	\
13661		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
13662		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
13663		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
13664		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
	13682	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	13683	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
	13684	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_32); \
	13685	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_32); \
13665	13686	__ret; \
13666	13687	})
13667	13688	#endif
...	...	@@ -13677,10 +13698,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13677	13698	int32x2x4_t __ret; \
13678	13699	__builtin_neon_vld1_x4_v(&__ret, __p0, 2); \
13679	13700	\
13680		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
13681		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
13682		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
13683		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
	13701	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	13702	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
	13703	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_32); \
	13704	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_32); \
13684	13705	__ret; \
13685	13706	})
13686	13707	#endif
...	...	@@ -13701,10 +13722,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13701	13722	int16x4x4_t __ret; \
13702	13723	__builtin_neon_vld1_x4_v(&__ret, __p0, 1); \
13703	13724	\
13704		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
13705		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
13706		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
13707		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	13725	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	13726	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	13727	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
	13728	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_16); \
13708	13729	__ret; \
13709	13730	})
13710	13731	#endif
...	...	@@ -13720,8 +13741,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13720	13741	poly8x8x2_t __ret; \
13721	13742	__builtin_neon_vld2_v(&__ret, __p0, 4); \
13722	13743	\
13723		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13724		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	13744	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	13745	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
13725	13746	__ret; \
13726	13747	})
13727	13748	#endif
...	...	@@ -13737,8 +13758,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13737	13758	poly16x4x2_t __ret; \
13738	13759	__builtin_neon_vld2_v(&__ret, __p0, 5); \
13739	13760	\
13740		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
13741		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	13761	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	13762	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
13742	13763	__ret; \
13743	13764	})
13744	13765	#endif
...	...	@@ -13754,8 +13775,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13754	13775	poly8x16x2_t __ret; \
13755	13776	__builtin_neon_vld2q_v(&__ret, __p0, 36); \
13756	13777	\
13757		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
13758		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	13778	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	13779	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
13759	13780	__ret; \
13760	13781	})
13761	13782	#endif
...	...	@@ -13771,8 +13792,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13771	13792	poly16x8x2_t __ret; \
13772	13793	__builtin_neon_vld2q_v(&__ret, __p0, 37); \
13773	13794	\
13774		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13775		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	13795	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	13796	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
13776	13797	__ret; \
13777	13798	})
13778	13799	#endif
...	...	@@ -13788,8 +13809,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13788	13809	uint8x16x2_t __ret; \
13789	13810	__builtin_neon_vld2q_v(&__ret, __p0, 48); \
13790	13811	\
13791		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
13792		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	13812	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	13813	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
13793	13814	__ret; \
13794	13815	})
13795	13816	#endif
...	...	@@ -13805,8 +13826,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13805	13826	uint32x4x2_t __ret; \
13806	13827	__builtin_neon_vld2q_v(&__ret, __p0, 50); \
13807	13828	\
13808		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
13809		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	13829	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	13830	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
13810	13831	__ret; \
13811	13832	})
13812	13833	#endif
...	...	@@ -13822,8 +13843,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13822	13843	uint16x8x2_t __ret; \
13823	13844	__builtin_neon_vld2q_v(&__ret, __p0, 49); \
13824	13845	\
13825		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13826		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	13846	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	13847	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
13827	13848	__ret; \
13828	13849	})
13829	13850	#endif
...	...	@@ -13839,8 +13860,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13839	13860	int8x16x2_t __ret; \
13840	13861	__builtin_neon_vld2q_v(&__ret, __p0, 32); \
13841	13862	\
13842		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
13843		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	13863	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	13864	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
13844	13865	__ret; \
13845	13866	})
13846	13867	#endif
...	...	@@ -13856,8 +13877,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13856	13877	float32x4x2_t __ret; \
13857	13878	__builtin_neon_vld2q_v(&__ret, __p0, 41); \
13858	13879	\
13859		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
13860		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	13880	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	13881	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
13861	13882	__ret; \
13862	13883	})
13863	13884	#endif
...	...	@@ -13873,8 +13894,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13873	13894	int32x4x2_t __ret; \
13874	13895	__builtin_neon_vld2q_v(&__ret, __p0, 34); \
13875	13896	\
13876		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
13877		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	13897	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	13898	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
13878	13899	__ret; \
13879	13900	})
13880	13901	#endif
...	...	@@ -13890,8 +13911,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13890	13911	int16x8x2_t __ret; \
13891	13912	__builtin_neon_vld2q_v(&__ret, __p0, 33); \
13892	13913	\
13893		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13894		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	13914	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	13915	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
13895	13916	__ret; \
13896	13917	})
13897	13918	#endif
...	...	@@ -13907,8 +13928,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13907	13928	uint8x8x2_t __ret; \
13908	13929	__builtin_neon_vld2_v(&__ret, __p0, 16); \
13909	13930	\
13910		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13911		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	13931	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	13932	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
13912	13933	__ret; \
13913	13934	})
13914	13935	#endif
...	...	@@ -13924,8 +13945,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13924	13945	uint32x2x2_t __ret; \
13925	13946	__builtin_neon_vld2_v(&__ret, __p0, 18); \
13926	13947	\
13927		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
13928		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
	13948	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	13949	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
13929	13950	__ret; \
13930	13951	})
13931	13952	#endif
...	...	@@ -13946,8 +13967,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13946	13967	uint16x4x2_t __ret; \
13947	13968	__builtin_neon_vld2_v(&__ret, __p0, 17); \
13948	13969	\
13949		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
13950		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	13970	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	13971	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
13951	13972	__ret; \
13952	13973	})
13953	13974	#endif
...	...	@@ -13963,8 +13984,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13963	13984	int8x8x2_t __ret; \
13964	13985	__builtin_neon_vld2_v(&__ret, __p0, 0); \
13965	13986	\
13966		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13967		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	13987	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	13988	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
13968	13989	__ret; \
13969	13990	})
13970	13991	#endif
...	...	@@ -13980,8 +14001,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13980	14001	float32x2x2_t __ret; \
13981	14002	__builtin_neon_vld2_v(&__ret, __p0, 9); \
13982	14003	\
13983		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
13984		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
	14004	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	14005	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
13985	14006	__ret; \
13986	14007	})
13987	14008	#endif
...	...	@@ -13997,8 +14018,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
13997	14018	int32x2x2_t __ret; \
13998	14019	__builtin_neon_vld2_v(&__ret, __p0, 2); \
13999	14020	\
14000		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
14001		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
	14021	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	14022	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
14002	14023	__ret; \
14003	14024	})
14004	14025	#endif
...	...	@@ -14019,8 +14040,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14019	14040	int16x4x2_t __ret; \
14020	14041	__builtin_neon_vld2_v(&__ret, __p0, 1); \
14021	14042	\
14022		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
14023		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	14043	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	14044	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
14024	14045	__ret; \
14025	14046	})
14026	14047	#endif
...	...	@@ -14036,8 +14057,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14036	14057	poly8x8x2_t __ret; \
14037	14058	__builtin_neon_vld2_dup_v(&__ret, __p0, 4); \
14038	14059	\
14039		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
14040		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	14060	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	14061	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
14041	14062	__ret; \
14042	14063	})
14043	14064	#endif
...	...	@@ -14053,8 +14074,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14053	14074	poly16x4x2_t __ret; \
14054	14075	__builtin_neon_vld2_dup_v(&__ret, __p0, 5); \
14055	14076	\
14056		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
14057		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	14077	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	14078	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
14058	14079	__ret; \
14059	14080	})
14060	14081	#endif
...	...	@@ -14070,8 +14091,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14070	14091	poly8x16x2_t __ret; \
14071	14092	__builtin_neon_vld2q_dup_v(&__ret, __p0, 36); \
14072	14093	\
14073		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
14074		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	14094	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	14095	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
14075	14096	__ret; \
14076	14097	})
14077	14098	#endif
...	...	@@ -14087,8 +14108,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14087	14108	poly16x8x2_t __ret; \
14088	14109	__builtin_neon_vld2q_dup_v(&__ret, __p0, 37); \
14089	14110	\
14090		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
14091		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	14111	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	14112	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
14092	14113	__ret; \
14093	14114	})
14094	14115	#endif
...	...	@@ -14104,8 +14125,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14104	14125	uint8x16x2_t __ret; \
14105	14126	__builtin_neon_vld2q_dup_v(&__ret, __p0, 48); \
14106	14127	\
14107		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
14108		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	14128	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	14129	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
14109	14130	__ret; \
14110	14131	})
14111	14132	#endif
...	...	@@ -14121,8 +14142,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14121	14142	uint32x4x2_t __ret; \
14122	14143	__builtin_neon_vld2q_dup_v(&__ret, __p0, 50); \
14123	14144	\
14124		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
14125		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	14145	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	14146	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
14126	14147	__ret; \
14127	14148	})
14128	14149	#endif
...	...	@@ -14138,8 +14159,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14138	14159	uint64x2x2_t __ret; \
14139	14160	__builtin_neon_vld2q_dup_v(&__ret, __p0, 51); \
14140	14161	\
14141		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
14142		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
	14162	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	14163	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
14143	14164	__ret; \
14144	14165	})
14145	14166	#endif
...	...	@@ -14155,8 +14176,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14155	14176	uint16x8x2_t __ret; \
14156	14177	__builtin_neon_vld2q_dup_v(&__ret, __p0, 49); \
14157	14178	\
14158		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
14159		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	14179	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	14180	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
14160	14181	__ret; \
14161	14182	})
14162	14183	#endif
...	...	@@ -14172,8 +14193,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14172	14193	int8x16x2_t __ret; \
14173	14194	__builtin_neon_vld2q_dup_v(&__ret, __p0, 32); \
14174	14195	\
14175		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
14176		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	14196	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	14197	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
14177	14198	__ret; \
14178	14199	})
14179	14200	#endif
...	...	@@ -14189,8 +14210,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14189	14210	float32x4x2_t __ret; \
14190	14211	__builtin_neon_vld2q_dup_v(&__ret, __p0, 41); \
14191	14212	\
14192		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
14193		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	14213	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	14214	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
14194	14215	__ret; \
14195	14216	})
14196	14217	#endif
...	...	@@ -14206,8 +14227,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14206	14227	int32x4x2_t __ret; \
14207	14228	__builtin_neon_vld2q_dup_v(&__ret, __p0, 34); \
14208	14229	\
14209		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
14210		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	14230	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	14231	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
14211	14232	__ret; \
14212	14233	})
14213	14234	#endif
...	...	@@ -14223,8 +14244,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14223	14244	int64x2x2_t __ret; \
14224	14245	__builtin_neon_vld2q_dup_v(&__ret, __p0, 35); \
14225	14246	\
14226		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
14227		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
	14247	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	14248	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
14228	14249	__ret; \
14229	14250	})
14230	14251	#endif
...	...	@@ -14240,8 +14261,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14240	14261	int16x8x2_t __ret; \
14241	14262	__builtin_neon_vld2q_dup_v(&__ret, __p0, 33); \
14242	14263	\
14243		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
14244		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	14264	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	14265	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
14245	14266	__ret; \
14246	14267	})
14247	14268	#endif
...	...	@@ -14257,8 +14278,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14257	14278	uint8x8x2_t __ret; \
14258	14279	__builtin_neon_vld2_dup_v(&__ret, __p0, 16); \
14259	14280	\
14260		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
14261		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	14281	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	14282	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
14262	14283	__ret; \
14263	14284	})
14264	14285	#endif
...	...	@@ -14274,8 +14295,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14274	14295	uint32x2x2_t __ret; \
14275	14296	__builtin_neon_vld2_dup_v(&__ret, __p0, 18); \
14276	14297	\
14277		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
14278		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
	14298	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	14299	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
14279	14300	__ret; \
14280	14301	})
14281	14302	#endif
...	...	@@ -14296,8 +14317,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14296	14317	uint16x4x2_t __ret; \
14297	14318	__builtin_neon_vld2_dup_v(&__ret, __p0, 17); \
14298	14319	\
14299		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
14300		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	14320	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	14321	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
14301	14322	__ret; \
14302	14323	})
14303	14324	#endif
...	...	@@ -14313,8 +14334,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14313	14334	int8x8x2_t __ret; \
14314	14335	__builtin_neon_vld2_dup_v(&__ret, __p0, 0); \
14315	14336	\
14316		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
14317		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	14337	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	14338	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
14318	14339	__ret; \
14319	14340	})
14320	14341	#endif
...	...	@@ -14330,8 +14351,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14330	14351	float32x2x2_t __ret; \
14331	14352	__builtin_neon_vld2_dup_v(&__ret, __p0, 9); \
14332	14353	\
14333		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
14334		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
	14354	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	14355	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
14335	14356	__ret; \
14336	14357	})
14337	14358	#endif
...	...	@@ -14347,8 +14368,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14347	14368	int32x2x2_t __ret; \
14348	14369	__builtin_neon_vld2_dup_v(&__ret, __p0, 2); \
14349	14370	\
14350		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
14351		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
	14371	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	14372	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
14352	14373	__ret; \
14353	14374	})
14354	14375	#endif
...	...	@@ -14369,8 +14390,8 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14369	14390	int16x4x2_t __ret; \
14370	14391	__builtin_neon_vld2_dup_v(&__ret, __p0, 1); \
14371	14392	\
14372		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
14373		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	14393	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	14394	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
14374	14395	__ret; \
14375	14396	})
14376	14397	#endif
...	...	@@ -14379,7 +14400,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14379	14400	#define vld2_lane_p8(__p0, __p1, __p2) __extension__ ({ \
14380	14401	poly8x8x2_t __ret; \
14381	14402	poly8x8x2_t __s1 = __p1; \
14382		__builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 4); \
	14403	__builtin_neon_vld2_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 4); \
14383	14404	__ret; \
14384	14405	})
14385	14406	#else
...	...	@@ -14387,12 +14408,12 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14387	14408	poly8x8x2_t __ret; \
14388	14409	poly8x8x2_t __s1 = __p1; \
14389	14410	poly8x8x2_t __rev1; \
14390		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
14391		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
14392		__builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 4); \
	14411	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	14412	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	14413	__builtin_neon_vld2_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __p2, 4); \
14393	14414	\
14394		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
14395		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	14415	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	14416	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
14396	14417	__ret; \
14397	14418	})
14398	14419	#endif
...	...	@@ -14401,7 +14422,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14401	14422	#define vld2_lane_p16(__p0, __p1, __p2) __extension__ ({ \
14402	14423	poly16x4x2_t __ret; \
14403	14424	poly16x4x2_t __s1 = __p1; \
14404		__builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 5); \
	14425	__builtin_neon_vld2_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 5); \
14405	14426	__ret; \
14406	14427	})
14407	14428	#else
...	...	@@ -14409,12 +14430,12 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14409	14430	poly16x4x2_t __ret; \
14410	14431	poly16x4x2_t __s1 = __p1; \
14411	14432	poly16x4x2_t __rev1; \
14412		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
14413		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
14414		__builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 5); \
	14433	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	14434	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	14435	__builtin_neon_vld2_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __p2, 5); \
14415	14436	\
14416		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
14417		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	14437	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	14438	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
14418	14439	__ret; \
14419	14440	})
14420	14441	#endif
...	...	@@ -14423,7 +14444,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14423	14444	#define vld2q_lane_p16(__p0, __p1, __p2) __extension__ ({ \
14424	14445	poly16x8x2_t __ret; \
14425	14446	poly16x8x2_t __s1 = __p1; \
14426		__builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 37); \
	14447	__builtin_neon_vld2q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 37); \
14427	14448	__ret; \
14428	14449	})
14429	14450	#else
...	...	@@ -14431,12 +14452,12 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14431	14452	poly16x8x2_t __ret; \
14432	14453	poly16x8x2_t __s1 = __p1; \
14433	14454	poly16x8x2_t __rev1; \
14434		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
14435		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
14436		__builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 37); \
	14455	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	14456	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	14457	__builtin_neon_vld2q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 37); \
14437	14458	\
14438		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
14439		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	14459	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	14460	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
14440	14461	__ret; \
14441	14462	})
14442	14463	#endif
...	...	@@ -14445,7 +14466,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14445	14466	#define vld2q_lane_u32(__p0, __p1, __p2) __extension__ ({ \
14446	14467	uint32x4x2_t __ret; \
14447	14468	uint32x4x2_t __s1 = __p1; \
14448		__builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 50); \
	14469	__builtin_neon_vld2q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 50); \
14449	14470	__ret; \
14450	14471	})
14451	14472	#else
...	...	@@ -14453,12 +14474,12 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14453	14474	uint32x4x2_t __ret; \
14454	14475	uint32x4x2_t __s1 = __p1; \
14455	14476	uint32x4x2_t __rev1; \
14456		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
14457		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
14458		__builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 50); \
	14477	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	14478	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	14479	__builtin_neon_vld2q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 50); \
14459	14480	\
14460		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
14461		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	14481	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	14482	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
14462	14483	__ret; \
14463	14484	})
14464	14485	#endif
...	...	@@ -14467,7 +14488,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14467	14488	#define vld2q_lane_u16(__p0, __p1, __p2) __extension__ ({ \
14468	14489	uint16x8x2_t __ret; \
14469	14490	uint16x8x2_t __s1 = __p1; \
14470		__builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 49); \
	14491	__builtin_neon_vld2q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 49); \
14471	14492	__ret; \
14472	14493	})
14473	14494	#else
...	...	@@ -14475,12 +14496,12 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14475	14496	uint16x8x2_t __ret; \
14476	14497	uint16x8x2_t __s1 = __p1; \
14477	14498	uint16x8x2_t __rev1; \
14478		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
14479		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
14480		__builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 49); \
	14499	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	14500	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	14501	__builtin_neon_vld2q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 49); \
14481	14502	\
14482		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
14483		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	14503	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	14504	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
14484	14505	__ret; \
14485	14506	})
14486	14507	#endif
...	...	@@ -14489,7 +14510,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14489	14510	#define vld2q_lane_f32(__p0, __p1, __p2) __extension__ ({ \
14490	14511	float32x4x2_t __ret; \
14491	14512	float32x4x2_t __s1 = __p1; \
14492		__builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 41); \
	14513	__builtin_neon_vld2q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 41); \
14493	14514	__ret; \
14494	14515	})
14495	14516	#else
...	...	@@ -14497,12 +14518,12 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14497	14518	float32x4x2_t __ret; \
14498	14519	float32x4x2_t __s1 = __p1; \
14499	14520	float32x4x2_t __rev1; \
14500		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
14501		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
14502		__builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 41); \
	14521	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	14522	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	14523	__builtin_neon_vld2q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 41); \
14503	14524	\
14504		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
14505		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	14525	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	14526	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
14506	14527	__ret; \
14507	14528	})
14508	14529	#endif
...	...	@@ -14511,7 +14532,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14511	14532	#define vld2q_lane_s32(__p0, __p1, __p2) __extension__ ({ \
14512	14533	int32x4x2_t __ret; \
14513	14534	int32x4x2_t __s1 = __p1; \
14514		__builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 34); \
	14535	__builtin_neon_vld2q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 34); \
14515	14536	__ret; \
14516	14537	})
14517	14538	#else
...	...	@@ -14519,12 +14540,12 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14519	14540	int32x4x2_t __ret; \
14520	14541	int32x4x2_t __s1 = __p1; \
14521	14542	int32x4x2_t __rev1; \
14522		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
14523		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
14524		__builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 34); \
	14543	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	14544	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	14545	__builtin_neon_vld2q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 34); \
14525	14546	\
14526		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
14527		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	14547	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	14548	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
14528	14549	__ret; \
14529	14550	})
14530	14551	#endif
...	...	@@ -14533,7 +14554,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14533	14554	#define vld2q_lane_s16(__p0, __p1, __p2) __extension__ ({ \
14534	14555	int16x8x2_t __ret; \
14535	14556	int16x8x2_t __s1 = __p1; \
14536		__builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 33); \
	14557	__builtin_neon_vld2q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 33); \
14537	14558	__ret; \
14538	14559	})
14539	14560	#else
...	...	@@ -14541,12 +14562,12 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14541	14562	int16x8x2_t __ret; \
14542	14563	int16x8x2_t __s1 = __p1; \
14543	14564	int16x8x2_t __rev1; \
14544		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
14545		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
14546		__builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 33); \
	14565	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	14566	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	14567	__builtin_neon_vld2q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 33); \
14547	14568	\
14548		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
14549		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	14569	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	14570	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
14550	14571	__ret; \
14551	14572	})
14552	14573	#endif
...	...	@@ -14555,7 +14576,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14555	14576	#define vld2_lane_u8(__p0, __p1, __p2) __extension__ ({ \
14556	14577	uint8x8x2_t __ret; \
14557	14578	uint8x8x2_t __s1 = __p1; \
14558		__builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 16); \
	14579	__builtin_neon_vld2_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 16); \
14559	14580	__ret; \
14560	14581	})
14561	14582	#else
...	...	@@ -14563,12 +14584,12 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14563	14584	uint8x8x2_t __ret; \
14564	14585	uint8x8x2_t __s1 = __p1; \
14565	14586	uint8x8x2_t __rev1; \
14566		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
14567		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
14568		__builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 16); \
	14587	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	14588	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	14589	__builtin_neon_vld2_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __p2, 16); \
14569	14590	\
14570		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
14571		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	14591	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	14592	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
14572	14593	__ret; \
14573	14594	})
14574	14595	#endif
...	...	@@ -14577,7 +14598,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14577	14598	#define vld2_lane_u32(__p0, __p1, __p2) __extension__ ({ \
14578	14599	uint32x2x2_t __ret; \
14579	14600	uint32x2x2_t __s1 = __p1; \
14580		__builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 18); \
	14601	__builtin_neon_vld2_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 18); \
14581	14602	__ret; \
14582	14603	})
14583	14604	#else
...	...	@@ -14585,12 +14606,12 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14585	14606	uint32x2x2_t __ret; \
14586	14607	uint32x2x2_t __s1 = __p1; \
14587	14608	uint32x2x2_t __rev1; \
14588		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
14589		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
14590		__builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 18); \
	14609	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	14610	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	14611	__builtin_neon_vld2_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __p2, 18); \
14591	14612	\
14592		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
14593		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
	14613	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	14614	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
14594	14615	__ret; \
14595	14616	})
14596	14617	#endif
...	...	@@ -14599,7 +14620,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14599	14620	#define vld2_lane_u16(__p0, __p1, __p2) __extension__ ({ \
14600	14621	uint16x4x2_t __ret; \
14601	14622	uint16x4x2_t __s1 = __p1; \
14602		__builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 17); \
	14623	__builtin_neon_vld2_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 17); \
14603	14624	__ret; \
14604	14625	})
14605	14626	#else
...	...	@@ -14607,12 +14628,12 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14607	14628	uint16x4x2_t __ret; \
14608	14629	uint16x4x2_t __s1 = __p1; \
14609	14630	uint16x4x2_t __rev1; \
14610		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
14611		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
14612		__builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 17); \
	14631	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	14632	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	14633	__builtin_neon_vld2_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __p2, 17); \
14613	14634	\
14614		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
14615		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	14635	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	14636	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
14616	14637	__ret; \
14617	14638	})
14618	14639	#endif
...	...	@@ -14621,7 +14642,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14621	14642	#define vld2_lane_s8(__p0, __p1, __p2) __extension__ ({ \
14622	14643	int8x8x2_t __ret; \
14623	14644	int8x8x2_t __s1 = __p1; \
14624		__builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 0); \
	14645	__builtin_neon_vld2_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 0); \
14625	14646	__ret; \
14626	14647	})
14627	14648	#else
...	...	@@ -14629,12 +14650,12 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14629	14650	int8x8x2_t __ret; \
14630	14651	int8x8x2_t __s1 = __p1; \
14631	14652	int8x8x2_t __rev1; \
14632		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
14633		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
14634		__builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 0); \
	14653	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	14654	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	14655	__builtin_neon_vld2_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __p2, 0); \
14635	14656	\
14636		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
14637		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	14657	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	14658	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
14638	14659	__ret; \
14639	14660	})
14640	14661	#endif
...	...	@@ -14643,7 +14664,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14643	14664	#define vld2_lane_f32(__p0, __p1, __p2) __extension__ ({ \
14644	14665	float32x2x2_t __ret; \
14645	14666	float32x2x2_t __s1 = __p1; \
14646		__builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 9); \
	14667	__builtin_neon_vld2_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 9); \
14647	14668	__ret; \
14648	14669	})
14649	14670	#else
...	...	@@ -14651,12 +14672,12 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14651	14672	float32x2x2_t __ret; \
14652	14673	float32x2x2_t __s1 = __p1; \
14653	14674	float32x2x2_t __rev1; \
14654		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
14655		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
14656		__builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 9); \
	14675	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	14676	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	14677	__builtin_neon_vld2_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __p2, 9); \
14657	14678	\
14658		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
14659		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
	14679	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	14680	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
14660	14681	__ret; \
14661	14682	})
14662	14683	#endif
...	...	@@ -14665,7 +14686,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14665	14686	#define vld2_lane_s32(__p0, __p1, __p2) __extension__ ({ \
14666	14687	int32x2x2_t __ret; \
14667	14688	int32x2x2_t __s1 = __p1; \
14668		__builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 2); \
	14689	__builtin_neon_vld2_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 2); \
14669	14690	__ret; \
14670	14691	})
14671	14692	#else
...	...	@@ -14673,12 +14694,12 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14673	14694	int32x2x2_t __ret; \
14674	14695	int32x2x2_t __s1 = __p1; \
14675	14696	int32x2x2_t __rev1; \
14676		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
14677		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
14678		__builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 2); \
	14697	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	14698	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	14699	__builtin_neon_vld2_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __p2, 2); \
14679	14700	\
14680		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
14681		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
	14701	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	14702	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
14682	14703	__ret; \
14683	14704	})
14684	14705	#endif
...	...	@@ -14687,7 +14708,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14687	14708	#define vld2_lane_s16(__p0, __p1, __p2) __extension__ ({ \
14688	14709	int16x4x2_t __ret; \
14689	14710	int16x4x2_t __s1 = __p1; \
14690		__builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 1); \
	14711	__builtin_neon_vld2_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 1); \
14691	14712	__ret; \
14692	14713	})
14693	14714	#else
...	...	@@ -14695,12 +14716,12 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14695	14716	int16x4x2_t __ret; \
14696	14717	int16x4x2_t __s1 = __p1; \
14697	14718	int16x4x2_t __rev1; \
14698		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
14699		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
14700		__builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 1); \
	14719	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	14720	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	14721	__builtin_neon_vld2_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __p2, 1); \
14701	14722	\
14702		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
14703		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	14723	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	14724	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
14704	14725	__ret; \
14705	14726	})
14706	14727	#endif
...	...	@@ -14716,9 +14737,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14716	14737	poly8x8x3_t __ret; \
14717	14738	__builtin_neon_vld3_v(&__ret, __p0, 4); \
14718	14739	\
14719		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
14720		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
14721		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	14740	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	14741	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	14742	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
14722	14743	__ret; \
14723	14744	})
14724	14745	#endif
...	...	@@ -14734,9 +14755,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14734	14755	poly16x4x3_t __ret; \
14735	14756	__builtin_neon_vld3_v(&__ret, __p0, 5); \
14736	14757	\
14737		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
14738		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
14739		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	14758	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	14759	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	14760	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
14740	14761	__ret; \
14741	14762	})
14742	14763	#endif
...	...	@@ -14752,9 +14773,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14752	14773	poly8x16x3_t __ret; \
14753	14774	__builtin_neon_vld3q_v(&__ret, __p0, 36); \
14754	14775	\
14755		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
14756		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
14757		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	14776	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	14777	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	14778	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
14758	14779	__ret; \
14759	14780	})
14760	14781	#endif
...	...	@@ -14770,9 +14791,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14770	14791	poly16x8x3_t __ret; \
14771	14792	__builtin_neon_vld3q_v(&__ret, __p0, 37); \
14772	14793	\
14773		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
14774		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
14775		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	14794	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	14795	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	14796	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
14776	14797	__ret; \
14777	14798	})
14778	14799	#endif
...	...	@@ -14788,9 +14809,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14788	14809	uint8x16x3_t __ret; \
14789	14810	__builtin_neon_vld3q_v(&__ret, __p0, 48); \
14790	14811	\
14791		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
14792		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
14793		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	14812	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	14813	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	14814	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
14794	14815	__ret; \
14795	14816	})
14796	14817	#endif
...	...	@@ -14806,9 +14827,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14806	14827	uint32x4x3_t __ret; \
14807	14828	__builtin_neon_vld3q_v(&__ret, __p0, 50); \
14808	14829	\
14809		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
14810		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
14811		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	14830	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	14831	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
	14832	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_32); \
14812	14833	__ret; \
14813	14834	})
14814	14835	#endif
...	...	@@ -14824,9 +14845,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14824	14845	uint16x8x3_t __ret; \
14825	14846	__builtin_neon_vld3q_v(&__ret, __p0, 49); \
14826	14847	\
14827		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
14828		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
14829		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	14848	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	14849	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	14850	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
14830	14851	__ret; \
14831	14852	})
14832	14853	#endif
...	...	@@ -14842,9 +14863,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14842	14863	int8x16x3_t __ret; \
14843	14864	__builtin_neon_vld3q_v(&__ret, __p0, 32); \
14844	14865	\
14845		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
14846		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
14847		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	14866	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	14867	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	14868	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
14848	14869	__ret; \
14849	14870	})
14850	14871	#endif
...	...	@@ -14860,9 +14881,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14860	14881	float32x4x3_t __ret; \
14861	14882	__builtin_neon_vld3q_v(&__ret, __p0, 41); \
14862	14883	\
14863		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
14864		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
14865		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	14884	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	14885	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
	14886	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_32); \
14866	14887	__ret; \
14867	14888	})
14868	14889	#endif
...	...	@@ -14878,9 +14899,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14878	14899	int32x4x3_t __ret; \
14879	14900	__builtin_neon_vld3q_v(&__ret, __p0, 34); \
14880	14901	\
14881		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
14882		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
14883		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	14902	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	14903	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
	14904	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_32); \
14884	14905	__ret; \
14885	14906	})
14886	14907	#endif
...	...	@@ -14896,9 +14917,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14896	14917	int16x8x3_t __ret; \
14897	14918	__builtin_neon_vld3q_v(&__ret, __p0, 33); \
14898	14919	\
14899		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
14900		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
14901		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	14920	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	14921	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	14922	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
14902	14923	__ret; \
14903	14924	})
14904	14925	#endif
...	...	@@ -14914,9 +14935,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14914	14935	uint8x8x3_t __ret; \
14915	14936	__builtin_neon_vld3_v(&__ret, __p0, 16); \
14916	14937	\
14917		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
14918		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
14919		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	14938	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	14939	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	14940	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
14920	14941	__ret; \
14921	14942	})
14922	14943	#endif
...	...	@@ -14932,9 +14953,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14932	14953	uint32x2x3_t __ret; \
14933	14954	__builtin_neon_vld3_v(&__ret, __p0, 18); \
14934	14955	\
14935		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
14936		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
14937		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
	14956	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	14957	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
	14958	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_32); \
14938	14959	__ret; \
14939	14960	})
14940	14961	#endif
...	...	@@ -14955,9 +14976,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14955	14976	uint16x4x3_t __ret; \
14956	14977	__builtin_neon_vld3_v(&__ret, __p0, 17); \
14957	14978	\
14958		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
14959		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
14960		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	14979	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	14980	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	14981	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
14961	14982	__ret; \
14962	14983	})
14963	14984	#endif
...	...	@@ -14973,9 +14994,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14973	14994	int8x8x3_t __ret; \
14974	14995	__builtin_neon_vld3_v(&__ret, __p0, 0); \
14975	14996	\
14976		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
14977		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
14978		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	14997	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	14998	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	14999	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
14979	15000	__ret; \
14980	15001	})
14981	15002	#endif
...	...	@@ -14991,9 +15012,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
14991	15012	float32x2x3_t __ret; \
14992	15013	__builtin_neon_vld3_v(&__ret, __p0, 9); \
14993	15014	\
14994		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
14995		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
14996		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
	15015	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	15016	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
	15017	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_32); \
14997	15018	__ret; \
14998	15019	})
14999	15020	#endif
...	...	@@ -15009,9 +15030,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15009	15030	int32x2x3_t __ret; \
15010	15031	__builtin_neon_vld3_v(&__ret, __p0, 2); \
15011	15032	\
15012		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
15013		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
15014		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
	15033	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	15034	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
	15035	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_32); \
15015	15036	__ret; \
15016	15037	})
15017	15038	#endif
...	...	@@ -15032,9 +15053,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15032	15053	int16x4x3_t __ret; \
15033	15054	__builtin_neon_vld3_v(&__ret, __p0, 1); \
15034	15055	\
15035		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
15036		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
15037		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	15056	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	15057	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	15058	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
15038	15059	__ret; \
15039	15060	})
15040	15061	#endif
...	...	@@ -15050,9 +15071,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15050	15071	poly8x8x3_t __ret; \
15051	15072	__builtin_neon_vld3_dup_v(&__ret, __p0, 4); \
15052	15073	\
15053		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
15054		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
15055		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	15074	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	15075	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	15076	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
15056	15077	__ret; \
15057	15078	})
15058	15079	#endif
...	...	@@ -15068,9 +15089,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15068	15089	poly16x4x3_t __ret; \
15069	15090	__builtin_neon_vld3_dup_v(&__ret, __p0, 5); \
15070	15091	\
15071		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
15072		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
15073		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	15092	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	15093	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	15094	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
15074	15095	__ret; \
15075	15096	})
15076	15097	#endif
...	...	@@ -15086,9 +15107,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15086	15107	poly8x16x3_t __ret; \
15087	15108	__builtin_neon_vld3q_dup_v(&__ret, __p0, 36); \
15088	15109	\
15089		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
15090		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
15091		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	15110	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	15111	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	15112	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
15092	15113	__ret; \
15093	15114	})
15094	15115	#endif
...	...	@@ -15104,9 +15125,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15104	15125	poly16x8x3_t __ret; \
15105	15126	__builtin_neon_vld3q_dup_v(&__ret, __p0, 37); \
15106	15127	\
15107		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
15108		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
15109		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	15128	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	15129	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	15130	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
15110	15131	__ret; \
15111	15132	})
15112	15133	#endif
...	...	@@ -15122,9 +15143,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15122	15143	uint8x16x3_t __ret; \
15123	15144	__builtin_neon_vld3q_dup_v(&__ret, __p0, 48); \
15124	15145	\
15125		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
15126		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
15127		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	15146	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	15147	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	15148	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
15128	15149	__ret; \
15129	15150	})
15130	15151	#endif
...	...	@@ -15140,9 +15161,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15140	15161	uint32x4x3_t __ret; \
15141	15162	__builtin_neon_vld3q_dup_v(&__ret, __p0, 50); \
15142	15163	\
15143		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
15144		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
15145		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	15164	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	15165	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
	15166	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_32); \
15146	15167	__ret; \
15147	15168	})
15148	15169	#endif
...	...	@@ -15158,9 +15179,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15158	15179	uint64x2x3_t __ret; \
15159	15180	__builtin_neon_vld3q_dup_v(&__ret, __p0, 51); \
15160	15181	\
15161		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
15162		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
15163		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
	15182	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	15183	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	15184	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
15164	15185	__ret; \
15165	15186	})
15166	15187	#endif
...	...	@@ -15176,9 +15197,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15176	15197	uint16x8x3_t __ret; \
15177	15198	__builtin_neon_vld3q_dup_v(&__ret, __p0, 49); \
15178	15199	\
15179		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
15180		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
15181		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	15200	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	15201	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	15202	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
15182	15203	__ret; \
15183	15204	})
15184	15205	#endif
...	...	@@ -15194,9 +15215,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15194	15215	int8x16x3_t __ret; \
15195	15216	__builtin_neon_vld3q_dup_v(&__ret, __p0, 32); \
15196	15217	\
15197		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
15198		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
15199		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	15218	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	15219	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	15220	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
15200	15221	__ret; \
15201	15222	})
15202	15223	#endif
...	...	@@ -15212,9 +15233,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15212	15233	float32x4x3_t __ret; \
15213	15234	__builtin_neon_vld3q_dup_v(&__ret, __p0, 41); \
15214	15235	\
15215		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
15216		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
15217		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	15236	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	15237	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
	15238	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_32); \
15218	15239	__ret; \
15219	15240	})
15220	15241	#endif
...	...	@@ -15230,9 +15251,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15230	15251	int32x4x3_t __ret; \
15231	15252	__builtin_neon_vld3q_dup_v(&__ret, __p0, 34); \
15232	15253	\
15233		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
15234		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
15235		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	15254	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	15255	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
	15256	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_32); \
15236	15257	__ret; \
15237	15258	})
15238	15259	#endif
...	...	@@ -15248,9 +15269,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15248	15269	int64x2x3_t __ret; \
15249	15270	__builtin_neon_vld3q_dup_v(&__ret, __p0, 35); \
15250	15271	\
15251		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
15252		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
15253		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
	15272	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	15273	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	15274	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
15254	15275	__ret; \
15255	15276	})
15256	15277	#endif
...	...	@@ -15266,9 +15287,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15266	15287	int16x8x3_t __ret; \
15267	15288	__builtin_neon_vld3q_dup_v(&__ret, __p0, 33); \
15268	15289	\
15269		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
15270		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
15271		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	15290	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	15291	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	15292	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
15272	15293	__ret; \
15273	15294	})
15274	15295	#endif
...	...	@@ -15284,9 +15305,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15284	15305	uint8x8x3_t __ret; \
15285	15306	__builtin_neon_vld3_dup_v(&__ret, __p0, 16); \
15286	15307	\
15287		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
15288		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
15289		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	15308	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	15309	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	15310	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
15290	15311	__ret; \
15291	15312	})
15292	15313	#endif
...	...	@@ -15302,9 +15323,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15302	15323	uint32x2x3_t __ret; \
15303	15324	__builtin_neon_vld3_dup_v(&__ret, __p0, 18); \
15304	15325	\
15305		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
15306		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
15307		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
	15326	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	15327	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
	15328	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_32); \
15308	15329	__ret; \
15309	15330	})
15310	15331	#endif
...	...	@@ -15325,9 +15346,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15325	15346	uint16x4x3_t __ret; \
15326	15347	__builtin_neon_vld3_dup_v(&__ret, __p0, 17); \
15327	15348	\
15328		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
15329		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
15330		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	15349	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	15350	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	15351	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
15331	15352	__ret; \
15332	15353	})
15333	15354	#endif
...	...	@@ -15343,9 +15364,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15343	15364	int8x8x3_t __ret; \
15344	15365	__builtin_neon_vld3_dup_v(&__ret, __p0, 0); \
15345	15366	\
15346		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
15347		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
15348		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	15367	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	15368	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	15369	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
15349	15370	__ret; \
15350	15371	})
15351	15372	#endif
...	...	@@ -15361,9 +15382,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15361	15382	float32x2x3_t __ret; \
15362	15383	__builtin_neon_vld3_dup_v(&__ret, __p0, 9); \
15363	15384	\
15364		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
15365		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
15366		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
	15385	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	15386	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
	15387	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_32); \
15367	15388	__ret; \
15368	15389	})
15369	15390	#endif
...	...	@@ -15379,9 +15400,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15379	15400	int32x2x3_t __ret; \
15380	15401	__builtin_neon_vld3_dup_v(&__ret, __p0, 2); \
15381	15402	\
15382		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
15383		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
15384		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
	15403	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	15404	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
	15405	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_32); \
15385	15406	__ret; \
15386	15407	})
15387	15408	#endif
...	...	@@ -15402,9 +15423,9 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15402	15423	int16x4x3_t __ret; \
15403	15424	__builtin_neon_vld3_dup_v(&__ret, __p0, 1); \
15404	15425	\
15405		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
15406		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
15407		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	15426	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	15427	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	15428	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
15408	15429	__ret; \
15409	15430	})
15410	15431	#endif
...	...	@@ -15413,7 +15434,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15413	15434	#define vld3_lane_p8(__p0, __p1, __p2) __extension__ ({ \
15414	15435	poly8x8x3_t __ret; \
15415	15436	poly8x8x3_t __s1 = __p1; \
15416		__builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 4); \
	15437	__builtin_neon_vld3_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 4); \
15417	15438	__ret; \
15418	15439	})
15419	15440	#else
...	...	@@ -15421,14 +15442,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15421	15442	poly8x8x3_t __ret; \
15422	15443	poly8x8x3_t __s1 = __p1; \
15423	15444	poly8x8x3_t __rev1; \
15424		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
15425		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
15426		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
15427		__builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 4); \
	15445	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	15446	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	15447	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	15448	__builtin_neon_vld3_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __p2, 4); \
15428	15449	\
15429		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
15430		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
15431		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	15450	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	15451	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	15452	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
15432	15453	__ret; \
15433	15454	})
15434	15455	#endif
...	...	@@ -15437,7 +15458,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15437	15458	#define vld3_lane_p16(__p0, __p1, __p2) __extension__ ({ \
15438	15459	poly16x4x3_t __ret; \
15439	15460	poly16x4x3_t __s1 = __p1; \
15440		__builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 5); \
	15461	__builtin_neon_vld3_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 5); \
15441	15462	__ret; \
15442	15463	})
15443	15464	#else
...	...	@@ -15445,14 +15466,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15445	15466	poly16x4x3_t __ret; \
15446	15467	poly16x4x3_t __s1 = __p1; \
15447	15468	poly16x4x3_t __rev1; \
15448		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
15449		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
15450		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
15451		__builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 5); \
	15469	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	15470	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	15471	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	15472	__builtin_neon_vld3_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __p2, 5); \
15452	15473	\
15453		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
15454		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
15455		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	15474	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	15475	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	15476	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
15456	15477	__ret; \
15457	15478	})
15458	15479	#endif
...	...	@@ -15461,7 +15482,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15461	15482	#define vld3q_lane_p16(__p0, __p1, __p2) __extension__ ({ \
15462	15483	poly16x8x3_t __ret; \
15463	15484	poly16x8x3_t __s1 = __p1; \
15464		__builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 37); \
	15485	__builtin_neon_vld3q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 37); \
15465	15486	__ret; \
15466	15487	})
15467	15488	#else
...	...	@@ -15469,14 +15490,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15469	15490	poly16x8x3_t __ret; \
15470	15491	poly16x8x3_t __s1 = __p1; \
15471	15492	poly16x8x3_t __rev1; \
15472		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
15473		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
15474		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
15475		__builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 37); \
	15493	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	15494	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	15495	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	15496	__builtin_neon_vld3q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 37); \
15476	15497	\
15477		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
15478		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
15479		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	15498	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	15499	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	15500	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
15480	15501	__ret; \
15481	15502	})
15482	15503	#endif
...	...	@@ -15485,7 +15506,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15485	15506	#define vld3q_lane_u32(__p0, __p1, __p2) __extension__ ({ \
15486	15507	uint32x4x3_t __ret; \
15487	15508	uint32x4x3_t __s1 = __p1; \
15488		__builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 50); \
	15509	__builtin_neon_vld3q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 50); \
15489	15510	__ret; \
15490	15511	})
15491	15512	#else
...	...	@@ -15493,14 +15514,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15493	15514	uint32x4x3_t __ret; \
15494	15515	uint32x4x3_t __s1 = __p1; \
15495	15516	uint32x4x3_t __rev1; \
15496		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
15497		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
15498		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
15499		__builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 50); \
	15517	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	15518	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	15519	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_32); \
	15520	__builtin_neon_vld3q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 50); \
15500	15521	\
15501		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
15502		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
15503		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	15522	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	15523	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
	15524	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_32); \
15504	15525	__ret; \
15505	15526	})
15506	15527	#endif
...	...	@@ -15509,7 +15530,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15509	15530	#define vld3q_lane_u16(__p0, __p1, __p2) __extension__ ({ \
15510	15531	uint16x8x3_t __ret; \
15511	15532	uint16x8x3_t __s1 = __p1; \
15512		__builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 49); \
	15533	__builtin_neon_vld3q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 49); \
15513	15534	__ret; \
15514	15535	})
15515	15536	#else
...	...	@@ -15517,14 +15538,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15517	15538	uint16x8x3_t __ret; \
15518	15539	uint16x8x3_t __s1 = __p1; \
15519	15540	uint16x8x3_t __rev1; \
15520		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
15521		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
15522		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
15523		__builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 49); \
	15541	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	15542	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	15543	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	15544	__builtin_neon_vld3q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 49); \
15524	15545	\
15525		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
15526		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
15527		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	15546	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	15547	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	15548	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
15528	15549	__ret; \
15529	15550	})
15530	15551	#endif
...	...	@@ -15533,7 +15554,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15533	15554	#define vld3q_lane_f32(__p0, __p1, __p2) __extension__ ({ \
15534	15555	float32x4x3_t __ret; \
15535	15556	float32x4x3_t __s1 = __p1; \
15536		__builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 41); \
	15557	__builtin_neon_vld3q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 41); \
15537	15558	__ret; \
15538	15559	})
15539	15560	#else
...	...	@@ -15541,14 +15562,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15541	15562	float32x4x3_t __ret; \
15542	15563	float32x4x3_t __s1 = __p1; \
15543	15564	float32x4x3_t __rev1; \
15544		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
15545		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
15546		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
15547		__builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 41); \
	15565	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	15566	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	15567	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_32); \
	15568	__builtin_neon_vld3q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 41); \
15548	15569	\
15549		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
15550		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
15551		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	15570	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	15571	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
	15572	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_32); \
15552	15573	__ret; \
15553	15574	})
15554	15575	#endif
...	...	@@ -15557,7 +15578,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15557	15578	#define vld3q_lane_s32(__p0, __p1, __p2) __extension__ ({ \
15558	15579	int32x4x3_t __ret; \
15559	15580	int32x4x3_t __s1 = __p1; \
15560		__builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 34); \
	15581	__builtin_neon_vld3q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 34); \
15561	15582	__ret; \
15562	15583	})
15563	15584	#else
...	...	@@ -15565,14 +15586,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15565	15586	int32x4x3_t __ret; \
15566	15587	int32x4x3_t __s1 = __p1; \
15567	15588	int32x4x3_t __rev1; \
15568		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
15569		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
15570		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
15571		__builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 34); \
	15589	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	15590	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	15591	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_32); \
	15592	__builtin_neon_vld3q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 34); \
15572	15593	\
15573		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
15574		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
15575		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	15594	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	15595	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
	15596	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_32); \
15576	15597	__ret; \
15577	15598	})
15578	15599	#endif
...	...	@@ -15581,7 +15602,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15581	15602	#define vld3q_lane_s16(__p0, __p1, __p2) __extension__ ({ \
15582	15603	int16x8x3_t __ret; \
15583	15604	int16x8x3_t __s1 = __p1; \
15584		__builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 33); \
	15605	__builtin_neon_vld3q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 33); \
15585	15606	__ret; \
15586	15607	})
15587	15608	#else
...	...	@@ -15589,14 +15610,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15589	15610	int16x8x3_t __ret; \
15590	15611	int16x8x3_t __s1 = __p1; \
15591	15612	int16x8x3_t __rev1; \
15592		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
15593		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
15594		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
15595		__builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 33); \
	15613	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	15614	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	15615	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	15616	__builtin_neon_vld3q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 33); \
15596	15617	\
15597		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
15598		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
15599		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	15618	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	15619	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	15620	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
15600	15621	__ret; \
15601	15622	})
15602	15623	#endif
...	...	@@ -15605,7 +15626,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15605	15626	#define vld3_lane_u8(__p0, __p1, __p2) __extension__ ({ \
15606	15627	uint8x8x3_t __ret; \
15607	15628	uint8x8x3_t __s1 = __p1; \
15608		__builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 16); \
	15629	__builtin_neon_vld3_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 16); \
15609	15630	__ret; \
15610	15631	})
15611	15632	#else
...	...	@@ -15613,14 +15634,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15613	15634	uint8x8x3_t __ret; \
15614	15635	uint8x8x3_t __s1 = __p1; \
15615	15636	uint8x8x3_t __rev1; \
15616		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
15617		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
15618		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
15619		__builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 16); \
	15637	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	15638	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	15639	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	15640	__builtin_neon_vld3_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __p2, 16); \
15620	15641	\
15621		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
15622		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
15623		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	15642	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	15643	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	15644	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
15624	15645	__ret; \
15625	15646	})
15626	15647	#endif
...	...	@@ -15629,7 +15650,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15629	15650	#define vld3_lane_u32(__p0, __p1, __p2) __extension__ ({ \
15630	15651	uint32x2x3_t __ret; \
15631	15652	uint32x2x3_t __s1 = __p1; \
15632		__builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 18); \
	15653	__builtin_neon_vld3_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 18); \
15633	15654	__ret; \
15634	15655	})
15635	15656	#else
...	...	@@ -15637,14 +15658,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15637	15658	uint32x2x3_t __ret; \
15638	15659	uint32x2x3_t __s1 = __p1; \
15639	15660	uint32x2x3_t __rev1; \
15640		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
15641		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
15642		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
15643		__builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 18); \
	15661	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	15662	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	15663	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_32); \
	15664	__builtin_neon_vld3_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __p2, 18); \
15644	15665	\
15645		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
15646		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
15647		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
	15666	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	15667	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
	15668	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_32); \
15648	15669	__ret; \
15649	15670	})
15650	15671	#endif
...	...	@@ -15653,7 +15674,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15653	15674	#define vld3_lane_u16(__p0, __p1, __p2) __extension__ ({ \
15654	15675	uint16x4x3_t __ret; \
15655	15676	uint16x4x3_t __s1 = __p1; \
15656		__builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 17); \
	15677	__builtin_neon_vld3_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 17); \
15657	15678	__ret; \
15658	15679	})
15659	15680	#else
...	...	@@ -15661,14 +15682,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15661	15682	uint16x4x3_t __ret; \
15662	15683	uint16x4x3_t __s1 = __p1; \
15663	15684	uint16x4x3_t __rev1; \
15664		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
15665		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
15666		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
15667		__builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 17); \
	15685	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	15686	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	15687	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	15688	__builtin_neon_vld3_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __p2, 17); \
15668	15689	\
15669		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
15670		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
15671		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	15690	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	15691	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	15692	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
15672	15693	__ret; \
15673	15694	})
15674	15695	#endif
...	...	@@ -15677,7 +15698,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15677	15698	#define vld3_lane_s8(__p0, __p1, __p2) __extension__ ({ \
15678	15699	int8x8x3_t __ret; \
15679	15700	int8x8x3_t __s1 = __p1; \
15680		__builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 0); \
	15701	__builtin_neon_vld3_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 0); \
15681	15702	__ret; \
15682	15703	})
15683	15704	#else
...	...	@@ -15685,14 +15706,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15685	15706	int8x8x3_t __ret; \
15686	15707	int8x8x3_t __s1 = __p1; \
15687	15708	int8x8x3_t __rev1; \
15688		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
15689		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
15690		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
15691		__builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 0); \
	15709	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	15710	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	15711	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	15712	__builtin_neon_vld3_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __p2, 0); \
15692	15713	\
15693		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
15694		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
15695		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	15714	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	15715	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	15716	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
15696	15717	__ret; \
15697	15718	})
15698	15719	#endif
...	...	@@ -15701,7 +15722,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15701	15722	#define vld3_lane_f32(__p0, __p1, __p2) __extension__ ({ \
15702	15723	float32x2x3_t __ret; \
15703	15724	float32x2x3_t __s1 = __p1; \
15704		__builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 9); \
	15725	__builtin_neon_vld3_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 9); \
15705	15726	__ret; \
15706	15727	})
15707	15728	#else
...	...	@@ -15709,14 +15730,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15709	15730	float32x2x3_t __ret; \
15710	15731	float32x2x3_t __s1 = __p1; \
15711	15732	float32x2x3_t __rev1; \
15712		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
15713		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
15714		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
15715		__builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 9); \
	15733	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	15734	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	15735	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_32); \
	15736	__builtin_neon_vld3_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __p2, 9); \
15716	15737	\
15717		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
15718		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
15719		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
	15738	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	15739	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
	15740	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_32); \
15720	15741	__ret; \
15721	15742	})
15722	15743	#endif
...	...	@@ -15725,7 +15746,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15725	15746	#define vld3_lane_s32(__p0, __p1, __p2) __extension__ ({ \
15726	15747	int32x2x3_t __ret; \
15727	15748	int32x2x3_t __s1 = __p1; \
15728		__builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 2); \
	15749	__builtin_neon_vld3_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 2); \
15729	15750	__ret; \
15730	15751	})
15731	15752	#else
...	...	@@ -15733,14 +15754,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15733	15754	int32x2x3_t __ret; \
15734	15755	int32x2x3_t __s1 = __p1; \
15735	15756	int32x2x3_t __rev1; \
15736		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
15737		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
15738		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
15739		__builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 2); \
	15757	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	15758	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	15759	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_32); \
	15760	__builtin_neon_vld3_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __p2, 2); \
15740	15761	\
15741		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
15742		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
15743		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
	15762	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	15763	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
	15764	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_32); \
15744	15765	__ret; \
15745	15766	})
15746	15767	#endif
...	...	@@ -15749,7 +15770,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15749	15770	#define vld3_lane_s16(__p0, __p1, __p2) __extension__ ({ \
15750	15771	int16x4x3_t __ret; \
15751	15772	int16x4x3_t __s1 = __p1; \
15752		__builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 1); \
	15773	__builtin_neon_vld3_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 1); \
15753	15774	__ret; \
15754	15775	})
15755	15776	#else
...	...	@@ -15757,14 +15778,14 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15757	15778	int16x4x3_t __ret; \
15758	15779	int16x4x3_t __s1 = __p1; \
15759	15780	int16x4x3_t __rev1; \
15760		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
15761		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
15762		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
15763		__builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 1); \
	15781	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	15782	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	15783	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	15784	__builtin_neon_vld3_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __p2, 1); \
15764	15785	\
15765		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
15766		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
15767		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	15786	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	15787	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	15788	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
15768	15789	__ret; \
15769	15790	})
15770	15791	#endif
...	...	@@ -15780,10 +15801,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15780	15801	poly8x8x4_t __ret; \
15781	15802	__builtin_neon_vld4_v(&__ret, __p0, 4); \
15782	15803	\
15783		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
15784		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
15785		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
15786		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	15804	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	15805	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	15806	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
	15807	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_8); \
15787	15808	__ret; \
15788	15809	})
15789	15810	#endif
...	...	@@ -15799,10 +15820,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15799	15820	poly16x4x4_t __ret; \
15800	15821	__builtin_neon_vld4_v(&__ret, __p0, 5); \
15801	15822	\
15802		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
15803		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
15804		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
15805		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	15823	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	15824	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	15825	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
	15826	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_16); \
15806	15827	__ret; \
15807	15828	})
15808	15829	#endif
...	...	@@ -15818,10 +15839,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15818	15839	poly8x16x4_t __ret; \
15819	15840	__builtin_neon_vld4q_v(&__ret, __p0, 36); \
15820	15841	\
15821		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
15822		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
15823		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
15824		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	15842	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	15843	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	15844	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
	15845	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_8); \
15825	15846	__ret; \
15826	15847	})
15827	15848	#endif
...	...	@@ -15837,10 +15858,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15837	15858	poly16x8x4_t __ret; \
15838	15859	__builtin_neon_vld4q_v(&__ret, __p0, 37); \
15839	15860	\
15840		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
15841		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
15842		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
15843		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	15861	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	15862	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	15863	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
	15864	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_16); \
15844	15865	__ret; \
15845	15866	})
15846	15867	#endif
...	...	@@ -15856,10 +15877,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15856	15877	uint8x16x4_t __ret; \
15857	15878	__builtin_neon_vld4q_v(&__ret, __p0, 48); \
15858	15879	\
15859		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
15860		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
15861		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
15862		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	15880	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	15881	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	15882	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
	15883	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_8); \
15863	15884	__ret; \
15864	15885	})
15865	15886	#endif
...	...	@@ -15875,10 +15896,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15875	15896	uint32x4x4_t __ret; \
15876	15897	__builtin_neon_vld4q_v(&__ret, __p0, 50); \
15877	15898	\
15878		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
15879		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
15880		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
15881		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	15899	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	15900	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
	15901	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_32); \
	15902	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_32); \
15882	15903	__ret; \
15883	15904	})
15884	15905	#endif
...	...	@@ -15894,10 +15915,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15894	15915	uint16x8x4_t __ret; \
15895	15916	__builtin_neon_vld4q_v(&__ret, __p0, 49); \
15896	15917	\
15897		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
15898		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
15899		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
15900		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	15918	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	15919	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	15920	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
	15921	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_16); \
15901	15922	__ret; \
15902	15923	})
15903	15924	#endif
...	...	@@ -15913,10 +15934,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15913	15934	int8x16x4_t __ret; \
15914	15935	__builtin_neon_vld4q_v(&__ret, __p0, 32); \
15915	15936	\
15916		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
15917		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
15918		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
15919		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	15937	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	15938	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	15939	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
	15940	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_8); \
15920	15941	__ret; \
15921	15942	})
15922	15943	#endif
...	...	@@ -15932,10 +15953,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15932	15953	float32x4x4_t __ret; \
15933	15954	__builtin_neon_vld4q_v(&__ret, __p0, 41); \
15934	15955	\
15935		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
15936		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
15937		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
15938		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	15956	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	15957	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
	15958	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_32); \
	15959	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_32); \
15939	15960	__ret; \
15940	15961	})
15941	15962	#endif
...	...	@@ -15951,10 +15972,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15951	15972	int32x4x4_t __ret; \
15952	15973	__builtin_neon_vld4q_v(&__ret, __p0, 34); \
15953	15974	\
15954		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
15955		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
15956		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
15957		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	15975	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	15976	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
	15977	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_32); \
	15978	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_32); \
15958	15979	__ret; \
15959	15980	})
15960	15981	#endif
...	...	@@ -15970,10 +15991,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15970	15991	int16x8x4_t __ret; \
15971	15992	__builtin_neon_vld4q_v(&__ret, __p0, 33); \
15972	15993	\
15973		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
15974		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
15975		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
15976		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	15994	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	15995	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	15996	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
	15997	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_16); \
15977	15998	__ret; \
15978	15999	})
15979	16000	#endif
...	...	@@ -15989,10 +16010,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
15989	16010	uint8x8x4_t __ret; \
15990	16011	__builtin_neon_vld4_v(&__ret, __p0, 16); \
15991	16012	\
15992		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
15993		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
15994		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
15995		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	16013	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	16014	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	16015	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
	16016	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_8); \
15996	16017	__ret; \
15997	16018	})
15998	16019	#endif
...	...	@@ -16008,10 +16029,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16008	16029	uint32x2x4_t __ret; \
16009	16030	__builtin_neon_vld4_v(&__ret, __p0, 18); \
16010	16031	\
16011		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
16012		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
16013		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
16014		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
	16032	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	16033	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
	16034	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_32); \
	16035	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_32); \
16015	16036	__ret; \
16016	16037	})
16017	16038	#endif
...	...	@@ -16032,10 +16053,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16032	16053	uint16x4x4_t __ret; \
16033	16054	__builtin_neon_vld4_v(&__ret, __p0, 17); \
16034	16055	\
16035		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
16036		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
16037		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
16038		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	16056	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	16057	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	16058	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
	16059	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_16); \
16039	16060	__ret; \
16040	16061	})
16041	16062	#endif
...	...	@@ -16051,10 +16072,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16051	16072	int8x8x4_t __ret; \
16052	16073	__builtin_neon_vld4_v(&__ret, __p0, 0); \
16053	16074	\
16054		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
16055		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
16056		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
16057		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	16075	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	16076	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	16077	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
	16078	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_8); \
16058	16079	__ret; \
16059	16080	})
16060	16081	#endif
...	...	@@ -16070,10 +16091,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16070	16091	float32x2x4_t __ret; \
16071	16092	__builtin_neon_vld4_v(&__ret, __p0, 9); \
16072	16093	\
16073		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
16074		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
16075		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
16076		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
	16094	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	16095	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
	16096	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_32); \
	16097	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_32); \
16077	16098	__ret; \
16078	16099	})
16079	16100	#endif
...	...	@@ -16089,10 +16110,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16089	16110	int32x2x4_t __ret; \
16090	16111	__builtin_neon_vld4_v(&__ret, __p0, 2); \
16091	16112	\
16092		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
16093		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
16094		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
16095		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
	16113	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	16114	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
	16115	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_32); \
	16116	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_32); \
16096	16117	__ret; \
16097	16118	})
16098	16119	#endif
...	...	@@ -16113,10 +16134,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16113	16134	int16x4x4_t __ret; \
16114	16135	__builtin_neon_vld4_v(&__ret, __p0, 1); \
16115	16136	\
16116		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
16117		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
16118		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
16119		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	16137	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	16138	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	16139	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
	16140	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_16); \
16120	16141	__ret; \
16121	16142	})
16122	16143	#endif
...	...	@@ -16132,10 +16153,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16132	16153	poly8x8x4_t __ret; \
16133	16154	__builtin_neon_vld4_dup_v(&__ret, __p0, 4); \
16134	16155	\
16135		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
16136		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
16137		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
16138		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	16156	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	16157	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	16158	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
	16159	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_8); \
16139	16160	__ret; \
16140	16161	})
16141	16162	#endif
...	...	@@ -16151,10 +16172,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16151	16172	poly16x4x4_t __ret; \
16152	16173	__builtin_neon_vld4_dup_v(&__ret, __p0, 5); \
16153	16174	\
16154		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
16155		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
16156		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
16157		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	16175	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	16176	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	16177	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
	16178	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_16); \
16158	16179	__ret; \
16159	16180	})
16160	16181	#endif
...	...	@@ -16170,10 +16191,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16170	16191	poly8x16x4_t __ret; \
16171	16192	__builtin_neon_vld4q_dup_v(&__ret, __p0, 36); \
16172	16193	\
16173		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
16174		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
16175		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
16176		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	16194	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	16195	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	16196	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
	16197	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_8); \
16177	16198	__ret; \
16178	16199	})
16179	16200	#endif
...	...	@@ -16189,10 +16210,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16189	16210	poly16x8x4_t __ret; \
16190	16211	__builtin_neon_vld4q_dup_v(&__ret, __p0, 37); \
16191	16212	\
16192		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
16193		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
16194		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
16195		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	16213	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	16214	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	16215	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
	16216	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_16); \
16196	16217	__ret; \
16197	16218	})
16198	16219	#endif
...	...	@@ -16208,10 +16229,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16208	16229	uint8x16x4_t __ret; \
16209	16230	__builtin_neon_vld4q_dup_v(&__ret, __p0, 48); \
16210	16231	\
16211		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
16212		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
16213		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
16214		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	16232	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	16233	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	16234	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
	16235	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_8); \
16215	16236	__ret; \
16216	16237	})
16217	16238	#endif
...	...	@@ -16227,10 +16248,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16227	16248	uint32x4x4_t __ret; \
16228	16249	__builtin_neon_vld4q_dup_v(&__ret, __p0, 50); \
16229	16250	\
16230		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
16231		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
16232		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
16233		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	16251	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	16252	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
	16253	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_32); \
	16254	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_32); \
16234	16255	__ret; \
16235	16256	})
16236	16257	#endif
...	...	@@ -16246,10 +16267,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16246	16267	uint64x2x4_t __ret; \
16247	16268	__builtin_neon_vld4q_dup_v(&__ret, __p0, 51); \
16248	16269	\
16249		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
16250		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
16251		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
16252		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
	16270	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	16271	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	16272	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
	16273	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_64); \
16253	16274	__ret; \
16254	16275	})
16255	16276	#endif
...	...	@@ -16265,10 +16286,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16265	16286	uint16x8x4_t __ret; \
16266	16287	__builtin_neon_vld4q_dup_v(&__ret, __p0, 49); \
16267	16288	\
16268		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
16269		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
16270		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
16271		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	16289	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	16290	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	16291	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
	16292	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_16); \
16272	16293	__ret; \
16273	16294	})
16274	16295	#endif
...	...	@@ -16284,10 +16305,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16284	16305	int8x16x4_t __ret; \
16285	16306	__builtin_neon_vld4q_dup_v(&__ret, __p0, 32); \
16286	16307	\
16287		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
16288		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
16289		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
16290		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	16308	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	16309	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	16310	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
	16311	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_8); \
16291	16312	__ret; \
16292	16313	})
16293	16314	#endif
...	...	@@ -16303,10 +16324,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16303	16324	float32x4x4_t __ret; \
16304	16325	__builtin_neon_vld4q_dup_v(&__ret, __p0, 41); \
16305	16326	\
16306		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
16307		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
16308		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
16309		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	16327	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	16328	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
	16329	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_32); \
	16330	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_32); \
16310	16331	__ret; \
16311	16332	})
16312	16333	#endif
...	...	@@ -16322,10 +16343,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16322	16343	int32x4x4_t __ret; \
16323	16344	__builtin_neon_vld4q_dup_v(&__ret, __p0, 34); \
16324	16345	\
16325		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
16326		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
16327		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
16328		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	16346	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	16347	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
	16348	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_32); \
	16349	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_32); \
16329	16350	__ret; \
16330	16351	})
16331	16352	#endif
...	...	@@ -16341,10 +16362,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16341	16362	int64x2x4_t __ret; \
16342	16363	__builtin_neon_vld4q_dup_v(&__ret, __p0, 35); \
16343	16364	\
16344		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
16345		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
16346		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
16347		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
	16365	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	16366	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	16367	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
	16368	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_64); \
16348	16369	__ret; \
16349	16370	})
16350	16371	#endif
...	...	@@ -16360,10 +16381,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16360	16381	int16x8x4_t __ret; \
16361	16382	__builtin_neon_vld4q_dup_v(&__ret, __p0, 33); \
16362	16383	\
16363		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
16364		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
16365		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
16366		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	16384	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	16385	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	16386	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
	16387	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_16); \
16367	16388	__ret; \
16368	16389	})
16369	16390	#endif
...	...	@@ -16379,10 +16400,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16379	16400	uint8x8x4_t __ret; \
16380	16401	__builtin_neon_vld4_dup_v(&__ret, __p0, 16); \
16381	16402	\
16382		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
16383		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
16384		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
16385		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	16403	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	16404	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	16405	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
	16406	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_8); \
16386	16407	__ret; \
16387	16408	})
16388	16409	#endif
...	...	@@ -16398,10 +16419,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16398	16419	uint32x2x4_t __ret; \
16399	16420	__builtin_neon_vld4_dup_v(&__ret, __p0, 18); \
16400	16421	\
16401		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
16402		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
16403		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
16404		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
	16422	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	16423	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
	16424	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_32); \
	16425	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_32); \
16405	16426	__ret; \
16406	16427	})
16407	16428	#endif
...	...	@@ -16422,10 +16443,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16422	16443	uint16x4x4_t __ret; \
16423	16444	__builtin_neon_vld4_dup_v(&__ret, __p0, 17); \
16424	16445	\
16425		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
16426		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
16427		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
16428		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	16446	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	16447	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	16448	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
	16449	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_16); \
16429	16450	__ret; \
16430	16451	})
16431	16452	#endif
...	...	@@ -16441,10 +16462,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16441	16462	int8x8x4_t __ret; \
16442	16463	__builtin_neon_vld4_dup_v(&__ret, __p0, 0); \
16443	16464	\
16444		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
16445		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
16446		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
16447		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	16465	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	16466	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	16467	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
	16468	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_8); \
16448	16469	__ret; \
16449	16470	})
16450	16471	#endif
...	...	@@ -16460,10 +16481,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16460	16481	float32x2x4_t __ret; \
16461	16482	__builtin_neon_vld4_dup_v(&__ret, __p0, 9); \
16462	16483	\
16463		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
16464		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
16465		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
16466		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
	16484	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	16485	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
	16486	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_32); \
	16487	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_32); \
16467	16488	__ret; \
16468	16489	})
16469	16490	#endif
...	...	@@ -16479,10 +16500,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16479	16500	int32x2x4_t __ret; \
16480	16501	__builtin_neon_vld4_dup_v(&__ret, __p0, 2); \
16481	16502	\
16482		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
16483		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
16484		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
16485		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
	16503	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	16504	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
	16505	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_32); \
	16506	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_32); \
16486	16507	__ret; \
16487	16508	})
16488	16509	#endif
...	...	@@ -16503,10 +16524,10 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16503	16524	int16x4x4_t __ret; \
16504	16525	__builtin_neon_vld4_dup_v(&__ret, __p0, 1); \
16505	16526	\
16506		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
16507		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
16508		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
16509		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	16527	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	16528	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	16529	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
	16530	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_16); \
16510	16531	__ret; \
16511	16532	})
16512	16533	#endif
...	...	@@ -16515,7 +16536,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16515	16536	#define vld4_lane_p8(__p0, __p1, __p2) __extension__ ({ \
16516	16537	poly8x8x4_t __ret; \
16517	16538	poly8x8x4_t __s1 = __p1; \
16518		__builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 4); \
	16539	__builtin_neon_vld4_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 4); \
16519	16540	__ret; \
16520	16541	})
16521	16542	#else
...	...	@@ -16523,16 +16544,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16523	16544	poly8x8x4_t __ret; \
16524	16545	poly8x8x4_t __s1 = __p1; \
16525	16546	poly8x8x4_t __rev1; \
16526		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
16527		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
16528		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
16529		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
16530		__builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 4); \
	16547	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	16548	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	16549	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	16550	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_8); \
	16551	__builtin_neon_vld4_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), __p2, 4); \
16531	16552	\
16532		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
16533		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
16534		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
16535		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	16553	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	16554	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	16555	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
	16556	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_8); \
16536	16557	__ret; \
16537	16558	})
16538	16559	#endif
...	...	@@ -16541,7 +16562,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16541	16562	#define vld4_lane_p16(__p0, __p1, __p2) __extension__ ({ \
16542	16563	poly16x4x4_t __ret; \
16543	16564	poly16x4x4_t __s1 = __p1; \
16544		__builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 5); \
	16565	__builtin_neon_vld4_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 5); \
16545	16566	__ret; \
16546	16567	})
16547	16568	#else
...	...	@@ -16549,16 +16570,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16549	16570	poly16x4x4_t __ret; \
16550	16571	poly16x4x4_t __s1 = __p1; \
16551	16572	poly16x4x4_t __rev1; \
16552		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
16553		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
16554		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
16555		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
16556		__builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 5); \
	16573	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	16574	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	16575	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	16576	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_16); \
	16577	__builtin_neon_vld4_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), __p2, 5); \
16557	16578	\
16558		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
16559		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
16560		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
16561		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	16579	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	16580	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	16581	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
	16582	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_16); \
16562	16583	__ret; \
16563	16584	})
16564	16585	#endif
...	...	@@ -16567,7 +16588,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16567	16588	#define vld4q_lane_p16(__p0, __p1, __p2) __extension__ ({ \
16568	16589	poly16x8x4_t __ret; \
16569	16590	poly16x8x4_t __s1 = __p1; \
16570		__builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 37); \
	16591	__builtin_neon_vld4q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 37); \
16571	16592	__ret; \
16572	16593	})
16573	16594	#else
...	...	@@ -16575,16 +16596,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16575	16596	poly16x8x4_t __ret; \
16576	16597	poly16x8x4_t __s1 = __p1; \
16577	16598	poly16x8x4_t __rev1; \
16578		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
16579		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
16580		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
16581		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
16582		__builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 37); \
	16599	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	16600	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	16601	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	16602	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_16); \
	16603	__builtin_neon_vld4q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 37); \
16583	16604	\
16584		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
16585		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
16586		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
16587		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	16605	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	16606	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	16607	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
	16608	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_16); \
16588	16609	__ret; \
16589	16610	})
16590	16611	#endif
...	...	@@ -16593,7 +16614,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16593	16614	#define vld4q_lane_u32(__p0, __p1, __p2) __extension__ ({ \
16594	16615	uint32x4x4_t __ret; \
16595	16616	uint32x4x4_t __s1 = __p1; \
16596		__builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 50); \
	16617	__builtin_neon_vld4q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 50); \
16597	16618	__ret; \
16598	16619	})
16599	16620	#else
...	...	@@ -16601,16 +16622,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16601	16622	uint32x4x4_t __ret; \
16602	16623	uint32x4x4_t __s1 = __p1; \
16603	16624	uint32x4x4_t __rev1; \
16604		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
16605		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
16606		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
16607		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
16608		__builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 50); \
	16625	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	16626	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	16627	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_32); \
	16628	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_32); \
	16629	__builtin_neon_vld4q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 50); \
16609	16630	\
16610		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
16611		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
16612		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
16613		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	16631	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	16632	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
	16633	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_32); \
	16634	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_32); \
16614	16635	__ret; \
16615	16636	})
16616	16637	#endif
...	...	@@ -16619,7 +16640,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16619	16640	#define vld4q_lane_u16(__p0, __p1, __p2) __extension__ ({ \
16620	16641	uint16x8x4_t __ret; \
16621	16642	uint16x8x4_t __s1 = __p1; \
16622		__builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 49); \
	16643	__builtin_neon_vld4q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 49); \
16623	16644	__ret; \
16624	16645	})
16625	16646	#else
...	...	@@ -16627,16 +16648,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16627	16648	uint16x8x4_t __ret; \
16628	16649	uint16x8x4_t __s1 = __p1; \
16629	16650	uint16x8x4_t __rev1; \
16630		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
16631		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
16632		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
16633		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
16634		__builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 49); \
	16651	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	16652	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	16653	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	16654	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_16); \
	16655	__builtin_neon_vld4q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 49); \
16635	16656	\
16636		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
16637		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
16638		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
16639		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	16657	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	16658	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	16659	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
	16660	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_16); \
16640	16661	__ret; \
16641	16662	})
16642	16663	#endif
...	...	@@ -16645,7 +16666,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16645	16666	#define vld4q_lane_f32(__p0, __p1, __p2) __extension__ ({ \
16646	16667	float32x4x4_t __ret; \
16647	16668	float32x4x4_t __s1 = __p1; \
16648		__builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 41); \
	16669	__builtin_neon_vld4q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 41); \
16649	16670	__ret; \
16650	16671	})
16651	16672	#else
...	...	@@ -16653,16 +16674,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16653	16674	float32x4x4_t __ret; \
16654	16675	float32x4x4_t __s1 = __p1; \
16655	16676	float32x4x4_t __rev1; \
16656		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
16657		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
16658		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
16659		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
16660		__builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 41); \
	16677	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	16678	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	16679	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_32); \
	16680	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_32); \
	16681	__builtin_neon_vld4q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 41); \
16661	16682	\
16662		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
16663		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
16664		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
16665		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	16683	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	16684	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
	16685	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_32); \
	16686	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_32); \
16666	16687	__ret; \
16667	16688	})
16668	16689	#endif
...	...	@@ -16671,7 +16692,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16671	16692	#define vld4q_lane_s32(__p0, __p1, __p2) __extension__ ({ \
16672	16693	int32x4x4_t __ret; \
16673	16694	int32x4x4_t __s1 = __p1; \
16674		__builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 34); \
	16695	__builtin_neon_vld4q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 34); \
16675	16696	__ret; \
16676	16697	})
16677	16698	#else
...	...	@@ -16679,16 +16700,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16679	16700	int32x4x4_t __ret; \
16680	16701	int32x4x4_t __s1 = __p1; \
16681	16702	int32x4x4_t __rev1; \
16682		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
16683		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
16684		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
16685		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
16686		__builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 34); \
	16703	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	16704	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	16705	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_32); \
	16706	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_32); \
	16707	__builtin_neon_vld4q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 34); \
16687	16708	\
16688		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
16689		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
16690		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
16691		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	16709	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32); \
	16710	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32); \
	16711	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_32); \
	16712	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_32); \
16692	16713	__ret; \
16693	16714	})
16694	16715	#endif
...	...	@@ -16697,7 +16718,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16697	16718	#define vld4q_lane_s16(__p0, __p1, __p2) __extension__ ({ \
16698	16719	int16x8x4_t __ret; \
16699	16720	int16x8x4_t __s1 = __p1; \
16700		__builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 33); \
	16721	__builtin_neon_vld4q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 33); \
16701	16722	__ret; \
16702	16723	})
16703	16724	#else
...	...	@@ -16705,16 +16726,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16705	16726	int16x8x4_t __ret; \
16706	16727	int16x8x4_t __s1 = __p1; \
16707	16728	int16x8x4_t __rev1; \
16708		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
16709		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
16710		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
16711		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
16712		__builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 33); \
	16729	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	16730	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	16731	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	16732	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_16); \
	16733	__builtin_neon_vld4q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 33); \
16713	16734	\
16714		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
16715		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
16716		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
16717		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	16735	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	16736	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	16737	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
	16738	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_16); \
16718	16739	__ret; \
16719	16740	})
16720	16741	#endif
...	...	@@ -16723,7 +16744,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16723	16744	#define vld4_lane_u8(__p0, __p1, __p2) __extension__ ({ \
16724	16745	uint8x8x4_t __ret; \
16725	16746	uint8x8x4_t __s1 = __p1; \
16726		__builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 16); \
	16747	__builtin_neon_vld4_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 16); \
16727	16748	__ret; \
16728	16749	})
16729	16750	#else
...	...	@@ -16731,16 +16752,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16731	16752	uint8x8x4_t __ret; \
16732	16753	uint8x8x4_t __s1 = __p1; \
16733	16754	uint8x8x4_t __rev1; \
16734		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
16735		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
16736		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
16737		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
16738		__builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 16); \
	16755	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	16756	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	16757	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	16758	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_8); \
	16759	__builtin_neon_vld4_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), __p2, 16); \
16739	16760	\
16740		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
16741		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
16742		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
16743		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	16761	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	16762	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	16763	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
	16764	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_8); \
16744	16765	__ret; \
16745	16766	})
16746	16767	#endif
...	...	@@ -16749,7 +16770,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16749	16770	#define vld4_lane_u32(__p0, __p1, __p2) __extension__ ({ \
16750	16771	uint32x2x4_t __ret; \
16751	16772	uint32x2x4_t __s1 = __p1; \
16752		__builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 18); \
	16773	__builtin_neon_vld4_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 18); \
16753	16774	__ret; \
16754	16775	})
16755	16776	#else
...	...	@@ -16757,16 +16778,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16757	16778	uint32x2x4_t __ret; \
16758	16779	uint32x2x4_t __s1 = __p1; \
16759	16780	uint32x2x4_t __rev1; \
16760		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
16761		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
16762		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
16763		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
16764		__builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 18); \
	16781	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	16782	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	16783	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_32); \
	16784	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_32); \
	16785	__builtin_neon_vld4_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), __p2, 18); \
16765	16786	\
16766		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
16767		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
16768		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
16769		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
	16787	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	16788	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
	16789	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_32); \
	16790	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_32); \
16770	16791	__ret; \
16771	16792	})
16772	16793	#endif
...	...	@@ -16775,7 +16796,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16775	16796	#define vld4_lane_u16(__p0, __p1, __p2) __extension__ ({ \
16776	16797	uint16x4x4_t __ret; \
16777	16798	uint16x4x4_t __s1 = __p1; \
16778		__builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 17); \
	16799	__builtin_neon_vld4_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 17); \
16779	16800	__ret; \
16780	16801	})
16781	16802	#else
...	...	@@ -16783,16 +16804,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16783	16804	uint16x4x4_t __ret; \
16784	16805	uint16x4x4_t __s1 = __p1; \
16785	16806	uint16x4x4_t __rev1; \
16786		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
16787		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
16788		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
16789		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
16790		__builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 17); \
	16807	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	16808	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	16809	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	16810	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_16); \
	16811	__builtin_neon_vld4_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), __p2, 17); \
16791	16812	\
16792		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
16793		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
16794		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
16795		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	16813	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	16814	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	16815	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
	16816	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_16); \
16796	16817	__ret; \
16797	16818	})
16798	16819	#endif
...	...	@@ -16801,7 +16822,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16801	16822	#define vld4_lane_s8(__p0, __p1, __p2) __extension__ ({ \
16802	16823	int8x8x4_t __ret; \
16803	16824	int8x8x4_t __s1 = __p1; \
16804		__builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 0); \
	16825	__builtin_neon_vld4_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 0); \
16805	16826	__ret; \
16806	16827	})
16807	16828	#else
...	...	@@ -16809,16 +16830,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16809	16830	int8x8x4_t __ret; \
16810	16831	int8x8x4_t __s1 = __p1; \
16811	16832	int8x8x4_t __rev1; \
16812		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
16813		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
16814		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
16815		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
16816		__builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 0); \
	16833	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	16834	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	16835	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	16836	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_8); \
	16837	__builtin_neon_vld4_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), __p2, 0); \
16817	16838	\
16818		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
16819		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
16820		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
16821		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	16839	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	16840	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	16841	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
	16842	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_8); \
16822	16843	__ret; \
16823	16844	})
16824	16845	#endif
...	...	@@ -16827,7 +16848,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16827	16848	#define vld4_lane_f32(__p0, __p1, __p2) __extension__ ({ \
16828	16849	float32x2x4_t __ret; \
16829	16850	float32x2x4_t __s1 = __p1; \
16830		__builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 9); \
	16851	__builtin_neon_vld4_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 9); \
16831	16852	__ret; \
16832	16853	})
16833	16854	#else
...	...	@@ -16835,16 +16856,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16835	16856	float32x2x4_t __ret; \
16836	16857	float32x2x4_t __s1 = __p1; \
16837	16858	float32x2x4_t __rev1; \
16838		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
16839		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
16840		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
16841		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
16842		__builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 9); \
	16859	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	16860	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	16861	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_32); \
	16862	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_32); \
	16863	__builtin_neon_vld4_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), __p2, 9); \
16843	16864	\
16844		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
16845		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
16846		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
16847		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
	16865	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	16866	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
	16867	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_32); \
	16868	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_32); \
16848	16869	__ret; \
16849	16870	})
16850	16871	#endif
...	...	@@ -16853,7 +16874,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16853	16874	#define vld4_lane_s32(__p0, __p1, __p2) __extension__ ({ \
16854	16875	int32x2x4_t __ret; \
16855	16876	int32x2x4_t __s1 = __p1; \
16856		__builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 2); \
	16877	__builtin_neon_vld4_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 2); \
16857	16878	__ret; \
16858	16879	})
16859	16880	#else
...	...	@@ -16861,16 +16882,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16861	16882	int32x2x4_t __ret; \
16862	16883	int32x2x4_t __s1 = __p1; \
16863	16884	int32x2x4_t __rev1; \
16864		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
16865		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
16866		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
16867		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
16868		__builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 2); \
	16885	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	16886	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	16887	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_32); \
	16888	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_32); \
	16889	__builtin_neon_vld4_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), __p2, 2); \
16869	16890	\
16870		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
16871		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
16872		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
16873		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
	16891	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32); \
	16892	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32); \
	16893	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_32); \
	16894	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_32); \
16874	16895	__ret; \
16875	16896	})
16876	16897	#endif
...	...	@@ -16879,7 +16900,7 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16879	16900	#define vld4_lane_s16(__p0, __p1, __p2) __extension__ ({ \
16880	16901	int16x4x4_t __ret; \
16881	16902	int16x4x4_t __s1 = __p1; \
16882		__builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 1); \
	16903	__builtin_neon_vld4_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 1); \
16883	16904	__ret; \
16884	16905	})
16885	16906	#else
...	...	@@ -16887,16 +16908,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16887	16908	int16x4x4_t __ret; \
16888	16909	int16x4x4_t __s1 = __p1; \
16889	16910	int16x4x4_t __rev1; \
16890		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
16891		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
16892		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
16893		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
16894		__builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 1); \
	16911	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	16912	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	16913	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	16914	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_16); \
	16915	__builtin_neon_vld4_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), __p2, 1); \
16895	16916	\
16896		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
16897		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
16898		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
16899		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	16917	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	16918	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	16919	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
	16920	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_16); \
16900	16921	__ret; \
16901	16922	})
16902	16923	#endif
...	...	@@ -16904,16 +16925,16 @@ __ai __attribute__((target("neon"))) int16x4_t vhsub_s16(int16x4_t __p0, int16x4
16904	16925	#ifdef __LITTLE_ENDIAN__
16905	16926	__ai __attribute__((target("neon"))) uint8x16_t vmaxq_u8(uint8x16_t __p0, uint8x16_t __p1) {
16906	16927	uint8x16_t __ret;
16907		__ret = (uint8x16_t) __builtin_neon_vmaxq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
	16928	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vmaxq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 48));
16908	16929	return __ret;
16909	16930	}
16910	16931	#else
16911	16932	__ai __attribute__((target("neon"))) uint8x16_t vmaxq_u8(uint8x16_t __p0, uint8x16_t __p1) {
16912	16933	uint8x16_t __ret;
16913		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
16914		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
16915		__ret = (uint8x16_t) __builtin_neon_vmaxq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
16916		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	16934	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	16935	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	16936	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vmaxq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 48));
	16937	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
16917	16938	return __ret;
16918	16939	}
16919	16940	#endif
...	...	@@ -16921,16 +16942,16 @@ __ai __attribute__((target("neon"))) uint8x16_t vmaxq_u8(uint8x16_t __p0, uint8x
16921	16942	#ifdef __LITTLE_ENDIAN__
16922	16943	__ai __attribute__((target("neon"))) uint32x4_t vmaxq_u32(uint32x4_t __p0, uint32x4_t __p1) {
16923	16944	uint32x4_t __ret;
16924		__ret = (uint32x4_t) __builtin_neon_vmaxq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
	16945	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vmaxq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50));
16925	16946	return __ret;
16926	16947	}
16927	16948	#else
16928	16949	__ai __attribute__((target("neon"))) uint32x4_t vmaxq_u32(uint32x4_t __p0, uint32x4_t __p1) {
16929	16950	uint32x4_t __ret;
16930		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
16931		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
16932		__ret = (uint32x4_t) __builtin_neon_vmaxq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
16933		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	16951	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	16952	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	16953	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vmaxq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 50));
	16954	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
16934	16955	return __ret;
16935	16956	}
16936	16957	#endif
...	...	@@ -16938,16 +16959,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vmaxq_u32(uint32x4_t __p0, uint3
16938	16959	#ifdef __LITTLE_ENDIAN__
16939	16960	__ai __attribute__((target("neon"))) uint16x8_t vmaxq_u16(uint16x8_t __p0, uint16x8_t __p1) {
16940	16961	uint16x8_t __ret;
16941		__ret = (uint16x8_t) __builtin_neon_vmaxq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
	16962	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vmaxq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 49));
16942	16963	return __ret;
16943	16964	}
16944	16965	#else
16945	16966	__ai __attribute__((target("neon"))) uint16x8_t vmaxq_u16(uint16x8_t __p0, uint16x8_t __p1) {
16946	16967	uint16x8_t __ret;
16947		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
16948		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
16949		__ret = (uint16x8_t) __builtin_neon_vmaxq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
16950		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	16968	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	16969	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	16970	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vmaxq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 49));
	16971	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
16951	16972	return __ret;
16952	16973	}
16953	16974	#endif
...	...	@@ -16955,16 +16976,16 @@ __ai __attribute__((target("neon"))) uint16x8_t vmaxq_u16(uint16x8_t __p0, uint1
16955	16976	#ifdef __LITTLE_ENDIAN__
16956	16977	__ai __attribute__((target("neon"))) int8x16_t vmaxq_s8(int8x16_t __p0, int8x16_t __p1) {
16957	16978	int8x16_t __ret;
16958		__ret = (int8x16_t) __builtin_neon_vmaxq_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
	16979	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vmaxq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 32));
16959	16980	return __ret;
16960	16981	}
16961	16982	#else
16962	16983	__ai __attribute__((target("neon"))) int8x16_t vmaxq_s8(int8x16_t __p0, int8x16_t __p1) {
16963	16984	int8x16_t __ret;
16964		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
16965		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
16966		__ret = (int8x16_t) __builtin_neon_vmaxq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 32);
16967		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	16985	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	16986	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	16987	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vmaxq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 32));
	16988	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
16968	16989	return __ret;
16969	16990	}
16970	16991	#endif
...	...	@@ -16972,16 +16993,16 @@ __ai __attribute__((target("neon"))) int8x16_t vmaxq_s8(int8x16_t __p0, int8x16_
16972	16993	#ifdef __LITTLE_ENDIAN__
16973	16994	__ai __attribute__((target("neon"))) float32x4_t vmaxq_f32(float32x4_t __p0, float32x4_t __p1) {
16974	16995	float32x4_t __ret;
16975		__ret = (float32x4_t) __builtin_neon_vmaxq_v((int8x16_t)__p0, (int8x16_t)__p1, 41);
	16996	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmaxq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 41));
16976	16997	return __ret;
16977	16998	}
16978	16999	#else
16979	17000	__ai __attribute__((target("neon"))) float32x4_t vmaxq_f32(float32x4_t __p0, float32x4_t __p1) {
16980	17001	float32x4_t __ret;
16981		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
16982		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
16983		__ret = (float32x4_t) __builtin_neon_vmaxq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 41);
16984		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	17002	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	17003	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	17004	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmaxq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 41));
	17005	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
16985	17006	return __ret;
16986	17007	}
16987	17008	#endif
...	...	@@ -16989,16 +17010,16 @@ __ai __attribute__((target("neon"))) float32x4_t vmaxq_f32(float32x4_t __p0, flo
16989	17010	#ifdef __LITTLE_ENDIAN__
16990	17011	__ai __attribute__((target("neon"))) int32x4_t vmaxq_s32(int32x4_t __p0, int32x4_t __p1) {
16991	17012	int32x4_t __ret;
16992		__ret = (int32x4_t) __builtin_neon_vmaxq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
	17013	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vmaxq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 34));
16993	17014	return __ret;
16994	17015	}
16995	17016	#else
16996	17017	__ai __attribute__((target("neon"))) int32x4_t vmaxq_s32(int32x4_t __p0, int32x4_t __p1) {
16997	17018	int32x4_t __ret;
16998		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
16999		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
17000		__ret = (int32x4_t) __builtin_neon_vmaxq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
17001		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	17019	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	17020	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	17021	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vmaxq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 34));
	17022	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
17002	17023	return __ret;
17003	17024	}
17004	17025	#endif
...	...	@@ -17006,16 +17027,16 @@ __ai __attribute__((target("neon"))) int32x4_t vmaxq_s32(int32x4_t __p0, int32x4
17006	17027	#ifdef __LITTLE_ENDIAN__
17007	17028	__ai __attribute__((target("neon"))) int16x8_t vmaxq_s16(int16x8_t __p0, int16x8_t __p1) {
17008	17029	int16x8_t __ret;
17009		__ret = (int16x8_t) __builtin_neon_vmaxq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
	17030	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vmaxq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 33));
17010	17031	return __ret;
17011	17032	}
17012	17033	#else
17013	17034	__ai __attribute__((target("neon"))) int16x8_t vmaxq_s16(int16x8_t __p0, int16x8_t __p1) {
17014	17035	int16x8_t __ret;
17015		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
17016		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
17017		__ret = (int16x8_t) __builtin_neon_vmaxq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
17018		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	17036	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	17037	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	17038	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vmaxq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 33));
	17039	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
17019	17040	return __ret;
17020	17041	}
17021	17042	#endif
...	...	@@ -17023,16 +17044,16 @@ __ai __attribute__((target("neon"))) int16x8_t vmaxq_s16(int16x8_t __p0, int16x8
17023	17044	#ifdef __LITTLE_ENDIAN__
17024	17045	__ai __attribute__((target("neon"))) uint8x8_t vmax_u8(uint8x8_t __p0, uint8x8_t __p1) {
17025	17046	uint8x8_t __ret;
17026		__ret = (uint8x8_t) __builtin_neon_vmax_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
	17047	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vmax_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 16));
17027	17048	return __ret;
17028	17049	}
17029	17050	#else
17030	17051	__ai __attribute__((target("neon"))) uint8x8_t vmax_u8(uint8x8_t __p0, uint8x8_t __p1) {
17031	17052	uint8x8_t __ret;
17032		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
17033		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
17034		__ret = (uint8x8_t) __builtin_neon_vmax_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
17035		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	17053	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	17054	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	17055	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vmax_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 16));
	17056	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
17036	17057	return __ret;
17037	17058	}
17038	17059	#endif
...	...	@@ -17040,16 +17061,16 @@ __ai __attribute__((target("neon"))) uint8x8_t vmax_u8(uint8x8_t __p0, uint8x8_t
17040	17061	#ifdef __LITTLE_ENDIAN__
17041	17062	__ai __attribute__((target("neon"))) uint32x2_t vmax_u32(uint32x2_t __p0, uint32x2_t __p1) {
17042	17063	uint32x2_t __ret;
17043		__ret = (uint32x2_t) __builtin_neon_vmax_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
	17064	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vmax_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 18));
17044	17065	return __ret;
17045	17066	}
17046	17067	#else
17047	17068	__ai __attribute__((target("neon"))) uint32x2_t vmax_u32(uint32x2_t __p0, uint32x2_t __p1) {
17048	17069	uint32x2_t __ret;
17049		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
17050		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
17051		__ret = (uint32x2_t) __builtin_neon_vmax_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
17052		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	17070	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	17071	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	17072	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vmax_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 18));
	17073	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
17053	17074	return __ret;
17054	17075	}
17055	17076	#endif
...	...	@@ -17057,16 +17078,16 @@ __ai __attribute__((target("neon"))) uint32x2_t vmax_u32(uint32x2_t __p0, uint32
17057	17078	#ifdef __LITTLE_ENDIAN__
17058	17079	__ai __attribute__((target("neon"))) uint16x4_t vmax_u16(uint16x4_t __p0, uint16x4_t __p1) {
17059	17080	uint16x4_t __ret;
17060		__ret = (uint16x4_t) __builtin_neon_vmax_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
	17081	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vmax_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 17));
17061	17082	return __ret;
17062	17083	}
17063	17084	#else
17064	17085	__ai __attribute__((target("neon"))) uint16x4_t vmax_u16(uint16x4_t __p0, uint16x4_t __p1) {
17065	17086	uint16x4_t __ret;
17066		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17067		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
17068		__ret = (uint16x4_t) __builtin_neon_vmax_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
17069		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	17087	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	17088	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	17089	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vmax_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 17));
	17090	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
17070	17091	return __ret;
17071	17092	}
17072	17093	#endif
...	...	@@ -17074,16 +17095,16 @@ __ai __attribute__((target("neon"))) uint16x4_t vmax_u16(uint16x4_t __p0, uint16
17074	17095	#ifdef __LITTLE_ENDIAN__
17075	17096	__ai __attribute__((target("neon"))) int8x8_t vmax_s8(int8x8_t __p0, int8x8_t __p1) {
17076	17097	int8x8_t __ret;
17077		__ret = (int8x8_t) __builtin_neon_vmax_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
	17098	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vmax_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 0));
17078	17099	return __ret;
17079	17100	}
17080	17101	#else
17081	17102	__ai __attribute__((target("neon"))) int8x8_t vmax_s8(int8x8_t __p0, int8x8_t __p1) {
17082	17103	int8x8_t __ret;
17083		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
17084		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
17085		__ret = (int8x8_t) __builtin_neon_vmax_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
17086		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	17104	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	17105	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	17106	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vmax_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 0));
	17107	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
17087	17108	return __ret;
17088	17109	}
17089	17110	#endif
...	...	@@ -17091,16 +17112,16 @@ __ai __attribute__((target("neon"))) int8x8_t vmax_s8(int8x8_t __p0, int8x8_t __
17091	17112	#ifdef __LITTLE_ENDIAN__
17092	17113	__ai __attribute__((target("neon"))) float32x2_t vmax_f32(float32x2_t __p0, float32x2_t __p1) {
17093	17114	float32x2_t __ret;
17094		__ret = (float32x2_t) __builtin_neon_vmax_v((int8x8_t)__p0, (int8x8_t)__p1, 9);
	17115	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vmax_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 9));
17095	17116	return __ret;
17096	17117	}
17097	17118	#else
17098	17119	__ai __attribute__((target("neon"))) float32x2_t vmax_f32(float32x2_t __p0, float32x2_t __p1) {
17099	17120	float32x2_t __ret;
17100		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
17101		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
17102		__ret = (float32x2_t) __builtin_neon_vmax_v((int8x8_t)__rev0, (int8x8_t)__rev1, 9);
17103		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	17121	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	17122	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	17123	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vmax_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 9));
	17124	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
17104	17125	return __ret;
17105	17126	}
17106	17127	#endif
...	...	@@ -17108,16 +17129,16 @@ __ai __attribute__((target("neon"))) float32x2_t vmax_f32(float32x2_t __p0, floa
17108	17129	#ifdef __LITTLE_ENDIAN__
17109	17130	__ai __attribute__((target("neon"))) int32x2_t vmax_s32(int32x2_t __p0, int32x2_t __p1) {
17110	17131	int32x2_t __ret;
17111		__ret = (int32x2_t) __builtin_neon_vmax_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
	17132	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vmax_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 2));
17112	17133	return __ret;
17113	17134	}
17114	17135	#else
17115	17136	__ai __attribute__((target("neon"))) int32x2_t vmax_s32(int32x2_t __p0, int32x2_t __p1) {
17116	17137	int32x2_t __ret;
17117		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
17118		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
17119		__ret = (int32x2_t) __builtin_neon_vmax_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
17120		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	17138	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	17139	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	17140	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vmax_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 2));
	17141	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
17121	17142	return __ret;
17122	17143	}
17123	17144	#endif
...	...	@@ -17125,16 +17146,16 @@ __ai __attribute__((target("neon"))) int32x2_t vmax_s32(int32x2_t __p0, int32x2_
17125	17146	#ifdef __LITTLE_ENDIAN__
17126	17147	__ai __attribute__((target("neon"))) int16x4_t vmax_s16(int16x4_t __p0, int16x4_t __p1) {
17127	17148	int16x4_t __ret;
17128		__ret = (int16x4_t) __builtin_neon_vmax_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
	17149	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vmax_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 1));
17129	17150	return __ret;
17130	17151	}
17131	17152	#else
17132	17153	__ai __attribute__((target("neon"))) int16x4_t vmax_s16(int16x4_t __p0, int16x4_t __p1) {
17133	17154	int16x4_t __ret;
17134		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17135		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
17136		__ret = (int16x4_t) __builtin_neon_vmax_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
17137		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	17155	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	17156	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	17157	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vmax_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 1));
	17158	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
17138	17159	return __ret;
17139	17160	}
17140	17161	#endif
...	...	@@ -17142,16 +17163,16 @@ __ai __attribute__((target("neon"))) int16x4_t vmax_s16(int16x4_t __p0, int16x4_
17142	17163	#ifdef __LITTLE_ENDIAN__
17143	17164	__ai __attribute__((target("neon"))) uint8x16_t vminq_u8(uint8x16_t __p0, uint8x16_t __p1) {
17144	17165	uint8x16_t __ret;
17145		__ret = (uint8x16_t) __builtin_neon_vminq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
	17166	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vminq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 48));
17146	17167	return __ret;
17147	17168	}
17148	17169	#else
17149	17170	__ai __attribute__((target("neon"))) uint8x16_t vminq_u8(uint8x16_t __p0, uint8x16_t __p1) {
17150	17171	uint8x16_t __ret;
17151		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
17152		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
17153		__ret = (uint8x16_t) __builtin_neon_vminq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
17154		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	17172	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	17173	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	17174	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vminq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 48));
	17175	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
17155	17176	return __ret;
17156	17177	}
17157	17178	#endif
...	...	@@ -17159,16 +17180,16 @@ __ai __attribute__((target("neon"))) uint8x16_t vminq_u8(uint8x16_t __p0, uint8x
17159	17180	#ifdef __LITTLE_ENDIAN__
17160	17181	__ai __attribute__((target("neon"))) uint32x4_t vminq_u32(uint32x4_t __p0, uint32x4_t __p1) {
17161	17182	uint32x4_t __ret;
17162		__ret = (uint32x4_t) __builtin_neon_vminq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
	17183	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vminq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50));
17163	17184	return __ret;
17164	17185	}
17165	17186	#else
17166	17187	__ai __attribute__((target("neon"))) uint32x4_t vminq_u32(uint32x4_t __p0, uint32x4_t __p1) {
17167	17188	uint32x4_t __ret;
17168		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17169		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
17170		__ret = (uint32x4_t) __builtin_neon_vminq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
17171		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	17189	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	17190	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	17191	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vminq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 50));
	17192	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
17172	17193	return __ret;
17173	17194	}
17174	17195	#endif
...	...	@@ -17176,16 +17197,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vminq_u32(uint32x4_t __p0, uint3
17176	17197	#ifdef __LITTLE_ENDIAN__
17177	17198	__ai __attribute__((target("neon"))) uint16x8_t vminq_u16(uint16x8_t __p0, uint16x8_t __p1) {
17178	17199	uint16x8_t __ret;
17179		__ret = (uint16x8_t) __builtin_neon_vminq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
	17200	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vminq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 49));
17180	17201	return __ret;
17181	17202	}
17182	17203	#else
17183	17204	__ai __attribute__((target("neon"))) uint16x8_t vminq_u16(uint16x8_t __p0, uint16x8_t __p1) {
17184	17205	uint16x8_t __ret;
17185		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
17186		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
17187		__ret = (uint16x8_t) __builtin_neon_vminq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
17188		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	17206	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	17207	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	17208	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vminq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 49));
	17209	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
17189	17210	return __ret;
17190	17211	}
17191	17212	#endif
...	...	@@ -17193,16 +17214,16 @@ __ai __attribute__((target("neon"))) uint16x8_t vminq_u16(uint16x8_t __p0, uint1
17193	17214	#ifdef __LITTLE_ENDIAN__
17194	17215	__ai __attribute__((target("neon"))) int8x16_t vminq_s8(int8x16_t __p0, int8x16_t __p1) {
17195	17216	int8x16_t __ret;
17196		__ret = (int8x16_t) __builtin_neon_vminq_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
	17217	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vminq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 32));
17197	17218	return __ret;
17198	17219	}
17199	17220	#else
17200	17221	__ai __attribute__((target("neon"))) int8x16_t vminq_s8(int8x16_t __p0, int8x16_t __p1) {
17201	17222	int8x16_t __ret;
17202		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
17203		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
17204		__ret = (int8x16_t) __builtin_neon_vminq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 32);
17205		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	17223	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	17224	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	17225	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vminq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 32));
	17226	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
17206	17227	return __ret;
17207	17228	}
17208	17229	#endif
...	...	@@ -17210,16 +17231,16 @@ __ai __attribute__((target("neon"))) int8x16_t vminq_s8(int8x16_t __p0, int8x16_
17210	17231	#ifdef __LITTLE_ENDIAN__
17211	17232	__ai __attribute__((target("neon"))) float32x4_t vminq_f32(float32x4_t __p0, float32x4_t __p1) {
17212	17233	float32x4_t __ret;
17213		__ret = (float32x4_t) __builtin_neon_vminq_v((int8x16_t)__p0, (int8x16_t)__p1, 41);
	17234	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vminq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 41));
17214	17235	return __ret;
17215	17236	}
17216	17237	#else
17217	17238	__ai __attribute__((target("neon"))) float32x4_t vminq_f32(float32x4_t __p0, float32x4_t __p1) {
17218	17239	float32x4_t __ret;
17219		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17220		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
17221		__ret = (float32x4_t) __builtin_neon_vminq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 41);
17222		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	17240	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	17241	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	17242	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vminq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 41));
	17243	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
17223	17244	return __ret;
17224	17245	}
17225	17246	#endif
...	...	@@ -17227,16 +17248,16 @@ __ai __attribute__((target("neon"))) float32x4_t vminq_f32(float32x4_t __p0, flo
17227	17248	#ifdef __LITTLE_ENDIAN__
17228	17249	__ai __attribute__((target("neon"))) int32x4_t vminq_s32(int32x4_t __p0, int32x4_t __p1) {
17229	17250	int32x4_t __ret;
17230		__ret = (int32x4_t) __builtin_neon_vminq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
	17251	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vminq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 34));
17231	17252	return __ret;
17232	17253	}
17233	17254	#else
17234	17255	__ai __attribute__((target("neon"))) int32x4_t vminq_s32(int32x4_t __p0, int32x4_t __p1) {
17235	17256	int32x4_t __ret;
17236		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17237		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
17238		__ret = (int32x4_t) __builtin_neon_vminq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
17239		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	17257	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	17258	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	17259	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vminq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 34));
	17260	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
17240	17261	return __ret;
17241	17262	}
17242	17263	#endif
...	...	@@ -17244,16 +17265,16 @@ __ai __attribute__((target("neon"))) int32x4_t vminq_s32(int32x4_t __p0, int32x4
17244	17265	#ifdef __LITTLE_ENDIAN__
17245	17266	__ai __attribute__((target("neon"))) int16x8_t vminq_s16(int16x8_t __p0, int16x8_t __p1) {
17246	17267	int16x8_t __ret;
17247		__ret = (int16x8_t) __builtin_neon_vminq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
	17268	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vminq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 33));
17248	17269	return __ret;
17249	17270	}
17250	17271	#else
17251	17272	__ai __attribute__((target("neon"))) int16x8_t vminq_s16(int16x8_t __p0, int16x8_t __p1) {
17252	17273	int16x8_t __ret;
17253		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
17254		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
17255		__ret = (int16x8_t) __builtin_neon_vminq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
17256		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	17274	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	17275	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	17276	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vminq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 33));
	17277	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
17257	17278	return __ret;
17258	17279	}
17259	17280	#endif
...	...	@@ -17261,16 +17282,16 @@ __ai __attribute__((target("neon"))) int16x8_t vminq_s16(int16x8_t __p0, int16x8
17261	17282	#ifdef __LITTLE_ENDIAN__
17262	17283	__ai __attribute__((target("neon"))) uint8x8_t vmin_u8(uint8x8_t __p0, uint8x8_t __p1) {
17263	17284	uint8x8_t __ret;
17264		__ret = (uint8x8_t) __builtin_neon_vmin_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
	17285	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vmin_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 16));
17265	17286	return __ret;
17266	17287	}
17267	17288	#else
17268	17289	__ai __attribute__((target("neon"))) uint8x8_t vmin_u8(uint8x8_t __p0, uint8x8_t __p1) {
17269	17290	uint8x8_t __ret;
17270		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
17271		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
17272		__ret = (uint8x8_t) __builtin_neon_vmin_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
17273		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	17291	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	17292	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	17293	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vmin_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 16));
	17294	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
17274	17295	return __ret;
17275	17296	}
17276	17297	#endif
...	...	@@ -17278,16 +17299,16 @@ __ai __attribute__((target("neon"))) uint8x8_t vmin_u8(uint8x8_t __p0, uint8x8_t
17278	17299	#ifdef __LITTLE_ENDIAN__
17279	17300	__ai __attribute__((target("neon"))) uint32x2_t vmin_u32(uint32x2_t __p0, uint32x2_t __p1) {
17280	17301	uint32x2_t __ret;
17281		__ret = (uint32x2_t) __builtin_neon_vmin_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
	17302	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vmin_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 18));
17282	17303	return __ret;
17283	17304	}
17284	17305	#else
17285	17306	__ai __attribute__((target("neon"))) uint32x2_t vmin_u32(uint32x2_t __p0, uint32x2_t __p1) {
17286	17307	uint32x2_t __ret;
17287		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
17288		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
17289		__ret = (uint32x2_t) __builtin_neon_vmin_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
17290		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	17308	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	17309	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	17310	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vmin_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 18));
	17311	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
17291	17312	return __ret;
17292	17313	}
17293	17314	#endif
...	...	@@ -17295,16 +17316,16 @@ __ai __attribute__((target("neon"))) uint32x2_t vmin_u32(uint32x2_t __p0, uint32
17295	17316	#ifdef __LITTLE_ENDIAN__
17296	17317	__ai __attribute__((target("neon"))) uint16x4_t vmin_u16(uint16x4_t __p0, uint16x4_t __p1) {
17297	17318	uint16x4_t __ret;
17298		__ret = (uint16x4_t) __builtin_neon_vmin_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
	17319	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vmin_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 17));
17299	17320	return __ret;
17300	17321	}
17301	17322	#else
17302	17323	__ai __attribute__((target("neon"))) uint16x4_t vmin_u16(uint16x4_t __p0, uint16x4_t __p1) {
17303	17324	uint16x4_t __ret;
17304		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17305		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
17306		__ret = (uint16x4_t) __builtin_neon_vmin_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
17307		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	17325	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	17326	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	17327	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vmin_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 17));
	17328	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
17308	17329	return __ret;
17309	17330	}
17310	17331	#endif
...	...	@@ -17312,16 +17333,16 @@ __ai __attribute__((target("neon"))) uint16x4_t vmin_u16(uint16x4_t __p0, uint16
17312	17333	#ifdef __LITTLE_ENDIAN__
17313	17334	__ai __attribute__((target("neon"))) int8x8_t vmin_s8(int8x8_t __p0, int8x8_t __p1) {
17314	17335	int8x8_t __ret;
17315		__ret = (int8x8_t) __builtin_neon_vmin_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
	17336	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vmin_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 0));
17316	17337	return __ret;
17317	17338	}
17318	17339	#else
17319	17340	__ai __attribute__((target("neon"))) int8x8_t vmin_s8(int8x8_t __p0, int8x8_t __p1) {
17320	17341	int8x8_t __ret;
17321		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
17322		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
17323		__ret = (int8x8_t) __builtin_neon_vmin_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
17324		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	17342	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	17343	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	17344	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vmin_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 0));
	17345	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
17325	17346	return __ret;
17326	17347	}
17327	17348	#endif
...	...	@@ -17329,16 +17350,16 @@ __ai __attribute__((target("neon"))) int8x8_t vmin_s8(int8x8_t __p0, int8x8_t __
17329	17350	#ifdef __LITTLE_ENDIAN__
17330	17351	__ai __attribute__((target("neon"))) float32x2_t vmin_f32(float32x2_t __p0, float32x2_t __p1) {
17331	17352	float32x2_t __ret;
17332		__ret = (float32x2_t) __builtin_neon_vmin_v((int8x8_t)__p0, (int8x8_t)__p1, 9);
	17353	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vmin_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 9));
17333	17354	return __ret;
17334	17355	}
17335	17356	#else
17336	17357	__ai __attribute__((target("neon"))) float32x2_t vmin_f32(float32x2_t __p0, float32x2_t __p1) {
17337	17358	float32x2_t __ret;
17338		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
17339		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
17340		__ret = (float32x2_t) __builtin_neon_vmin_v((int8x8_t)__rev0, (int8x8_t)__rev1, 9);
17341		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	17359	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	17360	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	17361	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vmin_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 9));
	17362	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
17342	17363	return __ret;
17343	17364	}
17344	17365	#endif
...	...	@@ -17346,16 +17367,16 @@ __ai __attribute__((target("neon"))) float32x2_t vmin_f32(float32x2_t __p0, floa
17346	17367	#ifdef __LITTLE_ENDIAN__
17347	17368	__ai __attribute__((target("neon"))) int32x2_t vmin_s32(int32x2_t __p0, int32x2_t __p1) {
17348	17369	int32x2_t __ret;
17349		__ret = (int32x2_t) __builtin_neon_vmin_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
	17370	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vmin_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 2));
17350	17371	return __ret;
17351	17372	}
17352	17373	#else
17353	17374	__ai __attribute__((target("neon"))) int32x2_t vmin_s32(int32x2_t __p0, int32x2_t __p1) {
17354	17375	int32x2_t __ret;
17355		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
17356		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
17357		__ret = (int32x2_t) __builtin_neon_vmin_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
17358		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	17376	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	17377	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	17378	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vmin_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 2));
	17379	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
17359	17380	return __ret;
17360	17381	}
17361	17382	#endif
...	...	@@ -17363,16 +17384,16 @@ __ai __attribute__((target("neon"))) int32x2_t vmin_s32(int32x2_t __p0, int32x2_
17363	17384	#ifdef __LITTLE_ENDIAN__
17364	17385	__ai __attribute__((target("neon"))) int16x4_t vmin_s16(int16x4_t __p0, int16x4_t __p1) {
17365	17386	int16x4_t __ret;
17366		__ret = (int16x4_t) __builtin_neon_vmin_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
	17387	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vmin_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 1));
17367	17388	return __ret;
17368	17389	}
17369	17390	#else
17370	17391	__ai __attribute__((target("neon"))) int16x4_t vmin_s16(int16x4_t __p0, int16x4_t __p1) {
17371	17392	int16x4_t __ret;
17372		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17373		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
17374		__ret = (int16x4_t) __builtin_neon_vmin_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
17375		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	17393	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	17394	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	17395	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vmin_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 1));
	17396	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
17376	17397	return __ret;
17377	17398	}
17378	17399	#endif
...	...	@@ -17386,11 +17407,11 @@ __ai __attribute__((target("neon"))) uint8x16_t vmlaq_u8(uint8x16_t __p0, uint8x
17386	17407	#else
17387	17408	__ai __attribute__((target("neon"))) uint8x16_t vmlaq_u8(uint8x16_t __p0, uint8x16_t __p1, uint8x16_t __p2) {
17388	17409	uint8x16_t __ret;
17389		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
17390		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
17391		uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	17410	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	17411	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	17412	uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
17392	17413	__ret = __rev0 + __rev1 * __rev2;
17393		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	17414	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
17394	17415	return __ret;
17395	17416	}
17396	17417	#endif
...	...	@@ -17404,11 +17425,11 @@ __ai __attribute__((target("neon"))) uint32x4_t vmlaq_u32(uint32x4_t __p0, uint3
17404	17425	#else
17405	17426	__ai __attribute__((target("neon"))) uint32x4_t vmlaq_u32(uint32x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
17406	17427	uint32x4_t __ret;
17407		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17408		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
17409		uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	17428	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	17429	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	17430	uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
17410	17431	__ret = __rev0 + __rev1 * __rev2;
17411		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	17432	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
17412	17433	return __ret;
17413	17434	}
17414	17435	#endif
...	...	@@ -17422,11 +17443,11 @@ __ai __attribute__((target("neon"))) uint16x8_t vmlaq_u16(uint16x8_t __p0, uint1
17422	17443	#else
17423	17444	__ai __attribute__((target("neon"))) uint16x8_t vmlaq_u16(uint16x8_t __p0, uint16x8_t __p1, uint16x8_t __p2) {
17424	17445	uint16x8_t __ret;
17425		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
17426		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
17427		uint16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	17446	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	17447	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	17448	uint16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
17428	17449	__ret = __rev0 + __rev1 * __rev2;
17429		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	17450	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
17430	17451	return __ret;
17431	17452	}
17432	17453	#endif
...	...	@@ -17440,11 +17461,11 @@ __ai __attribute__((target("neon"))) int8x16_t vmlaq_s8(int8x16_t __p0, int8x16_
17440	17461	#else
17441	17462	__ai __attribute__((target("neon"))) int8x16_t vmlaq_s8(int8x16_t __p0, int8x16_t __p1, int8x16_t __p2) {
17442	17463	int8x16_t __ret;
17443		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
17444		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
17445		int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	17464	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	17465	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	17466	int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
17446	17467	__ret = __rev0 + __rev1 * __rev2;
17447		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	17468	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
17448	17469	return __ret;
17449	17470	}
17450	17471	#endif
...	...	@@ -17458,11 +17479,11 @@ __ai __attribute__((target("neon"))) float32x4_t vmlaq_f32(float32x4_t __p0, flo
17458	17479	#else
17459	17480	__ai __attribute__((target("neon"))) float32x4_t vmlaq_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
17460	17481	float32x4_t __ret;
17461		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17462		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
17463		float32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	17482	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	17483	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	17484	float32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
17464	17485	__ret = __rev0 + __rev1 * __rev2;
17465		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	17486	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
17466	17487	return __ret;
17467	17488	}
17468	17489	#endif
...	...	@@ -17476,11 +17497,11 @@ __ai __attribute__((target("neon"))) int32x4_t vmlaq_s32(int32x4_t __p0, int32x4
17476	17497	#else
17477	17498	__ai __attribute__((target("neon"))) int32x4_t vmlaq_s32(int32x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
17478	17499	int32x4_t __ret;
17479		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17480		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
17481		int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	17500	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	17501	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	17502	int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
17482	17503	__ret = __rev0 + __rev1 * __rev2;
17483		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	17504	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
17484	17505	return __ret;
17485	17506	}
17486	17507	#endif
...	...	@@ -17494,11 +17515,11 @@ __ai __attribute__((target("neon"))) int16x8_t vmlaq_s16(int16x8_t __p0, int16x8
17494	17515	#else
17495	17516	__ai __attribute__((target("neon"))) int16x8_t vmlaq_s16(int16x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
17496	17517	int16x8_t __ret;
17497		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
17498		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
17499		int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	17518	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	17519	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	17520	int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
17500	17521	__ret = __rev0 + __rev1 * __rev2;
17501		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	17522	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
17502	17523	return __ret;
17503	17524	}
17504	17525	#endif
...	...	@@ -17512,11 +17533,11 @@ __ai __attribute__((target("neon"))) uint8x8_t vmla_u8(uint8x8_t __p0, uint8x8_t
17512	17533	#else
17513	17534	__ai __attribute__((target("neon"))) uint8x8_t vmla_u8(uint8x8_t __p0, uint8x8_t __p1, uint8x8_t __p2) {
17514	17535	uint8x8_t __ret;
17515		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
17516		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
17517		uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	17536	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	17537	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	17538	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
17518	17539	__ret = __rev0 + __rev1 * __rev2;
17519		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	17540	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
17520	17541	return __ret;
17521	17542	}
17522	17543	#endif
...	...	@@ -17530,11 +17551,11 @@ __ai __attribute__((target("neon"))) uint32x2_t vmla_u32(uint32x2_t __p0, uint32
17530	17551	#else
17531	17552	__ai __attribute__((target("neon"))) uint32x2_t vmla_u32(uint32x2_t __p0, uint32x2_t __p1, uint32x2_t __p2) {
17532	17553	uint32x2_t __ret;
17533		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
17534		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
17535		uint32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
	17554	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	17555	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	17556	uint32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_32);
17536	17557	__ret = __rev0 + __rev1 * __rev2;
17537		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	17558	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
17538	17559	return __ret;
17539	17560	}
17540	17561	#endif
...	...	@@ -17548,11 +17569,11 @@ __ai __attribute__((target("neon"))) uint16x4_t vmla_u16(uint16x4_t __p0, uint16
17548	17569	#else
17549	17570	__ai __attribute__((target("neon"))) uint16x4_t vmla_u16(uint16x4_t __p0, uint16x4_t __p1, uint16x4_t __p2) {
17550	17571	uint16x4_t __ret;
17551		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17552		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
17553		uint16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	17572	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	17573	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	17574	uint16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
17554	17575	__ret = __rev0 + __rev1 * __rev2;
17555		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	17576	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
17556	17577	return __ret;
17557	17578	}
17558	17579	#endif
...	...	@@ -17566,11 +17587,11 @@ __ai __attribute__((target("neon"))) int8x8_t vmla_s8(int8x8_t __p0, int8x8_t __
17566	17587	#else
17567	17588	__ai __attribute__((target("neon"))) int8x8_t vmla_s8(int8x8_t __p0, int8x8_t __p1, int8x8_t __p2) {
17568	17589	int8x8_t __ret;
17569		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
17570		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
17571		int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	17590	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	17591	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	17592	int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
17572	17593	__ret = __rev0 + __rev1 * __rev2;
17573		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	17594	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
17574	17595	return __ret;
17575	17596	}
17576	17597	#endif
...	...	@@ -17584,11 +17605,11 @@ __ai __attribute__((target("neon"))) float32x2_t vmla_f32(float32x2_t __p0, floa
17584	17605	#else
17585	17606	__ai __attribute__((target("neon"))) float32x2_t vmla_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
17586	17607	float32x2_t __ret;
17587		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
17588		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
17589		float32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
	17608	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	17609	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	17610	float32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_32);
17590	17611	__ret = __rev0 + __rev1 * __rev2;
17591		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	17612	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
17592	17613	return __ret;
17593	17614	}
17594	17615	#endif
...	...	@@ -17602,11 +17623,11 @@ __ai __attribute__((target("neon"))) int32x2_t vmla_s32(int32x2_t __p0, int32x2_
17602	17623	#else
17603	17624	__ai __attribute__((target("neon"))) int32x2_t vmla_s32(int32x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
17604	17625	int32x2_t __ret;
17605		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
17606		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
17607		int32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
	17626	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	17627	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	17628	int32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_32);
17608	17629	__ret = __rev0 + __rev1 * __rev2;
17609		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	17630	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
17610	17631	return __ret;
17611	17632	}
17612	17633	#endif
...	...	@@ -17620,11 +17641,11 @@ __ai __attribute__((target("neon"))) int16x4_t vmla_s16(int16x4_t __p0, int16x4_
17620	17641	#else
17621	17642	__ai __attribute__((target("neon"))) int16x4_t vmla_s16(int16x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
17622	17643	int16x4_t __ret;
17623		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17624		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
17625		int16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	17644	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	17645	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	17646	int16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
17626	17647	__ret = __rev0 + __rev1 * __rev2;
17627		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	17648	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
17628	17649	return __ret;
17629	17650	}
17630	17651	#endif
...	...	@@ -17644,11 +17665,11 @@ __ai __attribute__((target("neon"))) int16x4_t vmla_s16(int16x4_t __p0, int16x4_
17644	17665	uint32x4_t __s0_55 = __p0_55; \
17645	17666	uint32x4_t __s1_55 = __p1_55; \
17646	17667	uint32x2_t __s2_55 = __p2_55; \
17647		uint32x4_t __rev0_55; __rev0_55 = __builtin_shufflevector(__s0_55, __s0_55, 3, 2, 1, 0); \
17648		uint32x4_t __rev1_55; __rev1_55 = __builtin_shufflevector(__s1_55, __s1_55, 3, 2, 1, 0); \
17649		uint32x2_t __rev2_55; __rev2_55 = __builtin_shufflevector(__s2_55, __s2_55, 1, 0); \
	17668	uint32x4_t __rev0_55; __rev0_55 = __builtin_shufflevector(__s0_55, __s0_55, __lane_reverse_128_32); \
	17669	uint32x4_t __rev1_55; __rev1_55 = __builtin_shufflevector(__s1_55, __s1_55, __lane_reverse_128_32); \
	17670	uint32x2_t __rev2_55; __rev2_55 = __builtin_shufflevector(__s2_55, __s2_55, __lane_reverse_64_32); \
17650	17671	__ret_55 = __rev0_55 + __rev1_55 * __noswap_splatq_lane_u32(__rev2_55, __p3_55); \
17651		__ret_55 = __builtin_shufflevector(__ret_55, __ret_55, 3, 2, 1, 0); \
	17672	__ret_55 = __builtin_shufflevector(__ret_55, __ret_55, __lane_reverse_128_32); \
17652	17673	__ret_55; \
17653	17674	})
17654	17675	#endif
...	...	@@ -17668,11 +17689,11 @@ __ai __attribute__((target("neon"))) int16x4_t vmla_s16(int16x4_t __p0, int16x4_
17668	17689	uint16x8_t __s0_57 = __p0_57; \
17669	17690	uint16x8_t __s1_57 = __p1_57; \
17670	17691	uint16x4_t __s2_57 = __p2_57; \
17671		uint16x8_t __rev0_57; __rev0_57 = __builtin_shufflevector(__s0_57, __s0_57, 7, 6, 5, 4, 3, 2, 1, 0); \
17672		uint16x8_t __rev1_57; __rev1_57 = __builtin_shufflevector(__s1_57, __s1_57, 7, 6, 5, 4, 3, 2, 1, 0); \
17673		uint16x4_t __rev2_57; __rev2_57 = __builtin_shufflevector(__s2_57, __s2_57, 3, 2, 1, 0); \
	17692	uint16x8_t __rev0_57; __rev0_57 = __builtin_shufflevector(__s0_57, __s0_57, __lane_reverse_128_16); \
	17693	uint16x8_t __rev1_57; __rev1_57 = __builtin_shufflevector(__s1_57, __s1_57, __lane_reverse_128_16); \
	17694	uint16x4_t __rev2_57; __rev2_57 = __builtin_shufflevector(__s2_57, __s2_57, __lane_reverse_64_16); \
17674	17695	__ret_57 = __rev0_57 + __rev1_57 * __noswap_splatq_lane_u16(__rev2_57, __p3_57); \
17675		__ret_57 = __builtin_shufflevector(__ret_57, __ret_57, 7, 6, 5, 4, 3, 2, 1, 0); \
	17696	__ret_57 = __builtin_shufflevector(__ret_57, __ret_57, __lane_reverse_128_16); \
17676	17697	__ret_57; \
17677	17698	})
17678	17699	#endif
...	...	@@ -17692,11 +17713,11 @@ __ai __attribute__((target("neon"))) int16x4_t vmla_s16(int16x4_t __p0, int16x4_
17692	17713	float32x4_t __s0_59 = __p0_59; \
17693	17714	float32x4_t __s1_59 = __p1_59; \
17694	17715	float32x2_t __s2_59 = __p2_59; \
17695		float32x4_t __rev0_59; __rev0_59 = __builtin_shufflevector(__s0_59, __s0_59, 3, 2, 1, 0); \
17696		float32x4_t __rev1_59; __rev1_59 = __builtin_shufflevector(__s1_59, __s1_59, 3, 2, 1, 0); \
17697		float32x2_t __rev2_59; __rev2_59 = __builtin_shufflevector(__s2_59, __s2_59, 1, 0); \
	17716	float32x4_t __rev0_59; __rev0_59 = __builtin_shufflevector(__s0_59, __s0_59, __lane_reverse_128_32); \
	17717	float32x4_t __rev1_59; __rev1_59 = __builtin_shufflevector(__s1_59, __s1_59, __lane_reverse_128_32); \
	17718	float32x2_t __rev2_59; __rev2_59 = __builtin_shufflevector(__s2_59, __s2_59, __lane_reverse_64_32); \
17698	17719	__ret_59 = __rev0_59 + __rev1_59 * __noswap_splatq_lane_f32(__rev2_59, __p3_59); \
17699		__ret_59 = __builtin_shufflevector(__ret_59, __ret_59, 3, 2, 1, 0); \
	17720	__ret_59 = __builtin_shufflevector(__ret_59, __ret_59, __lane_reverse_128_32); \
17700	17721	__ret_59; \
17701	17722	})
17702	17723	#endif
...	...	@@ -17716,11 +17737,11 @@ __ai __attribute__((target("neon"))) int16x4_t vmla_s16(int16x4_t __p0, int16x4_
17716	17737	int32x4_t __s0_61 = __p0_61; \
17717	17738	int32x4_t __s1_61 = __p1_61; \
17718	17739	int32x2_t __s2_61 = __p2_61; \
17719		int32x4_t __rev0_61; __rev0_61 = __builtin_shufflevector(__s0_61, __s0_61, 3, 2, 1, 0); \
17720		int32x4_t __rev1_61; __rev1_61 = __builtin_shufflevector(__s1_61, __s1_61, 3, 2, 1, 0); \
17721		int32x2_t __rev2_61; __rev2_61 = __builtin_shufflevector(__s2_61, __s2_61, 1, 0); \
	17740	int32x4_t __rev0_61; __rev0_61 = __builtin_shufflevector(__s0_61, __s0_61, __lane_reverse_128_32); \
	17741	int32x4_t __rev1_61; __rev1_61 = __builtin_shufflevector(__s1_61, __s1_61, __lane_reverse_128_32); \
	17742	int32x2_t __rev2_61; __rev2_61 = __builtin_shufflevector(__s2_61, __s2_61, __lane_reverse_64_32); \
17722	17743	__ret_61 = __rev0_61 + __rev1_61 * __noswap_splatq_lane_s32(__rev2_61, __p3_61); \
17723		__ret_61 = __builtin_shufflevector(__ret_61, __ret_61, 3, 2, 1, 0); \
	17744	__ret_61 = __builtin_shufflevector(__ret_61, __ret_61, __lane_reverse_128_32); \
17724	17745	__ret_61; \
17725	17746	})
17726	17747	#endif
...	...	@@ -17740,11 +17761,11 @@ __ai __attribute__((target("neon"))) int16x4_t vmla_s16(int16x4_t __p0, int16x4_
17740	17761	int16x8_t __s0_63 = __p0_63; \
17741	17762	int16x8_t __s1_63 = __p1_63; \
17742	17763	int16x4_t __s2_63 = __p2_63; \
17743		int16x8_t __rev0_63; __rev0_63 = __builtin_shufflevector(__s0_63, __s0_63, 7, 6, 5, 4, 3, 2, 1, 0); \
17744		int16x8_t __rev1_63; __rev1_63 = __builtin_shufflevector(__s1_63, __s1_63, 7, 6, 5, 4, 3, 2, 1, 0); \
17745		int16x4_t __rev2_63; __rev2_63 = __builtin_shufflevector(__s2_63, __s2_63, 3, 2, 1, 0); \
	17764	int16x8_t __rev0_63; __rev0_63 = __builtin_shufflevector(__s0_63, __s0_63, __lane_reverse_128_16); \
	17765	int16x8_t __rev1_63; __rev1_63 = __builtin_shufflevector(__s1_63, __s1_63, __lane_reverse_128_16); \
	17766	int16x4_t __rev2_63; __rev2_63 = __builtin_shufflevector(__s2_63, __s2_63, __lane_reverse_64_16); \
17746	17767	__ret_63 = __rev0_63 + __rev1_63 * __noswap_splatq_lane_s16(__rev2_63, __p3_63); \
17747		__ret_63 = __builtin_shufflevector(__ret_63, __ret_63, 7, 6, 5, 4, 3, 2, 1, 0); \
	17768	__ret_63 = __builtin_shufflevector(__ret_63, __ret_63, __lane_reverse_128_16); \
17748	17769	__ret_63; \
17749	17770	})
17750	17771	#endif
...	...	@@ -17764,11 +17785,11 @@ __ai __attribute__((target("neon"))) int16x4_t vmla_s16(int16x4_t __p0, int16x4_
17764	17785	uint32x2_t __s0_65 = __p0_65; \
17765	17786	uint32x2_t __s1_65 = __p1_65; \
17766	17787	uint32x2_t __s2_65 = __p2_65; \
17767		uint32x2_t __rev0_65; __rev0_65 = __builtin_shufflevector(__s0_65, __s0_65, 1, 0); \
17768		uint32x2_t __rev1_65; __rev1_65 = __builtin_shufflevector(__s1_65, __s1_65, 1, 0); \
17769		uint32x2_t __rev2_65; __rev2_65 = __builtin_shufflevector(__s2_65, __s2_65, 1, 0); \
	17788	uint32x2_t __rev0_65; __rev0_65 = __builtin_shufflevector(__s0_65, __s0_65, __lane_reverse_64_32); \
	17789	uint32x2_t __rev1_65; __rev1_65 = __builtin_shufflevector(__s1_65, __s1_65, __lane_reverse_64_32); \
	17790	uint32x2_t __rev2_65; __rev2_65 = __builtin_shufflevector(__s2_65, __s2_65, __lane_reverse_64_32); \
17770	17791	__ret_65 = __rev0_65 + __rev1_65 * __noswap_splat_lane_u32(__rev2_65, __p3_65); \
17771		__ret_65 = __builtin_shufflevector(__ret_65, __ret_65, 1, 0); \
	17792	__ret_65 = __builtin_shufflevector(__ret_65, __ret_65, __lane_reverse_64_32); \
17772	17793	__ret_65; \
17773	17794	})
17774	17795	#endif
...	...	@@ -17788,11 +17809,11 @@ __ai __attribute__((target("neon"))) int16x4_t vmla_s16(int16x4_t __p0, int16x4_
17788	17809	uint16x4_t __s0_67 = __p0_67; \
17789	17810	uint16x4_t __s1_67 = __p1_67; \
17790	17811	uint16x4_t __s2_67 = __p2_67; \
17791		uint16x4_t __rev0_67; __rev0_67 = __builtin_shufflevector(__s0_67, __s0_67, 3, 2, 1, 0); \
17792		uint16x4_t __rev1_67; __rev1_67 = __builtin_shufflevector(__s1_67, __s1_67, 3, 2, 1, 0); \
17793		uint16x4_t __rev2_67; __rev2_67 = __builtin_shufflevector(__s2_67, __s2_67, 3, 2, 1, 0); \
	17812	uint16x4_t __rev0_67; __rev0_67 = __builtin_shufflevector(__s0_67, __s0_67, __lane_reverse_64_16); \
	17813	uint16x4_t __rev1_67; __rev1_67 = __builtin_shufflevector(__s1_67, __s1_67, __lane_reverse_64_16); \
	17814	uint16x4_t __rev2_67; __rev2_67 = __builtin_shufflevector(__s2_67, __s2_67, __lane_reverse_64_16); \
17794	17815	__ret_67 = __rev0_67 + __rev1_67 * __noswap_splat_lane_u16(__rev2_67, __p3_67); \
17795		__ret_67 = __builtin_shufflevector(__ret_67, __ret_67, 3, 2, 1, 0); \
	17816	__ret_67 = __builtin_shufflevector(__ret_67, __ret_67, __lane_reverse_64_16); \
17796	17817	__ret_67; \
17797	17818	})
17798	17819	#endif
...	...	@@ -17812,11 +17833,11 @@ __ai __attribute__((target("neon"))) int16x4_t vmla_s16(int16x4_t __p0, int16x4_
17812	17833	float32x2_t __s0_69 = __p0_69; \
17813	17834	float32x2_t __s1_69 = __p1_69; \
17814	17835	float32x2_t __s2_69 = __p2_69; \
17815		float32x2_t __rev0_69; __rev0_69 = __builtin_shufflevector(__s0_69, __s0_69, 1, 0); \
17816		float32x2_t __rev1_69; __rev1_69 = __builtin_shufflevector(__s1_69, __s1_69, 1, 0); \
17817		float32x2_t __rev2_69; __rev2_69 = __builtin_shufflevector(__s2_69, __s2_69, 1, 0); \
	17836	float32x2_t __rev0_69; __rev0_69 = __builtin_shufflevector(__s0_69, __s0_69, __lane_reverse_64_32); \
	17837	float32x2_t __rev1_69; __rev1_69 = __builtin_shufflevector(__s1_69, __s1_69, __lane_reverse_64_32); \
	17838	float32x2_t __rev2_69; __rev2_69 = __builtin_shufflevector(__s2_69, __s2_69, __lane_reverse_64_32); \
17818	17839	__ret_69 = __rev0_69 + __rev1_69 * __noswap_splat_lane_f32(__rev2_69, __p3_69); \
17819		__ret_69 = __builtin_shufflevector(__ret_69, __ret_69, 1, 0); \
	17840	__ret_69 = __builtin_shufflevector(__ret_69, __ret_69, __lane_reverse_64_32); \
17820	17841	__ret_69; \
17821	17842	})
17822	17843	#endif
...	...	@@ -17836,11 +17857,11 @@ __ai __attribute__((target("neon"))) int16x4_t vmla_s16(int16x4_t __p0, int16x4_
17836	17857	int32x2_t __s0_71 = __p0_71; \
17837	17858	int32x2_t __s1_71 = __p1_71; \
17838	17859	int32x2_t __s2_71 = __p2_71; \
17839		int32x2_t __rev0_71; __rev0_71 = __builtin_shufflevector(__s0_71, __s0_71, 1, 0); \
17840		int32x2_t __rev1_71; __rev1_71 = __builtin_shufflevector(__s1_71, __s1_71, 1, 0); \
17841		int32x2_t __rev2_71; __rev2_71 = __builtin_shufflevector(__s2_71, __s2_71, 1, 0); \
	17860	int32x2_t __rev0_71; __rev0_71 = __builtin_shufflevector(__s0_71, __s0_71, __lane_reverse_64_32); \
	17861	int32x2_t __rev1_71; __rev1_71 = __builtin_shufflevector(__s1_71, __s1_71, __lane_reverse_64_32); \
	17862	int32x2_t __rev2_71; __rev2_71 = __builtin_shufflevector(__s2_71, __s2_71, __lane_reverse_64_32); \
17842	17863	__ret_71 = __rev0_71 + __rev1_71 * __noswap_splat_lane_s32(__rev2_71, __p3_71); \
17843		__ret_71 = __builtin_shufflevector(__ret_71, __ret_71, 1, 0); \
	17864	__ret_71 = __builtin_shufflevector(__ret_71, __ret_71, __lane_reverse_64_32); \
17844	17865	__ret_71; \
17845	17866	})
17846	17867	#endif
...	...	@@ -17860,11 +17881,11 @@ __ai __attribute__((target("neon"))) int16x4_t vmla_s16(int16x4_t __p0, int16x4_
17860	17881	int16x4_t __s0_73 = __p0_73; \
17861	17882	int16x4_t __s1_73 = __p1_73; \
17862	17883	int16x4_t __s2_73 = __p2_73; \
17863		int16x4_t __rev0_73; __rev0_73 = __builtin_shufflevector(__s0_73, __s0_73, 3, 2, 1, 0); \
17864		int16x4_t __rev1_73; __rev1_73 = __builtin_shufflevector(__s1_73, __s1_73, 3, 2, 1, 0); \
17865		int16x4_t __rev2_73; __rev2_73 = __builtin_shufflevector(__s2_73, __s2_73, 3, 2, 1, 0); \
	17884	int16x4_t __rev0_73; __rev0_73 = __builtin_shufflevector(__s0_73, __s0_73, __lane_reverse_64_16); \
	17885	int16x4_t __rev1_73; __rev1_73 = __builtin_shufflevector(__s1_73, __s1_73, __lane_reverse_64_16); \
	17886	int16x4_t __rev2_73; __rev2_73 = __builtin_shufflevector(__s2_73, __s2_73, __lane_reverse_64_16); \
17866	17887	__ret_73 = __rev0_73 + __rev1_73 * __noswap_splat_lane_s16(__rev2_73, __p3_73); \
17867		__ret_73 = __builtin_shufflevector(__ret_73, __ret_73, 3, 2, 1, 0); \
	17888	__ret_73 = __builtin_shufflevector(__ret_73, __ret_73, __lane_reverse_64_16); \
17868	17889	__ret_73; \
17869	17890	})
17870	17891	#endif
...	...	@@ -17878,10 +17899,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vmlaq_n_u32(uint32x4_t __p0, uin
17878	17899	#else
17879	17900	__ai __attribute__((target("neon"))) uint32x4_t vmlaq_n_u32(uint32x4_t __p0, uint32x4_t __p1, uint32_t __p2) {
17880	17901	uint32x4_t __ret;
17881		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17882		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	17902	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	17903	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
17883	17904	__ret = __rev0 + __rev1 * (uint32x4_t) {__p2, __p2, __p2, __p2};
17884		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	17905	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
17885	17906	return __ret;
17886	17907	}
17887	17908	#endif
...	...	@@ -17895,10 +17916,10 @@ __ai __attribute__((target("neon"))) uint16x8_t vmlaq_n_u16(uint16x8_t __p0, uin
17895	17916	#else
17896	17917	__ai __attribute__((target("neon"))) uint16x8_t vmlaq_n_u16(uint16x8_t __p0, uint16x8_t __p1, uint16_t __p2) {
17897	17918	uint16x8_t __ret;
17898		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
17899		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	17919	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	17920	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
17900	17921	__ret = __rev0 + __rev1 * (uint16x8_t) {__p2, __p2, __p2, __p2, __p2, __p2, __p2, __p2};
17901		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	17922	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
17902	17923	return __ret;
17903	17924	}
17904	17925	#endif
...	...	@@ -17912,10 +17933,10 @@ __ai __attribute__((target("neon"))) float32x4_t vmlaq_n_f32(float32x4_t __p0, f
17912	17933	#else
17913	17934	__ai __attribute__((target("neon"))) float32x4_t vmlaq_n_f32(float32x4_t __p0, float32x4_t __p1, float32_t __p2) {
17914	17935	float32x4_t __ret;
17915		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17916		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	17936	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	17937	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
17917	17938	__ret = __rev0 + __rev1 * (float32x4_t) {__p2, __p2, __p2, __p2};
17918		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	17939	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
17919	17940	return __ret;
17920	17941	}
17921	17942	#endif
...	...	@@ -17929,10 +17950,10 @@ __ai __attribute__((target("neon"))) int32x4_t vmlaq_n_s32(int32x4_t __p0, int32
17929	17950	#else
17930	17951	__ai __attribute__((target("neon"))) int32x4_t vmlaq_n_s32(int32x4_t __p0, int32x4_t __p1, int32_t __p2) {
17931	17952	int32x4_t __ret;
17932		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17933		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	17953	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	17954	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
17934	17955	__ret = __rev0 + __rev1 * (int32x4_t) {__p2, __p2, __p2, __p2};
17935		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	17956	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
17936	17957	return __ret;
17937	17958	}
17938	17959	#endif
...	...	@@ -17946,10 +17967,10 @@ __ai __attribute__((target("neon"))) int16x8_t vmlaq_n_s16(int16x8_t __p0, int16
17946	17967	#else
17947	17968	__ai __attribute__((target("neon"))) int16x8_t vmlaq_n_s16(int16x8_t __p0, int16x8_t __p1, int16_t __p2) {
17948	17969	int16x8_t __ret;
17949		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
17950		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	17970	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	17971	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
17951	17972	__ret = __rev0 + __rev1 * (int16x8_t) {__p2, __p2, __p2, __p2, __p2, __p2, __p2, __p2};
17952		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	17973	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
17953	17974	return __ret;
17954	17975	}
17955	17976	#endif
...	...	@@ -17963,10 +17984,10 @@ __ai __attribute__((target("neon"))) uint32x2_t vmla_n_u32(uint32x2_t __p0, uint
17963	17984	#else
17964	17985	__ai __attribute__((target("neon"))) uint32x2_t vmla_n_u32(uint32x2_t __p0, uint32x2_t __p1, uint32_t __p2) {
17965	17986	uint32x2_t __ret;
17966		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
17967		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	17987	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	17988	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
17968	17989	__ret = __rev0 + __rev1 * (uint32x2_t) {__p2, __p2};
17969		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	17990	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
17970	17991	return __ret;
17971	17992	}
17972	17993	#endif
...	...	@@ -17980,10 +18001,10 @@ __ai __attribute__((target("neon"))) uint16x4_t vmla_n_u16(uint16x4_t __p0, uint
17980	18001	#else
17981	18002	__ai __attribute__((target("neon"))) uint16x4_t vmla_n_u16(uint16x4_t __p0, uint16x4_t __p1, uint16_t __p2) {
17982	18003	uint16x4_t __ret;
17983		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17984		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	18004	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	18005	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
17985	18006	__ret = __rev0 + __rev1 * (uint16x4_t) {__p2, __p2, __p2, __p2};
17986		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	18007	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
17987	18008	return __ret;
17988	18009	}
17989	18010	#endif
...	...	@@ -17997,10 +18018,10 @@ __ai __attribute__((target("neon"))) float32x2_t vmla_n_f32(float32x2_t __p0, fl
17997	18018	#else
17998	18019	__ai __attribute__((target("neon"))) float32x2_t vmla_n_f32(float32x2_t __p0, float32x2_t __p1, float32_t __p2) {
17999	18020	float32x2_t __ret;
18000		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
18001		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	18021	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	18022	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
18002	18023	__ret = __rev0 + __rev1 * (float32x2_t) {__p2, __p2};
18003		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	18024	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
18004	18025	return __ret;
18005	18026	}
18006	18027	#endif
...	...	@@ -18014,10 +18035,10 @@ __ai __attribute__((target("neon"))) int32x2_t vmla_n_s32(int32x2_t __p0, int32x
18014	18035	#else
18015	18036	__ai __attribute__((target("neon"))) int32x2_t vmla_n_s32(int32x2_t __p0, int32x2_t __p1, int32_t __p2) {
18016	18037	int32x2_t __ret;
18017		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
18018		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	18038	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	18039	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
18019	18040	__ret = __rev0 + __rev1 * (int32x2_t) {__p2, __p2};
18020		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	18041	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
18021	18042	return __ret;
18022	18043	}
18023	18044	#endif
...	...	@@ -18031,10 +18052,10 @@ __ai __attribute__((target("neon"))) int16x4_t vmla_n_s16(int16x4_t __p0, int16x
18031	18052	#else
18032	18053	__ai __attribute__((target("neon"))) int16x4_t vmla_n_s16(int16x4_t __p0, int16x4_t __p1, int16_t __p2) {
18033	18054	int16x4_t __ret;
18034		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
18035		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	18055	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	18056	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
18036	18057	__ret = __rev0 + __rev1 * (int16x4_t) {__p2, __p2, __p2, __p2};
18037		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	18058	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
18038	18059	return __ret;
18039	18060	}
18040	18061	#endif
...	...	@@ -18048,11 +18069,11 @@ __ai __attribute__((target("neon"))) uint8x16_t vmlsq_u8(uint8x16_t __p0, uint8x
18048	18069	#else
18049	18070	__ai __attribute__((target("neon"))) uint8x16_t vmlsq_u8(uint8x16_t __p0, uint8x16_t __p1, uint8x16_t __p2) {
18050	18071	uint8x16_t __ret;
18051		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
18052		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
18053		uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	18072	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	18073	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	18074	uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
18054	18075	__ret = __rev0 - __rev1 * __rev2;
18055		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	18076	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
18056	18077	return __ret;
18057	18078	}
18058	18079	#endif
...	...	@@ -18066,11 +18087,11 @@ __ai __attribute__((target("neon"))) uint32x4_t vmlsq_u32(uint32x4_t __p0, uint3
18066	18087	#else
18067	18088	__ai __attribute__((target("neon"))) uint32x4_t vmlsq_u32(uint32x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
18068	18089	uint32x4_t __ret;
18069		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
18070		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
18071		uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	18090	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	18091	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	18092	uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
18072	18093	__ret = __rev0 - __rev1 * __rev2;
18073		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	18094	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
18074	18095	return __ret;
18075	18096	}
18076	18097	#endif
...	...	@@ -18084,11 +18105,11 @@ __ai __attribute__((target("neon"))) uint16x8_t vmlsq_u16(uint16x8_t __p0, uint1
18084	18105	#else
18085	18106	__ai __attribute__((target("neon"))) uint16x8_t vmlsq_u16(uint16x8_t __p0, uint16x8_t __p1, uint16x8_t __p2) {
18086	18107	uint16x8_t __ret;
18087		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
18088		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
18089		uint16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	18108	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	18109	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	18110	uint16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
18090	18111	__ret = __rev0 - __rev1 * __rev2;
18091		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	18112	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
18092	18113	return __ret;
18093	18114	}
18094	18115	#endif
...	...	@@ -18102,11 +18123,11 @@ __ai __attribute__((target("neon"))) int8x16_t vmlsq_s8(int8x16_t __p0, int8x16_
18102	18123	#else
18103	18124	__ai __attribute__((target("neon"))) int8x16_t vmlsq_s8(int8x16_t __p0, int8x16_t __p1, int8x16_t __p2) {
18104	18125	int8x16_t __ret;
18105		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
18106		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
18107		int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	18126	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	18127	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	18128	int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
18108	18129	__ret = __rev0 - __rev1 * __rev2;
18109		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	18130	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
18110	18131	return __ret;
18111	18132	}
18112	18133	#endif
...	...	@@ -18120,11 +18141,11 @@ __ai __attribute__((target("neon"))) float32x4_t vmlsq_f32(float32x4_t __p0, flo
18120	18141	#else
18121	18142	__ai __attribute__((target("neon"))) float32x4_t vmlsq_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
18122	18143	float32x4_t __ret;
18123		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
18124		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
18125		float32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	18144	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	18145	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	18146	float32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
18126	18147	__ret = __rev0 - __rev1 * __rev2;
18127		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	18148	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
18128	18149	return __ret;
18129	18150	}
18130	18151	#endif
...	...	@@ -18138,11 +18159,11 @@ __ai __attribute__((target("neon"))) int32x4_t vmlsq_s32(int32x4_t __p0, int32x4
18138	18159	#else
18139	18160	__ai __attribute__((target("neon"))) int32x4_t vmlsq_s32(int32x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
18140	18161	int32x4_t __ret;
18141		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
18142		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
18143		int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	18162	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	18163	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	18164	int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
18144	18165	__ret = __rev0 - __rev1 * __rev2;
18145		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	18166	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
18146	18167	return __ret;
18147	18168	}
18148	18169	#endif
...	...	@@ -18156,11 +18177,11 @@ __ai __attribute__((target("neon"))) int16x8_t vmlsq_s16(int16x8_t __p0, int16x8
18156	18177	#else
18157	18178	__ai __attribute__((target("neon"))) int16x8_t vmlsq_s16(int16x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
18158	18179	int16x8_t __ret;
18159		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
18160		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
18161		int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	18180	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	18181	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	18182	int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
18162	18183	__ret = __rev0 - __rev1 * __rev2;
18163		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	18184	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
18164	18185	return __ret;
18165	18186	}
18166	18187	#endif
...	...	@@ -18174,11 +18195,11 @@ __ai __attribute__((target("neon"))) uint8x8_t vmls_u8(uint8x8_t __p0, uint8x8_t
18174	18195	#else
18175	18196	__ai __attribute__((target("neon"))) uint8x8_t vmls_u8(uint8x8_t __p0, uint8x8_t __p1, uint8x8_t __p2) {
18176	18197	uint8x8_t __ret;
18177		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
18178		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
18179		uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	18198	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	18199	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	18200	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
18180	18201	__ret = __rev0 - __rev1 * __rev2;
18181		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	18202	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
18182	18203	return __ret;
18183	18204	}
18184	18205	#endif
...	...	@@ -18192,11 +18213,11 @@ __ai __attribute__((target("neon"))) uint32x2_t vmls_u32(uint32x2_t __p0, uint32
18192	18213	#else
18193	18214	__ai __attribute__((target("neon"))) uint32x2_t vmls_u32(uint32x2_t __p0, uint32x2_t __p1, uint32x2_t __p2) {
18194	18215	uint32x2_t __ret;
18195		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
18196		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
18197		uint32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
	18216	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	18217	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	18218	uint32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_32);
18198	18219	__ret = __rev0 - __rev1 * __rev2;
18199		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	18220	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
18200	18221	return __ret;
18201	18222	}
18202	18223	#endif
...	...	@@ -18210,11 +18231,11 @@ __ai __attribute__((target("neon"))) uint16x4_t vmls_u16(uint16x4_t __p0, uint16
18210	18231	#else
18211	18232	__ai __attribute__((target("neon"))) uint16x4_t vmls_u16(uint16x4_t __p0, uint16x4_t __p1, uint16x4_t __p2) {
18212	18233	uint16x4_t __ret;
18213		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
18214		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
18215		uint16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	18234	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	18235	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	18236	uint16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
18216	18237	__ret = __rev0 - __rev1 * __rev2;
18217		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	18238	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
18218	18239	return __ret;
18219	18240	}
18220	18241	#endif
...	...	@@ -18228,11 +18249,11 @@ __ai __attribute__((target("neon"))) int8x8_t vmls_s8(int8x8_t __p0, int8x8_t __
18228	18249	#else
18229	18250	__ai __attribute__((target("neon"))) int8x8_t vmls_s8(int8x8_t __p0, int8x8_t __p1, int8x8_t __p2) {
18230	18251	int8x8_t __ret;
18231		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
18232		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
18233		int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	18252	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	18253	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	18254	int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
18234	18255	__ret = __rev0 - __rev1 * __rev2;
18235		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	18256	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
18236	18257	return __ret;
18237	18258	}
18238	18259	#endif
...	...	@@ -18246,11 +18267,11 @@ __ai __attribute__((target("neon"))) float32x2_t vmls_f32(float32x2_t __p0, floa
18246	18267	#else
18247	18268	__ai __attribute__((target("neon"))) float32x2_t vmls_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
18248	18269	float32x2_t __ret;
18249		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
18250		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
18251		float32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
	18270	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	18271	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	18272	float32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_32);
18252	18273	__ret = __rev0 - __rev1 * __rev2;
18253		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	18274	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
18254	18275	return __ret;
18255	18276	}
18256	18277	#endif
...	...	@@ -18264,11 +18285,11 @@ __ai __attribute__((target("neon"))) int32x2_t vmls_s32(int32x2_t __p0, int32x2_
18264	18285	#else
18265	18286	__ai __attribute__((target("neon"))) int32x2_t vmls_s32(int32x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
18266	18287	int32x2_t __ret;
18267		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
18268		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
18269		int32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
	18288	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	18289	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	18290	int32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_32);
18270	18291	__ret = __rev0 - __rev1 * __rev2;
18271		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	18292	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
18272	18293	return __ret;
18273	18294	}
18274	18295	#endif
...	...	@@ -18282,11 +18303,11 @@ __ai __attribute__((target("neon"))) int16x4_t vmls_s16(int16x4_t __p0, int16x4_
18282	18303	#else
18283	18304	__ai __attribute__((target("neon"))) int16x4_t vmls_s16(int16x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
18284	18305	int16x4_t __ret;
18285		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
18286		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
18287		int16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	18306	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	18307	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	18308	int16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
18288	18309	__ret = __rev0 - __rev1 * __rev2;
18289		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	18310	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
18290	18311	return __ret;
18291	18312	}
18292	18313	#endif
...	...	@@ -18306,11 +18327,11 @@ __ai __attribute__((target("neon"))) int16x4_t vmls_s16(int16x4_t __p0, int16x4_
18306	18327	uint32x4_t __s0_75 = __p0_75; \
18307	18328	uint32x4_t __s1_75 = __p1_75; \
18308	18329	uint32x2_t __s2_75 = __p2_75; \
18309		uint32x4_t __rev0_75; __rev0_75 = __builtin_shufflevector(__s0_75, __s0_75, 3, 2, 1, 0); \
18310		uint32x4_t __rev1_75; __rev1_75 = __builtin_shufflevector(__s1_75, __s1_75, 3, 2, 1, 0); \
18311		uint32x2_t __rev2_75; __rev2_75 = __builtin_shufflevector(__s2_75, __s2_75, 1, 0); \
	18330	uint32x4_t __rev0_75; __rev0_75 = __builtin_shufflevector(__s0_75, __s0_75, __lane_reverse_128_32); \
	18331	uint32x4_t __rev1_75; __rev1_75 = __builtin_shufflevector(__s1_75, __s1_75, __lane_reverse_128_32); \
	18332	uint32x2_t __rev2_75; __rev2_75 = __builtin_shufflevector(__s2_75, __s2_75, __lane_reverse_64_32); \
18312	18333	__ret_75 = __rev0_75 - __rev1_75 * __noswap_splatq_lane_u32(__rev2_75, __p3_75); \
18313		__ret_75 = __builtin_shufflevector(__ret_75, __ret_75, 3, 2, 1, 0); \
	18334	__ret_75 = __builtin_shufflevector(__ret_75, __ret_75, __lane_reverse_128_32); \
18314	18335	__ret_75; \
18315	18336	})
18316	18337	#endif
...	...	@@ -18330,11 +18351,11 @@ __ai __attribute__((target("neon"))) int16x4_t vmls_s16(int16x4_t __p0, int16x4_
18330	18351	uint16x8_t __s0_77 = __p0_77; \
18331	18352	uint16x8_t __s1_77 = __p1_77; \
18332	18353	uint16x4_t __s2_77 = __p2_77; \
18333		uint16x8_t __rev0_77; __rev0_77 = __builtin_shufflevector(__s0_77, __s0_77, 7, 6, 5, 4, 3, 2, 1, 0); \
18334		uint16x8_t __rev1_77; __rev1_77 = __builtin_shufflevector(__s1_77, __s1_77, 7, 6, 5, 4, 3, 2, 1, 0); \
18335		uint16x4_t __rev2_77; __rev2_77 = __builtin_shufflevector(__s2_77, __s2_77, 3, 2, 1, 0); \
	18354	uint16x8_t __rev0_77; __rev0_77 = __builtin_shufflevector(__s0_77, __s0_77, __lane_reverse_128_16); \
	18355	uint16x8_t __rev1_77; __rev1_77 = __builtin_shufflevector(__s1_77, __s1_77, __lane_reverse_128_16); \
	18356	uint16x4_t __rev2_77; __rev2_77 = __builtin_shufflevector(__s2_77, __s2_77, __lane_reverse_64_16); \
18336	18357	__ret_77 = __rev0_77 - __rev1_77 * __noswap_splatq_lane_u16(__rev2_77, __p3_77); \
18337		__ret_77 = __builtin_shufflevector(__ret_77, __ret_77, 7, 6, 5, 4, 3, 2, 1, 0); \
	18358	__ret_77 = __builtin_shufflevector(__ret_77, __ret_77, __lane_reverse_128_16); \
18338	18359	__ret_77; \
18339	18360	})
18340	18361	#endif
...	...	@@ -18354,11 +18375,11 @@ __ai __attribute__((target("neon"))) int16x4_t vmls_s16(int16x4_t __p0, int16x4_
18354	18375	float32x4_t __s0_79 = __p0_79; \
18355	18376	float32x4_t __s1_79 = __p1_79; \
18356	18377	float32x2_t __s2_79 = __p2_79; \
18357		float32x4_t __rev0_79; __rev0_79 = __builtin_shufflevector(__s0_79, __s0_79, 3, 2, 1, 0); \
18358		float32x4_t __rev1_79; __rev1_79 = __builtin_shufflevector(__s1_79, __s1_79, 3, 2, 1, 0); \
18359		float32x2_t __rev2_79; __rev2_79 = __builtin_shufflevector(__s2_79, __s2_79, 1, 0); \
	18378	float32x4_t __rev0_79; __rev0_79 = __builtin_shufflevector(__s0_79, __s0_79, __lane_reverse_128_32); \
	18379	float32x4_t __rev1_79; __rev1_79 = __builtin_shufflevector(__s1_79, __s1_79, __lane_reverse_128_32); \
	18380	float32x2_t __rev2_79; __rev2_79 = __builtin_shufflevector(__s2_79, __s2_79, __lane_reverse_64_32); \
18360	18381	__ret_79 = __rev0_79 - __rev1_79 * __noswap_splatq_lane_f32(__rev2_79, __p3_79); \
18361		__ret_79 = __builtin_shufflevector(__ret_79, __ret_79, 3, 2, 1, 0); \
	18382	__ret_79 = __builtin_shufflevector(__ret_79, __ret_79, __lane_reverse_128_32); \
18362	18383	__ret_79; \
18363	18384	})
18364	18385	#endif
...	...	@@ -18378,11 +18399,11 @@ __ai __attribute__((target("neon"))) int16x4_t vmls_s16(int16x4_t __p0, int16x4_
18378	18399	int32x4_t __s0_81 = __p0_81; \
18379	18400	int32x4_t __s1_81 = __p1_81; \
18380	18401	int32x2_t __s2_81 = __p2_81; \
18381		int32x4_t __rev0_81; __rev0_81 = __builtin_shufflevector(__s0_81, __s0_81, 3, 2, 1, 0); \
18382		int32x4_t __rev1_81; __rev1_81 = __builtin_shufflevector(__s1_81, __s1_81, 3, 2, 1, 0); \
18383		int32x2_t __rev2_81; __rev2_81 = __builtin_shufflevector(__s2_81, __s2_81, 1, 0); \
	18402	int32x4_t __rev0_81; __rev0_81 = __builtin_shufflevector(__s0_81, __s0_81, __lane_reverse_128_32); \
	18403	int32x4_t __rev1_81; __rev1_81 = __builtin_shufflevector(__s1_81, __s1_81, __lane_reverse_128_32); \
	18404	int32x2_t __rev2_81; __rev2_81 = __builtin_shufflevector(__s2_81, __s2_81, __lane_reverse_64_32); \
18384	18405	__ret_81 = __rev0_81 - __rev1_81 * __noswap_splatq_lane_s32(__rev2_81, __p3_81); \
18385		__ret_81 = __builtin_shufflevector(__ret_81, __ret_81, 3, 2, 1, 0); \
	18406	__ret_81 = __builtin_shufflevector(__ret_81, __ret_81, __lane_reverse_128_32); \
18386	18407	__ret_81; \
18387	18408	})
18388	18409	#endif
...	...	@@ -18402,11 +18423,11 @@ __ai __attribute__((target("neon"))) int16x4_t vmls_s16(int16x4_t __p0, int16x4_
18402	18423	int16x8_t __s0_83 = __p0_83; \
18403	18424	int16x8_t __s1_83 = __p1_83; \
18404	18425	int16x4_t __s2_83 = __p2_83; \
18405		int16x8_t __rev0_83; __rev0_83 = __builtin_shufflevector(__s0_83, __s0_83, 7, 6, 5, 4, 3, 2, 1, 0); \
18406		int16x8_t __rev1_83; __rev1_83 = __builtin_shufflevector(__s1_83, __s1_83, 7, 6, 5, 4, 3, 2, 1, 0); \
18407		int16x4_t __rev2_83; __rev2_83 = __builtin_shufflevector(__s2_83, __s2_83, 3, 2, 1, 0); \
	18426	int16x8_t __rev0_83; __rev0_83 = __builtin_shufflevector(__s0_83, __s0_83, __lane_reverse_128_16); \
	18427	int16x8_t __rev1_83; __rev1_83 = __builtin_shufflevector(__s1_83, __s1_83, __lane_reverse_128_16); \
	18428	int16x4_t __rev2_83; __rev2_83 = __builtin_shufflevector(__s2_83, __s2_83, __lane_reverse_64_16); \
18408	18429	__ret_83 = __rev0_83 - __rev1_83 * __noswap_splatq_lane_s16(__rev2_83, __p3_83); \
18409		__ret_83 = __builtin_shufflevector(__ret_83, __ret_83, 7, 6, 5, 4, 3, 2, 1, 0); \
	18430	__ret_83 = __builtin_shufflevector(__ret_83, __ret_83, __lane_reverse_128_16); \
18410	18431	__ret_83; \
18411	18432	})
18412	18433	#endif
...	...	@@ -18426,11 +18447,11 @@ __ai __attribute__((target("neon"))) int16x4_t vmls_s16(int16x4_t __p0, int16x4_
18426	18447	uint32x2_t __s0_85 = __p0_85; \
18427	18448	uint32x2_t __s1_85 = __p1_85; \
18428	18449	uint32x2_t __s2_85 = __p2_85; \
18429		uint32x2_t __rev0_85; __rev0_85 = __builtin_shufflevector(__s0_85, __s0_85, 1, 0); \
18430		uint32x2_t __rev1_85; __rev1_85 = __builtin_shufflevector(__s1_85, __s1_85, 1, 0); \
18431		uint32x2_t __rev2_85; __rev2_85 = __builtin_shufflevector(__s2_85, __s2_85, 1, 0); \
	18450	uint32x2_t __rev0_85; __rev0_85 = __builtin_shufflevector(__s0_85, __s0_85, __lane_reverse_64_32); \
	18451	uint32x2_t __rev1_85; __rev1_85 = __builtin_shufflevector(__s1_85, __s1_85, __lane_reverse_64_32); \
	18452	uint32x2_t __rev2_85; __rev2_85 = __builtin_shufflevector(__s2_85, __s2_85, __lane_reverse_64_32); \
18432	18453	__ret_85 = __rev0_85 - __rev1_85 * __noswap_splat_lane_u32(__rev2_85, __p3_85); \
18433		__ret_85 = __builtin_shufflevector(__ret_85, __ret_85, 1, 0); \
	18454	__ret_85 = __builtin_shufflevector(__ret_85, __ret_85, __lane_reverse_64_32); \
18434	18455	__ret_85; \
18435	18456	})
18436	18457	#endif
...	...	@@ -18450,11 +18471,11 @@ __ai __attribute__((target("neon"))) int16x4_t vmls_s16(int16x4_t __p0, int16x4_
18450	18471	uint16x4_t __s0_87 = __p0_87; \
18451	18472	uint16x4_t __s1_87 = __p1_87; \
18452	18473	uint16x4_t __s2_87 = __p2_87; \
18453		uint16x4_t __rev0_87; __rev0_87 = __builtin_shufflevector(__s0_87, __s0_87, 3, 2, 1, 0); \
18454		uint16x4_t __rev1_87; __rev1_87 = __builtin_shufflevector(__s1_87, __s1_87, 3, 2, 1, 0); \
18455		uint16x4_t __rev2_87; __rev2_87 = __builtin_shufflevector(__s2_87, __s2_87, 3, 2, 1, 0); \
	18474	uint16x4_t __rev0_87; __rev0_87 = __builtin_shufflevector(__s0_87, __s0_87, __lane_reverse_64_16); \
	18475	uint16x4_t __rev1_87; __rev1_87 = __builtin_shufflevector(__s1_87, __s1_87, __lane_reverse_64_16); \
	18476	uint16x4_t __rev2_87; __rev2_87 = __builtin_shufflevector(__s2_87, __s2_87, __lane_reverse_64_16); \
18456	18477	__ret_87 = __rev0_87 - __rev1_87 * __noswap_splat_lane_u16(__rev2_87, __p3_87); \
18457		__ret_87 = __builtin_shufflevector(__ret_87, __ret_87, 3, 2, 1, 0); \
	18478	__ret_87 = __builtin_shufflevector(__ret_87, __ret_87, __lane_reverse_64_16); \
18458	18479	__ret_87; \
18459	18480	})
18460	18481	#endif
...	...	@@ -18474,11 +18495,11 @@ __ai __attribute__((target("neon"))) int16x4_t vmls_s16(int16x4_t __p0, int16x4_
18474	18495	float32x2_t __s0_89 = __p0_89; \
18475	18496	float32x2_t __s1_89 = __p1_89; \
18476	18497	float32x2_t __s2_89 = __p2_89; \
18477		float32x2_t __rev0_89; __rev0_89 = __builtin_shufflevector(__s0_89, __s0_89, 1, 0); \
18478		float32x2_t __rev1_89; __rev1_89 = __builtin_shufflevector(__s1_89, __s1_89, 1, 0); \
18479		float32x2_t __rev2_89; __rev2_89 = __builtin_shufflevector(__s2_89, __s2_89, 1, 0); \
	18498	float32x2_t __rev0_89; __rev0_89 = __builtin_shufflevector(__s0_89, __s0_89, __lane_reverse_64_32); \
	18499	float32x2_t __rev1_89; __rev1_89 = __builtin_shufflevector(__s1_89, __s1_89, __lane_reverse_64_32); \
	18500	float32x2_t __rev2_89; __rev2_89 = __builtin_shufflevector(__s2_89, __s2_89, __lane_reverse_64_32); \
18480	18501	__ret_89 = __rev0_89 - __rev1_89 * __noswap_splat_lane_f32(__rev2_89, __p3_89); \
18481		__ret_89 = __builtin_shufflevector(__ret_89, __ret_89, 1, 0); \
	18502	__ret_89 = __builtin_shufflevector(__ret_89, __ret_89, __lane_reverse_64_32); \
18482	18503	__ret_89; \
18483	18504	})
18484	18505	#endif
...	...	@@ -18498,11 +18519,11 @@ __ai __attribute__((target("neon"))) int16x4_t vmls_s16(int16x4_t __p0, int16x4_
18498	18519	int32x2_t __s0_91 = __p0_91; \
18499	18520	int32x2_t __s1_91 = __p1_91; \
18500	18521	int32x2_t __s2_91 = __p2_91; \
18501		int32x2_t __rev0_91; __rev0_91 = __builtin_shufflevector(__s0_91, __s0_91, 1, 0); \
18502		int32x2_t __rev1_91; __rev1_91 = __builtin_shufflevector(__s1_91, __s1_91, 1, 0); \
18503		int32x2_t __rev2_91; __rev2_91 = __builtin_shufflevector(__s2_91, __s2_91, 1, 0); \
	18522	int32x2_t __rev0_91; __rev0_91 = __builtin_shufflevector(__s0_91, __s0_91, __lane_reverse_64_32); \
	18523	int32x2_t __rev1_91; __rev1_91 = __builtin_shufflevector(__s1_91, __s1_91, __lane_reverse_64_32); \
	18524	int32x2_t __rev2_91; __rev2_91 = __builtin_shufflevector(__s2_91, __s2_91, __lane_reverse_64_32); \
18504	18525	__ret_91 = __rev0_91 - __rev1_91 * __noswap_splat_lane_s32(__rev2_91, __p3_91); \
18505		__ret_91 = __builtin_shufflevector(__ret_91, __ret_91, 1, 0); \
	18526	__ret_91 = __builtin_shufflevector(__ret_91, __ret_91, __lane_reverse_64_32); \
18506	18527	__ret_91; \
18507	18528	})
18508	18529	#endif
...	...	@@ -18522,11 +18543,11 @@ __ai __attribute__((target("neon"))) int16x4_t vmls_s16(int16x4_t __p0, int16x4_
18522	18543	int16x4_t __s0_93 = __p0_93; \
18523	18544	int16x4_t __s1_93 = __p1_93; \
18524	18545	int16x4_t __s2_93 = __p2_93; \
18525		int16x4_t __rev0_93; __rev0_93 = __builtin_shufflevector(__s0_93, __s0_93, 3, 2, 1, 0); \
18526		int16x4_t __rev1_93; __rev1_93 = __builtin_shufflevector(__s1_93, __s1_93, 3, 2, 1, 0); \
18527		int16x4_t __rev2_93; __rev2_93 = __builtin_shufflevector(__s2_93, __s2_93, 3, 2, 1, 0); \
	18546	int16x4_t __rev0_93; __rev0_93 = __builtin_shufflevector(__s0_93, __s0_93, __lane_reverse_64_16); \
	18547	int16x4_t __rev1_93; __rev1_93 = __builtin_shufflevector(__s1_93, __s1_93, __lane_reverse_64_16); \
	18548	int16x4_t __rev2_93; __rev2_93 = __builtin_shufflevector(__s2_93, __s2_93, __lane_reverse_64_16); \
18528	18549	__ret_93 = __rev0_93 - __rev1_93 * __noswap_splat_lane_s16(__rev2_93, __p3_93); \
18529		__ret_93 = __builtin_shufflevector(__ret_93, __ret_93, 3, 2, 1, 0); \
	18550	__ret_93 = __builtin_shufflevector(__ret_93, __ret_93, __lane_reverse_64_16); \
18530	18551	__ret_93; \
18531	18552	})
18532	18553	#endif
...	...	@@ -18540,10 +18561,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vmlsq_n_u32(uint32x4_t __p0, uin
18540	18561	#else
18541	18562	__ai __attribute__((target("neon"))) uint32x4_t vmlsq_n_u32(uint32x4_t __p0, uint32x4_t __p1, uint32_t __p2) {
18542	18563	uint32x4_t __ret;
18543		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
18544		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	18564	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	18565	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
18545	18566	__ret = __rev0 - __rev1 * (uint32x4_t) {__p2, __p2, __p2, __p2};
18546		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	18567	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
18547	18568	return __ret;
18548	18569	}
18549	18570	#endif
...	...	@@ -18557,10 +18578,10 @@ __ai __attribute__((target("neon"))) uint16x8_t vmlsq_n_u16(uint16x8_t __p0, uin
18557	18578	#else
18558	18579	__ai __attribute__((target("neon"))) uint16x8_t vmlsq_n_u16(uint16x8_t __p0, uint16x8_t __p1, uint16_t __p2) {
18559	18580	uint16x8_t __ret;
18560		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
18561		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	18581	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	18582	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
18562	18583	__ret = __rev0 - __rev1 * (uint16x8_t) {__p2, __p2, __p2, __p2, __p2, __p2, __p2, __p2};
18563		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	18584	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
18564	18585	return __ret;
18565	18586	}
18566	18587	#endif
...	...	@@ -18574,10 +18595,10 @@ __ai __attribute__((target("neon"))) float32x4_t vmlsq_n_f32(float32x4_t __p0, f
18574	18595	#else
18575	18596	__ai __attribute__((target("neon"))) float32x4_t vmlsq_n_f32(float32x4_t __p0, float32x4_t __p1, float32_t __p2) {
18576	18597	float32x4_t __ret;
18577		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
18578		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	18598	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	18599	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
18579	18600	__ret = __rev0 - __rev1 * (float32x4_t) {__p2, __p2, __p2, __p2};
18580		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	18601	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
18581	18602	return __ret;
18582	18603	}
18583	18604	#endif
...	...	@@ -18591,10 +18612,10 @@ __ai __attribute__((target("neon"))) int32x4_t vmlsq_n_s32(int32x4_t __p0, int32
18591	18612	#else
18592	18613	__ai __attribute__((target("neon"))) int32x4_t vmlsq_n_s32(int32x4_t __p0, int32x4_t __p1, int32_t __p2) {
18593	18614	int32x4_t __ret;
18594		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
18595		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	18615	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	18616	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
18596	18617	__ret = __rev0 - __rev1 * (int32x4_t) {__p2, __p2, __p2, __p2};
18597		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	18618	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
18598	18619	return __ret;
18599	18620	}
18600	18621	#endif
...	...	@@ -18608,10 +18629,10 @@ __ai __attribute__((target("neon"))) int16x8_t vmlsq_n_s16(int16x8_t __p0, int16
18608	18629	#else
18609	18630	__ai __attribute__((target("neon"))) int16x8_t vmlsq_n_s16(int16x8_t __p0, int16x8_t __p1, int16_t __p2) {
18610	18631	int16x8_t __ret;
18611		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
18612		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	18632	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	18633	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
18613	18634	__ret = __rev0 - __rev1 * (int16x8_t) {__p2, __p2, __p2, __p2, __p2, __p2, __p2, __p2};
18614		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	18635	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
18615	18636	return __ret;
18616	18637	}
18617	18638	#endif
...	...	@@ -18625,10 +18646,10 @@ __ai __attribute__((target("neon"))) uint32x2_t vmls_n_u32(uint32x2_t __p0, uint
18625	18646	#else
18626	18647	__ai __attribute__((target("neon"))) uint32x2_t vmls_n_u32(uint32x2_t __p0, uint32x2_t __p1, uint32_t __p2) {
18627	18648	uint32x2_t __ret;
18628		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
18629		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	18649	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	18650	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
18630	18651	__ret = __rev0 - __rev1 * (uint32x2_t) {__p2, __p2};
18631		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	18652	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
18632	18653	return __ret;
18633	18654	}
18634	18655	#endif
...	...	@@ -18642,10 +18663,10 @@ __ai __attribute__((target("neon"))) uint16x4_t vmls_n_u16(uint16x4_t __p0, uint
18642	18663	#else
18643	18664	__ai __attribute__((target("neon"))) uint16x4_t vmls_n_u16(uint16x4_t __p0, uint16x4_t __p1, uint16_t __p2) {
18644	18665	uint16x4_t __ret;
18645		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
18646		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	18666	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	18667	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
18647	18668	__ret = __rev0 - __rev1 * (uint16x4_t) {__p2, __p2, __p2, __p2};
18648		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	18669	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
18649	18670	return __ret;
18650	18671	}
18651	18672	#endif
...	...	@@ -18659,10 +18680,10 @@ __ai __attribute__((target("neon"))) float32x2_t vmls_n_f32(float32x2_t __p0, fl
18659	18680	#else
18660	18681	__ai __attribute__((target("neon"))) float32x2_t vmls_n_f32(float32x2_t __p0, float32x2_t __p1, float32_t __p2) {
18661	18682	float32x2_t __ret;
18662		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
18663		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	18683	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	18684	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
18664	18685	__ret = __rev0 - __rev1 * (float32x2_t) {__p2, __p2};
18665		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	18686	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
18666	18687	return __ret;
18667	18688	}
18668	18689	#endif
...	...	@@ -18676,10 +18697,10 @@ __ai __attribute__((target("neon"))) int32x2_t vmls_n_s32(int32x2_t __p0, int32x
18676	18697	#else
18677	18698	__ai __attribute__((target("neon"))) int32x2_t vmls_n_s32(int32x2_t __p0, int32x2_t __p1, int32_t __p2) {
18678	18699	int32x2_t __ret;
18679		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
18680		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	18700	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	18701	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
18681	18702	__ret = __rev0 - __rev1 * (int32x2_t) {__p2, __p2};
18682		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	18703	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
18683	18704	return __ret;
18684	18705	}
18685	18706	#endif
...	...	@@ -18693,10 +18714,10 @@ __ai __attribute__((target("neon"))) int16x4_t vmls_n_s16(int16x4_t __p0, int16x
18693	18714	#else
18694	18715	__ai __attribute__((target("neon"))) int16x4_t vmls_n_s16(int16x4_t __p0, int16x4_t __p1, int16_t __p2) {
18695	18716	int16x4_t __ret;
18696		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
18697		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	18717	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	18718	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
18698	18719	__ret = __rev0 - __rev1 * (int16x4_t) {__p2, __p2, __p2, __p2};
18699		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	18720	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
18700	18721	return __ret;
18701	18722	}
18702	18723	#endif
...	...	@@ -18711,7 +18732,7 @@ __ai __attribute__((target("neon"))) poly8x8_t vmov_n_p8(poly8_t __p0) {
18711	18732	__ai __attribute__((target("neon"))) poly8x8_t vmov_n_p8(poly8_t __p0) {
18712	18733	poly8x8_t __ret;
18713	18734	__ret = (poly8x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
18714		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	18735	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
18715	18736	return __ret;
18716	18737	}
18717	18738	#endif
...	...	@@ -18726,7 +18747,7 @@ __ai __attribute__((target("neon"))) poly16x4_t vmov_n_p16(poly16_t __p0) {
18726	18747	__ai __attribute__((target("neon"))) poly16x4_t vmov_n_p16(poly16_t __p0) {
18727	18748	poly16x4_t __ret;
18728	18749	__ret = (poly16x4_t) {__p0, __p0, __p0, __p0};
18729		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	18750	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
18730	18751	return __ret;
18731	18752	}
18732	18753	#endif
...	...	@@ -18741,7 +18762,7 @@ __ai __attribute__((target("neon"))) poly8x16_t vmovq_n_p8(poly8_t __p0) {
18741	18762	__ai __attribute__((target("neon"))) poly8x16_t vmovq_n_p8(poly8_t __p0) {
18742	18763	poly8x16_t __ret;
18743	18764	__ret = (poly8x16_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
18744		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	18765	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
18745	18766	return __ret;
18746	18767	}
18747	18768	#endif
...	...	@@ -18756,7 +18777,7 @@ __ai __attribute__((target("neon"))) poly16x8_t vmovq_n_p16(poly16_t __p0) {
18756	18777	__ai __attribute__((target("neon"))) poly16x8_t vmovq_n_p16(poly16_t __p0) {
18757	18778	poly16x8_t __ret;
18758	18779	__ret = (poly16x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
18759		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	18780	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
18760	18781	return __ret;
18761	18782	}
18762	18783	#endif
...	...	@@ -18771,7 +18792,7 @@ __ai __attribute__((target("neon"))) uint8x16_t vmovq_n_u8(uint8_t __p0) {
18771	18792	__ai __attribute__((target("neon"))) uint8x16_t vmovq_n_u8(uint8_t __p0) {
18772	18793	uint8x16_t __ret;
18773	18794	__ret = (uint8x16_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
18774		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	18795	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
18775	18796	return __ret;
18776	18797	}
18777	18798	#endif
...	...	@@ -18786,7 +18807,7 @@ __ai __attribute__((target("neon"))) uint32x4_t vmovq_n_u32(uint32_t __p0) {
18786	18807	__ai __attribute__((target("neon"))) uint32x4_t vmovq_n_u32(uint32_t __p0) {
18787	18808	uint32x4_t __ret;
18788	18809	__ret = (uint32x4_t) {__p0, __p0, __p0, __p0};
18789		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	18810	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
18790	18811	return __ret;
18791	18812	}
18792	18813	#endif
...	...	@@ -18801,7 +18822,7 @@ __ai __attribute__((target("neon"))) uint64x2_t vmovq_n_u64(uint64_t __p0) {
18801	18822	__ai __attribute__((target("neon"))) uint64x2_t vmovq_n_u64(uint64_t __p0) {
18802	18823	uint64x2_t __ret;
18803	18824	__ret = (uint64x2_t) {__p0, __p0};
18804		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	18825	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
18805	18826	return __ret;
18806	18827	}
18807	18828	#endif
...	...	@@ -18816,7 +18837,7 @@ __ai __attribute__((target("neon"))) uint16x8_t vmovq_n_u16(uint16_t __p0) {
18816	18837	__ai __attribute__((target("neon"))) uint16x8_t vmovq_n_u16(uint16_t __p0) {
18817	18838	uint16x8_t __ret;
18818	18839	__ret = (uint16x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
18819		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	18840	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
18820	18841	return __ret;
18821	18842	}
18822	18843	#endif
...	...	@@ -18831,7 +18852,7 @@ __ai __attribute__((target("neon"))) int8x16_t vmovq_n_s8(int8_t __p0) {
18831	18852	__ai __attribute__((target("neon"))) int8x16_t vmovq_n_s8(int8_t __p0) {
18832	18853	int8x16_t __ret;
18833	18854	__ret = (int8x16_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
18834		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	18855	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
18835	18856	return __ret;
18836	18857	}
18837	18858	#endif
...	...	@@ -18846,7 +18867,7 @@ __ai __attribute__((target("neon"))) float32x4_t vmovq_n_f32(float32_t __p0) {
18846	18867	__ai __attribute__((target("neon"))) float32x4_t vmovq_n_f32(float32_t __p0) {
18847	18868	float32x4_t __ret;
18848	18869	__ret = (float32x4_t) {__p0, __p0, __p0, __p0};
18849		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	18870	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
18850	18871	return __ret;
18851	18872	}
18852	18873	#endif
...	...	@@ -18863,7 +18884,7 @@ __ai __attribute__((target("neon"))) float32x4_t vmovq_n_f32(float32_t __p0) {
18863	18884	float16x8_t __ret; \
18864	18885	float16_t __s0 = __p0; \
18865	18886	__ret = (float16x8_t) {__s0, __s0, __s0, __s0, __s0, __s0, __s0, __s0}; \
18866		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	18887	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
18867	18888	__ret; \
18868	18889	})
18869	18890	#endif
...	...	@@ -18878,7 +18899,7 @@ __ai __attribute__((target("neon"))) int32x4_t vmovq_n_s32(int32_t __p0) {
18878	18899	__ai __attribute__((target("neon"))) int32x4_t vmovq_n_s32(int32_t __p0) {
18879	18900	int32x4_t __ret;
18880	18901	__ret = (int32x4_t) {__p0, __p0, __p0, __p0};
18881		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	18902	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
18882	18903	return __ret;
18883	18904	}
18884	18905	#endif
...	...	@@ -18893,7 +18914,7 @@ __ai __attribute__((target("neon"))) int64x2_t vmovq_n_s64(int64_t __p0) {
18893	18914	__ai __attribute__((target("neon"))) int64x2_t vmovq_n_s64(int64_t __p0) {
18894	18915	int64x2_t __ret;
18895	18916	__ret = (int64x2_t) {__p0, __p0};
18896		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	18917	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
18897	18918	return __ret;
18898	18919	}
18899	18920	#endif
...	...	@@ -18908,7 +18929,7 @@ __ai __attribute__((target("neon"))) int16x8_t vmovq_n_s16(int16_t __p0) {
18908	18929	__ai __attribute__((target("neon"))) int16x8_t vmovq_n_s16(int16_t __p0) {
18909	18930	int16x8_t __ret;
18910	18931	__ret = (int16x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
18911		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	18932	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
18912	18933	return __ret;
18913	18934	}
18914	18935	#endif
...	...	@@ -18923,7 +18944,7 @@ __ai __attribute__((target("neon"))) uint8x8_t vmov_n_u8(uint8_t __p0) {
18923	18944	__ai __attribute__((target("neon"))) uint8x8_t vmov_n_u8(uint8_t __p0) {
18924	18945	uint8x8_t __ret;
18925	18946	__ret = (uint8x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
18926		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	18947	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
18927	18948	return __ret;
18928	18949	}
18929	18950	#endif
...	...	@@ -18938,7 +18959,7 @@ __ai __attribute__((target("neon"))) uint32x2_t vmov_n_u32(uint32_t __p0) {
18938	18959	__ai __attribute__((target("neon"))) uint32x2_t vmov_n_u32(uint32_t __p0) {
18939	18960	uint32x2_t __ret;
18940	18961	__ret = (uint32x2_t) {__p0, __p0};
18941		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	18962	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
18942	18963	return __ret;
18943	18964	}
18944	18965	#endif
...	...	@@ -18958,7 +18979,7 @@ __ai __attribute__((target("neon"))) uint16x4_t vmov_n_u16(uint16_t __p0) {
18958	18979	__ai __attribute__((target("neon"))) uint16x4_t vmov_n_u16(uint16_t __p0) {
18959	18980	uint16x4_t __ret;
18960	18981	__ret = (uint16x4_t) {__p0, __p0, __p0, __p0};
18961		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	18982	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
18962	18983	return __ret;
18963	18984	}
18964	18985	#endif
...	...	@@ -18973,7 +18994,7 @@ __ai __attribute__((target("neon"))) int8x8_t vmov_n_s8(int8_t __p0) {
18973	18994	__ai __attribute__((target("neon"))) int8x8_t vmov_n_s8(int8_t __p0) {
18974	18995	int8x8_t __ret;
18975	18996	__ret = (int8x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
18976		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	18997	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
18977	18998	return __ret;
18978	18999	}
18979	19000	#endif
...	...	@@ -18988,7 +19009,7 @@ __ai __attribute__((target("neon"))) float32x2_t vmov_n_f32(float32_t __p0) {
18988	19009	__ai __attribute__((target("neon"))) float32x2_t vmov_n_f32(float32_t __p0) {
18989	19010	float32x2_t __ret;
18990	19011	__ret = (float32x2_t) {__p0, __p0};
18991		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	19012	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
18992	19013	return __ret;
18993	19014	}
18994	19015	#endif
...	...	@@ -19005,7 +19026,7 @@ __ai __attribute__((target("neon"))) float32x2_t vmov_n_f32(float32_t __p0) {
19005	19026	float16x4_t __ret; \
19006	19027	float16_t __s0 = __p0; \
19007	19028	__ret = (float16x4_t) {__s0, __s0, __s0, __s0}; \
19008		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	19029	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
19009	19030	__ret; \
19010	19031	})
19011	19032	#endif
...	...	@@ -19020,7 +19041,7 @@ __ai __attribute__((target("neon"))) int32x2_t vmov_n_s32(int32_t __p0) {
19020	19041	__ai __attribute__((target("neon"))) int32x2_t vmov_n_s32(int32_t __p0) {
19021	19042	int32x2_t __ret;
19022	19043	__ret = (int32x2_t) {__p0, __p0};
19023		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	19044	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
19024	19045	return __ret;
19025	19046	}
19026	19047	#endif
...	...	@@ -19040,7 +19061,7 @@ __ai __attribute__((target("neon"))) int16x4_t vmov_n_s16(int16_t __p0) {
19040	19061	__ai __attribute__((target("neon"))) int16x4_t vmov_n_s16(int16_t __p0) {
19041	19062	int16x4_t __ret;
19042	19063	__ret = (int16x4_t) {__p0, __p0, __p0, __p0};
19043		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	19064	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
19044	19065	return __ret;
19045	19066	}
19046	19067	#endif
...	...	@@ -19048,20 +19069,20 @@ __ai __attribute__((target("neon"))) int16x4_t vmov_n_s16(int16_t __p0) {
19048	19069	#ifdef __LITTLE_ENDIAN__
19049	19070	__ai __attribute__((target("neon"))) uint16x8_t vmovl_u8(uint8x8_t __p0) {
19050	19071	uint16x8_t __ret;
19051		__ret = (uint16x8_t) __builtin_neon_vmovl_v((int8x8_t)__p0, 49);
	19072	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vmovl_v(__builtin_bit_cast(int8x8_t, __p0), 49));
19052	19073	return __ret;
19053	19074	}
19054	19075	#else
19055	19076	__ai __attribute__((target("neon"))) uint16x8_t vmovl_u8(uint8x8_t __p0) {
19056	19077	uint16x8_t __ret;
19057		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
19058		__ret = (uint16x8_t) __builtin_neon_vmovl_v((int8x8_t)__rev0, 49);
19059		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	19078	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	19079	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vmovl_v(__builtin_bit_cast(int8x8_t, __rev0), 49));
	19080	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
19060	19081	return __ret;
19061	19082	}
19062	19083	__ai __attribute__((target("neon"))) uint16x8_t __noswap_vmovl_u8(uint8x8_t __p0) {
19063	19084	uint16x8_t __ret;
19064		__ret = (uint16x8_t) __builtin_neon_vmovl_v((int8x8_t)__p0, 49);
	19085	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vmovl_v(__builtin_bit_cast(int8x8_t, __p0), 49));
19065	19086	return __ret;
19066	19087	}
19067	19088	#endif
...	...	@@ -19069,20 +19090,20 @@ __ai __attribute__((target("neon"))) uint16x8_t __noswap_vmovl_u8(uint8x8_t __p0
19069	19090	#ifdef __LITTLE_ENDIAN__
19070	19091	__ai __attribute__((target("neon"))) uint64x2_t vmovl_u32(uint32x2_t __p0) {
19071	19092	uint64x2_t __ret;
19072		__ret = (uint64x2_t) __builtin_neon_vmovl_v((int8x8_t)__p0, 51);
	19093	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vmovl_v(__builtin_bit_cast(int8x8_t, __p0), 51));
19073	19094	return __ret;
19074	19095	}
19075	19096	#else
19076	19097	__ai __attribute__((target("neon"))) uint64x2_t vmovl_u32(uint32x2_t __p0) {
19077	19098	uint64x2_t __ret;
19078		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
19079		__ret = (uint64x2_t) __builtin_neon_vmovl_v((int8x8_t)__rev0, 51);
19080		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	19099	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	19100	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vmovl_v(__builtin_bit_cast(int8x8_t, __rev0), 51));
	19101	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
19081	19102	return __ret;
19082	19103	}
19083	19104	__ai __attribute__((target("neon"))) uint64x2_t __noswap_vmovl_u32(uint32x2_t __p0) {
19084	19105	uint64x2_t __ret;
19085		__ret = (uint64x2_t) __builtin_neon_vmovl_v((int8x8_t)__p0, 51);
	19106	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vmovl_v(__builtin_bit_cast(int8x8_t, __p0), 51));
19086	19107	return __ret;
19087	19108	}
19088	19109	#endif
...	...	@@ -19090,20 +19111,20 @@ __ai __attribute__((target("neon"))) uint64x2_t __noswap_vmovl_u32(uint32x2_t __
19090	19111	#ifdef __LITTLE_ENDIAN__
19091	19112	__ai __attribute__((target("neon"))) uint32x4_t vmovl_u16(uint16x4_t __p0) {
19092	19113	uint32x4_t __ret;
19093		__ret = (uint32x4_t) __builtin_neon_vmovl_v((int8x8_t)__p0, 50);
	19114	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vmovl_v(__builtin_bit_cast(int8x8_t, __p0), 50));
19094	19115	return __ret;
19095	19116	}
19096	19117	#else
19097	19118	__ai __attribute__((target("neon"))) uint32x4_t vmovl_u16(uint16x4_t __p0) {
19098	19119	uint32x4_t __ret;
19099		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
19100		__ret = (uint32x4_t) __builtin_neon_vmovl_v((int8x8_t)__rev0, 50);
19101		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	19120	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	19121	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vmovl_v(__builtin_bit_cast(int8x8_t, __rev0), 50));
	19122	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
19102	19123	return __ret;
19103	19124	}
19104	19125	__ai __attribute__((target("neon"))) uint32x4_t __noswap_vmovl_u16(uint16x4_t __p0) {
19105	19126	uint32x4_t __ret;
19106		__ret = (uint32x4_t) __builtin_neon_vmovl_v((int8x8_t)__p0, 50);
	19127	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vmovl_v(__builtin_bit_cast(int8x8_t, __p0), 50));
19107	19128	return __ret;
19108	19129	}
19109	19130	#endif
...	...	@@ -19111,20 +19132,20 @@ __ai __attribute__((target("neon"))) uint32x4_t __noswap_vmovl_u16(uint16x4_t __
19111	19132	#ifdef __LITTLE_ENDIAN__
19112	19133	__ai __attribute__((target("neon"))) int16x8_t vmovl_s8(int8x8_t __p0) {
19113	19134	int16x8_t __ret;
19114		__ret = (int16x8_t) __builtin_neon_vmovl_v((int8x8_t)__p0, 33);
	19135	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vmovl_v(__builtin_bit_cast(int8x8_t, __p0), 33));
19115	19136	return __ret;
19116	19137	}
19117	19138	#else
19118	19139	__ai __attribute__((target("neon"))) int16x8_t vmovl_s8(int8x8_t __p0) {
19119	19140	int16x8_t __ret;
19120		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
19121		__ret = (int16x8_t) __builtin_neon_vmovl_v((int8x8_t)__rev0, 33);
19122		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	19141	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	19142	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vmovl_v(__builtin_bit_cast(int8x8_t, __rev0), 33));
	19143	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
19123	19144	return __ret;
19124	19145	}
19125	19146	__ai __attribute__((target("neon"))) int16x8_t __noswap_vmovl_s8(int8x8_t __p0) {
19126	19147	int16x8_t __ret;
19127		__ret = (int16x8_t) __builtin_neon_vmovl_v((int8x8_t)__p0, 33);
	19148	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vmovl_v(__builtin_bit_cast(int8x8_t, __p0), 33));
19128	19149	return __ret;
19129	19150	}
19130	19151	#endif
...	...	@@ -19132,20 +19153,20 @@ __ai __attribute__((target("neon"))) int16x8_t __noswap_vmovl_s8(int8x8_t __p0)
19132	19153	#ifdef __LITTLE_ENDIAN__
19133	19154	__ai __attribute__((target("neon"))) int64x2_t vmovl_s32(int32x2_t __p0) {
19134	19155	int64x2_t __ret;
19135		__ret = (int64x2_t) __builtin_neon_vmovl_v((int8x8_t)__p0, 35);
	19156	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vmovl_v(__builtin_bit_cast(int8x8_t, __p0), 35));
19136	19157	return __ret;
19137	19158	}
19138	19159	#else
19139	19160	__ai __attribute__((target("neon"))) int64x2_t vmovl_s32(int32x2_t __p0) {
19140	19161	int64x2_t __ret;
19141		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
19142		__ret = (int64x2_t) __builtin_neon_vmovl_v((int8x8_t)__rev0, 35);
19143		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	19162	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	19163	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vmovl_v(__builtin_bit_cast(int8x8_t, __rev0), 35));
	19164	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
19144	19165	return __ret;
19145	19166	}
19146	19167	__ai __attribute__((target("neon"))) int64x2_t __noswap_vmovl_s32(int32x2_t __p0) {
19147	19168	int64x2_t __ret;
19148		__ret = (int64x2_t) __builtin_neon_vmovl_v((int8x8_t)__p0, 35);
	19169	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vmovl_v(__builtin_bit_cast(int8x8_t, __p0), 35));
19149	19170	return __ret;
19150	19171	}
19151	19172	#endif
...	...	@@ -19153,20 +19174,20 @@ __ai __attribute__((target("neon"))) int64x2_t __noswap_vmovl_s32(int32x2_t __p0
19153	19174	#ifdef __LITTLE_ENDIAN__
19154	19175	__ai __attribute__((target("neon"))) int32x4_t vmovl_s16(int16x4_t __p0) {
19155	19176	int32x4_t __ret;
19156		__ret = (int32x4_t) __builtin_neon_vmovl_v((int8x8_t)__p0, 34);
	19177	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vmovl_v(__builtin_bit_cast(int8x8_t, __p0), 34));
19157	19178	return __ret;
19158	19179	}
19159	19180	#else
19160	19181	__ai __attribute__((target("neon"))) int32x4_t vmovl_s16(int16x4_t __p0) {
19161	19182	int32x4_t __ret;
19162		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
19163		__ret = (int32x4_t) __builtin_neon_vmovl_v((int8x8_t)__rev0, 34);
19164		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	19183	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	19184	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vmovl_v(__builtin_bit_cast(int8x8_t, __rev0), 34));
	19185	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
19165	19186	return __ret;
19166	19187	}
19167	19188	__ai __attribute__((target("neon"))) int32x4_t __noswap_vmovl_s16(int16x4_t __p0) {
19168	19189	int32x4_t __ret;
19169		__ret = (int32x4_t) __builtin_neon_vmovl_v((int8x8_t)__p0, 34);
	19190	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vmovl_v(__builtin_bit_cast(int8x8_t, __p0), 34));
19170	19191	return __ret;
19171	19192	}
19172	19193	#endif
...	...	@@ -19174,20 +19195,20 @@ __ai __attribute__((target("neon"))) int32x4_t __noswap_vmovl_s16(int16x4_t __p0
19174	19195	#ifdef __LITTLE_ENDIAN__
19175	19196	__ai __attribute__((target("neon"))) uint16x4_t vmovn_u32(uint32x4_t __p0) {
19176	19197	uint16x4_t __ret;
19177		__ret = (uint16x4_t) __builtin_neon_vmovn_v((int8x16_t)__p0, 17);
	19198	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vmovn_v(__builtin_bit_cast(int8x16_t, __p0), 17));
19178	19199	return __ret;
19179	19200	}
19180	19201	#else
19181	19202	__ai __attribute__((target("neon"))) uint16x4_t vmovn_u32(uint32x4_t __p0) {
19182	19203	uint16x4_t __ret;
19183		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
19184		__ret = (uint16x4_t) __builtin_neon_vmovn_v((int8x16_t)__rev0, 17);
19185		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	19204	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	19205	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vmovn_v(__builtin_bit_cast(int8x16_t, __rev0), 17));
	19206	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
19186	19207	return __ret;
19187	19208	}
19188	19209	__ai __attribute__((target("neon"))) uint16x4_t __noswap_vmovn_u32(uint32x4_t __p0) {
19189	19210	uint16x4_t __ret;
19190		__ret = (uint16x4_t) __builtin_neon_vmovn_v((int8x16_t)__p0, 17);
	19211	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vmovn_v(__builtin_bit_cast(int8x16_t, __p0), 17));
19191	19212	return __ret;
19192	19213	}
19193	19214	#endif
...	...	@@ -19195,20 +19216,20 @@ __ai __attribute__((target("neon"))) uint16x4_t __noswap_vmovn_u32(uint32x4_t __
19195	19216	#ifdef __LITTLE_ENDIAN__
19196	19217	__ai __attribute__((target("neon"))) uint32x2_t vmovn_u64(uint64x2_t __p0) {
19197	19218	uint32x2_t __ret;
19198		__ret = (uint32x2_t) __builtin_neon_vmovn_v((int8x16_t)__p0, 18);
	19219	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vmovn_v(__builtin_bit_cast(int8x16_t, __p0), 18));
19199	19220	return __ret;
19200	19221	}
19201	19222	#else
19202	19223	__ai __attribute__((target("neon"))) uint32x2_t vmovn_u64(uint64x2_t __p0) {
19203	19224	uint32x2_t __ret;
19204		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
19205		__ret = (uint32x2_t) __builtin_neon_vmovn_v((int8x16_t)__rev0, 18);
19206		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	19225	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	19226	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vmovn_v(__builtin_bit_cast(int8x16_t, __rev0), 18));
	19227	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
19207	19228	return __ret;
19208	19229	}
19209	19230	__ai __attribute__((target("neon"))) uint32x2_t __noswap_vmovn_u64(uint64x2_t __p0) {
19210	19231	uint32x2_t __ret;
19211		__ret = (uint32x2_t) __builtin_neon_vmovn_v((int8x16_t)__p0, 18);
	19232	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vmovn_v(__builtin_bit_cast(int8x16_t, __p0), 18));
19212	19233	return __ret;
19213	19234	}
19214	19235	#endif
...	...	@@ -19216,20 +19237,20 @@ __ai __attribute__((target("neon"))) uint32x2_t __noswap_vmovn_u64(uint64x2_t __
19216	19237	#ifdef __LITTLE_ENDIAN__
19217	19238	__ai __attribute__((target("neon"))) uint8x8_t vmovn_u16(uint16x8_t __p0) {
19218	19239	uint8x8_t __ret;
19219		__ret = (uint8x8_t) __builtin_neon_vmovn_v((int8x16_t)__p0, 16);
	19240	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vmovn_v(__builtin_bit_cast(int8x16_t, __p0), 16));
19220	19241	return __ret;
19221	19242	}
19222	19243	#else
19223	19244	__ai __attribute__((target("neon"))) uint8x8_t vmovn_u16(uint16x8_t __p0) {
19224	19245	uint8x8_t __ret;
19225		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
19226		__ret = (uint8x8_t) __builtin_neon_vmovn_v((int8x16_t)__rev0, 16);
19227		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	19246	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	19247	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vmovn_v(__builtin_bit_cast(int8x16_t, __rev0), 16));
	19248	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
19228	19249	return __ret;
19229	19250	}
19230	19251	__ai __attribute__((target("neon"))) uint8x8_t __noswap_vmovn_u16(uint16x8_t __p0) {
19231	19252	uint8x8_t __ret;
19232		__ret = (uint8x8_t) __builtin_neon_vmovn_v((int8x16_t)__p0, 16);
	19253	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vmovn_v(__builtin_bit_cast(int8x16_t, __p0), 16));
19233	19254	return __ret;
19234	19255	}
19235	19256	#endif
...	...	@@ -19237,20 +19258,20 @@ __ai __attribute__((target("neon"))) uint8x8_t __noswap_vmovn_u16(uint16x8_t __p
19237	19258	#ifdef __LITTLE_ENDIAN__
19238	19259	__ai __attribute__((target("neon"))) int16x4_t vmovn_s32(int32x4_t __p0) {
19239	19260	int16x4_t __ret;
19240		__ret = (int16x4_t) __builtin_neon_vmovn_v((int8x16_t)__p0, 1);
	19261	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vmovn_v(__builtin_bit_cast(int8x16_t, __p0), 1));
19241	19262	return __ret;
19242	19263	}
19243	19264	#else
19244	19265	__ai __attribute__((target("neon"))) int16x4_t vmovn_s32(int32x4_t __p0) {
19245	19266	int16x4_t __ret;
19246		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
19247		__ret = (int16x4_t) __builtin_neon_vmovn_v((int8x16_t)__rev0, 1);
19248		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	19267	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	19268	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vmovn_v(__builtin_bit_cast(int8x16_t, __rev0), 1));
	19269	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
19249	19270	return __ret;
19250	19271	}
19251	19272	__ai __attribute__((target("neon"))) int16x4_t __noswap_vmovn_s32(int32x4_t __p0) {
19252	19273	int16x4_t __ret;
19253		__ret = (int16x4_t) __builtin_neon_vmovn_v((int8x16_t)__p0, 1);
	19274	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vmovn_v(__builtin_bit_cast(int8x16_t, __p0), 1));
19254	19275	return __ret;
19255	19276	}
19256	19277	#endif
...	...	@@ -19258,20 +19279,20 @@ __ai __attribute__((target("neon"))) int16x4_t __noswap_vmovn_s32(int32x4_t __p0
19258	19279	#ifdef __LITTLE_ENDIAN__
19259	19280	__ai __attribute__((target("neon"))) int32x2_t vmovn_s64(int64x2_t __p0) {
19260	19281	int32x2_t __ret;
19261		__ret = (int32x2_t) __builtin_neon_vmovn_v((int8x16_t)__p0, 2);
	19282	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vmovn_v(__builtin_bit_cast(int8x16_t, __p0), 2));
19262	19283	return __ret;
19263	19284	}
19264	19285	#else
19265	19286	__ai __attribute__((target("neon"))) int32x2_t vmovn_s64(int64x2_t __p0) {
19266	19287	int32x2_t __ret;
19267		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
19268		__ret = (int32x2_t) __builtin_neon_vmovn_v((int8x16_t)__rev0, 2);
19269		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	19288	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	19289	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vmovn_v(__builtin_bit_cast(int8x16_t, __rev0), 2));
	19290	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
19270	19291	return __ret;
19271	19292	}
19272	19293	__ai __attribute__((target("neon"))) int32x2_t __noswap_vmovn_s64(int64x2_t __p0) {
19273	19294	int32x2_t __ret;
19274		__ret = (int32x2_t) __builtin_neon_vmovn_v((int8x16_t)__p0, 2);
	19295	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vmovn_v(__builtin_bit_cast(int8x16_t, __p0), 2));
19275	19296	return __ret;
19276	19297	}
19277	19298	#endif
...	...	@@ -19279,20 +19300,20 @@ __ai __attribute__((target("neon"))) int32x2_t __noswap_vmovn_s64(int64x2_t __p0
19279	19300	#ifdef __LITTLE_ENDIAN__
19280	19301	__ai __attribute__((target("neon"))) int8x8_t vmovn_s16(int16x8_t __p0) {
19281	19302	int8x8_t __ret;
19282		__ret = (int8x8_t) __builtin_neon_vmovn_v((int8x16_t)__p0, 0);
	19303	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vmovn_v(__builtin_bit_cast(int8x16_t, __p0), 0));
19283	19304	return __ret;
19284	19305	}
19285	19306	#else
19286	19307	__ai __attribute__((target("neon"))) int8x8_t vmovn_s16(int16x8_t __p0) {
19287	19308	int8x8_t __ret;
19288		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
19289		__ret = (int8x8_t) __builtin_neon_vmovn_v((int8x16_t)__rev0, 0);
19290		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	19309	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	19310	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vmovn_v(__builtin_bit_cast(int8x16_t, __rev0), 0));
	19311	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
19291	19312	return __ret;
19292	19313	}
19293	19314	__ai __attribute__((target("neon"))) int8x8_t __noswap_vmovn_s16(int16x8_t __p0) {
19294	19315	int8x8_t __ret;
19295		__ret = (int8x8_t) __builtin_neon_vmovn_v((int8x16_t)__p0, 0);
	19316	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vmovn_v(__builtin_bit_cast(int8x16_t, __p0), 0));
19296	19317	return __ret;
19297	19318	}
19298	19319	#endif
...	...	@@ -19306,10 +19327,10 @@ __ai __attribute__((target("neon"))) uint8x16_t vmulq_u8(uint8x16_t __p0, uint8x
19306	19327	#else
19307	19328	__ai __attribute__((target("neon"))) uint8x16_t vmulq_u8(uint8x16_t __p0, uint8x16_t __p1) {
19308	19329	uint8x16_t __ret;
19309		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
19310		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	19330	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	19331	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
19311	19332	__ret = __rev0 * __rev1;
19312		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	19333	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
19313	19334	return __ret;
19314	19335	}
19315	19336	#endif
...	...	@@ -19323,10 +19344,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vmulq_u32(uint32x4_t __p0, uint3
19323	19344	#else
19324	19345	__ai __attribute__((target("neon"))) uint32x4_t vmulq_u32(uint32x4_t __p0, uint32x4_t __p1) {
19325	19346	uint32x4_t __ret;
19326		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
19327		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	19347	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	19348	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
19328	19349	__ret = __rev0 * __rev1;
19329		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	19350	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
19330	19351	return __ret;
19331	19352	}
19332	19353	#endif
...	...	@@ -19340,10 +19361,10 @@ __ai __attribute__((target("neon"))) uint16x8_t vmulq_u16(uint16x8_t __p0, uint1
19340	19361	#else
19341	19362	__ai __attribute__((target("neon"))) uint16x8_t vmulq_u16(uint16x8_t __p0, uint16x8_t __p1) {
19342	19363	uint16x8_t __ret;
19343		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
19344		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	19364	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	19365	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
19345	19366	__ret = __rev0 * __rev1;
19346		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	19367	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
19347	19368	return __ret;
19348	19369	}
19349	19370	#endif
...	...	@@ -19357,10 +19378,10 @@ __ai __attribute__((target("neon"))) int8x16_t vmulq_s8(int8x16_t __p0, int8x16_
19357	19378	#else
19358	19379	__ai __attribute__((target("neon"))) int8x16_t vmulq_s8(int8x16_t __p0, int8x16_t __p1) {
19359	19380	int8x16_t __ret;
19360		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
19361		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	19381	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	19382	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
19362	19383	__ret = __rev0 * __rev1;
19363		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	19384	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
19364	19385	return __ret;
19365	19386	}
19366	19387	#endif
...	...	@@ -19374,10 +19395,10 @@ __ai __attribute__((target("neon"))) float32x4_t vmulq_f32(float32x4_t __p0, flo
19374	19395	#else
19375	19396	__ai __attribute__((target("neon"))) float32x4_t vmulq_f32(float32x4_t __p0, float32x4_t __p1) {
19376	19397	float32x4_t __ret;
19377		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
19378		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	19398	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	19399	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
19379	19400	__ret = __rev0 * __rev1;
19380		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	19401	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
19381	19402	return __ret;
19382	19403	}
19383	19404	#endif
...	...	@@ -19391,10 +19412,10 @@ __ai __attribute__((target("neon"))) int32x4_t vmulq_s32(int32x4_t __p0, int32x4
19391	19412	#else
19392	19413	__ai __attribute__((target("neon"))) int32x4_t vmulq_s32(int32x4_t __p0, int32x4_t __p1) {
19393	19414	int32x4_t __ret;
19394		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
19395		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	19415	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	19416	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
19396	19417	__ret = __rev0 * __rev1;
19397		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	19418	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
19398	19419	return __ret;
19399	19420	}
19400	19421	#endif
...	...	@@ -19408,10 +19429,10 @@ __ai __attribute__((target("neon"))) int16x8_t vmulq_s16(int16x8_t __p0, int16x8
19408	19429	#else
19409	19430	__ai __attribute__((target("neon"))) int16x8_t vmulq_s16(int16x8_t __p0, int16x8_t __p1) {
19410	19431	int16x8_t __ret;
19411		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
19412		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	19432	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	19433	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
19413	19434	__ret = __rev0 * __rev1;
19414		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	19435	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
19415	19436	return __ret;
19416	19437	}
19417	19438	#endif
...	...	@@ -19425,10 +19446,10 @@ __ai __attribute__((target("neon"))) uint8x8_t vmul_u8(uint8x8_t __p0, uint8x8_t
19425	19446	#else
19426	19447	__ai __attribute__((target("neon"))) uint8x8_t vmul_u8(uint8x8_t __p0, uint8x8_t __p1) {
19427	19448	uint8x8_t __ret;
19428		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
19429		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	19449	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	19450	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
19430	19451	__ret = __rev0 * __rev1;
19431		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	19452	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
19432	19453	return __ret;
19433	19454	}
19434	19455	#endif
...	...	@@ -19442,10 +19463,10 @@ __ai __attribute__((target("neon"))) uint32x2_t vmul_u32(uint32x2_t __p0, uint32
19442	19463	#else
19443	19464	__ai __attribute__((target("neon"))) uint32x2_t vmul_u32(uint32x2_t __p0, uint32x2_t __p1) {
19444	19465	uint32x2_t __ret;
19445		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
19446		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	19466	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	19467	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
19447	19468	__ret = __rev0 * __rev1;
19448		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	19469	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
19449	19470	return __ret;
19450	19471	}
19451	19472	#endif
...	...	@@ -19459,10 +19480,10 @@ __ai __attribute__((target("neon"))) uint16x4_t vmul_u16(uint16x4_t __p0, uint16
19459	19480	#else
19460	19481	__ai __attribute__((target("neon"))) uint16x4_t vmul_u16(uint16x4_t __p0, uint16x4_t __p1) {
19461	19482	uint16x4_t __ret;
19462		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
19463		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	19483	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	19484	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
19464	19485	__ret = __rev0 * __rev1;
19465		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	19486	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
19466	19487	return __ret;
19467	19488	}
19468	19489	#endif
...	...	@@ -19476,10 +19497,10 @@ __ai __attribute__((target("neon"))) int8x8_t vmul_s8(int8x8_t __p0, int8x8_t __
19476	19497	#else
19477	19498	__ai __attribute__((target("neon"))) int8x8_t vmul_s8(int8x8_t __p0, int8x8_t __p1) {
19478	19499	int8x8_t __ret;
19479		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
19480		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	19500	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	19501	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
19481	19502	__ret = __rev0 * __rev1;
19482		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	19503	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
19483	19504	return __ret;
19484	19505	}
19485	19506	#endif
...	...	@@ -19493,10 +19514,10 @@ __ai __attribute__((target("neon"))) float32x2_t vmul_f32(float32x2_t __p0, floa
19493	19514	#else
19494	19515	__ai __attribute__((target("neon"))) float32x2_t vmul_f32(float32x2_t __p0, float32x2_t __p1) {
19495	19516	float32x2_t __ret;
19496		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
19497		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	19517	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	19518	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
19498	19519	__ret = __rev0 * __rev1;
19499		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	19520	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
19500	19521	return __ret;
19501	19522	}
19502	19523	#endif
...	...	@@ -19510,10 +19531,10 @@ __ai __attribute__((target("neon"))) int32x2_t vmul_s32(int32x2_t __p0, int32x2_
19510	19531	#else
19511	19532	__ai __attribute__((target("neon"))) int32x2_t vmul_s32(int32x2_t __p0, int32x2_t __p1) {
19512	19533	int32x2_t __ret;
19513		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
19514		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	19534	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	19535	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
19515	19536	__ret = __rev0 * __rev1;
19516		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	19537	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
19517	19538	return __ret;
19518	19539	}
19519	19540	#endif
...	...	@@ -19527,10 +19548,10 @@ __ai __attribute__((target("neon"))) int16x4_t vmul_s16(int16x4_t __p0, int16x4_
19527	19548	#else
19528	19549	__ai __attribute__((target("neon"))) int16x4_t vmul_s16(int16x4_t __p0, int16x4_t __p1) {
19529	19550	int16x4_t __ret;
19530		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
19531		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	19551	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	19552	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
19532	19553	__ret = __rev0 * __rev1;
19533		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	19554	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
19534	19555	return __ret;
19535	19556	}
19536	19557	#endif
...	...	@@ -19538,16 +19559,16 @@ __ai __attribute__((target("neon"))) int16x4_t vmul_s16(int16x4_t __p0, int16x4_
19538	19559	#ifdef __LITTLE_ENDIAN__
19539	19560	__ai __attribute__((target("neon"))) poly8x8_t vmul_p8(poly8x8_t __p0, poly8x8_t __p1) {
19540	19561	poly8x8_t __ret;
19541		__ret = (poly8x8_t) __builtin_neon_vmul_v((int8x8_t)__p0, (int8x8_t)__p1, 4);
	19562	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vmul_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 4));
19542	19563	return __ret;
19543	19564	}
19544	19565	#else
19545	19566	__ai __attribute__((target("neon"))) poly8x8_t vmul_p8(poly8x8_t __p0, poly8x8_t __p1) {
19546	19567	poly8x8_t __ret;
19547		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
19548		poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
19549		__ret = (poly8x8_t) __builtin_neon_vmul_v((int8x8_t)__rev0, (int8x8_t)__rev1, 4);
19550		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	19568	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	19569	poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	19570	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vmul_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 4));
	19571	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
19551	19572	return __ret;
19552	19573	}
19553	19574	#endif
...	...	@@ -19555,16 +19576,16 @@ __ai __attribute__((target("neon"))) poly8x8_t vmul_p8(poly8x8_t __p0, poly8x8_t
19555	19576	#ifdef __LITTLE_ENDIAN__
19556	19577	__ai __attribute__((target("neon"))) poly8x16_t vmulq_p8(poly8x16_t __p0, poly8x16_t __p1) {
19557	19578	poly8x16_t __ret;
19558		__ret = (poly8x16_t) __builtin_neon_vmulq_v((int8x16_t)__p0, (int8x16_t)__p1, 36);
	19579	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vmulq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 36));
19559	19580	return __ret;
19560	19581	}
19561	19582	#else
19562	19583	__ai __attribute__((target("neon"))) poly8x16_t vmulq_p8(poly8x16_t __p0, poly8x16_t __p1) {
19563	19584	poly8x16_t __ret;
19564		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
19565		poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
19566		__ret = (poly8x16_t) __builtin_neon_vmulq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 36);
19567		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	19585	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	19586	poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	19587	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vmulq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 36));
	19588	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
19568	19589	return __ret;
19569	19590	}
19570	19591	#endif
...	...	@@ -19582,10 +19603,10 @@ __ai __attribute__((target("neon"))) poly8x16_t vmulq_p8(poly8x16_t __p0, poly8x
19582	19603	uint32x4_t __ret_95; \
19583	19604	uint32x4_t __s0_95 = __p0_95; \
19584	19605	uint32x2_t __s1_95 = __p1_95; \
19585		uint32x4_t __rev0_95; __rev0_95 = __builtin_shufflevector(__s0_95, __s0_95, 3, 2, 1, 0); \
19586		uint32x2_t __rev1_95; __rev1_95 = __builtin_shufflevector(__s1_95, __s1_95, 1, 0); \
	19606	uint32x4_t __rev0_95; __rev0_95 = __builtin_shufflevector(__s0_95, __s0_95, __lane_reverse_128_32); \
	19607	uint32x2_t __rev1_95; __rev1_95 = __builtin_shufflevector(__s1_95, __s1_95, __lane_reverse_64_32); \
19587	19608	__ret_95 = __rev0_95 * __noswap_splatq_lane_u32(__rev1_95, __p2_95); \
19588		__ret_95 = __builtin_shufflevector(__ret_95, __ret_95, 3, 2, 1, 0); \
	19609	__ret_95 = __builtin_shufflevector(__ret_95, __ret_95, __lane_reverse_128_32); \
19589	19610	__ret_95; \
19590	19611	})
19591	19612	#endif
...	...	@@ -19603,10 +19624,10 @@ __ai __attribute__((target("neon"))) poly8x16_t vmulq_p8(poly8x16_t __p0, poly8x
19603	19624	uint16x8_t __ret_97; \
19604	19625	uint16x8_t __s0_97 = __p0_97; \
19605	19626	uint16x4_t __s1_97 = __p1_97; \
19606		uint16x8_t __rev0_97; __rev0_97 = __builtin_shufflevector(__s0_97, __s0_97, 7, 6, 5, 4, 3, 2, 1, 0); \
19607		uint16x4_t __rev1_97; __rev1_97 = __builtin_shufflevector(__s1_97, __s1_97, 3, 2, 1, 0); \
	19627	uint16x8_t __rev0_97; __rev0_97 = __builtin_shufflevector(__s0_97, __s0_97, __lane_reverse_128_16); \
	19628	uint16x4_t __rev1_97; __rev1_97 = __builtin_shufflevector(__s1_97, __s1_97, __lane_reverse_64_16); \
19608	19629	__ret_97 = __rev0_97 * __noswap_splatq_lane_u16(__rev1_97, __p2_97); \
19609		__ret_97 = __builtin_shufflevector(__ret_97, __ret_97, 7, 6, 5, 4, 3, 2, 1, 0); \
	19630	__ret_97 = __builtin_shufflevector(__ret_97, __ret_97, __lane_reverse_128_16); \
19610	19631	__ret_97; \
19611	19632	})
19612	19633	#endif
...	...	@@ -19624,10 +19645,10 @@ __ai __attribute__((target("neon"))) poly8x16_t vmulq_p8(poly8x16_t __p0, poly8x
19624	19645	float32x4_t __ret_99; \
19625	19646	float32x4_t __s0_99 = __p0_99; \
19626	19647	float32x2_t __s1_99 = __p1_99; \
19627		float32x4_t __rev0_99; __rev0_99 = __builtin_shufflevector(__s0_99, __s0_99, 3, 2, 1, 0); \
19628		float32x2_t __rev1_99; __rev1_99 = __builtin_shufflevector(__s1_99, __s1_99, 1, 0); \
	19648	float32x4_t __rev0_99; __rev0_99 = __builtin_shufflevector(__s0_99, __s0_99, __lane_reverse_128_32); \
	19649	float32x2_t __rev1_99; __rev1_99 = __builtin_shufflevector(__s1_99, __s1_99, __lane_reverse_64_32); \
19629	19650	__ret_99 = __rev0_99 * __noswap_splatq_lane_f32(__rev1_99, __p2_99); \
19630		__ret_99 = __builtin_shufflevector(__ret_99, __ret_99, 3, 2, 1, 0); \
	19651	__ret_99 = __builtin_shufflevector(__ret_99, __ret_99, __lane_reverse_128_32); \
19631	19652	__ret_99; \
19632	19653	})
19633	19654	#endif
...	...	@@ -19645,10 +19666,10 @@ __ai __attribute__((target("neon"))) poly8x16_t vmulq_p8(poly8x16_t __p0, poly8x
19645	19666	int32x4_t __ret_101; \
19646	19667	int32x4_t __s0_101 = __p0_101; \
19647	19668	int32x2_t __s1_101 = __p1_101; \
19648		int32x4_t __rev0_101; __rev0_101 = __builtin_shufflevector(__s0_101, __s0_101, 3, 2, 1, 0); \
19649		int32x2_t __rev1_101; __rev1_101 = __builtin_shufflevector(__s1_101, __s1_101, 1, 0); \
	19669	int32x4_t __rev0_101; __rev0_101 = __builtin_shufflevector(__s0_101, __s0_101, __lane_reverse_128_32); \
	19670	int32x2_t __rev1_101; __rev1_101 = __builtin_shufflevector(__s1_101, __s1_101, __lane_reverse_64_32); \
19650	19671	__ret_101 = __rev0_101 * __noswap_splatq_lane_s32(__rev1_101, __p2_101); \
19651		__ret_101 = __builtin_shufflevector(__ret_101, __ret_101, 3, 2, 1, 0); \
	19672	__ret_101 = __builtin_shufflevector(__ret_101, __ret_101, __lane_reverse_128_32); \
19652	19673	__ret_101; \
19653	19674	})
19654	19675	#endif
...	...	@@ -19666,10 +19687,10 @@ __ai __attribute__((target("neon"))) poly8x16_t vmulq_p8(poly8x16_t __p0, poly8x
19666	19687	int16x8_t __ret_103; \
19667	19688	int16x8_t __s0_103 = __p0_103; \
19668	19689	int16x4_t __s1_103 = __p1_103; \
19669		int16x8_t __rev0_103; __rev0_103 = __builtin_shufflevector(__s0_103, __s0_103, 7, 6, 5, 4, 3, 2, 1, 0); \
19670		int16x4_t __rev1_103; __rev1_103 = __builtin_shufflevector(__s1_103, __s1_103, 3, 2, 1, 0); \
	19690	int16x8_t __rev0_103; __rev0_103 = __builtin_shufflevector(__s0_103, __s0_103, __lane_reverse_128_16); \
	19691	int16x4_t __rev1_103; __rev1_103 = __builtin_shufflevector(__s1_103, __s1_103, __lane_reverse_64_16); \
19671	19692	__ret_103 = __rev0_103 * __noswap_splatq_lane_s16(__rev1_103, __p2_103); \
19672		__ret_103 = __builtin_shufflevector(__ret_103, __ret_103, 7, 6, 5, 4, 3, 2, 1, 0); \
	19693	__ret_103 = __builtin_shufflevector(__ret_103, __ret_103, __lane_reverse_128_16); \
19673	19694	__ret_103; \
19674	19695	})
19675	19696	#endif
...	...	@@ -19687,10 +19708,10 @@ __ai __attribute__((target("neon"))) poly8x16_t vmulq_p8(poly8x16_t __p0, poly8x
19687	19708	uint32x2_t __ret_105; \
19688	19709	uint32x2_t __s0_105 = __p0_105; \
19689	19710	uint32x2_t __s1_105 = __p1_105; \
19690		uint32x2_t __rev0_105; __rev0_105 = __builtin_shufflevector(__s0_105, __s0_105, 1, 0); \
19691		uint32x2_t __rev1_105; __rev1_105 = __builtin_shufflevector(__s1_105, __s1_105, 1, 0); \
	19711	uint32x2_t __rev0_105; __rev0_105 = __builtin_shufflevector(__s0_105, __s0_105, __lane_reverse_64_32); \
	19712	uint32x2_t __rev1_105; __rev1_105 = __builtin_shufflevector(__s1_105, __s1_105, __lane_reverse_64_32); \
19692	19713	__ret_105 = __rev0_105 * __noswap_splat_lane_u32(__rev1_105, __p2_105); \
19693		__ret_105 = __builtin_shufflevector(__ret_105, __ret_105, 1, 0); \
	19714	__ret_105 = __builtin_shufflevector(__ret_105, __ret_105, __lane_reverse_64_32); \
19694	19715	__ret_105; \
19695	19716	})
19696	19717	#endif
...	...	@@ -19708,10 +19729,10 @@ __ai __attribute__((target("neon"))) poly8x16_t vmulq_p8(poly8x16_t __p0, poly8x
19708	19729	uint16x4_t __ret_107; \
19709	19730	uint16x4_t __s0_107 = __p0_107; \
19710	19731	uint16x4_t __s1_107 = __p1_107; \
19711		uint16x4_t __rev0_107; __rev0_107 = __builtin_shufflevector(__s0_107, __s0_107, 3, 2, 1, 0); \
19712		uint16x4_t __rev1_107; __rev1_107 = __builtin_shufflevector(__s1_107, __s1_107, 3, 2, 1, 0); \
	19732	uint16x4_t __rev0_107; __rev0_107 = __builtin_shufflevector(__s0_107, __s0_107, __lane_reverse_64_16); \
	19733	uint16x4_t __rev1_107; __rev1_107 = __builtin_shufflevector(__s1_107, __s1_107, __lane_reverse_64_16); \
19713	19734	__ret_107 = __rev0_107 * __noswap_splat_lane_u16(__rev1_107, __p2_107); \
19714		__ret_107 = __builtin_shufflevector(__ret_107, __ret_107, 3, 2, 1, 0); \
	19735	__ret_107 = __builtin_shufflevector(__ret_107, __ret_107, __lane_reverse_64_16); \
19715	19736	__ret_107; \
19716	19737	})
19717	19738	#endif
...	...	@@ -19729,10 +19750,10 @@ __ai __attribute__((target("neon"))) poly8x16_t vmulq_p8(poly8x16_t __p0, poly8x
19729	19750	float32x2_t __ret_109; \
19730	19751	float32x2_t __s0_109 = __p0_109; \
19731	19752	float32x2_t __s1_109 = __p1_109; \
19732		float32x2_t __rev0_109; __rev0_109 = __builtin_shufflevector(__s0_109, __s0_109, 1, 0); \
19733		float32x2_t __rev1_109; __rev1_109 = __builtin_shufflevector(__s1_109, __s1_109, 1, 0); \
	19753	float32x2_t __rev0_109; __rev0_109 = __builtin_shufflevector(__s0_109, __s0_109, __lane_reverse_64_32); \
	19754	float32x2_t __rev1_109; __rev1_109 = __builtin_shufflevector(__s1_109, __s1_109, __lane_reverse_64_32); \
19734	19755	__ret_109 = __rev0_109 * __noswap_splat_lane_f32(__rev1_109, __p2_109); \
19735		__ret_109 = __builtin_shufflevector(__ret_109, __ret_109, 1, 0); \
	19756	__ret_109 = __builtin_shufflevector(__ret_109, __ret_109, __lane_reverse_64_32); \
19736	19757	__ret_109; \
19737	19758	})
19738	19759	#endif
...	...	@@ -19750,10 +19771,10 @@ __ai __attribute__((target("neon"))) poly8x16_t vmulq_p8(poly8x16_t __p0, poly8x
19750	19771	int32x2_t __ret_111; \
19751	19772	int32x2_t __s0_111 = __p0_111; \
19752	19773	int32x2_t __s1_111 = __p1_111; \
19753		int32x2_t __rev0_111; __rev0_111 = __builtin_shufflevector(__s0_111, __s0_111, 1, 0); \
19754		int32x2_t __rev1_111; __rev1_111 = __builtin_shufflevector(__s1_111, __s1_111, 1, 0); \
	19774	int32x2_t __rev0_111; __rev0_111 = __builtin_shufflevector(__s0_111, __s0_111, __lane_reverse_64_32); \
	19775	int32x2_t __rev1_111; __rev1_111 = __builtin_shufflevector(__s1_111, __s1_111, __lane_reverse_64_32); \
19755	19776	__ret_111 = __rev0_111 * __noswap_splat_lane_s32(__rev1_111, __p2_111); \
19756		__ret_111 = __builtin_shufflevector(__ret_111, __ret_111, 1, 0); \
	19777	__ret_111 = __builtin_shufflevector(__ret_111, __ret_111, __lane_reverse_64_32); \
19757	19778	__ret_111; \
19758	19779	})
19759	19780	#endif
...	...	@@ -19771,10 +19792,10 @@ __ai __attribute__((target("neon"))) poly8x16_t vmulq_p8(poly8x16_t __p0, poly8x
19771	19792	int16x4_t __ret_113; \
19772	19793	int16x4_t __s0_113 = __p0_113; \
19773	19794	int16x4_t __s1_113 = __p1_113; \
19774		int16x4_t __rev0_113; __rev0_113 = __builtin_shufflevector(__s0_113, __s0_113, 3, 2, 1, 0); \
19775		int16x4_t __rev1_113; __rev1_113 = __builtin_shufflevector(__s1_113, __s1_113, 3, 2, 1, 0); \
	19795	int16x4_t __rev0_113; __rev0_113 = __builtin_shufflevector(__s0_113, __s0_113, __lane_reverse_64_16); \
	19796	int16x4_t __rev1_113; __rev1_113 = __builtin_shufflevector(__s1_113, __s1_113, __lane_reverse_64_16); \
19776	19797	__ret_113 = __rev0_113 * __noswap_splat_lane_s16(__rev1_113, __p2_113); \
19777		__ret_113 = __builtin_shufflevector(__ret_113, __ret_113, 3, 2, 1, 0); \
	19798	__ret_113 = __builtin_shufflevector(__ret_113, __ret_113, __lane_reverse_64_16); \
19778	19799	__ret_113; \
19779	19800	})
19780	19801	#endif
...	...	@@ -19788,9 +19809,9 @@ __ai __attribute__((target("neon"))) uint32x4_t vmulq_n_u32(uint32x4_t __p0, uin
19788	19809	#else
19789	19810	__ai __attribute__((target("neon"))) uint32x4_t vmulq_n_u32(uint32x4_t __p0, uint32_t __p1) {
19790	19811	uint32x4_t __ret;
19791		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	19812	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
19792	19813	__ret = __rev0 * (uint32x4_t) {__p1, __p1, __p1, __p1};
19793		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	19814	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
19794	19815	return __ret;
19795	19816	}
19796	19817	#endif
...	...	@@ -19804,9 +19825,9 @@ __ai __attribute__((target("neon"))) uint16x8_t vmulq_n_u16(uint16x8_t __p0, uin
19804	19825	#else
19805	19826	__ai __attribute__((target("neon"))) uint16x8_t vmulq_n_u16(uint16x8_t __p0, uint16_t __p1) {
19806	19827	uint16x8_t __ret;
19807		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	19828	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
19808	19829	__ret = __rev0 * (uint16x8_t) {__p1, __p1, __p1, __p1, __p1, __p1, __p1, __p1};
19809		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	19830	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
19810	19831	return __ret;
19811	19832	}
19812	19833	#endif
...	...	@@ -19820,9 +19841,9 @@ __ai __attribute__((target("neon"))) float32x4_t vmulq_n_f32(float32x4_t __p0, f
19820	19841	#else
19821	19842	__ai __attribute__((target("neon"))) float32x4_t vmulq_n_f32(float32x4_t __p0, float32_t __p1) {
19822	19843	float32x4_t __ret;
19823		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	19844	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
19824	19845	__ret = __rev0 * (float32x4_t) {__p1, __p1, __p1, __p1};
19825		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	19846	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
19826	19847	return __ret;
19827	19848	}
19828	19849	#endif
...	...	@@ -19836,9 +19857,9 @@ __ai __attribute__((target("neon"))) int32x4_t vmulq_n_s32(int32x4_t __p0, int32
19836	19857	#else
19837	19858	__ai __attribute__((target("neon"))) int32x4_t vmulq_n_s32(int32x4_t __p0, int32_t __p1) {
19838	19859	int32x4_t __ret;
19839		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	19860	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
19840	19861	__ret = __rev0 * (int32x4_t) {__p1, __p1, __p1, __p1};
19841		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	19862	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
19842	19863	return __ret;
19843	19864	}
19844	19865	#endif
...	...	@@ -19852,9 +19873,9 @@ __ai __attribute__((target("neon"))) int16x8_t vmulq_n_s16(int16x8_t __p0, int16
19852	19873	#else
19853	19874	__ai __attribute__((target("neon"))) int16x8_t vmulq_n_s16(int16x8_t __p0, int16_t __p1) {
19854	19875	int16x8_t __ret;
19855		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	19876	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
19856	19877	__ret = __rev0 * (int16x8_t) {__p1, __p1, __p1, __p1, __p1, __p1, __p1, __p1};
19857		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	19878	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
19858	19879	return __ret;
19859	19880	}
19860	19881	#endif
...	...	@@ -19868,9 +19889,9 @@ __ai __attribute__((target("neon"))) uint32x2_t vmul_n_u32(uint32x2_t __p0, uint
19868	19889	#else
19869	19890	__ai __attribute__((target("neon"))) uint32x2_t vmul_n_u32(uint32x2_t __p0, uint32_t __p1) {
19870	19891	uint32x2_t __ret;
19871		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
	19892	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
19872	19893	__ret = __rev0 * (uint32x2_t) {__p1, __p1};
19873		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	19894	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
19874	19895	return __ret;
19875	19896	}
19876	19897	#endif
...	...	@@ -19884,9 +19905,9 @@ __ai __attribute__((target("neon"))) uint16x4_t vmul_n_u16(uint16x4_t __p0, uint
19884	19905	#else
19885	19906	__ai __attribute__((target("neon"))) uint16x4_t vmul_n_u16(uint16x4_t __p0, uint16_t __p1) {
19886	19907	uint16x4_t __ret;
19887		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	19908	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
19888	19909	__ret = __rev0 * (uint16x4_t) {__p1, __p1, __p1, __p1};
19889		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	19910	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
19890	19911	return __ret;
19891	19912	}
19892	19913	#endif
...	...	@@ -19900,9 +19921,9 @@ __ai __attribute__((target("neon"))) float32x2_t vmul_n_f32(float32x2_t __p0, fl
19900	19921	#else
19901	19922	__ai __attribute__((target("neon"))) float32x2_t vmul_n_f32(float32x2_t __p0, float32_t __p1) {
19902	19923	float32x2_t __ret;
19903		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
	19924	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
19904	19925	__ret = __rev0 * (float32x2_t) {__p1, __p1};
19905		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	19926	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
19906	19927	return __ret;
19907	19928	}
19908	19929	#endif
...	...	@@ -19916,9 +19937,9 @@ __ai __attribute__((target("neon"))) int32x2_t vmul_n_s32(int32x2_t __p0, int32_
19916	19937	#else
19917	19938	__ai __attribute__((target("neon"))) int32x2_t vmul_n_s32(int32x2_t __p0, int32_t __p1) {
19918	19939	int32x2_t __ret;
19919		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
	19940	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
19920	19941	__ret = __rev0 * (int32x2_t) {__p1, __p1};
19921		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	19942	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
19922	19943	return __ret;
19923	19944	}
19924	19945	#endif
...	...	@@ -19932,9 +19953,9 @@ __ai __attribute__((target("neon"))) int16x4_t vmul_n_s16(int16x4_t __p0, int16_
19932	19953	#else
19933	19954	__ai __attribute__((target("neon"))) int16x4_t vmul_n_s16(int16x4_t __p0, int16_t __p1) {
19934	19955	int16x4_t __ret;
19935		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	19956	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
19936	19957	__ret = __rev0 * (int16x4_t) {__p1, __p1, __p1, __p1};
19937		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	19958	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
19938	19959	return __ret;
19939	19960	}
19940	19961	#endif
...	...	@@ -19942,21 +19963,21 @@ __ai __attribute__((target("neon"))) int16x4_t vmul_n_s16(int16x4_t __p0, int16_
19942	19963	#ifdef __LITTLE_ENDIAN__
19943	19964	__ai __attribute__((target("neon"))) poly16x8_t vmull_p8(poly8x8_t __p0, poly8x8_t __p1) {
19944	19965	poly16x8_t __ret;
19945		__ret = (poly16x8_t) __builtin_neon_vmull_v((int8x8_t)__p0, (int8x8_t)__p1, 37);
	19966	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vmull_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 37));
19946	19967	return __ret;
19947	19968	}
19948	19969	#else
19949	19970	__ai __attribute__((target("neon"))) poly16x8_t vmull_p8(poly8x8_t __p0, poly8x8_t __p1) {
19950	19971	poly16x8_t __ret;
19951		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
19952		poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
19953		__ret = (poly16x8_t) __builtin_neon_vmull_v((int8x8_t)__rev0, (int8x8_t)__rev1, 37);
19954		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	19972	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	19973	poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	19974	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vmull_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 37));
	19975	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
19955	19976	return __ret;
19956	19977	}
19957	19978	__ai __attribute__((target("neon"))) poly16x8_t __noswap_vmull_p8(poly8x8_t __p0, poly8x8_t __p1) {
19958	19979	poly16x8_t __ret;
19959		__ret = (poly16x8_t) __builtin_neon_vmull_v((int8x8_t)__p0, (int8x8_t)__p1, 37);
	19980	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vmull_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 37));
19960	19981	return __ret;
19961	19982	}
19962	19983	#endif
...	...	@@ -19964,21 +19985,21 @@ __ai __attribute__((target("neon"))) poly16x8_t __noswap_vmull_p8(poly8x8_t __p0
19964	19985	#ifdef __LITTLE_ENDIAN__
19965	19986	__ai __attribute__((target("neon"))) uint16x8_t vmull_u8(uint8x8_t __p0, uint8x8_t __p1) {
19966	19987	uint16x8_t __ret;
19967		__ret = (uint16x8_t) __builtin_neon_vmull_v((int8x8_t)__p0, (int8x8_t)__p1, 49);
	19988	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vmull_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 49));
19968	19989	return __ret;
19969	19990	}
19970	19991	#else
19971	19992	__ai __attribute__((target("neon"))) uint16x8_t vmull_u8(uint8x8_t __p0, uint8x8_t __p1) {
19972	19993	uint16x8_t __ret;
19973		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
19974		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
19975		__ret = (uint16x8_t) __builtin_neon_vmull_v((int8x8_t)__rev0, (int8x8_t)__rev1, 49);
19976		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	19994	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	19995	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	19996	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vmull_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 49));
	19997	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
19977	19998	return __ret;
19978	19999	}
19979	20000	__ai __attribute__((target("neon"))) uint16x8_t __noswap_vmull_u8(uint8x8_t __p0, uint8x8_t __p1) {
19980	20001	uint16x8_t __ret;
19981		__ret = (uint16x8_t) __builtin_neon_vmull_v((int8x8_t)__p0, (int8x8_t)__p1, 49);
	20002	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vmull_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 49));
19982	20003	return __ret;
19983	20004	}
19984	20005	#endif
...	...	@@ -19986,21 +20007,21 @@ __ai __attribute__((target("neon"))) uint16x8_t __noswap_vmull_u8(uint8x8_t __p0
19986	20007	#ifdef __LITTLE_ENDIAN__
19987	20008	__ai __attribute__((target("neon"))) uint64x2_t vmull_u32(uint32x2_t __p0, uint32x2_t __p1) {
19988	20009	uint64x2_t __ret;
19989		__ret = (uint64x2_t) __builtin_neon_vmull_v((int8x8_t)__p0, (int8x8_t)__p1, 51);
	20010	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vmull_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 51));
19990	20011	return __ret;
19991	20012	}
19992	20013	#else
19993	20014	__ai __attribute__((target("neon"))) uint64x2_t vmull_u32(uint32x2_t __p0, uint32x2_t __p1) {
19994	20015	uint64x2_t __ret;
19995		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
19996		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
19997		__ret = (uint64x2_t) __builtin_neon_vmull_v((int8x8_t)__rev0, (int8x8_t)__rev1, 51);
19998		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	20016	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	20017	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	20018	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vmull_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 51));
	20019	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
19999	20020	return __ret;
20000	20021	}
20001	20022	__ai __attribute__((target("neon"))) uint64x2_t __noswap_vmull_u32(uint32x2_t __p0, uint32x2_t __p1) {
20002	20023	uint64x2_t __ret;
20003		__ret = (uint64x2_t) __builtin_neon_vmull_v((int8x8_t)__p0, (int8x8_t)__p1, 51);
	20024	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vmull_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 51));
20004	20025	return __ret;
20005	20026	}
20006	20027	#endif
...	...	@@ -20008,21 +20029,21 @@ __ai __attribute__((target("neon"))) uint64x2_t __noswap_vmull_u32(uint32x2_t __
20008	20029	#ifdef __LITTLE_ENDIAN__
20009	20030	__ai __attribute__((target("neon"))) uint32x4_t vmull_u16(uint16x4_t __p0, uint16x4_t __p1) {
20010	20031	uint32x4_t __ret;
20011		__ret = (uint32x4_t) __builtin_neon_vmull_v((int8x8_t)__p0, (int8x8_t)__p1, 50);
	20032	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vmull_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 50));
20012	20033	return __ret;
20013	20034	}
20014	20035	#else
20015	20036	__ai __attribute__((target("neon"))) uint32x4_t vmull_u16(uint16x4_t __p0, uint16x4_t __p1) {
20016	20037	uint32x4_t __ret;
20017		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
20018		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
20019		__ret = (uint32x4_t) __builtin_neon_vmull_v((int8x8_t)__rev0, (int8x8_t)__rev1, 50);
20020		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	20038	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	20039	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	20040	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vmull_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 50));
	20041	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
20021	20042	return __ret;
20022	20043	}
20023	20044	__ai __attribute__((target("neon"))) uint32x4_t __noswap_vmull_u16(uint16x4_t __p0, uint16x4_t __p1) {
20024	20045	uint32x4_t __ret;
20025		__ret = (uint32x4_t) __builtin_neon_vmull_v((int8x8_t)__p0, (int8x8_t)__p1, 50);
	20046	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vmull_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 50));
20026	20047	return __ret;
20027	20048	}
20028	20049	#endif
...	...	@@ -20030,21 +20051,21 @@ __ai __attribute__((target("neon"))) uint32x4_t __noswap_vmull_u16(uint16x4_t __
20030	20051	#ifdef __LITTLE_ENDIAN__
20031	20052	__ai __attribute__((target("neon"))) int16x8_t vmull_s8(int8x8_t __p0, int8x8_t __p1) {
20032	20053	int16x8_t __ret;
20033		__ret = (int16x8_t) __builtin_neon_vmull_v((int8x8_t)__p0, (int8x8_t)__p1, 33);
	20054	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vmull_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 33));
20034	20055	return __ret;
20035	20056	}
20036	20057	#else
20037	20058	__ai __attribute__((target("neon"))) int16x8_t vmull_s8(int8x8_t __p0, int8x8_t __p1) {
20038	20059	int16x8_t __ret;
20039		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
20040		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
20041		__ret = (int16x8_t) __builtin_neon_vmull_v((int8x8_t)__rev0, (int8x8_t)__rev1, 33);
20042		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	20060	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	20061	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	20062	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vmull_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 33));
	20063	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
20043	20064	return __ret;
20044	20065	}
20045	20066	__ai __attribute__((target("neon"))) int16x8_t __noswap_vmull_s8(int8x8_t __p0, int8x8_t __p1) {
20046	20067	int16x8_t __ret;
20047		__ret = (int16x8_t) __builtin_neon_vmull_v((int8x8_t)__p0, (int8x8_t)__p1, 33);
	20068	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vmull_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 33));
20048	20069	return __ret;
20049	20070	}
20050	20071	#endif
...	...	@@ -20052,21 +20073,21 @@ __ai __attribute__((target("neon"))) int16x8_t __noswap_vmull_s8(int8x8_t __p0,
20052	20073	#ifdef __LITTLE_ENDIAN__
20053	20074	__ai __attribute__((target("neon"))) int64x2_t vmull_s32(int32x2_t __p0, int32x2_t __p1) {
20054	20075	int64x2_t __ret;
20055		__ret = (int64x2_t) __builtin_neon_vmull_v((int8x8_t)__p0, (int8x8_t)__p1, 35);
	20076	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vmull_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 35));
20056	20077	return __ret;
20057	20078	}
20058	20079	#else
20059	20080	__ai __attribute__((target("neon"))) int64x2_t vmull_s32(int32x2_t __p0, int32x2_t __p1) {
20060	20081	int64x2_t __ret;
20061		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
20062		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
20063		__ret = (int64x2_t) __builtin_neon_vmull_v((int8x8_t)__rev0, (int8x8_t)__rev1, 35);
20064		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	20082	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	20083	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	20084	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vmull_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 35));
	20085	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
20065	20086	return __ret;
20066	20087	}
20067	20088	__ai __attribute__((target("neon"))) int64x2_t __noswap_vmull_s32(int32x2_t __p0, int32x2_t __p1) {
20068	20089	int64x2_t __ret;
20069		__ret = (int64x2_t) __builtin_neon_vmull_v((int8x8_t)__p0, (int8x8_t)__p1, 35);
	20090	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vmull_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 35));
20070	20091	return __ret;
20071	20092	}
20072	20093	#endif
...	...	@@ -20074,21 +20095,21 @@ __ai __attribute__((target("neon"))) int64x2_t __noswap_vmull_s32(int32x2_t __p0
20074	20095	#ifdef __LITTLE_ENDIAN__
20075	20096	__ai __attribute__((target("neon"))) int32x4_t vmull_s16(int16x4_t __p0, int16x4_t __p1) {
20076	20097	int32x4_t __ret;
20077		__ret = (int32x4_t) __builtin_neon_vmull_v((int8x8_t)__p0, (int8x8_t)__p1, 34);
	20098	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vmull_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 34));
20078	20099	return __ret;
20079	20100	}
20080	20101	#else
20081	20102	__ai __attribute__((target("neon"))) int32x4_t vmull_s16(int16x4_t __p0, int16x4_t __p1) {
20082	20103	int32x4_t __ret;
20083		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
20084		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
20085		__ret = (int32x4_t) __builtin_neon_vmull_v((int8x8_t)__rev0, (int8x8_t)__rev1, 34);
20086		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	20104	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	20105	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	20106	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vmull_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 34));
	20107	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
20087	20108	return __ret;
20088	20109	}
20089	20110	__ai __attribute__((target("neon"))) int32x4_t __noswap_vmull_s16(int16x4_t __p0, int16x4_t __p1) {
20090	20111	int32x4_t __ret;
20091		__ret = (int32x4_t) __builtin_neon_vmull_v((int8x8_t)__p0, (int8x8_t)__p1, 34);
	20112	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vmull_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 34));
20092	20113	return __ret;
20093	20114	}
20094	20115	#endif
...	...	@@ -20106,10 +20127,10 @@ __ai __attribute__((target("neon"))) int32x4_t __noswap_vmull_s16(int16x4_t __p0
20106	20127	uint64x2_t __ret_115; \
20107	20128	uint32x2_t __s0_115 = __p0_115; \
20108	20129	uint32x2_t __s1_115 = __p1_115; \
20109		uint32x2_t __rev0_115; __rev0_115 = __builtin_shufflevector(__s0_115, __s0_115, 1, 0); \
20110		uint32x2_t __rev1_115; __rev1_115 = __builtin_shufflevector(__s1_115, __s1_115, 1, 0); \
	20130	uint32x2_t __rev0_115; __rev0_115 = __builtin_shufflevector(__s0_115, __s0_115, __lane_reverse_64_32); \
	20131	uint32x2_t __rev1_115; __rev1_115 = __builtin_shufflevector(__s1_115, __s1_115, __lane_reverse_64_32); \
20111	20132	__ret_115 = __noswap_vmull_u32(__rev0_115, __noswap_splat_lane_u32(__rev1_115, __p2_115)); \
20112		__ret_115 = __builtin_shufflevector(__ret_115, __ret_115, 1, 0); \
	20133	__ret_115 = __builtin_shufflevector(__ret_115, __ret_115, __lane_reverse_128_64); \
20113	20134	__ret_115; \
20114	20135	})
20115	20136	#endif
...	...	@@ -20127,10 +20148,10 @@ __ai __attribute__((target("neon"))) int32x4_t __noswap_vmull_s16(int16x4_t __p0
20127	20148	uint32x4_t __ret_117; \
20128	20149	uint16x4_t __s0_117 = __p0_117; \
20129	20150	uint16x4_t __s1_117 = __p1_117; \
20130		uint16x4_t __rev0_117; __rev0_117 = __builtin_shufflevector(__s0_117, __s0_117, 3, 2, 1, 0); \
20131		uint16x4_t __rev1_117; __rev1_117 = __builtin_shufflevector(__s1_117, __s1_117, 3, 2, 1, 0); \
	20151	uint16x4_t __rev0_117; __rev0_117 = __builtin_shufflevector(__s0_117, __s0_117, __lane_reverse_64_16); \
	20152	uint16x4_t __rev1_117; __rev1_117 = __builtin_shufflevector(__s1_117, __s1_117, __lane_reverse_64_16); \
20132	20153	__ret_117 = __noswap_vmull_u16(__rev0_117, __noswap_splat_lane_u16(__rev1_117, __p2_117)); \
20133		__ret_117 = __builtin_shufflevector(__ret_117, __ret_117, 3, 2, 1, 0); \
	20154	__ret_117 = __builtin_shufflevector(__ret_117, __ret_117, __lane_reverse_128_32); \
20134	20155	__ret_117; \
20135	20156	})
20136	20157	#endif
...	...	@@ -20148,10 +20169,10 @@ __ai __attribute__((target("neon"))) int32x4_t __noswap_vmull_s16(int16x4_t __p0
20148	20169	int64x2_t __ret_119; \
20149	20170	int32x2_t __s0_119 = __p0_119; \
20150	20171	int32x2_t __s1_119 = __p1_119; \
20151		int32x2_t __rev0_119; __rev0_119 = __builtin_shufflevector(__s0_119, __s0_119, 1, 0); \
20152		int32x2_t __rev1_119; __rev1_119 = __builtin_shufflevector(__s1_119, __s1_119, 1, 0); \
	20172	int32x2_t __rev0_119; __rev0_119 = __builtin_shufflevector(__s0_119, __s0_119, __lane_reverse_64_32); \
	20173	int32x2_t __rev1_119; __rev1_119 = __builtin_shufflevector(__s1_119, __s1_119, __lane_reverse_64_32); \
20153	20174	__ret_119 = __noswap_vmull_s32(__rev0_119, __noswap_splat_lane_s32(__rev1_119, __p2_119)); \
20154		__ret_119 = __builtin_shufflevector(__ret_119, __ret_119, 1, 0); \
	20175	__ret_119 = __builtin_shufflevector(__ret_119, __ret_119, __lane_reverse_128_64); \
20155	20176	__ret_119; \
20156	20177	})
20157	20178	#endif
...	...	@@ -20169,10 +20190,10 @@ __ai __attribute__((target("neon"))) int32x4_t __noswap_vmull_s16(int16x4_t __p0
20169	20190	int32x4_t __ret_121; \
20170	20191	int16x4_t __s0_121 = __p0_121; \
20171	20192	int16x4_t __s1_121 = __p1_121; \
20172		int16x4_t __rev0_121; __rev0_121 = __builtin_shufflevector(__s0_121, __s0_121, 3, 2, 1, 0); \
20173		int16x4_t __rev1_121; __rev1_121 = __builtin_shufflevector(__s1_121, __s1_121, 3, 2, 1, 0); \
	20193	int16x4_t __rev0_121; __rev0_121 = __builtin_shufflevector(__s0_121, __s0_121, __lane_reverse_64_16); \
	20194	int16x4_t __rev1_121; __rev1_121 = __builtin_shufflevector(__s1_121, __s1_121, __lane_reverse_64_16); \
20174	20195	__ret_121 = __noswap_vmull_s16(__rev0_121, __noswap_splat_lane_s16(__rev1_121, __p2_121)); \
20175		__ret_121 = __builtin_shufflevector(__ret_121, __ret_121, 3, 2, 1, 0); \
	20196	__ret_121 = __builtin_shufflevector(__ret_121, __ret_121, __lane_reverse_128_32); \
20176	20197	__ret_121; \
20177	20198	})
20178	20199	#endif
...	...	@@ -20186,9 +20207,9 @@ __ai __attribute__((target("neon"))) uint64x2_t vmull_n_u32(uint32x2_t __p0, uin
20186	20207	#else
20187	20208	__ai __attribute__((target("neon"))) uint64x2_t vmull_n_u32(uint32x2_t __p0, uint32_t __p1) {
20188	20209	uint64x2_t __ret;
20189		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
	20210	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
20190	20211	__ret = __noswap_vmull_u32(__rev0, (uint32x2_t) {__p1, __p1});
20191		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	20212	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
20192	20213	return __ret;
20193	20214	}
20194	20215	__ai __attribute__((target("neon"))) uint64x2_t __noswap_vmull_n_u32(uint32x2_t __p0, uint32_t __p1) {
...	...	@@ -20207,9 +20228,9 @@ __ai __attribute__((target("neon"))) uint32x4_t vmull_n_u16(uint16x4_t __p0, uin
20207	20228	#else
20208	20229	__ai __attribute__((target("neon"))) uint32x4_t vmull_n_u16(uint16x4_t __p0, uint16_t __p1) {
20209	20230	uint32x4_t __ret;
20210		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	20231	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
20211	20232	__ret = __noswap_vmull_u16(__rev0, (uint16x4_t) {__p1, __p1, __p1, __p1});
20212		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	20233	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
20213	20234	return __ret;
20214	20235	}
20215	20236	__ai __attribute__((target("neon"))) uint32x4_t __noswap_vmull_n_u16(uint16x4_t __p0, uint16_t __p1) {
...	...	@@ -20228,9 +20249,9 @@ __ai __attribute__((target("neon"))) int64x2_t vmull_n_s32(int32x2_t __p0, int32
20228	20249	#else
20229	20250	__ai __attribute__((target("neon"))) int64x2_t vmull_n_s32(int32x2_t __p0, int32_t __p1) {
20230	20251	int64x2_t __ret;
20231		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
	20252	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
20232	20253	__ret = __noswap_vmull_s32(__rev0, (int32x2_t) {__p1, __p1});
20233		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	20254	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
20234	20255	return __ret;
20235	20256	}
20236	20257	__ai __attribute__((target("neon"))) int64x2_t __noswap_vmull_n_s32(int32x2_t __p0, int32_t __p1) {
...	...	@@ -20249,9 +20270,9 @@ __ai __attribute__((target("neon"))) int32x4_t vmull_n_s16(int16x4_t __p0, int16
20249	20270	#else
20250	20271	__ai __attribute__((target("neon"))) int32x4_t vmull_n_s16(int16x4_t __p0, int16_t __p1) {
20251	20272	int32x4_t __ret;
20252		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	20273	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
20253	20274	__ret = __noswap_vmull_s16(__rev0, (int16x4_t) {__p1, __p1, __p1, __p1});
20254		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	20275	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
20255	20276	return __ret;
20256	20277	}
20257	20278	__ai __attribute__((target("neon"))) int32x4_t __noswap_vmull_n_s16(int16x4_t __p0, int16_t __p1) {
...	...	@@ -20270,9 +20291,9 @@ __ai __attribute__((target("neon"))) poly8x8_t vmvn_p8(poly8x8_t __p0) {
20270	20291	#else
20271	20292	__ai __attribute__((target("neon"))) poly8x8_t vmvn_p8(poly8x8_t __p0) {
20272	20293	poly8x8_t __ret;
20273		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	20294	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
20274	20295	__ret = ~__rev0;
20275		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	20296	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
20276	20297	return __ret;
20277	20298	}
20278	20299	#endif
...	...	@@ -20286,9 +20307,9 @@ __ai __attribute__((target("neon"))) poly8x16_t vmvnq_p8(poly8x16_t __p0) {
20286	20307	#else
20287	20308	__ai __attribute__((target("neon"))) poly8x16_t vmvnq_p8(poly8x16_t __p0) {
20288	20309	poly8x16_t __ret;
20289		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	20310	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
20290	20311	__ret = ~__rev0;
20291		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	20312	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
20292	20313	return __ret;
20293	20314	}
20294	20315	#endif
...	...	@@ -20302,9 +20323,9 @@ __ai __attribute__((target("neon"))) uint8x16_t vmvnq_u8(uint8x16_t __p0) {
20302	20323	#else
20303	20324	__ai __attribute__((target("neon"))) uint8x16_t vmvnq_u8(uint8x16_t __p0) {
20304	20325	uint8x16_t __ret;
20305		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	20326	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
20306	20327	__ret = ~__rev0;
20307		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	20328	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
20308	20329	return __ret;
20309	20330	}
20310	20331	#endif
...	...	@@ -20318,9 +20339,9 @@ __ai __attribute__((target("neon"))) uint32x4_t vmvnq_u32(uint32x4_t __p0) {
20318	20339	#else
20319	20340	__ai __attribute__((target("neon"))) uint32x4_t vmvnq_u32(uint32x4_t __p0) {
20320	20341	uint32x4_t __ret;
20321		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	20342	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
20322	20343	__ret = ~__rev0;
20323		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	20344	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
20324	20345	return __ret;
20325	20346	}
20326	20347	#endif
...	...	@@ -20334,9 +20355,9 @@ __ai __attribute__((target("neon"))) uint16x8_t vmvnq_u16(uint16x8_t __p0) {
20334	20355	#else
20335	20356	__ai __attribute__((target("neon"))) uint16x8_t vmvnq_u16(uint16x8_t __p0) {
20336	20357	uint16x8_t __ret;
20337		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	20358	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
20338	20359	__ret = ~__rev0;
20339		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	20360	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
20340	20361	return __ret;
20341	20362	}
20342	20363	#endif
...	...	@@ -20350,9 +20371,9 @@ __ai __attribute__((target("neon"))) int8x16_t vmvnq_s8(int8x16_t __p0) {
20350	20371	#else
20351	20372	__ai __attribute__((target("neon"))) int8x16_t vmvnq_s8(int8x16_t __p0) {
20352	20373	int8x16_t __ret;
20353		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	20374	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
20354	20375	__ret = ~__rev0;
20355		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	20376	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
20356	20377	return __ret;
20357	20378	}
20358	20379	#endif
...	...	@@ -20366,9 +20387,9 @@ __ai __attribute__((target("neon"))) int32x4_t vmvnq_s32(int32x4_t __p0) {
20366	20387	#else
20367	20388	__ai __attribute__((target("neon"))) int32x4_t vmvnq_s32(int32x4_t __p0) {
20368	20389	int32x4_t __ret;
20369		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	20390	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
20370	20391	__ret = ~__rev0;
20371		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	20392	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
20372	20393	return __ret;
20373	20394	}
20374	20395	#endif
...	...	@@ -20382,9 +20403,9 @@ __ai __attribute__((target("neon"))) int16x8_t vmvnq_s16(int16x8_t __p0) {
20382	20403	#else
20383	20404	__ai __attribute__((target("neon"))) int16x8_t vmvnq_s16(int16x8_t __p0) {
20384	20405	int16x8_t __ret;
20385		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	20406	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
20386	20407	__ret = ~__rev0;
20387		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	20408	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
20388	20409	return __ret;
20389	20410	}
20390	20411	#endif
...	...	@@ -20398,9 +20419,9 @@ __ai __attribute__((target("neon"))) uint8x8_t vmvn_u8(uint8x8_t __p0) {
20398	20419	#else
20399	20420	__ai __attribute__((target("neon"))) uint8x8_t vmvn_u8(uint8x8_t __p0) {
20400	20421	uint8x8_t __ret;
20401		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	20422	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
20402	20423	__ret = ~__rev0;
20403		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	20424	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
20404	20425	return __ret;
20405	20426	}
20406	20427	#endif
...	...	@@ -20414,9 +20435,9 @@ __ai __attribute__((target("neon"))) uint32x2_t vmvn_u32(uint32x2_t __p0) {
20414	20435	#else
20415	20436	__ai __attribute__((target("neon"))) uint32x2_t vmvn_u32(uint32x2_t __p0) {
20416	20437	uint32x2_t __ret;
20417		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
	20438	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
20418	20439	__ret = ~__rev0;
20419		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	20440	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
20420	20441	return __ret;
20421	20442	}
20422	20443	#endif
...	...	@@ -20430,9 +20451,9 @@ __ai __attribute__((target("neon"))) uint16x4_t vmvn_u16(uint16x4_t __p0) {
20430	20451	#else
20431	20452	__ai __attribute__((target("neon"))) uint16x4_t vmvn_u16(uint16x4_t __p0) {
20432	20453	uint16x4_t __ret;
20433		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	20454	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
20434	20455	__ret = ~__rev0;
20435		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	20456	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
20436	20457	return __ret;
20437	20458	}
20438	20459	#endif
...	...	@@ -20446,9 +20467,9 @@ __ai __attribute__((target("neon"))) int8x8_t vmvn_s8(int8x8_t __p0) {
20446	20467	#else
20447	20468	__ai __attribute__((target("neon"))) int8x8_t vmvn_s8(int8x8_t __p0) {
20448	20469	int8x8_t __ret;
20449		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	20470	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
20450	20471	__ret = ~__rev0;
20451		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	20472	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
20452	20473	return __ret;
20453	20474	}
20454	20475	#endif
...	...	@@ -20462,9 +20483,9 @@ __ai __attribute__((target("neon"))) int32x2_t vmvn_s32(int32x2_t __p0) {
20462	20483	#else
20463	20484	__ai __attribute__((target("neon"))) int32x2_t vmvn_s32(int32x2_t __p0) {
20464	20485	int32x2_t __ret;
20465		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
	20486	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
20466	20487	__ret = ~__rev0;
20467		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	20488	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
20468	20489	return __ret;
20469	20490	}
20470	20491	#endif
...	...	@@ -20478,9 +20499,9 @@ __ai __attribute__((target("neon"))) int16x4_t vmvn_s16(int16x4_t __p0) {
20478	20499	#else
20479	20500	__ai __attribute__((target("neon"))) int16x4_t vmvn_s16(int16x4_t __p0) {
20480	20501	int16x4_t __ret;
20481		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	20502	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
20482	20503	__ret = ~__rev0;
20483		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	20504	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
20484	20505	return __ret;
20485	20506	}
20486	20507	#endif
...	...	@@ -20494,9 +20515,9 @@ __ai __attribute__((target("neon"))) int8x16_t vnegq_s8(int8x16_t __p0) {
20494	20515	#else
20495	20516	__ai __attribute__((target("neon"))) int8x16_t vnegq_s8(int8x16_t __p0) {
20496	20517	int8x16_t __ret;
20497		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	20518	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
20498	20519	__ret = -__rev0;
20499		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	20520	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
20500	20521	return __ret;
20501	20522	}
20502	20523	#endif
...	...	@@ -20510,9 +20531,9 @@ __ai __attribute__((target("neon"))) float32x4_t vnegq_f32(float32x4_t __p0) {
20510	20531	#else
20511	20532	__ai __attribute__((target("neon"))) float32x4_t vnegq_f32(float32x4_t __p0) {
20512	20533	float32x4_t __ret;
20513		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	20534	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
20514	20535	__ret = -__rev0;
20515		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	20536	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
20516	20537	return __ret;
20517	20538	}
20518	20539	#endif
...	...	@@ -20526,9 +20547,9 @@ __ai __attribute__((target("neon"))) int32x4_t vnegq_s32(int32x4_t __p0) {
20526	20547	#else
20527	20548	__ai __attribute__((target("neon"))) int32x4_t vnegq_s32(int32x4_t __p0) {
20528	20549	int32x4_t __ret;
20529		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	20550	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
20530	20551	__ret = -__rev0;
20531		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	20552	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
20532	20553	return __ret;
20533	20554	}
20534	20555	#endif
...	...	@@ -20542,9 +20563,9 @@ __ai __attribute__((target("neon"))) int16x8_t vnegq_s16(int16x8_t __p0) {
20542	20563	#else
20543	20564	__ai __attribute__((target("neon"))) int16x8_t vnegq_s16(int16x8_t __p0) {
20544	20565	int16x8_t __ret;
20545		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	20566	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
20546	20567	__ret = -__rev0;
20547		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	20568	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
20548	20569	return __ret;
20549	20570	}
20550	20571	#endif
...	...	@@ -20558,9 +20579,9 @@ __ai __attribute__((target("neon"))) int8x8_t vneg_s8(int8x8_t __p0) {
20558	20579	#else
20559	20580	__ai __attribute__((target("neon"))) int8x8_t vneg_s8(int8x8_t __p0) {
20560	20581	int8x8_t __ret;
20561		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	20582	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
20562	20583	__ret = -__rev0;
20563		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	20584	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
20564	20585	return __ret;
20565	20586	}
20566	20587	#endif
...	...	@@ -20574,9 +20595,9 @@ __ai __attribute__((target("neon"))) float32x2_t vneg_f32(float32x2_t __p0) {
20574	20595	#else
20575	20596	__ai __attribute__((target("neon"))) float32x2_t vneg_f32(float32x2_t __p0) {
20576	20597	float32x2_t __ret;
20577		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
	20598	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
20578	20599	__ret = -__rev0;
20579		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	20600	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
20580	20601	return __ret;
20581	20602	}
20582	20603	#endif
...	...	@@ -20590,9 +20611,9 @@ __ai __attribute__((target("neon"))) int32x2_t vneg_s32(int32x2_t __p0) {
20590	20611	#else
20591	20612	__ai __attribute__((target("neon"))) int32x2_t vneg_s32(int32x2_t __p0) {
20592	20613	int32x2_t __ret;
20593		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
	20614	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
20594	20615	__ret = -__rev0;
20595		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	20616	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
20596	20617	return __ret;
20597	20618	}
20598	20619	#endif
...	...	@@ -20606,9 +20627,9 @@ __ai __attribute__((target("neon"))) int16x4_t vneg_s16(int16x4_t __p0) {
20606	20627	#else
20607	20628	__ai __attribute__((target("neon"))) int16x4_t vneg_s16(int16x4_t __p0) {
20608	20629	int16x4_t __ret;
20609		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	20630	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
20610	20631	__ret = -__rev0;
20611		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	20632	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
20612	20633	return __ret;
20613	20634	}
20614	20635	#endif
...	...	@@ -20622,10 +20643,10 @@ __ai __attribute__((target("neon"))) uint8x16_t vornq_u8(uint8x16_t __p0, uint8x
20622	20643	#else
20623	20644	__ai __attribute__((target("neon"))) uint8x16_t vornq_u8(uint8x16_t __p0, uint8x16_t __p1) {
20624	20645	uint8x16_t __ret;
20625		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
20626		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	20646	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	20647	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
20627	20648	__ret = __rev0 \| ~__rev1;
20628		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	20649	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
20629	20650	return __ret;
20630	20651	}
20631	20652	#endif
...	...	@@ -20639,10 +20660,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vornq_u32(uint32x4_t __p0, uint3
20639	20660	#else
20640	20661	__ai __attribute__((target("neon"))) uint32x4_t vornq_u32(uint32x4_t __p0, uint32x4_t __p1) {
20641	20662	uint32x4_t __ret;
20642		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
20643		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	20663	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	20664	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
20644	20665	__ret = __rev0 \| ~__rev1;
20645		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	20666	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
20646	20667	return __ret;
20647	20668	}
20648	20669	#endif
...	...	@@ -20656,10 +20677,10 @@ __ai __attribute__((target("neon"))) uint64x2_t vornq_u64(uint64x2_t __p0, uint6
20656	20677	#else
20657	20678	__ai __attribute__((target("neon"))) uint64x2_t vornq_u64(uint64x2_t __p0, uint64x2_t __p1) {
20658	20679	uint64x2_t __ret;
20659		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
20660		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	20680	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	20681	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
20661	20682	__ret = __rev0 \| ~__rev1;
20662		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	20683	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
20663	20684	return __ret;
20664	20685	}
20665	20686	#endif
...	...	@@ -20673,10 +20694,10 @@ __ai __attribute__((target("neon"))) uint16x8_t vornq_u16(uint16x8_t __p0, uint1
20673	20694	#else
20674	20695	__ai __attribute__((target("neon"))) uint16x8_t vornq_u16(uint16x8_t __p0, uint16x8_t __p1) {
20675	20696	uint16x8_t __ret;
20676		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
20677		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	20697	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	20698	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
20678	20699	__ret = __rev0 \| ~__rev1;
20679		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	20700	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
20680	20701	return __ret;
20681	20702	}
20682	20703	#endif
...	...	@@ -20690,10 +20711,10 @@ __ai __attribute__((target("neon"))) int8x16_t vornq_s8(int8x16_t __p0, int8x16_
20690	20711	#else
20691	20712	__ai __attribute__((target("neon"))) int8x16_t vornq_s8(int8x16_t __p0, int8x16_t __p1) {
20692	20713	int8x16_t __ret;
20693		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
20694		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	20714	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	20715	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
20695	20716	__ret = __rev0 \| ~__rev1;
20696		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	20717	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
20697	20718	return __ret;
20698	20719	}
20699	20720	#endif
...	...	@@ -20707,10 +20728,10 @@ __ai __attribute__((target("neon"))) int32x4_t vornq_s32(int32x4_t __p0, int32x4
20707	20728	#else
20708	20729	__ai __attribute__((target("neon"))) int32x4_t vornq_s32(int32x4_t __p0, int32x4_t __p1) {
20709	20730	int32x4_t __ret;
20710		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
20711		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	20731	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	20732	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
20712	20733	__ret = __rev0 \| ~__rev1;
20713		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	20734	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
20714	20735	return __ret;
20715	20736	}
20716	20737	#endif
...	...	@@ -20724,10 +20745,10 @@ __ai __attribute__((target("neon"))) int64x2_t vornq_s64(int64x2_t __p0, int64x2
20724	20745	#else
20725	20746	__ai __attribute__((target("neon"))) int64x2_t vornq_s64(int64x2_t __p0, int64x2_t __p1) {
20726	20747	int64x2_t __ret;
20727		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
20728		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	20748	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	20749	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
20729	20750	__ret = __rev0 \| ~__rev1;
20730		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	20751	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
20731	20752	return __ret;
20732	20753	}
20733	20754	#endif
...	...	@@ -20741,10 +20762,10 @@ __ai __attribute__((target("neon"))) int16x8_t vornq_s16(int16x8_t __p0, int16x8
20741	20762	#else
20742	20763	__ai __attribute__((target("neon"))) int16x8_t vornq_s16(int16x8_t __p0, int16x8_t __p1) {
20743	20764	int16x8_t __ret;
20744		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
20745		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	20765	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	20766	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
20746	20767	__ret = __rev0 \| ~__rev1;
20747		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	20768	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
20748	20769	return __ret;
20749	20770	}
20750	20771	#endif
...	...	@@ -20758,10 +20779,10 @@ __ai __attribute__((target("neon"))) uint8x8_t vorn_u8(uint8x8_t __p0, uint8x8_t
20758	20779	#else
20759	20780	__ai __attribute__((target("neon"))) uint8x8_t vorn_u8(uint8x8_t __p0, uint8x8_t __p1) {
20760	20781	uint8x8_t __ret;
20761		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
20762		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	20782	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	20783	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
20763	20784	__ret = __rev0 \| ~__rev1;
20764		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	20785	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
20765	20786	return __ret;
20766	20787	}
20767	20788	#endif
...	...	@@ -20775,10 +20796,10 @@ __ai __attribute__((target("neon"))) uint32x2_t vorn_u32(uint32x2_t __p0, uint32
20775	20796	#else
20776	20797	__ai __attribute__((target("neon"))) uint32x2_t vorn_u32(uint32x2_t __p0, uint32x2_t __p1) {
20777	20798	uint32x2_t __ret;
20778		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
20779		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	20799	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	20800	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
20780	20801	__ret = __rev0 \| ~__rev1;
20781		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	20802	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
20782	20803	return __ret;
20783	20804	}
20784	20805	#endif
...	...	@@ -20797,10 +20818,10 @@ __ai __attribute__((target("neon"))) uint16x4_t vorn_u16(uint16x4_t __p0, uint16
20797	20818	#else
20798	20819	__ai __attribute__((target("neon"))) uint16x4_t vorn_u16(uint16x4_t __p0, uint16x4_t __p1) {
20799	20820	uint16x4_t __ret;
20800		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
20801		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	20821	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	20822	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
20802	20823	__ret = __rev0 \| ~__rev1;
20803		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	20824	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
20804	20825	return __ret;
20805	20826	}
20806	20827	#endif
...	...	@@ -20814,10 +20835,10 @@ __ai __attribute__((target("neon"))) int8x8_t vorn_s8(int8x8_t __p0, int8x8_t __
20814	20835	#else
20815	20836	__ai __attribute__((target("neon"))) int8x8_t vorn_s8(int8x8_t __p0, int8x8_t __p1) {
20816	20837	int8x8_t __ret;
20817		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
20818		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	20838	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	20839	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
20819	20840	__ret = __rev0 \| ~__rev1;
20820		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	20841	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
20821	20842	return __ret;
20822	20843	}
20823	20844	#endif
...	...	@@ -20831,10 +20852,10 @@ __ai __attribute__((target("neon"))) int32x2_t vorn_s32(int32x2_t __p0, int32x2_
20831	20852	#else
20832	20853	__ai __attribute__((target("neon"))) int32x2_t vorn_s32(int32x2_t __p0, int32x2_t __p1) {
20833	20854	int32x2_t __ret;
20834		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
20835		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	20855	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	20856	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
20836	20857	__ret = __rev0 \| ~__rev1;
20837		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	20858	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
20838	20859	return __ret;
20839	20860	}
20840	20861	#endif
...	...	@@ -20853,10 +20874,10 @@ __ai __attribute__((target("neon"))) int16x4_t vorn_s16(int16x4_t __p0, int16x4_
20853	20874	#else
20854	20875	__ai __attribute__((target("neon"))) int16x4_t vorn_s16(int16x4_t __p0, int16x4_t __p1) {
20855	20876	int16x4_t __ret;
20856		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
20857		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	20877	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	20878	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
20858	20879	__ret = __rev0 \| ~__rev1;
20859		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	20880	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
20860	20881	return __ret;
20861	20882	}
20862	20883	#endif
...	...	@@ -20870,10 +20891,10 @@ __ai __attribute__((target("neon"))) uint8x16_t vorrq_u8(uint8x16_t __p0, uint8x
20870	20891	#else
20871	20892	__ai __attribute__((target("neon"))) uint8x16_t vorrq_u8(uint8x16_t __p0, uint8x16_t __p1) {
20872	20893	uint8x16_t __ret;
20873		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
20874		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	20894	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	20895	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
20875	20896	__ret = __rev0 \| __rev1;
20876		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	20897	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
20877	20898	return __ret;
20878	20899	}
20879	20900	#endif
...	...	@@ -20887,10 +20908,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vorrq_u32(uint32x4_t __p0, uint3
20887	20908	#else
20888	20909	__ai __attribute__((target("neon"))) uint32x4_t vorrq_u32(uint32x4_t __p0, uint32x4_t __p1) {
20889	20910	uint32x4_t __ret;
20890		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
20891		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	20911	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	20912	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
20892	20913	__ret = __rev0 \| __rev1;
20893		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	20914	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
20894	20915	return __ret;
20895	20916	}
20896	20917	#endif
...	...	@@ -20904,10 +20925,10 @@ __ai __attribute__((target("neon"))) uint64x2_t vorrq_u64(uint64x2_t __p0, uint6
20904	20925	#else
20905	20926	__ai __attribute__((target("neon"))) uint64x2_t vorrq_u64(uint64x2_t __p0, uint64x2_t __p1) {
20906	20927	uint64x2_t __ret;
20907		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
20908		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	20928	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	20929	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
20909	20930	__ret = __rev0 \| __rev1;
20910		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	20931	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
20911	20932	return __ret;
20912	20933	}
20913	20934	#endif
...	...	@@ -20921,10 +20942,10 @@ __ai __attribute__((target("neon"))) uint16x8_t vorrq_u16(uint16x8_t __p0, uint1
20921	20942	#else
20922	20943	__ai __attribute__((target("neon"))) uint16x8_t vorrq_u16(uint16x8_t __p0, uint16x8_t __p1) {
20923	20944	uint16x8_t __ret;
20924		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
20925		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	20945	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	20946	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
20926	20947	__ret = __rev0 \| __rev1;
20927		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	20948	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
20928	20949	return __ret;
20929	20950	}
20930	20951	#endif
...	...	@@ -20938,10 +20959,10 @@ __ai __attribute__((target("neon"))) int8x16_t vorrq_s8(int8x16_t __p0, int8x16_
20938	20959	#else
20939	20960	__ai __attribute__((target("neon"))) int8x16_t vorrq_s8(int8x16_t __p0, int8x16_t __p1) {
20940	20961	int8x16_t __ret;
20941		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
20942		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	20962	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	20963	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
20943	20964	__ret = __rev0 \| __rev1;
20944		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	20965	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
20945	20966	return __ret;
20946	20967	}
20947	20968	#endif
...	...	@@ -20955,10 +20976,10 @@ __ai __attribute__((target("neon"))) int32x4_t vorrq_s32(int32x4_t __p0, int32x4
20955	20976	#else
20956	20977	__ai __attribute__((target("neon"))) int32x4_t vorrq_s32(int32x4_t __p0, int32x4_t __p1) {
20957	20978	int32x4_t __ret;
20958		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
20959		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	20979	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	20980	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
20960	20981	__ret = __rev0 \| __rev1;
20961		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	20982	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
20962	20983	return __ret;
20963	20984	}
20964	20985	#endif
...	...	@@ -20972,10 +20993,10 @@ __ai __attribute__((target("neon"))) int64x2_t vorrq_s64(int64x2_t __p0, int64x2
20972	20993	#else
20973	20994	__ai __attribute__((target("neon"))) int64x2_t vorrq_s64(int64x2_t __p0, int64x2_t __p1) {
20974	20995	int64x2_t __ret;
20975		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
20976		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	20996	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	20997	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
20977	20998	__ret = __rev0 \| __rev1;
20978		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	20999	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
20979	21000	return __ret;
20980	21001	}
20981	21002	#endif
...	...	@@ -20989,10 +21010,10 @@ __ai __attribute__((target("neon"))) int16x8_t vorrq_s16(int16x8_t __p0, int16x8
20989	21010	#else
20990	21011	__ai __attribute__((target("neon"))) int16x8_t vorrq_s16(int16x8_t __p0, int16x8_t __p1) {
20991	21012	int16x8_t __ret;
20992		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
20993		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	21013	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	21014	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
20994	21015	__ret = __rev0 \| __rev1;
20995		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	21016	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
20996	21017	return __ret;
20997	21018	}
20998	21019	#endif
...	...	@@ -21006,10 +21027,10 @@ __ai __attribute__((target("neon"))) uint8x8_t vorr_u8(uint8x8_t __p0, uint8x8_t
21006	21027	#else
21007	21028	__ai __attribute__((target("neon"))) uint8x8_t vorr_u8(uint8x8_t __p0, uint8x8_t __p1) {
21008	21029	uint8x8_t __ret;
21009		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
21010		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	21030	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	21031	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
21011	21032	__ret = __rev0 \| __rev1;
21012		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	21033	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
21013	21034	return __ret;
21014	21035	}
21015	21036	#endif
...	...	@@ -21023,10 +21044,10 @@ __ai __attribute__((target("neon"))) uint32x2_t vorr_u32(uint32x2_t __p0, uint32
21023	21044	#else
21024	21045	__ai __attribute__((target("neon"))) uint32x2_t vorr_u32(uint32x2_t __p0, uint32x2_t __p1) {
21025	21046	uint32x2_t __ret;
21026		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
21027		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	21047	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	21048	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
21028	21049	__ret = __rev0 \| __rev1;
21029		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	21050	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
21030	21051	return __ret;
21031	21052	}
21032	21053	#endif
...	...	@@ -21045,10 +21066,10 @@ __ai __attribute__((target("neon"))) uint16x4_t vorr_u16(uint16x4_t __p0, uint16
21045	21066	#else
21046	21067	__ai __attribute__((target("neon"))) uint16x4_t vorr_u16(uint16x4_t __p0, uint16x4_t __p1) {
21047	21068	uint16x4_t __ret;
21048		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
21049		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	21069	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	21070	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
21050	21071	__ret = __rev0 \| __rev1;
21051		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	21072	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
21052	21073	return __ret;
21053	21074	}
21054	21075	#endif
...	...	@@ -21062,10 +21083,10 @@ __ai __attribute__((target("neon"))) int8x8_t vorr_s8(int8x8_t __p0, int8x8_t __
21062	21083	#else
21063	21084	__ai __attribute__((target("neon"))) int8x8_t vorr_s8(int8x8_t __p0, int8x8_t __p1) {
21064	21085	int8x8_t __ret;
21065		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
21066		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	21086	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	21087	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
21067	21088	__ret = __rev0 \| __rev1;
21068		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	21089	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
21069	21090	return __ret;
21070	21091	}
21071	21092	#endif
...	...	@@ -21079,10 +21100,10 @@ __ai __attribute__((target("neon"))) int32x2_t vorr_s32(int32x2_t __p0, int32x2_
21079	21100	#else
21080	21101	__ai __attribute__((target("neon"))) int32x2_t vorr_s32(int32x2_t __p0, int32x2_t __p1) {
21081	21102	int32x2_t __ret;
21082		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
21083		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	21103	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	21104	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
21084	21105	__ret = __rev0 \| __rev1;
21085		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	21106	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
21086	21107	return __ret;
21087	21108	}
21088	21109	#endif
...	...	@@ -21101,10 +21122,10 @@ __ai __attribute__((target("neon"))) int16x4_t vorr_s16(int16x4_t __p0, int16x4_
21101	21122	#else
21102	21123	__ai __attribute__((target("neon"))) int16x4_t vorr_s16(int16x4_t __p0, int16x4_t __p1) {
21103	21124	int16x4_t __ret;
21104		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
21105		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	21125	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	21126	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
21106	21127	__ret = __rev0 \| __rev1;
21107		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	21128	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
21108	21129	return __ret;
21109	21130	}
21110	21131	#endif
...	...	@@ -21112,16 +21133,16 @@ __ai __attribute__((target("neon"))) int16x4_t vorr_s16(int16x4_t __p0, int16x4_
21112	21133	#ifdef __LITTLE_ENDIAN__
21113	21134	__ai __attribute__((target("neon"))) uint16x8_t vpadalq_u8(uint16x8_t __p0, uint8x16_t __p1) {
21114	21135	uint16x8_t __ret;
21115		__ret = (uint16x8_t) __builtin_neon_vpadalq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
	21136	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vpadalq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 49));
21116	21137	return __ret;
21117	21138	}
21118	21139	#else
21119	21140	__ai __attribute__((target("neon"))) uint16x8_t vpadalq_u8(uint16x8_t __p0, uint8x16_t __p1) {
21120	21141	uint16x8_t __ret;
21121		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
21122		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
21123		__ret = (uint16x8_t) __builtin_neon_vpadalq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
21124		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	21142	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	21143	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	21144	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vpadalq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 49));
	21145	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
21125	21146	return __ret;
21126	21147	}
21127	21148	#endif
...	...	@@ -21129,16 +21150,16 @@ __ai __attribute__((target("neon"))) uint16x8_t vpadalq_u8(uint16x8_t __p0, uint
21129	21150	#ifdef __LITTLE_ENDIAN__
21130	21151	__ai __attribute__((target("neon"))) uint64x2_t vpadalq_u32(uint64x2_t __p0, uint32x4_t __p1) {
21131	21152	uint64x2_t __ret;
21132		__ret = (uint64x2_t) __builtin_neon_vpadalq_v((int8x16_t)__p0, (int8x16_t)__p1, 51);
	21153	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vpadalq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 51));
21133	21154	return __ret;
21134	21155	}
21135	21156	#else
21136	21157	__ai __attribute__((target("neon"))) uint64x2_t vpadalq_u32(uint64x2_t __p0, uint32x4_t __p1) {
21137	21158	uint64x2_t __ret;
21138		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
21139		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
21140		__ret = (uint64x2_t) __builtin_neon_vpadalq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 51);
21141		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	21159	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	21160	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	21161	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vpadalq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 51));
	21162	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
21142	21163	return __ret;
21143	21164	}
21144	21165	#endif
...	...	@@ -21146,16 +21167,16 @@ __ai __attribute__((target("neon"))) uint64x2_t vpadalq_u32(uint64x2_t __p0, uin
21146	21167	#ifdef __LITTLE_ENDIAN__
21147	21168	__ai __attribute__((target("neon"))) uint32x4_t vpadalq_u16(uint32x4_t __p0, uint16x8_t __p1) {
21148	21169	uint32x4_t __ret;
21149		__ret = (uint32x4_t) __builtin_neon_vpadalq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
	21170	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vpadalq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50));
21150	21171	return __ret;
21151	21172	}
21152	21173	#else
21153	21174	__ai __attribute__((target("neon"))) uint32x4_t vpadalq_u16(uint32x4_t __p0, uint16x8_t __p1) {
21154	21175	uint32x4_t __ret;
21155		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
21156		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
21157		__ret = (uint32x4_t) __builtin_neon_vpadalq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
21158		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	21176	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	21177	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	21178	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vpadalq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 50));
	21179	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
21159	21180	return __ret;
21160	21181	}
21161	21182	#endif
...	...	@@ -21163,16 +21184,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vpadalq_u16(uint32x4_t __p0, uin
21163	21184	#ifdef __LITTLE_ENDIAN__
21164	21185	__ai __attribute__((target("neon"))) int16x8_t vpadalq_s8(int16x8_t __p0, int8x16_t __p1) {
21165	21186	int16x8_t __ret;
21166		__ret = (int16x8_t) __builtin_neon_vpadalq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
	21187	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vpadalq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 33));
21167	21188	return __ret;
21168	21189	}
21169	21190	#else
21170	21191	__ai __attribute__((target("neon"))) int16x8_t vpadalq_s8(int16x8_t __p0, int8x16_t __p1) {
21171	21192	int16x8_t __ret;
21172		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
21173		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
21174		__ret = (int16x8_t) __builtin_neon_vpadalq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
21175		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	21193	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	21194	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	21195	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vpadalq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 33));
	21196	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
21176	21197	return __ret;
21177	21198	}
21178	21199	#endif
...	...	@@ -21180,16 +21201,16 @@ __ai __attribute__((target("neon"))) int16x8_t vpadalq_s8(int16x8_t __p0, int8x1
21180	21201	#ifdef __LITTLE_ENDIAN__
21181	21202	__ai __attribute__((target("neon"))) int64x2_t vpadalq_s32(int64x2_t __p0, int32x4_t __p1) {
21182	21203	int64x2_t __ret;
21183		__ret = (int64x2_t) __builtin_neon_vpadalq_v((int8x16_t)__p0, (int8x16_t)__p1, 35);
	21204	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vpadalq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 35));
21184	21205	return __ret;
21185	21206	}
21186	21207	#else
21187	21208	__ai __attribute__((target("neon"))) int64x2_t vpadalq_s32(int64x2_t __p0, int32x4_t __p1) {
21188	21209	int64x2_t __ret;
21189		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
21190		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
21191		__ret = (int64x2_t) __builtin_neon_vpadalq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 35);
21192		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	21210	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	21211	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	21212	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vpadalq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 35));
	21213	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
21193	21214	return __ret;
21194	21215	}
21195	21216	#endif
...	...	@@ -21197,16 +21218,16 @@ __ai __attribute__((target("neon"))) int64x2_t vpadalq_s32(int64x2_t __p0, int32
21197	21218	#ifdef __LITTLE_ENDIAN__
21198	21219	__ai __attribute__((target("neon"))) int32x4_t vpadalq_s16(int32x4_t __p0, int16x8_t __p1) {
21199	21220	int32x4_t __ret;
21200		__ret = (int32x4_t) __builtin_neon_vpadalq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
	21221	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vpadalq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 34));
21201	21222	return __ret;
21202	21223	}
21203	21224	#else
21204	21225	__ai __attribute__((target("neon"))) int32x4_t vpadalq_s16(int32x4_t __p0, int16x8_t __p1) {
21205	21226	int32x4_t __ret;
21206		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
21207		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
21208		__ret = (int32x4_t) __builtin_neon_vpadalq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
21209		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	21227	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	21228	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	21229	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vpadalq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 34));
	21230	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
21210	21231	return __ret;
21211	21232	}
21212	21233	#endif
...	...	@@ -21214,16 +21235,16 @@ __ai __attribute__((target("neon"))) int32x4_t vpadalq_s16(int32x4_t __p0, int16
21214	21235	#ifdef __LITTLE_ENDIAN__
21215	21236	__ai __attribute__((target("neon"))) uint16x4_t vpadal_u8(uint16x4_t __p0, uint8x8_t __p1) {
21216	21237	uint16x4_t __ret;
21217		__ret = (uint16x4_t) __builtin_neon_vpadal_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
	21238	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vpadal_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 17));
21218	21239	return __ret;
21219	21240	}
21220	21241	#else
21221	21242	__ai __attribute__((target("neon"))) uint16x4_t vpadal_u8(uint16x4_t __p0, uint8x8_t __p1) {
21222	21243	uint16x4_t __ret;
21223		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
21224		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
21225		__ret = (uint16x4_t) __builtin_neon_vpadal_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
21226		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	21244	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	21245	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	21246	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vpadal_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 17));
	21247	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
21227	21248	return __ret;
21228	21249	}
21229	21250	#endif
...	...	@@ -21231,14 +21252,14 @@ __ai __attribute__((target("neon"))) uint16x4_t vpadal_u8(uint16x4_t __p0, uint8
21231	21252	#ifdef __LITTLE_ENDIAN__
21232	21253	__ai __attribute__((target("neon"))) uint64x1_t vpadal_u32(uint64x1_t __p0, uint32x2_t __p1) {
21233	21254	uint64x1_t __ret;
21234		__ret = (uint64x1_t) __builtin_neon_vpadal_v((int8x8_t)__p0, (int8x8_t)__p1, 19);
	21255	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vpadal_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 19));
21235	21256	return __ret;
21236	21257	}
21237	21258	#else
21238	21259	__ai __attribute__((target("neon"))) uint64x1_t vpadal_u32(uint64x1_t __p0, uint32x2_t __p1) {
21239	21260	uint64x1_t __ret;
21240		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
21241		__ret = (uint64x1_t) __builtin_neon_vpadal_v((int8x8_t)__p0, (int8x8_t)__rev1, 19);
	21261	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	21262	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vpadal_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __rev1), 19));
21242	21263	return __ret;
21243	21264	}
21244	21265	#endif
...	...	@@ -21246,16 +21267,16 @@ __ai __attribute__((target("neon"))) uint64x1_t vpadal_u32(uint64x1_t __p0, uint
21246	21267	#ifdef __LITTLE_ENDIAN__
21247	21268	__ai __attribute__((target("neon"))) uint32x2_t vpadal_u16(uint32x2_t __p0, uint16x4_t __p1) {
21248	21269	uint32x2_t __ret;
21249		__ret = (uint32x2_t) __builtin_neon_vpadal_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
	21270	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vpadal_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 18));
21250	21271	return __ret;
21251	21272	}
21252	21273	#else
21253	21274	__ai __attribute__((target("neon"))) uint32x2_t vpadal_u16(uint32x2_t __p0, uint16x4_t __p1) {
21254	21275	uint32x2_t __ret;
21255		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
21256		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
21257		__ret = (uint32x2_t) __builtin_neon_vpadal_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
21258		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	21276	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	21277	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	21278	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vpadal_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 18));
	21279	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
21259	21280	return __ret;
21260	21281	}
21261	21282	#endif
...	...	@@ -21263,16 +21284,16 @@ __ai __attribute__((target("neon"))) uint32x2_t vpadal_u16(uint32x2_t __p0, uint
21263	21284	#ifdef __LITTLE_ENDIAN__
21264	21285	__ai __attribute__((target("neon"))) int16x4_t vpadal_s8(int16x4_t __p0, int8x8_t __p1) {
21265	21286	int16x4_t __ret;
21266		__ret = (int16x4_t) __builtin_neon_vpadal_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
	21287	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vpadal_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 1));
21267	21288	return __ret;
21268	21289	}
21269	21290	#else
21270	21291	__ai __attribute__((target("neon"))) int16x4_t vpadal_s8(int16x4_t __p0, int8x8_t __p1) {
21271	21292	int16x4_t __ret;
21272		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
21273		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
21274		__ret = (int16x4_t) __builtin_neon_vpadal_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
21275		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	21293	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	21294	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	21295	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vpadal_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 1));
	21296	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
21276	21297	return __ret;
21277	21298	}
21278	21299	#endif
...	...	@@ -21280,14 +21301,14 @@ __ai __attribute__((target("neon"))) int16x4_t vpadal_s8(int16x4_t __p0, int8x8_
21280	21301	#ifdef __LITTLE_ENDIAN__
21281	21302	__ai __attribute__((target("neon"))) int64x1_t vpadal_s32(int64x1_t __p0, int32x2_t __p1) {
21282	21303	int64x1_t __ret;
21283		__ret = (int64x1_t) __builtin_neon_vpadal_v((int8x8_t)__p0, (int8x8_t)__p1, 3);
	21304	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vpadal_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 3));
21284	21305	return __ret;
21285	21306	}
21286	21307	#else
21287	21308	__ai __attribute__((target("neon"))) int64x1_t vpadal_s32(int64x1_t __p0, int32x2_t __p1) {
21288	21309	int64x1_t __ret;
21289		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
21290		__ret = (int64x1_t) __builtin_neon_vpadal_v((int8x8_t)__p0, (int8x8_t)__rev1, 3);
	21310	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	21311	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vpadal_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __rev1), 3));
21291	21312	return __ret;
21292	21313	}
21293	21314	#endif
...	...	@@ -21295,16 +21316,16 @@ __ai __attribute__((target("neon"))) int64x1_t vpadal_s32(int64x1_t __p0, int32x
21295	21316	#ifdef __LITTLE_ENDIAN__
21296	21317	__ai __attribute__((target("neon"))) int32x2_t vpadal_s16(int32x2_t __p0, int16x4_t __p1) {
21297	21318	int32x2_t __ret;
21298		__ret = (int32x2_t) __builtin_neon_vpadal_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
	21319	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vpadal_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 2));
21299	21320	return __ret;
21300	21321	}
21301	21322	#else
21302	21323	__ai __attribute__((target("neon"))) int32x2_t vpadal_s16(int32x2_t __p0, int16x4_t __p1) {
21303	21324	int32x2_t __ret;
21304		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
21305		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
21306		__ret = (int32x2_t) __builtin_neon_vpadal_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
21307		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	21325	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	21326	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	21327	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vpadal_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 2));
	21328	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
21308	21329	return __ret;
21309	21330	}
21310	21331	#endif
...	...	@@ -21312,16 +21333,16 @@ __ai __attribute__((target("neon"))) int32x2_t vpadal_s16(int32x2_t __p0, int16x
21312	21333	#ifdef __LITTLE_ENDIAN__
21313	21334	__ai __attribute__((target("neon"))) uint8x8_t vpadd_u8(uint8x8_t __p0, uint8x8_t __p1) {
21314	21335	uint8x8_t __ret;
21315		__ret = (uint8x8_t) __builtin_neon_vpadd_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
	21336	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vpadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 16));
21316	21337	return __ret;
21317	21338	}
21318	21339	#else
21319	21340	__ai __attribute__((target("neon"))) uint8x8_t vpadd_u8(uint8x8_t __p0, uint8x8_t __p1) {
21320	21341	uint8x8_t __ret;
21321		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
21322		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
21323		__ret = (uint8x8_t) __builtin_neon_vpadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
21324		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	21342	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	21343	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	21344	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vpadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 16));
	21345	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
21325	21346	return __ret;
21326	21347	}
21327	21348	#endif
...	...	@@ -21329,16 +21350,16 @@ __ai __attribute__((target("neon"))) uint8x8_t vpadd_u8(uint8x8_t __p0, uint8x8_
21329	21350	#ifdef __LITTLE_ENDIAN__
21330	21351	__ai __attribute__((target("neon"))) uint32x2_t vpadd_u32(uint32x2_t __p0, uint32x2_t __p1) {
21331	21352	uint32x2_t __ret;
21332		__ret = (uint32x2_t) __builtin_neon_vpadd_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
	21353	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vpadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 18));
21333	21354	return __ret;
21334	21355	}
21335	21356	#else
21336	21357	__ai __attribute__((target("neon"))) uint32x2_t vpadd_u32(uint32x2_t __p0, uint32x2_t __p1) {
21337	21358	uint32x2_t __ret;
21338		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
21339		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
21340		__ret = (uint32x2_t) __builtin_neon_vpadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
21341		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	21359	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	21360	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	21361	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vpadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 18));
	21362	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
21342	21363	return __ret;
21343	21364	}
21344	21365	#endif
...	...	@@ -21346,16 +21367,16 @@ __ai __attribute__((target("neon"))) uint32x2_t vpadd_u32(uint32x2_t __p0, uint3
21346	21367	#ifdef __LITTLE_ENDIAN__
21347	21368	__ai __attribute__((target("neon"))) uint16x4_t vpadd_u16(uint16x4_t __p0, uint16x4_t __p1) {
21348	21369	uint16x4_t __ret;
21349		__ret = (uint16x4_t) __builtin_neon_vpadd_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
	21370	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vpadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 17));
21350	21371	return __ret;
21351	21372	}
21352	21373	#else
21353	21374	__ai __attribute__((target("neon"))) uint16x4_t vpadd_u16(uint16x4_t __p0, uint16x4_t __p1) {
21354	21375	uint16x4_t __ret;
21355		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
21356		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
21357		__ret = (uint16x4_t) __builtin_neon_vpadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
21358		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	21376	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	21377	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	21378	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vpadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 17));
	21379	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
21359	21380	return __ret;
21360	21381	}
21361	21382	#endif
...	...	@@ -21363,16 +21384,16 @@ __ai __attribute__((target("neon"))) uint16x4_t vpadd_u16(uint16x4_t __p0, uint1
21363	21384	#ifdef __LITTLE_ENDIAN__
21364	21385	__ai __attribute__((target("neon"))) int8x8_t vpadd_s8(int8x8_t __p0, int8x8_t __p1) {
21365	21386	int8x8_t __ret;
21366		__ret = (int8x8_t) __builtin_neon_vpadd_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
	21387	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vpadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 0));
21367	21388	return __ret;
21368	21389	}
21369	21390	#else
21370	21391	__ai __attribute__((target("neon"))) int8x8_t vpadd_s8(int8x8_t __p0, int8x8_t __p1) {
21371	21392	int8x8_t __ret;
21372		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
21373		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
21374		__ret = (int8x8_t) __builtin_neon_vpadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
21375		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	21393	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	21394	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	21395	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vpadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 0));
	21396	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
21376	21397	return __ret;
21377	21398	}
21378	21399	#endif
...	...	@@ -21380,16 +21401,16 @@ __ai __attribute__((target("neon"))) int8x8_t vpadd_s8(int8x8_t __p0, int8x8_t _
21380	21401	#ifdef __LITTLE_ENDIAN__
21381	21402	__ai __attribute__((target("neon"))) float32x2_t vpadd_f32(float32x2_t __p0, float32x2_t __p1) {
21382	21403	float32x2_t __ret;
21383		__ret = (float32x2_t) __builtin_neon_vpadd_v((int8x8_t)__p0, (int8x8_t)__p1, 9);
	21404	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vpadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 9));
21384	21405	return __ret;
21385	21406	}
21386	21407	#else
21387	21408	__ai __attribute__((target("neon"))) float32x2_t vpadd_f32(float32x2_t __p0, float32x2_t __p1) {
21388	21409	float32x2_t __ret;
21389		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
21390		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
21391		__ret = (float32x2_t) __builtin_neon_vpadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 9);
21392		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	21410	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	21411	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	21412	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vpadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 9));
	21413	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
21393	21414	return __ret;
21394	21415	}
21395	21416	#endif
...	...	@@ -21397,16 +21418,16 @@ __ai __attribute__((target("neon"))) float32x2_t vpadd_f32(float32x2_t __p0, flo
21397	21418	#ifdef __LITTLE_ENDIAN__
21398	21419	__ai __attribute__((target("neon"))) int32x2_t vpadd_s32(int32x2_t __p0, int32x2_t __p1) {
21399	21420	int32x2_t __ret;
21400		__ret = (int32x2_t) __builtin_neon_vpadd_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
	21421	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vpadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 2));
21401	21422	return __ret;
21402	21423	}
21403	21424	#else
21404	21425	__ai __attribute__((target("neon"))) int32x2_t vpadd_s32(int32x2_t __p0, int32x2_t __p1) {
21405	21426	int32x2_t __ret;
21406		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
21407		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
21408		__ret = (int32x2_t) __builtin_neon_vpadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
21409		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	21427	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	21428	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	21429	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vpadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 2));
	21430	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
21410	21431	return __ret;
21411	21432	}
21412	21433	#endif
...	...	@@ -21414,16 +21435,16 @@ __ai __attribute__((target("neon"))) int32x2_t vpadd_s32(int32x2_t __p0, int32x2
21414	21435	#ifdef __LITTLE_ENDIAN__
21415	21436	__ai __attribute__((target("neon"))) int16x4_t vpadd_s16(int16x4_t __p0, int16x4_t __p1) {
21416	21437	int16x4_t __ret;
21417		__ret = (int16x4_t) __builtin_neon_vpadd_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
	21438	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vpadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 1));
21418	21439	return __ret;
21419	21440	}
21420	21441	#else
21421	21442	__ai __attribute__((target("neon"))) int16x4_t vpadd_s16(int16x4_t __p0, int16x4_t __p1) {
21422	21443	int16x4_t __ret;
21423		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
21424		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
21425		__ret = (int16x4_t) __builtin_neon_vpadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
21426		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	21444	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	21445	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	21446	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vpadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 1));
	21447	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
21427	21448	return __ret;
21428	21449	}
21429	21450	#endif
...	...	@@ -21431,15 +21452,15 @@ __ai __attribute__((target("neon"))) int16x4_t vpadd_s16(int16x4_t __p0, int16x4
21431	21452	#ifdef __LITTLE_ENDIAN__
21432	21453	__ai __attribute__((target("neon"))) uint16x8_t vpaddlq_u8(uint8x16_t __p0) {
21433	21454	uint16x8_t __ret;
21434		__ret = (uint16x8_t) __builtin_neon_vpaddlq_v((int8x16_t)__p0, 49);
	21455	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vpaddlq_v(__builtin_bit_cast(int8x16_t, __p0), 49));
21435	21456	return __ret;
21436	21457	}
21437	21458	#else
21438	21459	__ai __attribute__((target("neon"))) uint16x8_t vpaddlq_u8(uint8x16_t __p0) {
21439	21460	uint16x8_t __ret;
21440		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
21441		__ret = (uint16x8_t) __builtin_neon_vpaddlq_v((int8x16_t)__rev0, 49);
21442		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	21461	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	21462	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vpaddlq_v(__builtin_bit_cast(int8x16_t, __rev0), 49));
	21463	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
21443	21464	return __ret;
21444	21465	}
21445	21466	#endif
...	...	@@ -21447,15 +21468,15 @@ __ai __attribute__((target("neon"))) uint16x8_t vpaddlq_u8(uint8x16_t __p0) {
21447	21468	#ifdef __LITTLE_ENDIAN__
21448	21469	__ai __attribute__((target("neon"))) uint64x2_t vpaddlq_u32(uint32x4_t __p0) {
21449	21470	uint64x2_t __ret;
21450		__ret = (uint64x2_t) __builtin_neon_vpaddlq_v((int8x16_t)__p0, 51);
	21471	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vpaddlq_v(__builtin_bit_cast(int8x16_t, __p0), 51));
21451	21472	return __ret;
21452	21473	}
21453	21474	#else
21454	21475	__ai __attribute__((target("neon"))) uint64x2_t vpaddlq_u32(uint32x4_t __p0) {
21455	21476	uint64x2_t __ret;
21456		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
21457		__ret = (uint64x2_t) __builtin_neon_vpaddlq_v((int8x16_t)__rev0, 51);
21458		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	21477	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	21478	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vpaddlq_v(__builtin_bit_cast(int8x16_t, __rev0), 51));
	21479	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
21459	21480	return __ret;
21460	21481	}
21461	21482	#endif
...	...	@@ -21463,15 +21484,15 @@ __ai __attribute__((target("neon"))) uint64x2_t vpaddlq_u32(uint32x4_t __p0) {
21463	21484	#ifdef __LITTLE_ENDIAN__
21464	21485	__ai __attribute__((target("neon"))) uint32x4_t vpaddlq_u16(uint16x8_t __p0) {
21465	21486	uint32x4_t __ret;
21466		__ret = (uint32x4_t) __builtin_neon_vpaddlq_v((int8x16_t)__p0, 50);
	21487	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vpaddlq_v(__builtin_bit_cast(int8x16_t, __p0), 50));
21467	21488	return __ret;
21468	21489	}
21469	21490	#else
21470	21491	__ai __attribute__((target("neon"))) uint32x4_t vpaddlq_u16(uint16x8_t __p0) {
21471	21492	uint32x4_t __ret;
21472		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
21473		__ret = (uint32x4_t) __builtin_neon_vpaddlq_v((int8x16_t)__rev0, 50);
21474		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	21493	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	21494	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vpaddlq_v(__builtin_bit_cast(int8x16_t, __rev0), 50));
	21495	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
21475	21496	return __ret;
21476	21497	}
21477	21498	#endif
...	...	@@ -21479,15 +21500,15 @@ __ai __attribute__((target("neon"))) uint32x4_t vpaddlq_u16(uint16x8_t __p0) {
21479	21500	#ifdef __LITTLE_ENDIAN__
21480	21501	__ai __attribute__((target("neon"))) int16x8_t vpaddlq_s8(int8x16_t __p0) {
21481	21502	int16x8_t __ret;
21482		__ret = (int16x8_t) __builtin_neon_vpaddlq_v((int8x16_t)__p0, 33);
	21503	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vpaddlq_v(__builtin_bit_cast(int8x16_t, __p0), 33));
21483	21504	return __ret;
21484	21505	}
21485	21506	#else
21486	21507	__ai __attribute__((target("neon"))) int16x8_t vpaddlq_s8(int8x16_t __p0) {
21487	21508	int16x8_t __ret;
21488		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
21489		__ret = (int16x8_t) __builtin_neon_vpaddlq_v((int8x16_t)__rev0, 33);
21490		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	21509	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	21510	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vpaddlq_v(__builtin_bit_cast(int8x16_t, __rev0), 33));
	21511	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
21491	21512	return __ret;
21492	21513	}
21493	21514	#endif
...	...	@@ -21495,15 +21516,15 @@ __ai __attribute__((target("neon"))) int16x8_t vpaddlq_s8(int8x16_t __p0) {
21495	21516	#ifdef __LITTLE_ENDIAN__
21496	21517	__ai __attribute__((target("neon"))) int64x2_t vpaddlq_s32(int32x4_t __p0) {
21497	21518	int64x2_t __ret;
21498		__ret = (int64x2_t) __builtin_neon_vpaddlq_v((int8x16_t)__p0, 35);
	21519	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vpaddlq_v(__builtin_bit_cast(int8x16_t, __p0), 35));
21499	21520	return __ret;
21500	21521	}
21501	21522	#else
21502	21523	__ai __attribute__((target("neon"))) int64x2_t vpaddlq_s32(int32x4_t __p0) {
21503	21524	int64x2_t __ret;
21504		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
21505		__ret = (int64x2_t) __builtin_neon_vpaddlq_v((int8x16_t)__rev0, 35);
21506		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	21525	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	21526	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vpaddlq_v(__builtin_bit_cast(int8x16_t, __rev0), 35));
	21527	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
21507	21528	return __ret;
21508	21529	}
21509	21530	#endif
...	...	@@ -21511,15 +21532,15 @@ __ai __attribute__((target("neon"))) int64x2_t vpaddlq_s32(int32x4_t __p0) {
21511	21532	#ifdef __LITTLE_ENDIAN__
21512	21533	__ai __attribute__((target("neon"))) int32x4_t vpaddlq_s16(int16x8_t __p0) {
21513	21534	int32x4_t __ret;
21514		__ret = (int32x4_t) __builtin_neon_vpaddlq_v((int8x16_t)__p0, 34);
	21535	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vpaddlq_v(__builtin_bit_cast(int8x16_t, __p0), 34));
21515	21536	return __ret;
21516	21537	}
21517	21538	#else
21518	21539	__ai __attribute__((target("neon"))) int32x4_t vpaddlq_s16(int16x8_t __p0) {
21519	21540	int32x4_t __ret;
21520		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
21521		__ret = (int32x4_t) __builtin_neon_vpaddlq_v((int8x16_t)__rev0, 34);
21522		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	21541	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	21542	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vpaddlq_v(__builtin_bit_cast(int8x16_t, __rev0), 34));
	21543	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
21523	21544	return __ret;
21524	21545	}
21525	21546	#endif
...	...	@@ -21527,15 +21548,15 @@ __ai __attribute__((target("neon"))) int32x4_t vpaddlq_s16(int16x8_t __p0) {
21527	21548	#ifdef __LITTLE_ENDIAN__
21528	21549	__ai __attribute__((target("neon"))) uint16x4_t vpaddl_u8(uint8x8_t __p0) {
21529	21550	uint16x4_t __ret;
21530		__ret = (uint16x4_t) __builtin_neon_vpaddl_v((int8x8_t)__p0, 17);
	21551	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vpaddl_v(__builtin_bit_cast(int8x8_t, __p0), 17));
21531	21552	return __ret;
21532	21553	}
21533	21554	#else
21534	21555	__ai __attribute__((target("neon"))) uint16x4_t vpaddl_u8(uint8x8_t __p0) {
21535	21556	uint16x4_t __ret;
21536		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
21537		__ret = (uint16x4_t) __builtin_neon_vpaddl_v((int8x8_t)__rev0, 17);
21538		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	21557	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	21558	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vpaddl_v(__builtin_bit_cast(int8x8_t, __rev0), 17));
	21559	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
21539	21560	return __ret;
21540	21561	}
21541	21562	#endif
...	...	@@ -21543,14 +21564,14 @@ __ai __attribute__((target("neon"))) uint16x4_t vpaddl_u8(uint8x8_t __p0) {
21543	21564	#ifdef __LITTLE_ENDIAN__
21544	21565	__ai __attribute__((target("neon"))) uint64x1_t vpaddl_u32(uint32x2_t __p0) {
21545	21566	uint64x1_t __ret;
21546		__ret = (uint64x1_t) __builtin_neon_vpaddl_v((int8x8_t)__p0, 19);
	21567	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vpaddl_v(__builtin_bit_cast(int8x8_t, __p0), 19));
21547	21568	return __ret;
21548	21569	}
21549	21570	#else
21550	21571	__ai __attribute__((target("neon"))) uint64x1_t vpaddl_u32(uint32x2_t __p0) {
21551	21572	uint64x1_t __ret;
21552		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
21553		__ret = (uint64x1_t) __builtin_neon_vpaddl_v((int8x8_t)__rev0, 19);
	21573	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	21574	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vpaddl_v(__builtin_bit_cast(int8x8_t, __rev0), 19));
21554	21575	return __ret;
21555	21576	}
21556	21577	#endif
...	...	@@ -21558,15 +21579,15 @@ __ai __attribute__((target("neon"))) uint64x1_t vpaddl_u32(uint32x2_t __p0) {
21558	21579	#ifdef __LITTLE_ENDIAN__
21559	21580	__ai __attribute__((target("neon"))) uint32x2_t vpaddl_u16(uint16x4_t __p0) {
21560	21581	uint32x2_t __ret;
21561		__ret = (uint32x2_t) __builtin_neon_vpaddl_v((int8x8_t)__p0, 18);
	21582	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vpaddl_v(__builtin_bit_cast(int8x8_t, __p0), 18));
21562	21583	return __ret;
21563	21584	}
21564	21585	#else
21565	21586	__ai __attribute__((target("neon"))) uint32x2_t vpaddl_u16(uint16x4_t __p0) {
21566	21587	uint32x2_t __ret;
21567		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
21568		__ret = (uint32x2_t) __builtin_neon_vpaddl_v((int8x8_t)__rev0, 18);
21569		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	21588	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	21589	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vpaddl_v(__builtin_bit_cast(int8x8_t, __rev0), 18));
	21590	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
21570	21591	return __ret;
21571	21592	}
21572	21593	#endif
...	...	@@ -21574,15 +21595,15 @@ __ai __attribute__((target("neon"))) uint32x2_t vpaddl_u16(uint16x4_t __p0) {
21574	21595	#ifdef __LITTLE_ENDIAN__
21575	21596	__ai __attribute__((target("neon"))) int16x4_t vpaddl_s8(int8x8_t __p0) {
21576	21597	int16x4_t __ret;
21577		__ret = (int16x4_t) __builtin_neon_vpaddl_v((int8x8_t)__p0, 1);
	21598	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vpaddl_v(__builtin_bit_cast(int8x8_t, __p0), 1));
21578	21599	return __ret;
21579	21600	}
21580	21601	#else
21581	21602	__ai __attribute__((target("neon"))) int16x4_t vpaddl_s8(int8x8_t __p0) {
21582	21603	int16x4_t __ret;
21583		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
21584		__ret = (int16x4_t) __builtin_neon_vpaddl_v((int8x8_t)__rev0, 1);
21585		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	21604	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	21605	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vpaddl_v(__builtin_bit_cast(int8x8_t, __rev0), 1));
	21606	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
21586	21607	return __ret;
21587	21608	}
21588	21609	#endif
...	...	@@ -21590,14 +21611,14 @@ __ai __attribute__((target("neon"))) int16x4_t vpaddl_s8(int8x8_t __p0) {
21590	21611	#ifdef __LITTLE_ENDIAN__
21591	21612	__ai __attribute__((target("neon"))) int64x1_t vpaddl_s32(int32x2_t __p0) {
21592	21613	int64x1_t __ret;
21593		__ret = (int64x1_t) __builtin_neon_vpaddl_v((int8x8_t)__p0, 3);
	21614	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vpaddl_v(__builtin_bit_cast(int8x8_t, __p0), 3));
21594	21615	return __ret;
21595	21616	}
21596	21617	#else
21597	21618	__ai __attribute__((target("neon"))) int64x1_t vpaddl_s32(int32x2_t __p0) {
21598	21619	int64x1_t __ret;
21599		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
21600		__ret = (int64x1_t) __builtin_neon_vpaddl_v((int8x8_t)__rev0, 3);
	21620	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	21621	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vpaddl_v(__builtin_bit_cast(int8x8_t, __rev0), 3));
21601	21622	return __ret;
21602	21623	}
21603	21624	#endif
...	...	@@ -21605,15 +21626,15 @@ __ai __attribute__((target("neon"))) int64x1_t vpaddl_s32(int32x2_t __p0) {
21605	21626	#ifdef __LITTLE_ENDIAN__
21606	21627	__ai __attribute__((target("neon"))) int32x2_t vpaddl_s16(int16x4_t __p0) {
21607	21628	int32x2_t __ret;
21608		__ret = (int32x2_t) __builtin_neon_vpaddl_v((int8x8_t)__p0, 2);
	21629	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vpaddl_v(__builtin_bit_cast(int8x8_t, __p0), 2));
21609	21630	return __ret;
21610	21631	}
21611	21632	#else
21612	21633	__ai __attribute__((target("neon"))) int32x2_t vpaddl_s16(int16x4_t __p0) {
21613	21634	int32x2_t __ret;
21614		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
21615		__ret = (int32x2_t) __builtin_neon_vpaddl_v((int8x8_t)__rev0, 2);
21616		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	21635	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	21636	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vpaddl_v(__builtin_bit_cast(int8x8_t, __rev0), 2));
	21637	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
21617	21638	return __ret;
21618	21639	}
21619	21640	#endif
...	...	@@ -21621,16 +21642,16 @@ __ai __attribute__((target("neon"))) int32x2_t vpaddl_s16(int16x4_t __p0) {
21621	21642	#ifdef __LITTLE_ENDIAN__
21622	21643	__ai __attribute__((target("neon"))) uint8x8_t vpmax_u8(uint8x8_t __p0, uint8x8_t __p1) {
21623	21644	uint8x8_t __ret;
21624		__ret = (uint8x8_t) __builtin_neon_vpmax_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
	21645	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vpmax_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 16));
21625	21646	return __ret;
21626	21647	}
21627	21648	#else
21628	21649	__ai __attribute__((target("neon"))) uint8x8_t vpmax_u8(uint8x8_t __p0, uint8x8_t __p1) {
21629	21650	uint8x8_t __ret;
21630		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
21631		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
21632		__ret = (uint8x8_t) __builtin_neon_vpmax_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
21633		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	21651	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	21652	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	21653	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vpmax_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 16));
	21654	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
21634	21655	return __ret;
21635	21656	}
21636	21657	#endif
...	...	@@ -21638,16 +21659,16 @@ __ai __attribute__((target("neon"))) uint8x8_t vpmax_u8(uint8x8_t __p0, uint8x8_
21638	21659	#ifdef __LITTLE_ENDIAN__
21639	21660	__ai __attribute__((target("neon"))) uint32x2_t vpmax_u32(uint32x2_t __p0, uint32x2_t __p1) {
21640	21661	uint32x2_t __ret;
21641		__ret = (uint32x2_t) __builtin_neon_vpmax_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
	21662	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vpmax_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 18));
21642	21663	return __ret;
21643	21664	}
21644	21665	#else
21645	21666	__ai __attribute__((target("neon"))) uint32x2_t vpmax_u32(uint32x2_t __p0, uint32x2_t __p1) {
21646	21667	uint32x2_t __ret;
21647		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
21648		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
21649		__ret = (uint32x2_t) __builtin_neon_vpmax_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
21650		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	21668	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	21669	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	21670	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vpmax_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 18));
	21671	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
21651	21672	return __ret;
21652	21673	}
21653	21674	#endif
...	...	@@ -21655,16 +21676,16 @@ __ai __attribute__((target("neon"))) uint32x2_t vpmax_u32(uint32x2_t __p0, uint3
21655	21676	#ifdef __LITTLE_ENDIAN__
21656	21677	__ai __attribute__((target("neon"))) uint16x4_t vpmax_u16(uint16x4_t __p0, uint16x4_t __p1) {
21657	21678	uint16x4_t __ret;
21658		__ret = (uint16x4_t) __builtin_neon_vpmax_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
	21679	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vpmax_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 17));
21659	21680	return __ret;
21660	21681	}
21661	21682	#else
21662	21683	__ai __attribute__((target("neon"))) uint16x4_t vpmax_u16(uint16x4_t __p0, uint16x4_t __p1) {
21663	21684	uint16x4_t __ret;
21664		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
21665		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
21666		__ret = (uint16x4_t) __builtin_neon_vpmax_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
21667		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	21685	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	21686	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	21687	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vpmax_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 17));
	21688	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
21668	21689	return __ret;
21669	21690	}
21670	21691	#endif
...	...	@@ -21672,16 +21693,16 @@ __ai __attribute__((target("neon"))) uint16x4_t vpmax_u16(uint16x4_t __p0, uint1
21672	21693	#ifdef __LITTLE_ENDIAN__
21673	21694	__ai __attribute__((target("neon"))) int8x8_t vpmax_s8(int8x8_t __p0, int8x8_t __p1) {
21674	21695	int8x8_t __ret;
21675		__ret = (int8x8_t) __builtin_neon_vpmax_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
	21696	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vpmax_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 0));
21676	21697	return __ret;
21677	21698	}
21678	21699	#else
21679	21700	__ai __attribute__((target("neon"))) int8x8_t vpmax_s8(int8x8_t __p0, int8x8_t __p1) {
21680	21701	int8x8_t __ret;
21681		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
21682		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
21683		__ret = (int8x8_t) __builtin_neon_vpmax_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
21684		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	21702	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	21703	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	21704	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vpmax_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 0));
	21705	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
21685	21706	return __ret;
21686	21707	}
21687	21708	#endif
...	...	@@ -21689,16 +21710,16 @@ __ai __attribute__((target("neon"))) int8x8_t vpmax_s8(int8x8_t __p0, int8x8_t _
21689	21710	#ifdef __LITTLE_ENDIAN__
21690	21711	__ai __attribute__((target("neon"))) float32x2_t vpmax_f32(float32x2_t __p0, float32x2_t __p1) {
21691	21712	float32x2_t __ret;
21692		__ret = (float32x2_t) __builtin_neon_vpmax_v((int8x8_t)__p0, (int8x8_t)__p1, 9);
	21713	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vpmax_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 9));
21693	21714	return __ret;
21694	21715	}
21695	21716	#else
21696	21717	__ai __attribute__((target("neon"))) float32x2_t vpmax_f32(float32x2_t __p0, float32x2_t __p1) {
21697	21718	float32x2_t __ret;
21698		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
21699		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
21700		__ret = (float32x2_t) __builtin_neon_vpmax_v((int8x8_t)__rev0, (int8x8_t)__rev1, 9);
21701		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	21719	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	21720	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	21721	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vpmax_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 9));
	21722	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
21702	21723	return __ret;
21703	21724	}
21704	21725	#endif
...	...	@@ -21706,16 +21727,16 @@ __ai __attribute__((target("neon"))) float32x2_t vpmax_f32(float32x2_t __p0, flo
21706	21727	#ifdef __LITTLE_ENDIAN__
21707	21728	__ai __attribute__((target("neon"))) int32x2_t vpmax_s32(int32x2_t __p0, int32x2_t __p1) {
21708	21729	int32x2_t __ret;
21709		__ret = (int32x2_t) __builtin_neon_vpmax_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
	21730	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vpmax_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 2));
21710	21731	return __ret;
21711	21732	}
21712	21733	#else
21713	21734	__ai __attribute__((target("neon"))) int32x2_t vpmax_s32(int32x2_t __p0, int32x2_t __p1) {
21714	21735	int32x2_t __ret;
21715		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
21716		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
21717		__ret = (int32x2_t) __builtin_neon_vpmax_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
21718		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	21736	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	21737	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	21738	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vpmax_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 2));
	21739	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
21719	21740	return __ret;
21720	21741	}
21721	21742	#endif
...	...	@@ -21723,16 +21744,16 @@ __ai __attribute__((target("neon"))) int32x2_t vpmax_s32(int32x2_t __p0, int32x2
21723	21744	#ifdef __LITTLE_ENDIAN__
21724	21745	__ai __attribute__((target("neon"))) int16x4_t vpmax_s16(int16x4_t __p0, int16x4_t __p1) {
21725	21746	int16x4_t __ret;
21726		__ret = (int16x4_t) __builtin_neon_vpmax_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
	21747	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vpmax_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 1));
21727	21748	return __ret;
21728	21749	}
21729	21750	#else
21730	21751	__ai __attribute__((target("neon"))) int16x4_t vpmax_s16(int16x4_t __p0, int16x4_t __p1) {
21731	21752	int16x4_t __ret;
21732		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
21733		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
21734		__ret = (int16x4_t) __builtin_neon_vpmax_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
21735		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	21753	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	21754	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	21755	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vpmax_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 1));
	21756	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
21736	21757	return __ret;
21737	21758	}
21738	21759	#endif
...	...	@@ -21740,16 +21761,16 @@ __ai __attribute__((target("neon"))) int16x4_t vpmax_s16(int16x4_t __p0, int16x4
21740	21761	#ifdef __LITTLE_ENDIAN__
21741	21762	__ai __attribute__((target("neon"))) uint8x8_t vpmin_u8(uint8x8_t __p0, uint8x8_t __p1) {
21742	21763	uint8x8_t __ret;
21743		__ret = (uint8x8_t) __builtin_neon_vpmin_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
	21764	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vpmin_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 16));
21744	21765	return __ret;
21745	21766	}
21746	21767	#else
21747	21768	__ai __attribute__((target("neon"))) uint8x8_t vpmin_u8(uint8x8_t __p0, uint8x8_t __p1) {
21748	21769	uint8x8_t __ret;
21749		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
21750		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
21751		__ret = (uint8x8_t) __builtin_neon_vpmin_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
21752		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	21770	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	21771	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	21772	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vpmin_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 16));
	21773	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
21753	21774	return __ret;
21754	21775	}
21755	21776	#endif
...	...	@@ -21757,16 +21778,16 @@ __ai __attribute__((target("neon"))) uint8x8_t vpmin_u8(uint8x8_t __p0, uint8x8_
21757	21778	#ifdef __LITTLE_ENDIAN__
21758	21779	__ai __attribute__((target("neon"))) uint32x2_t vpmin_u32(uint32x2_t __p0, uint32x2_t __p1) {
21759	21780	uint32x2_t __ret;
21760		__ret = (uint32x2_t) __builtin_neon_vpmin_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
	21781	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vpmin_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 18));
21761	21782	return __ret;
21762	21783	}
21763	21784	#else
21764	21785	__ai __attribute__((target("neon"))) uint32x2_t vpmin_u32(uint32x2_t __p0, uint32x2_t __p1) {
21765	21786	uint32x2_t __ret;
21766		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
21767		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
21768		__ret = (uint32x2_t) __builtin_neon_vpmin_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
21769		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	21787	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	21788	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	21789	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vpmin_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 18));
	21790	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
21770	21791	return __ret;
21771	21792	}
21772	21793	#endif
...	...	@@ -21774,16 +21795,16 @@ __ai __attribute__((target("neon"))) uint32x2_t vpmin_u32(uint32x2_t __p0, uint3
21774	21795	#ifdef __LITTLE_ENDIAN__
21775	21796	__ai __attribute__((target("neon"))) uint16x4_t vpmin_u16(uint16x4_t __p0, uint16x4_t __p1) {
21776	21797	uint16x4_t __ret;
21777		__ret = (uint16x4_t) __builtin_neon_vpmin_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
	21798	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vpmin_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 17));
21778	21799	return __ret;
21779	21800	}
21780	21801	#else
21781	21802	__ai __attribute__((target("neon"))) uint16x4_t vpmin_u16(uint16x4_t __p0, uint16x4_t __p1) {
21782	21803	uint16x4_t __ret;
21783		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
21784		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
21785		__ret = (uint16x4_t) __builtin_neon_vpmin_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
21786		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	21804	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	21805	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	21806	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vpmin_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 17));
	21807	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
21787	21808	return __ret;
21788	21809	}
21789	21810	#endif
...	...	@@ -21791,16 +21812,16 @@ __ai __attribute__((target("neon"))) uint16x4_t vpmin_u16(uint16x4_t __p0, uint1
21791	21812	#ifdef __LITTLE_ENDIAN__
21792	21813	__ai __attribute__((target("neon"))) int8x8_t vpmin_s8(int8x8_t __p0, int8x8_t __p1) {
21793	21814	int8x8_t __ret;
21794		__ret = (int8x8_t) __builtin_neon_vpmin_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
	21815	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vpmin_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 0));
21795	21816	return __ret;
21796	21817	}
21797	21818	#else
21798	21819	__ai __attribute__((target("neon"))) int8x8_t vpmin_s8(int8x8_t __p0, int8x8_t __p1) {
21799	21820	int8x8_t __ret;
21800		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
21801		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
21802		__ret = (int8x8_t) __builtin_neon_vpmin_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
21803		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	21821	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	21822	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	21823	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vpmin_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 0));
	21824	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
21804	21825	return __ret;
21805	21826	}
21806	21827	#endif
...	...	@@ -21808,16 +21829,16 @@ __ai __attribute__((target("neon"))) int8x8_t vpmin_s8(int8x8_t __p0, int8x8_t _
21808	21829	#ifdef __LITTLE_ENDIAN__
21809	21830	__ai __attribute__((target("neon"))) float32x2_t vpmin_f32(float32x2_t __p0, float32x2_t __p1) {
21810	21831	float32x2_t __ret;
21811		__ret = (float32x2_t) __builtin_neon_vpmin_v((int8x8_t)__p0, (int8x8_t)__p1, 9);
	21832	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vpmin_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 9));
21812	21833	return __ret;
21813	21834	}
21814	21835	#else
21815	21836	__ai __attribute__((target("neon"))) float32x2_t vpmin_f32(float32x2_t __p0, float32x2_t __p1) {
21816	21837	float32x2_t __ret;
21817		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
21818		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
21819		__ret = (float32x2_t) __builtin_neon_vpmin_v((int8x8_t)__rev0, (int8x8_t)__rev1, 9);
21820		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	21838	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	21839	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	21840	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vpmin_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 9));
	21841	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
21821	21842	return __ret;
21822	21843	}
21823	21844	#endif
...	...	@@ -21825,16 +21846,16 @@ __ai __attribute__((target("neon"))) float32x2_t vpmin_f32(float32x2_t __p0, flo
21825	21846	#ifdef __LITTLE_ENDIAN__
21826	21847	__ai __attribute__((target("neon"))) int32x2_t vpmin_s32(int32x2_t __p0, int32x2_t __p1) {
21827	21848	int32x2_t __ret;
21828		__ret = (int32x2_t) __builtin_neon_vpmin_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
	21849	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vpmin_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 2));
21829	21850	return __ret;
21830	21851	}
21831	21852	#else
21832	21853	__ai __attribute__((target("neon"))) int32x2_t vpmin_s32(int32x2_t __p0, int32x2_t __p1) {
21833	21854	int32x2_t __ret;
21834		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
21835		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
21836		__ret = (int32x2_t) __builtin_neon_vpmin_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
21837		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	21855	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	21856	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	21857	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vpmin_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 2));
	21858	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
21838	21859	return __ret;
21839	21860	}
21840	21861	#endif
...	...	@@ -21842,16 +21863,16 @@ __ai __attribute__((target("neon"))) int32x2_t vpmin_s32(int32x2_t __p0, int32x2
21842	21863	#ifdef __LITTLE_ENDIAN__
21843	21864	__ai __attribute__((target("neon"))) int16x4_t vpmin_s16(int16x4_t __p0, int16x4_t __p1) {
21844	21865	int16x4_t __ret;
21845		__ret = (int16x4_t) __builtin_neon_vpmin_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
	21866	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vpmin_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 1));
21846	21867	return __ret;
21847	21868	}
21848	21869	#else
21849	21870	__ai __attribute__((target("neon"))) int16x4_t vpmin_s16(int16x4_t __p0, int16x4_t __p1) {
21850	21871	int16x4_t __ret;
21851		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
21852		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
21853		__ret = (int16x4_t) __builtin_neon_vpmin_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
21854		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	21872	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	21873	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	21874	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vpmin_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 1));
	21875	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
21855	21876	return __ret;
21856	21877	}
21857	21878	#endif
...	...	@@ -21859,15 +21880,15 @@ __ai __attribute__((target("neon"))) int16x4_t vpmin_s16(int16x4_t __p0, int16x4
21859	21880	#ifdef __LITTLE_ENDIAN__
21860	21881	__ai __attribute__((target("neon"))) int8x16_t vqabsq_s8(int8x16_t __p0) {
21861	21882	int8x16_t __ret;
21862		__ret = (int8x16_t) __builtin_neon_vqabsq_v((int8x16_t)__p0, 32);
	21883	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vqabsq_v(__builtin_bit_cast(int8x16_t, __p0), 32));
21863	21884	return __ret;
21864	21885	}
21865	21886	#else
21866	21887	__ai __attribute__((target("neon"))) int8x16_t vqabsq_s8(int8x16_t __p0) {
21867	21888	int8x16_t __ret;
21868		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
21869		__ret = (int8x16_t) __builtin_neon_vqabsq_v((int8x16_t)__rev0, 32);
21870		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	21889	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	21890	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vqabsq_v(__builtin_bit_cast(int8x16_t, __rev0), 32));
	21891	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
21871	21892	return __ret;
21872	21893	}
21873	21894	#endif
...	...	@@ -21875,15 +21896,15 @@ __ai __attribute__((target("neon"))) int8x16_t vqabsq_s8(int8x16_t __p0) {
21875	21896	#ifdef __LITTLE_ENDIAN__
21876	21897	__ai __attribute__((target("neon"))) int32x4_t vqabsq_s32(int32x4_t __p0) {
21877	21898	int32x4_t __ret;
21878		__ret = (int32x4_t) __builtin_neon_vqabsq_v((int8x16_t)__p0, 34);
	21899	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqabsq_v(__builtin_bit_cast(int8x16_t, __p0), 34));
21879	21900	return __ret;
21880	21901	}
21881	21902	#else
21882	21903	__ai __attribute__((target("neon"))) int32x4_t vqabsq_s32(int32x4_t __p0) {
21883	21904	int32x4_t __ret;
21884		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
21885		__ret = (int32x4_t) __builtin_neon_vqabsq_v((int8x16_t)__rev0, 34);
21886		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	21905	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	21906	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqabsq_v(__builtin_bit_cast(int8x16_t, __rev0), 34));
	21907	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
21887	21908	return __ret;
21888	21909	}
21889	21910	#endif
...	...	@@ -21891,15 +21912,15 @@ __ai __attribute__((target("neon"))) int32x4_t vqabsq_s32(int32x4_t __p0) {
21891	21912	#ifdef __LITTLE_ENDIAN__
21892	21913	__ai __attribute__((target("neon"))) int16x8_t vqabsq_s16(int16x8_t __p0) {
21893	21914	int16x8_t __ret;
21894		__ret = (int16x8_t) __builtin_neon_vqabsq_v((int8x16_t)__p0, 33);
	21915	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqabsq_v(__builtin_bit_cast(int8x16_t, __p0), 33));
21895	21916	return __ret;
21896	21917	}
21897	21918	#else
21898	21919	__ai __attribute__((target("neon"))) int16x8_t vqabsq_s16(int16x8_t __p0) {
21899	21920	int16x8_t __ret;
21900		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
21901		__ret = (int16x8_t) __builtin_neon_vqabsq_v((int8x16_t)__rev0, 33);
21902		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	21921	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	21922	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqabsq_v(__builtin_bit_cast(int8x16_t, __rev0), 33));
	21923	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
21903	21924	return __ret;
21904	21925	}
21905	21926	#endif
...	...	@@ -21907,15 +21928,15 @@ __ai __attribute__((target("neon"))) int16x8_t vqabsq_s16(int16x8_t __p0) {
21907	21928	#ifdef __LITTLE_ENDIAN__
21908	21929	__ai __attribute__((target("neon"))) int8x8_t vqabs_s8(int8x8_t __p0) {
21909	21930	int8x8_t __ret;
21910		__ret = (int8x8_t) __builtin_neon_vqabs_v((int8x8_t)__p0, 0);
	21931	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqabs_v(__builtin_bit_cast(int8x8_t, __p0), 0));
21911	21932	return __ret;
21912	21933	}
21913	21934	#else
21914	21935	__ai __attribute__((target("neon"))) int8x8_t vqabs_s8(int8x8_t __p0) {
21915	21936	int8x8_t __ret;
21916		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
21917		__ret = (int8x8_t) __builtin_neon_vqabs_v((int8x8_t)__rev0, 0);
21918		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	21937	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	21938	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqabs_v(__builtin_bit_cast(int8x8_t, __rev0), 0));
	21939	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
21919	21940	return __ret;
21920	21941	}
21921	21942	#endif
...	...	@@ -21923,15 +21944,15 @@ __ai __attribute__((target("neon"))) int8x8_t vqabs_s8(int8x8_t __p0) {
21923	21944	#ifdef __LITTLE_ENDIAN__
21924	21945	__ai __attribute__((target("neon"))) int32x2_t vqabs_s32(int32x2_t __p0) {
21925	21946	int32x2_t __ret;
21926		__ret = (int32x2_t) __builtin_neon_vqabs_v((int8x8_t)__p0, 2);
	21947	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqabs_v(__builtin_bit_cast(int8x8_t, __p0), 2));
21927	21948	return __ret;
21928	21949	}
21929	21950	#else
21930	21951	__ai __attribute__((target("neon"))) int32x2_t vqabs_s32(int32x2_t __p0) {
21931	21952	int32x2_t __ret;
21932		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
21933		__ret = (int32x2_t) __builtin_neon_vqabs_v((int8x8_t)__rev0, 2);
21934		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	21953	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	21954	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqabs_v(__builtin_bit_cast(int8x8_t, __rev0), 2));
	21955	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
21935	21956	return __ret;
21936	21957	}
21937	21958	#endif
...	...	@@ -21939,15 +21960,15 @@ __ai __attribute__((target("neon"))) int32x2_t vqabs_s32(int32x2_t __p0) {
21939	21960	#ifdef __LITTLE_ENDIAN__
21940	21961	__ai __attribute__((target("neon"))) int16x4_t vqabs_s16(int16x4_t __p0) {
21941	21962	int16x4_t __ret;
21942		__ret = (int16x4_t) __builtin_neon_vqabs_v((int8x8_t)__p0, 1);
	21963	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqabs_v(__builtin_bit_cast(int8x8_t, __p0), 1));
21943	21964	return __ret;
21944	21965	}
21945	21966	#else
21946	21967	__ai __attribute__((target("neon"))) int16x4_t vqabs_s16(int16x4_t __p0) {
21947	21968	int16x4_t __ret;
21948		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
21949		__ret = (int16x4_t) __builtin_neon_vqabs_v((int8x8_t)__rev0, 1);
21950		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	21969	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	21970	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqabs_v(__builtin_bit_cast(int8x8_t, __rev0), 1));
	21971	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
21951	21972	return __ret;
21952	21973	}
21953	21974	#endif
...	...	@@ -21955,16 +21976,16 @@ __ai __attribute__((target("neon"))) int16x4_t vqabs_s16(int16x4_t __p0) {
21955	21976	#ifdef __LITTLE_ENDIAN__
21956	21977	__ai __attribute__((target("neon"))) uint8x16_t vqaddq_u8(uint8x16_t __p0, uint8x16_t __p1) {
21957	21978	uint8x16_t __ret;
21958		__ret = (uint8x16_t) __builtin_neon_vqaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
	21979	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vqaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 48));
21959	21980	return __ret;
21960	21981	}
21961	21982	#else
21962	21983	__ai __attribute__((target("neon"))) uint8x16_t vqaddq_u8(uint8x16_t __p0, uint8x16_t __p1) {
21963	21984	uint8x16_t __ret;
21964		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
21965		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
21966		__ret = (uint8x16_t) __builtin_neon_vqaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
21967		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	21985	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	21986	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	21987	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vqaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 48));
	21988	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
21968	21989	return __ret;
21969	21990	}
21970	21991	#endif
...	...	@@ -21972,16 +21993,16 @@ __ai __attribute__((target("neon"))) uint8x16_t vqaddq_u8(uint8x16_t __p0, uint8
21972	21993	#ifdef __LITTLE_ENDIAN__
21973	21994	__ai __attribute__((target("neon"))) uint32x4_t vqaddq_u32(uint32x4_t __p0, uint32x4_t __p1) {
21974	21995	uint32x4_t __ret;
21975		__ret = (uint32x4_t) __builtin_neon_vqaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
	21996	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vqaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50));
21976	21997	return __ret;
21977	21998	}
21978	21999	#else
21979	22000	__ai __attribute__((target("neon"))) uint32x4_t vqaddq_u32(uint32x4_t __p0, uint32x4_t __p1) {
21980	22001	uint32x4_t __ret;
21981		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
21982		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
21983		__ret = (uint32x4_t) __builtin_neon_vqaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
21984		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	22002	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	22003	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	22004	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vqaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 50));
	22005	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
21985	22006	return __ret;
21986	22007	}
21987	22008	#endif
...	...	@@ -21989,16 +22010,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vqaddq_u32(uint32x4_t __p0, uint
21989	22010	#ifdef __LITTLE_ENDIAN__
21990	22011	__ai __attribute__((target("neon"))) uint64x2_t vqaddq_u64(uint64x2_t __p0, uint64x2_t __p1) {
21991	22012	uint64x2_t __ret;
21992		__ret = (uint64x2_t) __builtin_neon_vqaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 51);
	22013	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vqaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 51));
21993	22014	return __ret;
21994	22015	}
21995	22016	#else
21996	22017	__ai __attribute__((target("neon"))) uint64x2_t vqaddq_u64(uint64x2_t __p0, uint64x2_t __p1) {
21997	22018	uint64x2_t __ret;
21998		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
21999		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
22000		__ret = (uint64x2_t) __builtin_neon_vqaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 51);
22001		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	22019	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	22020	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	22021	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vqaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 51));
	22022	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
22002	22023	return __ret;
22003	22024	}
22004	22025	#endif
...	...	@@ -22006,16 +22027,16 @@ __ai __attribute__((target("neon"))) uint64x2_t vqaddq_u64(uint64x2_t __p0, uint
22006	22027	#ifdef __LITTLE_ENDIAN__
22007	22028	__ai __attribute__((target("neon"))) uint16x8_t vqaddq_u16(uint16x8_t __p0, uint16x8_t __p1) {
22008	22029	uint16x8_t __ret;
22009		__ret = (uint16x8_t) __builtin_neon_vqaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
	22030	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vqaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 49));
22010	22031	return __ret;
22011	22032	}
22012	22033	#else
22013	22034	__ai __attribute__((target("neon"))) uint16x8_t vqaddq_u16(uint16x8_t __p0, uint16x8_t __p1) {
22014	22035	uint16x8_t __ret;
22015		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
22016		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
22017		__ret = (uint16x8_t) __builtin_neon_vqaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
22018		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	22036	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	22037	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	22038	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vqaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 49));
	22039	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
22019	22040	return __ret;
22020	22041	}
22021	22042	#endif
...	...	@@ -22023,16 +22044,16 @@ __ai __attribute__((target("neon"))) uint16x8_t vqaddq_u16(uint16x8_t __p0, uint
22023	22044	#ifdef __LITTLE_ENDIAN__
22024	22045	__ai __attribute__((target("neon"))) int8x16_t vqaddq_s8(int8x16_t __p0, int8x16_t __p1) {
22025	22046	int8x16_t __ret;
22026		__ret = (int8x16_t) __builtin_neon_vqaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
	22047	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vqaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 32));
22027	22048	return __ret;
22028	22049	}
22029	22050	#else
22030	22051	__ai __attribute__((target("neon"))) int8x16_t vqaddq_s8(int8x16_t __p0, int8x16_t __p1) {
22031	22052	int8x16_t __ret;
22032		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
22033		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
22034		__ret = (int8x16_t) __builtin_neon_vqaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 32);
22035		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	22053	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	22054	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	22055	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vqaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 32));
	22056	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
22036	22057	return __ret;
22037	22058	}
22038	22059	#endif
...	...	@@ -22040,16 +22061,16 @@ __ai __attribute__((target("neon"))) int8x16_t vqaddq_s8(int8x16_t __p0, int8x16
22040	22061	#ifdef __LITTLE_ENDIAN__
22041	22062	__ai __attribute__((target("neon"))) int32x4_t vqaddq_s32(int32x4_t __p0, int32x4_t __p1) {
22042	22063	int32x4_t __ret;
22043		__ret = (int32x4_t) __builtin_neon_vqaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
	22064	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 34));
22044	22065	return __ret;
22045	22066	}
22046	22067	#else
22047	22068	__ai __attribute__((target("neon"))) int32x4_t vqaddq_s32(int32x4_t __p0, int32x4_t __p1) {
22048	22069	int32x4_t __ret;
22049		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22050		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
22051		__ret = (int32x4_t) __builtin_neon_vqaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
22052		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	22070	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	22071	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	22072	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 34));
	22073	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
22053	22074	return __ret;
22054	22075	}
22055	22076	#endif
...	...	@@ -22057,16 +22078,16 @@ __ai __attribute__((target("neon"))) int32x4_t vqaddq_s32(int32x4_t __p0, int32x
22057	22078	#ifdef __LITTLE_ENDIAN__
22058	22079	__ai __attribute__((target("neon"))) int64x2_t vqaddq_s64(int64x2_t __p0, int64x2_t __p1) {
22059	22080	int64x2_t __ret;
22060		__ret = (int64x2_t) __builtin_neon_vqaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 35);
	22081	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vqaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 35));
22061	22082	return __ret;
22062	22083	}
22063	22084	#else
22064	22085	__ai __attribute__((target("neon"))) int64x2_t vqaddq_s64(int64x2_t __p0, int64x2_t __p1) {
22065	22086	int64x2_t __ret;
22066		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
22067		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
22068		__ret = (int64x2_t) __builtin_neon_vqaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 35);
22069		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	22087	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	22088	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	22089	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vqaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 35));
	22090	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
22070	22091	return __ret;
22071	22092	}
22072	22093	#endif
...	...	@@ -22074,16 +22095,16 @@ __ai __attribute__((target("neon"))) int64x2_t vqaddq_s64(int64x2_t __p0, int64x
22074	22095	#ifdef __LITTLE_ENDIAN__
22075	22096	__ai __attribute__((target("neon"))) int16x8_t vqaddq_s16(int16x8_t __p0, int16x8_t __p1) {
22076	22097	int16x8_t __ret;
22077		__ret = (int16x8_t) __builtin_neon_vqaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
	22098	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 33));
22078	22099	return __ret;
22079	22100	}
22080	22101	#else
22081	22102	__ai __attribute__((target("neon"))) int16x8_t vqaddq_s16(int16x8_t __p0, int16x8_t __p1) {
22082	22103	int16x8_t __ret;
22083		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
22084		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
22085		__ret = (int16x8_t) __builtin_neon_vqaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
22086		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	22104	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	22105	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	22106	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 33));
	22107	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
22087	22108	return __ret;
22088	22109	}
22089	22110	#endif
...	...	@@ -22091,16 +22112,16 @@ __ai __attribute__((target("neon"))) int16x8_t vqaddq_s16(int16x8_t __p0, int16x
22091	22112	#ifdef __LITTLE_ENDIAN__
22092	22113	__ai __attribute__((target("neon"))) uint8x8_t vqadd_u8(uint8x8_t __p0, uint8x8_t __p1) {
22093	22114	uint8x8_t __ret;
22094		__ret = (uint8x8_t) __builtin_neon_vqadd_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
	22115	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 16));
22095	22116	return __ret;
22096	22117	}
22097	22118	#else
22098	22119	__ai __attribute__((target("neon"))) uint8x8_t vqadd_u8(uint8x8_t __p0, uint8x8_t __p1) {
22099	22120	uint8x8_t __ret;
22100		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
22101		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
22102		__ret = (uint8x8_t) __builtin_neon_vqadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
22103		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	22121	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	22122	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	22123	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 16));
	22124	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
22104	22125	return __ret;
22105	22126	}
22106	22127	#endif
...	...	@@ -22108,38 +22129,38 @@ __ai __attribute__((target("neon"))) uint8x8_t vqadd_u8(uint8x8_t __p0, uint8x8_
22108	22129	#ifdef __LITTLE_ENDIAN__
22109	22130	__ai __attribute__((target("neon"))) uint32x2_t vqadd_u32(uint32x2_t __p0, uint32x2_t __p1) {
22110	22131	uint32x2_t __ret;
22111		__ret = (uint32x2_t) __builtin_neon_vqadd_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
	22132	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vqadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 18));
22112	22133	return __ret;
22113	22134	}
22114	22135	#else
22115	22136	__ai __attribute__((target("neon"))) uint32x2_t vqadd_u32(uint32x2_t __p0, uint32x2_t __p1) {
22116	22137	uint32x2_t __ret;
22117		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
22118		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
22119		__ret = (uint32x2_t) __builtin_neon_vqadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
22120		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	22138	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	22139	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	22140	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vqadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 18));
	22141	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
22121	22142	return __ret;
22122	22143	}
22123	22144	#endif
22124	22145
22125	22146	__ai __attribute__((target("neon"))) uint64x1_t vqadd_u64(uint64x1_t __p0, uint64x1_t __p1) {
22126	22147	uint64x1_t __ret;
22127		__ret = (uint64x1_t) __builtin_neon_vqadd_v((int8x8_t)__p0, (int8x8_t)__p1, 19);
	22148	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vqadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 19));
22128	22149	return __ret;
22129	22150	}
22130	22151	#ifdef __LITTLE_ENDIAN__
22131	22152	__ai __attribute__((target("neon"))) uint16x4_t vqadd_u16(uint16x4_t __p0, uint16x4_t __p1) {
22132	22153	uint16x4_t __ret;
22133		__ret = (uint16x4_t) __builtin_neon_vqadd_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
	22154	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vqadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 17));
22134	22155	return __ret;
22135	22156	}
22136	22157	#else
22137	22158	__ai __attribute__((target("neon"))) uint16x4_t vqadd_u16(uint16x4_t __p0, uint16x4_t __p1) {
22138	22159	uint16x4_t __ret;
22139		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22140		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
22141		__ret = (uint16x4_t) __builtin_neon_vqadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
22142		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	22160	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	22161	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	22162	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vqadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 17));
	22163	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
22143	22164	return __ret;
22144	22165	}
22145	22166	#endif
...	...	@@ -22147,16 +22168,16 @@ __ai __attribute__((target("neon"))) uint16x4_t vqadd_u16(uint16x4_t __p0, uint1
22147	22168	#ifdef __LITTLE_ENDIAN__
22148	22169	__ai __attribute__((target("neon"))) int8x8_t vqadd_s8(int8x8_t __p0, int8x8_t __p1) {
22149	22170	int8x8_t __ret;
22150		__ret = (int8x8_t) __builtin_neon_vqadd_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
	22171	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 0));
22151	22172	return __ret;
22152	22173	}
22153	22174	#else
22154	22175	__ai __attribute__((target("neon"))) int8x8_t vqadd_s8(int8x8_t __p0, int8x8_t __p1) {
22155	22176	int8x8_t __ret;
22156		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
22157		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
22158		__ret = (int8x8_t) __builtin_neon_vqadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
22159		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	22177	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	22178	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	22179	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 0));
	22180	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
22160	22181	return __ret;
22161	22182	}
22162	22183	#endif
...	...	@@ -22164,38 +22185,38 @@ __ai __attribute__((target("neon"))) int8x8_t vqadd_s8(int8x8_t __p0, int8x8_t _
22164	22185	#ifdef __LITTLE_ENDIAN__
22165	22186	__ai __attribute__((target("neon"))) int32x2_t vqadd_s32(int32x2_t __p0, int32x2_t __p1) {
22166	22187	int32x2_t __ret;
22167		__ret = (int32x2_t) __builtin_neon_vqadd_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
	22188	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 2));
22168	22189	return __ret;
22169	22190	}
22170	22191	#else
22171	22192	__ai __attribute__((target("neon"))) int32x2_t vqadd_s32(int32x2_t __p0, int32x2_t __p1) {
22172	22193	int32x2_t __ret;
22173		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
22174		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
22175		__ret = (int32x2_t) __builtin_neon_vqadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
22176		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	22194	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	22195	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	22196	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 2));
	22197	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
22177	22198	return __ret;
22178	22199	}
22179	22200	#endif
22180	22201
22181	22202	__ai __attribute__((target("neon"))) int64x1_t vqadd_s64(int64x1_t __p0, int64x1_t __p1) {
22182	22203	int64x1_t __ret;
22183		__ret = (int64x1_t) __builtin_neon_vqadd_v((int8x8_t)__p0, (int8x8_t)__p1, 3);
	22204	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vqadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 3));
22184	22205	return __ret;
22185	22206	}
22186	22207	#ifdef __LITTLE_ENDIAN__
22187	22208	__ai __attribute__((target("neon"))) int16x4_t vqadd_s16(int16x4_t __p0, int16x4_t __p1) {
22188	22209	int16x4_t __ret;
22189		__ret = (int16x4_t) __builtin_neon_vqadd_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
	22210	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 1));
22190	22211	return __ret;
22191	22212	}
22192	22213	#else
22193	22214	__ai __attribute__((target("neon"))) int16x4_t vqadd_s16(int16x4_t __p0, int16x4_t __p1) {
22194	22215	int16x4_t __ret;
22195		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22196		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
22197		__ret = (int16x4_t) __builtin_neon_vqadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
22198		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	22216	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	22217	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	22218	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 1));
	22219	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
22199	22220	return __ret;
22200	22221	}
22201	22222	#endif
...	...	@@ -22203,22 +22224,22 @@ __ai __attribute__((target("neon"))) int16x4_t vqadd_s16(int16x4_t __p0, int16x4
22203	22224	#ifdef __LITTLE_ENDIAN__
22204	22225	__ai __attribute__((target("neon"))) int64x2_t vqdmlal_s32(int64x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
22205	22226	int64x2_t __ret;
22206		__ret = (int64x2_t) __builtin_neon_vqdmlal_v((int8x16_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 35);
	22227	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vqdmlal_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 35));
22207	22228	return __ret;
22208	22229	}
22209	22230	#else
22210	22231	__ai __attribute__((target("neon"))) int64x2_t vqdmlal_s32(int64x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
22211	22232	int64x2_t __ret;
22212		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
22213		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
22214		int32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
22215		__ret = (int64x2_t) __builtin_neon_vqdmlal_v((int8x16_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 35);
22216		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	22233	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	22234	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	22235	int32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_32);
	22236	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vqdmlal_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 35));
	22237	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
22217	22238	return __ret;
22218	22239	}
22219	22240	__ai __attribute__((target("neon"))) int64x2_t __noswap_vqdmlal_s32(int64x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
22220	22241	int64x2_t __ret;
22221		__ret = (int64x2_t) __builtin_neon_vqdmlal_v((int8x16_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 35);
	22242	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vqdmlal_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 35));
22222	22243	return __ret;
22223	22244	}
22224	22245	#endif
...	...	@@ -22226,22 +22247,22 @@ __ai __attribute__((target("neon"))) int64x2_t __noswap_vqdmlal_s32(int64x2_t __
22226	22247	#ifdef __LITTLE_ENDIAN__
22227	22248	__ai __attribute__((target("neon"))) int32x4_t vqdmlal_s16(int32x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
22228	22249	int32x4_t __ret;
22229		__ret = (int32x4_t) __builtin_neon_vqdmlal_v((int8x16_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 34);
	22250	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqdmlal_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 34));
22230	22251	return __ret;
22231	22252	}
22232	22253	#else
22233	22254	__ai __attribute__((target("neon"))) int32x4_t vqdmlal_s16(int32x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
22234	22255	int32x4_t __ret;
22235		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22236		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
22237		int16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
22238		__ret = (int32x4_t) __builtin_neon_vqdmlal_v((int8x16_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 34);
22239		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	22256	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	22257	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	22258	int16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
	22259	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqdmlal_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 34));
	22260	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
22240	22261	return __ret;
22241	22262	}
22242	22263	__ai __attribute__((target("neon"))) int32x4_t __noswap_vqdmlal_s16(int32x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
22243	22264	int32x4_t __ret;
22244		__ret = (int32x4_t) __builtin_neon_vqdmlal_v((int8x16_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 34);
	22265	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqdmlal_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 34));
22245	22266	return __ret;
22246	22267	}
22247	22268	#endif
...	...	@@ -22261,11 +22282,11 @@ __ai __attribute__((target("neon"))) int32x4_t __noswap_vqdmlal_s16(int32x4_t __
22261	22282	int64x2_t __s0_123 = __p0_123; \
22262	22283	int32x2_t __s1_123 = __p1_123; \
22263	22284	int32x2_t __s2_123 = __p2_123; \
22264		int64x2_t __rev0_123; __rev0_123 = __builtin_shufflevector(__s0_123, __s0_123, 1, 0); \
22265		int32x2_t __rev1_123; __rev1_123 = __builtin_shufflevector(__s1_123, __s1_123, 1, 0); \
22266		int32x2_t __rev2_123; __rev2_123 = __builtin_shufflevector(__s2_123, __s2_123, 1, 0); \
	22285	int64x2_t __rev0_123; __rev0_123 = __builtin_shufflevector(__s0_123, __s0_123, __lane_reverse_128_64); \
	22286	int32x2_t __rev1_123; __rev1_123 = __builtin_shufflevector(__s1_123, __s1_123, __lane_reverse_64_32); \
	22287	int32x2_t __rev2_123; __rev2_123 = __builtin_shufflevector(__s2_123, __s2_123, __lane_reverse_64_32); \
22267	22288	__ret_123 = __noswap_vqdmlal_s32(__rev0_123, __rev1_123, __noswap_splat_lane_s32(__rev2_123, __p3_123)); \
22268		__ret_123 = __builtin_shufflevector(__ret_123, __ret_123, 1, 0); \
	22289	__ret_123 = __builtin_shufflevector(__ret_123, __ret_123, __lane_reverse_128_64); \
22269	22290	__ret_123; \
22270	22291	})
22271	22292	#endif
...	...	@@ -22285,11 +22306,11 @@ __ai __attribute__((target("neon"))) int32x4_t __noswap_vqdmlal_s16(int32x4_t __
22285	22306	int32x4_t __s0_125 = __p0_125; \
22286	22307	int16x4_t __s1_125 = __p1_125; \
22287	22308	int16x4_t __s2_125 = __p2_125; \
22288		int32x4_t __rev0_125; __rev0_125 = __builtin_shufflevector(__s0_125, __s0_125, 3, 2, 1, 0); \
22289		int16x4_t __rev1_125; __rev1_125 = __builtin_shufflevector(__s1_125, __s1_125, 3, 2, 1, 0); \
22290		int16x4_t __rev2_125; __rev2_125 = __builtin_shufflevector(__s2_125, __s2_125, 3, 2, 1, 0); \
	22309	int32x4_t __rev0_125; __rev0_125 = __builtin_shufflevector(__s0_125, __s0_125, __lane_reverse_128_32); \
	22310	int16x4_t __rev1_125; __rev1_125 = __builtin_shufflevector(__s1_125, __s1_125, __lane_reverse_64_16); \
	22311	int16x4_t __rev2_125; __rev2_125 = __builtin_shufflevector(__s2_125, __s2_125, __lane_reverse_64_16); \
22291	22312	__ret_125 = __noswap_vqdmlal_s16(__rev0_125, __rev1_125, __noswap_splat_lane_s16(__rev2_125, __p3_125)); \
22292		__ret_125 = __builtin_shufflevector(__ret_125, __ret_125, 3, 2, 1, 0); \
	22313	__ret_125 = __builtin_shufflevector(__ret_125, __ret_125, __lane_reverse_128_32); \
22293	22314	__ret_125; \
22294	22315	})
22295	22316	#endif
...	...	@@ -22303,10 +22324,10 @@ __ai __attribute__((target("neon"))) int64x2_t vqdmlal_n_s32(int64x2_t __p0, int
22303	22324	#else
22304	22325	__ai __attribute__((target("neon"))) int64x2_t vqdmlal_n_s32(int64x2_t __p0, int32x2_t __p1, int32_t __p2) {
22305	22326	int64x2_t __ret;
22306		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
22307		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	22327	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	22328	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
22308	22329	__ret = __noswap_vqdmlal_s32(__rev0, __rev1, (int32x2_t) {__p2, __p2});
22309		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	22330	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
22310	22331	return __ret;
22311	22332	}
22312	22333	__ai __attribute__((target("neon"))) int64x2_t __noswap_vqdmlal_n_s32(int64x2_t __p0, int32x2_t __p1, int32_t __p2) {
...	...	@@ -22325,10 +22346,10 @@ __ai __attribute__((target("neon"))) int32x4_t vqdmlal_n_s16(int32x4_t __p0, int
22325	22346	#else
22326	22347	__ai __attribute__((target("neon"))) int32x4_t vqdmlal_n_s16(int32x4_t __p0, int16x4_t __p1, int16_t __p2) {
22327	22348	int32x4_t __ret;
22328		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22329		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	22349	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	22350	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
22330	22351	__ret = __noswap_vqdmlal_s16(__rev0, __rev1, (int16x4_t) {__p2, __p2, __p2, __p2});
22331		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	22352	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
22332	22353	return __ret;
22333	22354	}
22334	22355	__ai __attribute__((target("neon"))) int32x4_t __noswap_vqdmlal_n_s16(int32x4_t __p0, int16x4_t __p1, int16_t __p2) {
...	...	@@ -22341,22 +22362,22 @@ __ai __attribute__((target("neon"))) int32x4_t __noswap_vqdmlal_n_s16(int32x4_t
22341	22362	#ifdef __LITTLE_ENDIAN__
22342	22363	__ai __attribute__((target("neon"))) int64x2_t vqdmlsl_s32(int64x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
22343	22364	int64x2_t __ret;
22344		__ret = (int64x2_t) __builtin_neon_vqdmlsl_v((int8x16_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 35);
	22365	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vqdmlsl_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 35));
22345	22366	return __ret;
22346	22367	}
22347	22368	#else
22348	22369	__ai __attribute__((target("neon"))) int64x2_t vqdmlsl_s32(int64x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
22349	22370	int64x2_t __ret;
22350		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
22351		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
22352		int32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
22353		__ret = (int64x2_t) __builtin_neon_vqdmlsl_v((int8x16_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 35);
22354		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	22371	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	22372	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	22373	int32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_32);
	22374	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vqdmlsl_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 35));
	22375	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
22355	22376	return __ret;
22356	22377	}
22357	22378	__ai __attribute__((target("neon"))) int64x2_t __noswap_vqdmlsl_s32(int64x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
22358	22379	int64x2_t __ret;
22359		__ret = (int64x2_t) __builtin_neon_vqdmlsl_v((int8x16_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 35);
	22380	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vqdmlsl_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 35));
22360	22381	return __ret;
22361	22382	}
22362	22383	#endif
...	...	@@ -22364,22 +22385,22 @@ __ai __attribute__((target("neon"))) int64x2_t __noswap_vqdmlsl_s32(int64x2_t __
22364	22385	#ifdef __LITTLE_ENDIAN__
22365	22386	__ai __attribute__((target("neon"))) int32x4_t vqdmlsl_s16(int32x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
22366	22387	int32x4_t __ret;
22367		__ret = (int32x4_t) __builtin_neon_vqdmlsl_v((int8x16_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 34);
	22388	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqdmlsl_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 34));
22368	22389	return __ret;
22369	22390	}
22370	22391	#else
22371	22392	__ai __attribute__((target("neon"))) int32x4_t vqdmlsl_s16(int32x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
22372	22393	int32x4_t __ret;
22373		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22374		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
22375		int16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
22376		__ret = (int32x4_t) __builtin_neon_vqdmlsl_v((int8x16_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 34);
22377		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	22394	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	22395	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	22396	int16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
	22397	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqdmlsl_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 34));
	22398	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
22378	22399	return __ret;
22379	22400	}
22380	22401	__ai __attribute__((target("neon"))) int32x4_t __noswap_vqdmlsl_s16(int32x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
22381	22402	int32x4_t __ret;
22382		__ret = (int32x4_t) __builtin_neon_vqdmlsl_v((int8x16_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 34);
	22403	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqdmlsl_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 34));
22383	22404	return __ret;
22384	22405	}
22385	22406	#endif
...	...	@@ -22399,11 +22420,11 @@ __ai __attribute__((target("neon"))) int32x4_t __noswap_vqdmlsl_s16(int32x4_t __
22399	22420	int64x2_t __s0_127 = __p0_127; \
22400	22421	int32x2_t __s1_127 = __p1_127; \
22401	22422	int32x2_t __s2_127 = __p2_127; \
22402		int64x2_t __rev0_127; __rev0_127 = __builtin_shufflevector(__s0_127, __s0_127, 1, 0); \
22403		int32x2_t __rev1_127; __rev1_127 = __builtin_shufflevector(__s1_127, __s1_127, 1, 0); \
22404		int32x2_t __rev2_127; __rev2_127 = __builtin_shufflevector(__s2_127, __s2_127, 1, 0); \
	22423	int64x2_t __rev0_127; __rev0_127 = __builtin_shufflevector(__s0_127, __s0_127, __lane_reverse_128_64); \
	22424	int32x2_t __rev1_127; __rev1_127 = __builtin_shufflevector(__s1_127, __s1_127, __lane_reverse_64_32); \
	22425	int32x2_t __rev2_127; __rev2_127 = __builtin_shufflevector(__s2_127, __s2_127, __lane_reverse_64_32); \
22405	22426	__ret_127 = __noswap_vqdmlsl_s32(__rev0_127, __rev1_127, __noswap_splat_lane_s32(__rev2_127, __p3_127)); \
22406		__ret_127 = __builtin_shufflevector(__ret_127, __ret_127, 1, 0); \
	22427	__ret_127 = __builtin_shufflevector(__ret_127, __ret_127, __lane_reverse_128_64); \
22407	22428	__ret_127; \
22408	22429	})
22409	22430	#endif
...	...	@@ -22423,11 +22444,11 @@ __ai __attribute__((target("neon"))) int32x4_t __noswap_vqdmlsl_s16(int32x4_t __
22423	22444	int32x4_t __s0_129 = __p0_129; \
22424	22445	int16x4_t __s1_129 = __p1_129; \
22425	22446	int16x4_t __s2_129 = __p2_129; \
22426		int32x4_t __rev0_129; __rev0_129 = __builtin_shufflevector(__s0_129, __s0_129, 3, 2, 1, 0); \
22427		int16x4_t __rev1_129; __rev1_129 = __builtin_shufflevector(__s1_129, __s1_129, 3, 2, 1, 0); \
22428		int16x4_t __rev2_129; __rev2_129 = __builtin_shufflevector(__s2_129, __s2_129, 3, 2, 1, 0); \
	22447	int32x4_t __rev0_129; __rev0_129 = __builtin_shufflevector(__s0_129, __s0_129, __lane_reverse_128_32); \
	22448	int16x4_t __rev1_129; __rev1_129 = __builtin_shufflevector(__s1_129, __s1_129, __lane_reverse_64_16); \
	22449	int16x4_t __rev2_129; __rev2_129 = __builtin_shufflevector(__s2_129, __s2_129, __lane_reverse_64_16); \
22429	22450	__ret_129 = __noswap_vqdmlsl_s16(__rev0_129, __rev1_129, __noswap_splat_lane_s16(__rev2_129, __p3_129)); \
22430		__ret_129 = __builtin_shufflevector(__ret_129, __ret_129, 3, 2, 1, 0); \
	22451	__ret_129 = __builtin_shufflevector(__ret_129, __ret_129, __lane_reverse_128_32); \
22431	22452	__ret_129; \
22432	22453	})
22433	22454	#endif
...	...	@@ -22441,10 +22462,10 @@ __ai __attribute__((target("neon"))) int64x2_t vqdmlsl_n_s32(int64x2_t __p0, int
22441	22462	#else
22442	22463	__ai __attribute__((target("neon"))) int64x2_t vqdmlsl_n_s32(int64x2_t __p0, int32x2_t __p1, int32_t __p2) {
22443	22464	int64x2_t __ret;
22444		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
22445		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	22465	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	22466	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
22446	22467	__ret = __noswap_vqdmlsl_s32(__rev0, __rev1, (int32x2_t) {__p2, __p2});
22447		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	22468	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
22448	22469	return __ret;
22449	22470	}
22450	22471	__ai __attribute__((target("neon"))) int64x2_t __noswap_vqdmlsl_n_s32(int64x2_t __p0, int32x2_t __p1, int32_t __p2) {
...	...	@@ -22463,10 +22484,10 @@ __ai __attribute__((target("neon"))) int32x4_t vqdmlsl_n_s16(int32x4_t __p0, int
22463	22484	#else
22464	22485	__ai __attribute__((target("neon"))) int32x4_t vqdmlsl_n_s16(int32x4_t __p0, int16x4_t __p1, int16_t __p2) {
22465	22486	int32x4_t __ret;
22466		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22467		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	22487	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	22488	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
22468	22489	__ret = __noswap_vqdmlsl_s16(__rev0, __rev1, (int16x4_t) {__p2, __p2, __p2, __p2});
22469		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	22490	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
22470	22491	return __ret;
22471	22492	}
22472	22493	__ai __attribute__((target("neon"))) int32x4_t __noswap_vqdmlsl_n_s16(int32x4_t __p0, int16x4_t __p1, int16_t __p2) {
...	...	@@ -22479,21 +22500,21 @@ __ai __attribute__((target("neon"))) int32x4_t __noswap_vqdmlsl_n_s16(int32x4_t
22479	22500	#ifdef __LITTLE_ENDIAN__
22480	22501	__ai __attribute__((target("neon"))) int32x4_t vqdmulhq_s32(int32x4_t __p0, int32x4_t __p1) {
22481	22502	int32x4_t __ret;
22482		__ret = (int32x4_t) __builtin_neon_vqdmulhq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
	22503	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqdmulhq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 34));
22483	22504	return __ret;
22484	22505	}
22485	22506	#else
22486	22507	__ai __attribute__((target("neon"))) int32x4_t vqdmulhq_s32(int32x4_t __p0, int32x4_t __p1) {
22487	22508	int32x4_t __ret;
22488		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22489		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
22490		__ret = (int32x4_t) __builtin_neon_vqdmulhq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
22491		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	22509	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	22510	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	22511	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqdmulhq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 34));
	22512	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
22492	22513	return __ret;
22493	22514	}
22494	22515	__ai __attribute__((target("neon"))) int32x4_t __noswap_vqdmulhq_s32(int32x4_t __p0, int32x4_t __p1) {
22495	22516	int32x4_t __ret;
22496		__ret = (int32x4_t) __builtin_neon_vqdmulhq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
	22517	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqdmulhq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 34));
22497	22518	return __ret;
22498	22519	}
22499	22520	#endif
...	...	@@ -22501,21 +22522,21 @@ __ai __attribute__((target("neon"))) int32x4_t __noswap_vqdmulhq_s32(int32x4_t _
22501	22522	#ifdef __LITTLE_ENDIAN__
22502	22523	__ai __attribute__((target("neon"))) int16x8_t vqdmulhq_s16(int16x8_t __p0, int16x8_t __p1) {
22503	22524	int16x8_t __ret;
22504		__ret = (int16x8_t) __builtin_neon_vqdmulhq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
	22525	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqdmulhq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 33));
22505	22526	return __ret;
22506	22527	}
22507	22528	#else
22508	22529	__ai __attribute__((target("neon"))) int16x8_t vqdmulhq_s16(int16x8_t __p0, int16x8_t __p1) {
22509	22530	int16x8_t __ret;
22510		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
22511		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
22512		__ret = (int16x8_t) __builtin_neon_vqdmulhq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
22513		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	22531	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	22532	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	22533	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqdmulhq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 33));
	22534	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
22514	22535	return __ret;
22515	22536	}
22516	22537	__ai __attribute__((target("neon"))) int16x8_t __noswap_vqdmulhq_s16(int16x8_t __p0, int16x8_t __p1) {
22517	22538	int16x8_t __ret;
22518		__ret = (int16x8_t) __builtin_neon_vqdmulhq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
	22539	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqdmulhq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 33));
22519	22540	return __ret;
22520	22541	}
22521	22542	#endif
...	...	@@ -22523,21 +22544,21 @@ __ai __attribute__((target("neon"))) int16x8_t __noswap_vqdmulhq_s16(int16x8_t _
22523	22544	#ifdef __LITTLE_ENDIAN__
22524	22545	__ai __attribute__((target("neon"))) int32x2_t vqdmulh_s32(int32x2_t __p0, int32x2_t __p1) {
22525	22546	int32x2_t __ret;
22526		__ret = (int32x2_t) __builtin_neon_vqdmulh_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
	22547	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqdmulh_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 2));
22527	22548	return __ret;
22528	22549	}
22529	22550	#else
22530	22551	__ai __attribute__((target("neon"))) int32x2_t vqdmulh_s32(int32x2_t __p0, int32x2_t __p1) {
22531	22552	int32x2_t __ret;
22532		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
22533		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
22534		__ret = (int32x2_t) __builtin_neon_vqdmulh_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
22535		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	22553	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	22554	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	22555	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqdmulh_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 2));
	22556	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
22536	22557	return __ret;
22537	22558	}
22538	22559	__ai __attribute__((target("neon"))) int32x2_t __noswap_vqdmulh_s32(int32x2_t __p0, int32x2_t __p1) {
22539	22560	int32x2_t __ret;
22540		__ret = (int32x2_t) __builtin_neon_vqdmulh_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
	22561	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqdmulh_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 2));
22541	22562	return __ret;
22542	22563	}
22543	22564	#endif
...	...	@@ -22545,21 +22566,21 @@ __ai __attribute__((target("neon"))) int32x2_t __noswap_vqdmulh_s32(int32x2_t __
22545	22566	#ifdef __LITTLE_ENDIAN__
22546	22567	__ai __attribute__((target("neon"))) int16x4_t vqdmulh_s16(int16x4_t __p0, int16x4_t __p1) {
22547	22568	int16x4_t __ret;
22548		__ret = (int16x4_t) __builtin_neon_vqdmulh_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
	22569	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqdmulh_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 1));
22549	22570	return __ret;
22550	22571	}
22551	22572	#else
22552	22573	__ai __attribute__((target("neon"))) int16x4_t vqdmulh_s16(int16x4_t __p0, int16x4_t __p1) {
22553	22574	int16x4_t __ret;
22554		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22555		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
22556		__ret = (int16x4_t) __builtin_neon_vqdmulh_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
22557		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	22575	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	22576	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	22577	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqdmulh_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 1));
	22578	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
22558	22579	return __ret;
22559	22580	}
22560	22581	__ai __attribute__((target("neon"))) int16x4_t __noswap_vqdmulh_s16(int16x4_t __p0, int16x4_t __p1) {
22561	22582	int16x4_t __ret;
22562		__ret = (int16x4_t) __builtin_neon_vqdmulh_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
	22583	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqdmulh_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 1));
22563	22584	return __ret;
22564	22585	}
22565	22586	#endif
...	...	@@ -22573,9 +22594,9 @@ __ai __attribute__((target("neon"))) int32x4_t vqdmulhq_n_s32(int32x4_t __p0, in
22573	22594	#else
22574	22595	__ai __attribute__((target("neon"))) int32x4_t vqdmulhq_n_s32(int32x4_t __p0, int32_t __p1) {
22575	22596	int32x4_t __ret;
22576		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	22597	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
22577	22598	__ret = __noswap_vqdmulhq_s32(__rev0, (int32x4_t) {__p1, __p1, __p1, __p1});
22578		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	22599	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
22579	22600	return __ret;
22580	22601	}
22581	22602	#endif
...	...	@@ -22589,9 +22610,9 @@ __ai __attribute__((target("neon"))) int16x8_t vqdmulhq_n_s16(int16x8_t __p0, in
22589	22610	#else
22590	22611	__ai __attribute__((target("neon"))) int16x8_t vqdmulhq_n_s16(int16x8_t __p0, int16_t __p1) {
22591	22612	int16x8_t __ret;
22592		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	22613	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
22593	22614	__ret = __noswap_vqdmulhq_s16(__rev0, (int16x8_t) {__p1, __p1, __p1, __p1, __p1, __p1, __p1, __p1});
22594		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	22615	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
22595	22616	return __ret;
22596	22617	}
22597	22618	#endif
...	...	@@ -22605,9 +22626,9 @@ __ai __attribute__((target("neon"))) int32x2_t vqdmulh_n_s32(int32x2_t __p0, int
22605	22626	#else
22606	22627	__ai __attribute__((target("neon"))) int32x2_t vqdmulh_n_s32(int32x2_t __p0, int32_t __p1) {
22607	22628	int32x2_t __ret;
22608		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
	22629	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
22609	22630	__ret = __noswap_vqdmulh_s32(__rev0, (int32x2_t) {__p1, __p1});
22610		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	22631	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
22611	22632	return __ret;
22612	22633	}
22613	22634	#endif
...	...	@@ -22621,9 +22642,9 @@ __ai __attribute__((target("neon"))) int16x4_t vqdmulh_n_s16(int16x4_t __p0, int
22621	22642	#else
22622	22643	__ai __attribute__((target("neon"))) int16x4_t vqdmulh_n_s16(int16x4_t __p0, int16_t __p1) {
22623	22644	int16x4_t __ret;
22624		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	22645	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
22625	22646	__ret = __noswap_vqdmulh_s16(__rev0, (int16x4_t) {__p1, __p1, __p1, __p1});
22626		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	22647	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
22627	22648	return __ret;
22628	22649	}
22629	22650	#endif
...	...	@@ -22631,21 +22652,21 @@ __ai __attribute__((target("neon"))) int16x4_t vqdmulh_n_s16(int16x4_t __p0, int
22631	22652	#ifdef __LITTLE_ENDIAN__
22632	22653	__ai __attribute__((target("neon"))) int64x2_t vqdmull_s32(int32x2_t __p0, int32x2_t __p1) {
22633	22654	int64x2_t __ret;
22634		__ret = (int64x2_t) __builtin_neon_vqdmull_v((int8x8_t)__p0, (int8x8_t)__p1, 35);
	22655	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vqdmull_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 35));
22635	22656	return __ret;
22636	22657	}
22637	22658	#else
22638	22659	__ai __attribute__((target("neon"))) int64x2_t vqdmull_s32(int32x2_t __p0, int32x2_t __p1) {
22639	22660	int64x2_t __ret;
22640		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
22641		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
22642		__ret = (int64x2_t) __builtin_neon_vqdmull_v((int8x8_t)__rev0, (int8x8_t)__rev1, 35);
22643		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	22661	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	22662	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	22663	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vqdmull_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 35));
	22664	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
22644	22665	return __ret;
22645	22666	}
22646	22667	__ai __attribute__((target("neon"))) int64x2_t __noswap_vqdmull_s32(int32x2_t __p0, int32x2_t __p1) {
22647	22668	int64x2_t __ret;
22648		__ret = (int64x2_t) __builtin_neon_vqdmull_v((int8x8_t)__p0, (int8x8_t)__p1, 35);
	22669	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vqdmull_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 35));
22649	22670	return __ret;
22650	22671	}
22651	22672	#endif
...	...	@@ -22653,21 +22674,21 @@ __ai __attribute__((target("neon"))) int64x2_t __noswap_vqdmull_s32(int32x2_t __
22653	22674	#ifdef __LITTLE_ENDIAN__
22654	22675	__ai __attribute__((target("neon"))) int32x4_t vqdmull_s16(int16x4_t __p0, int16x4_t __p1) {
22655	22676	int32x4_t __ret;
22656		__ret = (int32x4_t) __builtin_neon_vqdmull_v((int8x8_t)__p0, (int8x8_t)__p1, 34);
	22677	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqdmull_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 34));
22657	22678	return __ret;
22658	22679	}
22659	22680	#else
22660	22681	__ai __attribute__((target("neon"))) int32x4_t vqdmull_s16(int16x4_t __p0, int16x4_t __p1) {
22661	22682	int32x4_t __ret;
22662		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22663		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
22664		__ret = (int32x4_t) __builtin_neon_vqdmull_v((int8x8_t)__rev0, (int8x8_t)__rev1, 34);
22665		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	22683	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	22684	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	22685	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqdmull_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 34));
	22686	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
22666	22687	return __ret;
22667	22688	}
22668	22689	__ai __attribute__((target("neon"))) int32x4_t __noswap_vqdmull_s16(int16x4_t __p0, int16x4_t __p1) {
22669	22690	int32x4_t __ret;
22670		__ret = (int32x4_t) __builtin_neon_vqdmull_v((int8x8_t)__p0, (int8x8_t)__p1, 34);
	22691	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqdmull_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 34));
22671	22692	return __ret;
22672	22693	}
22673	22694	#endif
...	...	@@ -22685,10 +22706,10 @@ __ai __attribute__((target("neon"))) int32x4_t __noswap_vqdmull_s16(int16x4_t __
22685	22706	int64x2_t __ret_131; \
22686	22707	int32x2_t __s0_131 = __p0_131; \
22687	22708	int32x2_t __s1_131 = __p1_131; \
22688		int32x2_t __rev0_131; __rev0_131 = __builtin_shufflevector(__s0_131, __s0_131, 1, 0); \
22689		int32x2_t __rev1_131; __rev1_131 = __builtin_shufflevector(__s1_131, __s1_131, 1, 0); \
	22709	int32x2_t __rev0_131; __rev0_131 = __builtin_shufflevector(__s0_131, __s0_131, __lane_reverse_64_32); \
	22710	int32x2_t __rev1_131; __rev1_131 = __builtin_shufflevector(__s1_131, __s1_131, __lane_reverse_64_32); \
22690	22711	__ret_131 = __noswap_vqdmull_s32(__rev0_131, __noswap_splat_lane_s32(__rev1_131, __p2_131)); \
22691		__ret_131 = __builtin_shufflevector(__ret_131, __ret_131, 1, 0); \
	22712	__ret_131 = __builtin_shufflevector(__ret_131, __ret_131, __lane_reverse_128_64); \
22692	22713	__ret_131; \
22693	22714	})
22694	22715	#endif
...	...	@@ -22706,10 +22727,10 @@ __ai __attribute__((target("neon"))) int32x4_t __noswap_vqdmull_s16(int16x4_t __
22706	22727	int32x4_t __ret_133; \
22707	22728	int16x4_t __s0_133 = __p0_133; \
22708	22729	int16x4_t __s1_133 = __p1_133; \
22709		int16x4_t __rev0_133; __rev0_133 = __builtin_shufflevector(__s0_133, __s0_133, 3, 2, 1, 0); \
22710		int16x4_t __rev1_133; __rev1_133 = __builtin_shufflevector(__s1_133, __s1_133, 3, 2, 1, 0); \
	22730	int16x4_t __rev0_133; __rev0_133 = __builtin_shufflevector(__s0_133, __s0_133, __lane_reverse_64_16); \
	22731	int16x4_t __rev1_133; __rev1_133 = __builtin_shufflevector(__s1_133, __s1_133, __lane_reverse_64_16); \
22711	22732	__ret_133 = __noswap_vqdmull_s16(__rev0_133, __noswap_splat_lane_s16(__rev1_133, __p2_133)); \
22712		__ret_133 = __builtin_shufflevector(__ret_133, __ret_133, 3, 2, 1, 0); \
	22733	__ret_133 = __builtin_shufflevector(__ret_133, __ret_133, __lane_reverse_128_32); \
22713	22734	__ret_133; \
22714	22735	})
22715	22736	#endif
...	...	@@ -22723,9 +22744,9 @@ __ai __attribute__((target("neon"))) int64x2_t vqdmull_n_s32(int32x2_t __p0, int
22723	22744	#else
22724	22745	__ai __attribute__((target("neon"))) int64x2_t vqdmull_n_s32(int32x2_t __p0, int32_t __p1) {
22725	22746	int64x2_t __ret;
22726		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
	22747	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
22727	22748	__ret = __noswap_vqdmull_s32(__rev0, (int32x2_t) {__p1, __p1});
22728		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	22749	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
22729	22750	return __ret;
22730	22751	}
22731	22752	__ai __attribute__((target("neon"))) int64x2_t __noswap_vqdmull_n_s32(int32x2_t __p0, int32_t __p1) {
...	...	@@ -22744,9 +22765,9 @@ __ai __attribute__((target("neon"))) int32x4_t vqdmull_n_s16(int16x4_t __p0, int
22744	22765	#else
22745	22766	__ai __attribute__((target("neon"))) int32x4_t vqdmull_n_s16(int16x4_t __p0, int16_t __p1) {
22746	22767	int32x4_t __ret;
22747		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	22768	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
22748	22769	__ret = __noswap_vqdmull_s16(__rev0, (int16x4_t) {__p1, __p1, __p1, __p1});
22749		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	22770	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
22750	22771	return __ret;
22751	22772	}
22752	22773	__ai __attribute__((target("neon"))) int32x4_t __noswap_vqdmull_n_s16(int16x4_t __p0, int16_t __p1) {
...	...	@@ -22759,20 +22780,20 @@ __ai __attribute__((target("neon"))) int32x4_t __noswap_vqdmull_n_s16(int16x4_t
22759	22780	#ifdef __LITTLE_ENDIAN__
22760	22781	__ai __attribute__((target("neon"))) uint16x4_t vqmovn_u32(uint32x4_t __p0) {
22761	22782	uint16x4_t __ret;
22762		__ret = (uint16x4_t) __builtin_neon_vqmovn_v((int8x16_t)__p0, 17);
	22783	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vqmovn_v(__builtin_bit_cast(int8x16_t, __p0), 17));
22763	22784	return __ret;
22764	22785	}
22765	22786	#else
22766	22787	__ai __attribute__((target("neon"))) uint16x4_t vqmovn_u32(uint32x4_t __p0) {
22767	22788	uint16x4_t __ret;
22768		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22769		__ret = (uint16x4_t) __builtin_neon_vqmovn_v((int8x16_t)__rev0, 17);
22770		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	22789	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	22790	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vqmovn_v(__builtin_bit_cast(int8x16_t, __rev0), 17));
	22791	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
22771	22792	return __ret;
22772	22793	}
22773	22794	__ai __attribute__((target("neon"))) uint16x4_t __noswap_vqmovn_u32(uint32x4_t __p0) {
22774	22795	uint16x4_t __ret;
22775		__ret = (uint16x4_t) __builtin_neon_vqmovn_v((int8x16_t)__p0, 17);
	22796	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vqmovn_v(__builtin_bit_cast(int8x16_t, __p0), 17));
22776	22797	return __ret;
22777	22798	}
22778	22799	#endif
...	...	@@ -22780,20 +22801,20 @@ __ai __attribute__((target("neon"))) uint16x4_t __noswap_vqmovn_u32(uint32x4_t _
22780	22801	#ifdef __LITTLE_ENDIAN__
22781	22802	__ai __attribute__((target("neon"))) uint32x2_t vqmovn_u64(uint64x2_t __p0) {
22782	22803	uint32x2_t __ret;
22783		__ret = (uint32x2_t) __builtin_neon_vqmovn_v((int8x16_t)__p0, 18);
	22804	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vqmovn_v(__builtin_bit_cast(int8x16_t, __p0), 18));
22784	22805	return __ret;
22785	22806	}
22786	22807	#else
22787	22808	__ai __attribute__((target("neon"))) uint32x2_t vqmovn_u64(uint64x2_t __p0) {
22788	22809	uint32x2_t __ret;
22789		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
22790		__ret = (uint32x2_t) __builtin_neon_vqmovn_v((int8x16_t)__rev0, 18);
22791		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	22810	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	22811	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vqmovn_v(__builtin_bit_cast(int8x16_t, __rev0), 18));
	22812	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
22792	22813	return __ret;
22793	22814	}
22794	22815	__ai __attribute__((target("neon"))) uint32x2_t __noswap_vqmovn_u64(uint64x2_t __p0) {
22795	22816	uint32x2_t __ret;
22796		__ret = (uint32x2_t) __builtin_neon_vqmovn_v((int8x16_t)__p0, 18);
	22817	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vqmovn_v(__builtin_bit_cast(int8x16_t, __p0), 18));
22797	22818	return __ret;
22798	22819	}
22799	22820	#endif
...	...	@@ -22801,20 +22822,20 @@ __ai __attribute__((target("neon"))) uint32x2_t __noswap_vqmovn_u64(uint64x2_t _
22801	22822	#ifdef __LITTLE_ENDIAN__
22802	22823	__ai __attribute__((target("neon"))) uint8x8_t vqmovn_u16(uint16x8_t __p0) {
22803	22824	uint8x8_t __ret;
22804		__ret = (uint8x8_t) __builtin_neon_vqmovn_v((int8x16_t)__p0, 16);
	22825	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqmovn_v(__builtin_bit_cast(int8x16_t, __p0), 16));
22805	22826	return __ret;
22806	22827	}
22807	22828	#else
22808	22829	__ai __attribute__((target("neon"))) uint8x8_t vqmovn_u16(uint16x8_t __p0) {
22809	22830	uint8x8_t __ret;
22810		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
22811		__ret = (uint8x8_t) __builtin_neon_vqmovn_v((int8x16_t)__rev0, 16);
22812		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	22831	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	22832	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqmovn_v(__builtin_bit_cast(int8x16_t, __rev0), 16));
	22833	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
22813	22834	return __ret;
22814	22835	}
22815	22836	__ai __attribute__((target("neon"))) uint8x8_t __noswap_vqmovn_u16(uint16x8_t __p0) {
22816	22837	uint8x8_t __ret;
22817		__ret = (uint8x8_t) __builtin_neon_vqmovn_v((int8x16_t)__p0, 16);
	22838	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqmovn_v(__builtin_bit_cast(int8x16_t, __p0), 16));
22818	22839	return __ret;
22819	22840	}
22820	22841	#endif
...	...	@@ -22822,20 +22843,20 @@ __ai __attribute__((target("neon"))) uint8x8_t __noswap_vqmovn_u16(uint16x8_t __
22822	22843	#ifdef __LITTLE_ENDIAN__
22823	22844	__ai __attribute__((target("neon"))) int16x4_t vqmovn_s32(int32x4_t __p0) {
22824	22845	int16x4_t __ret;
22825		__ret = (int16x4_t) __builtin_neon_vqmovn_v((int8x16_t)__p0, 1);
	22846	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqmovn_v(__builtin_bit_cast(int8x16_t, __p0), 1));
22826	22847	return __ret;
22827	22848	}
22828	22849	#else
22829	22850	__ai __attribute__((target("neon"))) int16x4_t vqmovn_s32(int32x4_t __p0) {
22830	22851	int16x4_t __ret;
22831		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22832		__ret = (int16x4_t) __builtin_neon_vqmovn_v((int8x16_t)__rev0, 1);
22833		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	22852	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	22853	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqmovn_v(__builtin_bit_cast(int8x16_t, __rev0), 1));
	22854	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
22834	22855	return __ret;
22835	22856	}
22836	22857	__ai __attribute__((target("neon"))) int16x4_t __noswap_vqmovn_s32(int32x4_t __p0) {
22837	22858	int16x4_t __ret;
22838		__ret = (int16x4_t) __builtin_neon_vqmovn_v((int8x16_t)__p0, 1);
	22859	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqmovn_v(__builtin_bit_cast(int8x16_t, __p0), 1));
22839	22860	return __ret;
22840	22861	}
22841	22862	#endif
...	...	@@ -22843,20 +22864,20 @@ __ai __attribute__((target("neon"))) int16x4_t __noswap_vqmovn_s32(int32x4_t __p
22843	22864	#ifdef __LITTLE_ENDIAN__
22844	22865	__ai __attribute__((target("neon"))) int32x2_t vqmovn_s64(int64x2_t __p0) {
22845	22866	int32x2_t __ret;
22846		__ret = (int32x2_t) __builtin_neon_vqmovn_v((int8x16_t)__p0, 2);
	22867	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqmovn_v(__builtin_bit_cast(int8x16_t, __p0), 2));
22847	22868	return __ret;
22848	22869	}
22849	22870	#else
22850	22871	__ai __attribute__((target("neon"))) int32x2_t vqmovn_s64(int64x2_t __p0) {
22851	22872	int32x2_t __ret;
22852		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
22853		__ret = (int32x2_t) __builtin_neon_vqmovn_v((int8x16_t)__rev0, 2);
22854		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	22873	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	22874	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqmovn_v(__builtin_bit_cast(int8x16_t, __rev0), 2));
	22875	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
22855	22876	return __ret;
22856	22877	}
22857	22878	__ai __attribute__((target("neon"))) int32x2_t __noswap_vqmovn_s64(int64x2_t __p0) {
22858	22879	int32x2_t __ret;
22859		__ret = (int32x2_t) __builtin_neon_vqmovn_v((int8x16_t)__p0, 2);
	22880	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqmovn_v(__builtin_bit_cast(int8x16_t, __p0), 2));
22860	22881	return __ret;
22861	22882	}
22862	22883	#endif
...	...	@@ -22864,20 +22885,20 @@ __ai __attribute__((target("neon"))) int32x2_t __noswap_vqmovn_s64(int64x2_t __p
22864	22885	#ifdef __LITTLE_ENDIAN__
22865	22886	__ai __attribute__((target("neon"))) int8x8_t vqmovn_s16(int16x8_t __p0) {
22866	22887	int8x8_t __ret;
22867		__ret = (int8x8_t) __builtin_neon_vqmovn_v((int8x16_t)__p0, 0);
	22888	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqmovn_v(__builtin_bit_cast(int8x16_t, __p0), 0));
22868	22889	return __ret;
22869	22890	}
22870	22891	#else
22871	22892	__ai __attribute__((target("neon"))) int8x8_t vqmovn_s16(int16x8_t __p0) {
22872	22893	int8x8_t __ret;
22873		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
22874		__ret = (int8x8_t) __builtin_neon_vqmovn_v((int8x16_t)__rev0, 0);
22875		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	22894	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	22895	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqmovn_v(__builtin_bit_cast(int8x16_t, __rev0), 0));
	22896	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
22876	22897	return __ret;
22877	22898	}
22878	22899	__ai __attribute__((target("neon"))) int8x8_t __noswap_vqmovn_s16(int16x8_t __p0) {
22879	22900	int8x8_t __ret;
22880		__ret = (int8x8_t) __builtin_neon_vqmovn_v((int8x16_t)__p0, 0);
	22901	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqmovn_v(__builtin_bit_cast(int8x16_t, __p0), 0));
22881	22902	return __ret;
22882	22903	}
22883	22904	#endif
...	...	@@ -22885,20 +22906,20 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vqmovn_s16(int16x8_t __p0
22885	22906	#ifdef __LITTLE_ENDIAN__
22886	22907	__ai __attribute__((target("neon"))) uint16x4_t vqmovun_s32(int32x4_t __p0) {
22887	22908	uint16x4_t __ret;
22888		__ret = (uint16x4_t) __builtin_neon_vqmovun_v((int8x16_t)__p0, 17);
	22909	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vqmovun_v(__builtin_bit_cast(int8x16_t, __p0), 17));
22889	22910	return __ret;
22890	22911	}
22891	22912	#else
22892	22913	__ai __attribute__((target("neon"))) uint16x4_t vqmovun_s32(int32x4_t __p0) {
22893	22914	uint16x4_t __ret;
22894		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22895		__ret = (uint16x4_t) __builtin_neon_vqmovun_v((int8x16_t)__rev0, 17);
22896		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	22915	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	22916	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vqmovun_v(__builtin_bit_cast(int8x16_t, __rev0), 17));
	22917	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
22897	22918	return __ret;
22898	22919	}
22899	22920	__ai __attribute__((target("neon"))) uint16x4_t __noswap_vqmovun_s32(int32x4_t __p0) {
22900	22921	uint16x4_t __ret;
22901		__ret = (uint16x4_t) __builtin_neon_vqmovun_v((int8x16_t)__p0, 17);
	22922	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vqmovun_v(__builtin_bit_cast(int8x16_t, __p0), 17));
22902	22923	return __ret;
22903	22924	}
22904	22925	#endif
...	...	@@ -22906,20 +22927,20 @@ __ai __attribute__((target("neon"))) uint16x4_t __noswap_vqmovun_s32(int32x4_t _
22906	22927	#ifdef __LITTLE_ENDIAN__
22907	22928	__ai __attribute__((target("neon"))) uint32x2_t vqmovun_s64(int64x2_t __p0) {
22908	22929	uint32x2_t __ret;
22909		__ret = (uint32x2_t) __builtin_neon_vqmovun_v((int8x16_t)__p0, 18);
	22930	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vqmovun_v(__builtin_bit_cast(int8x16_t, __p0), 18));
22910	22931	return __ret;
22911	22932	}
22912	22933	#else
22913	22934	__ai __attribute__((target("neon"))) uint32x2_t vqmovun_s64(int64x2_t __p0) {
22914	22935	uint32x2_t __ret;
22915		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
22916		__ret = (uint32x2_t) __builtin_neon_vqmovun_v((int8x16_t)__rev0, 18);
22917		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	22936	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	22937	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vqmovun_v(__builtin_bit_cast(int8x16_t, __rev0), 18));
	22938	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
22918	22939	return __ret;
22919	22940	}
22920	22941	__ai __attribute__((target("neon"))) uint32x2_t __noswap_vqmovun_s64(int64x2_t __p0) {
22921	22942	uint32x2_t __ret;
22922		__ret = (uint32x2_t) __builtin_neon_vqmovun_v((int8x16_t)__p0, 18);
	22943	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vqmovun_v(__builtin_bit_cast(int8x16_t, __p0), 18));
22923	22944	return __ret;
22924	22945	}
22925	22946	#endif
...	...	@@ -22927,20 +22948,20 @@ __ai __attribute__((target("neon"))) uint32x2_t __noswap_vqmovun_s64(int64x2_t _
22927	22948	#ifdef __LITTLE_ENDIAN__
22928	22949	__ai __attribute__((target("neon"))) uint8x8_t vqmovun_s16(int16x8_t __p0) {
22929	22950	uint8x8_t __ret;
22930		__ret = (uint8x8_t) __builtin_neon_vqmovun_v((int8x16_t)__p0, 16);
	22951	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqmovun_v(__builtin_bit_cast(int8x16_t, __p0), 16));
22931	22952	return __ret;
22932	22953	}
22933	22954	#else
22934	22955	__ai __attribute__((target("neon"))) uint8x8_t vqmovun_s16(int16x8_t __p0) {
22935	22956	uint8x8_t __ret;
22936		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
22937		__ret = (uint8x8_t) __builtin_neon_vqmovun_v((int8x16_t)__rev0, 16);
22938		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	22957	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	22958	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqmovun_v(__builtin_bit_cast(int8x16_t, __rev0), 16));
	22959	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
22939	22960	return __ret;
22940	22961	}
22941	22962	__ai __attribute__((target("neon"))) uint8x8_t __noswap_vqmovun_s16(int16x8_t __p0) {
22942	22963	uint8x8_t __ret;
22943		__ret = (uint8x8_t) __builtin_neon_vqmovun_v((int8x16_t)__p0, 16);
	22964	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqmovun_v(__builtin_bit_cast(int8x16_t, __p0), 16));
22944	22965	return __ret;
22945	22966	}
22946	22967	#endif
...	...	@@ -22948,15 +22969,15 @@ __ai __attribute__((target("neon"))) uint8x8_t __noswap_vqmovun_s16(int16x8_t __
22948	22969	#ifdef __LITTLE_ENDIAN__
22949	22970	__ai __attribute__((target("neon"))) int8x16_t vqnegq_s8(int8x16_t __p0) {
22950	22971	int8x16_t __ret;
22951		__ret = (int8x16_t) __builtin_neon_vqnegq_v((int8x16_t)__p0, 32);
	22972	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vqnegq_v(__builtin_bit_cast(int8x16_t, __p0), 32));
22952	22973	return __ret;
22953	22974	}
22954	22975	#else
22955	22976	__ai __attribute__((target("neon"))) int8x16_t vqnegq_s8(int8x16_t __p0) {
22956	22977	int8x16_t __ret;
22957		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
22958		__ret = (int8x16_t) __builtin_neon_vqnegq_v((int8x16_t)__rev0, 32);
22959		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	22978	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	22979	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vqnegq_v(__builtin_bit_cast(int8x16_t, __rev0), 32));
	22980	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
22960	22981	return __ret;
22961	22982	}
22962	22983	#endif
...	...	@@ -22964,15 +22985,15 @@ __ai __attribute__((target("neon"))) int8x16_t vqnegq_s8(int8x16_t __p0) {
22964	22985	#ifdef __LITTLE_ENDIAN__
22965	22986	__ai __attribute__((target("neon"))) int32x4_t vqnegq_s32(int32x4_t __p0) {
22966	22987	int32x4_t __ret;
22967		__ret = (int32x4_t) __builtin_neon_vqnegq_v((int8x16_t)__p0, 34);
	22988	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqnegq_v(__builtin_bit_cast(int8x16_t, __p0), 34));
22968	22989	return __ret;
22969	22990	}
22970	22991	#else
22971	22992	__ai __attribute__((target("neon"))) int32x4_t vqnegq_s32(int32x4_t __p0) {
22972	22993	int32x4_t __ret;
22973		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22974		__ret = (int32x4_t) __builtin_neon_vqnegq_v((int8x16_t)__rev0, 34);
22975		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	22994	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	22995	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqnegq_v(__builtin_bit_cast(int8x16_t, __rev0), 34));
	22996	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
22976	22997	return __ret;
22977	22998	}
22978	22999	#endif
...	...	@@ -22980,15 +23001,15 @@ __ai __attribute__((target("neon"))) int32x4_t vqnegq_s32(int32x4_t __p0) {
22980	23001	#ifdef __LITTLE_ENDIAN__
22981	23002	__ai __attribute__((target("neon"))) int16x8_t vqnegq_s16(int16x8_t __p0) {
22982	23003	int16x8_t __ret;
22983		__ret = (int16x8_t) __builtin_neon_vqnegq_v((int8x16_t)__p0, 33);
	23004	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqnegq_v(__builtin_bit_cast(int8x16_t, __p0), 33));
22984	23005	return __ret;
22985	23006	}
22986	23007	#else
22987	23008	__ai __attribute__((target("neon"))) int16x8_t vqnegq_s16(int16x8_t __p0) {
22988	23009	int16x8_t __ret;
22989		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
22990		__ret = (int16x8_t) __builtin_neon_vqnegq_v((int8x16_t)__rev0, 33);
22991		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	23010	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	23011	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqnegq_v(__builtin_bit_cast(int8x16_t, __rev0), 33));
	23012	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
22992	23013	return __ret;
22993	23014	}
22994	23015	#endif
...	...	@@ -22996,15 +23017,15 @@ __ai __attribute__((target("neon"))) int16x8_t vqnegq_s16(int16x8_t __p0) {
22996	23017	#ifdef __LITTLE_ENDIAN__
22997	23018	__ai __attribute__((target("neon"))) int8x8_t vqneg_s8(int8x8_t __p0) {
22998	23019	int8x8_t __ret;
22999		__ret = (int8x8_t) __builtin_neon_vqneg_v((int8x8_t)__p0, 0);
	23020	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqneg_v(__builtin_bit_cast(int8x8_t, __p0), 0));
23000	23021	return __ret;
23001	23022	}
23002	23023	#else
23003	23024	__ai __attribute__((target("neon"))) int8x8_t vqneg_s8(int8x8_t __p0) {
23004	23025	int8x8_t __ret;
23005		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
23006		__ret = (int8x8_t) __builtin_neon_vqneg_v((int8x8_t)__rev0, 0);
23007		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	23026	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	23027	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqneg_v(__builtin_bit_cast(int8x8_t, __rev0), 0));
	23028	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
23008	23029	return __ret;
23009	23030	}
23010	23031	#endif
...	...	@@ -23012,15 +23033,15 @@ __ai __attribute__((target("neon"))) int8x8_t vqneg_s8(int8x8_t __p0) {
23012	23033	#ifdef __LITTLE_ENDIAN__
23013	23034	__ai __attribute__((target("neon"))) int32x2_t vqneg_s32(int32x2_t __p0) {
23014	23035	int32x2_t __ret;
23015		__ret = (int32x2_t) __builtin_neon_vqneg_v((int8x8_t)__p0, 2);
	23036	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqneg_v(__builtin_bit_cast(int8x8_t, __p0), 2));
23016	23037	return __ret;
23017	23038	}
23018	23039	#else
23019	23040	__ai __attribute__((target("neon"))) int32x2_t vqneg_s32(int32x2_t __p0) {
23020	23041	int32x2_t __ret;
23021		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
23022		__ret = (int32x2_t) __builtin_neon_vqneg_v((int8x8_t)__rev0, 2);
23023		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	23042	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	23043	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqneg_v(__builtin_bit_cast(int8x8_t, __rev0), 2));
	23044	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
23024	23045	return __ret;
23025	23046	}
23026	23047	#endif
...	...	@@ -23028,15 +23049,15 @@ __ai __attribute__((target("neon"))) int32x2_t vqneg_s32(int32x2_t __p0) {
23028	23049	#ifdef __LITTLE_ENDIAN__
23029	23050	__ai __attribute__((target("neon"))) int16x4_t vqneg_s16(int16x4_t __p0) {
23030	23051	int16x4_t __ret;
23031		__ret = (int16x4_t) __builtin_neon_vqneg_v((int8x8_t)__p0, 1);
	23052	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqneg_v(__builtin_bit_cast(int8x8_t, __p0), 1));
23032	23053	return __ret;
23033	23054	}
23034	23055	#else
23035	23056	__ai __attribute__((target("neon"))) int16x4_t vqneg_s16(int16x4_t __p0) {
23036	23057	int16x4_t __ret;
23037		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
23038		__ret = (int16x4_t) __builtin_neon_vqneg_v((int8x8_t)__rev0, 1);
23039		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	23058	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	23059	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqneg_v(__builtin_bit_cast(int8x8_t, __rev0), 1));
	23060	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
23040	23061	return __ret;
23041	23062	}
23042	23063	#endif
...	...	@@ -23044,21 +23065,21 @@ __ai __attribute__((target("neon"))) int16x4_t vqneg_s16(int16x4_t __p0) {
23044	23065	#ifdef __LITTLE_ENDIAN__
23045	23066	__ai __attribute__((target("neon"))) int32x4_t vqrdmulhq_s32(int32x4_t __p0, int32x4_t __p1) {
23046	23067	int32x4_t __ret;
23047		__ret = (int32x4_t) __builtin_neon_vqrdmulhq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
	23068	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqrdmulhq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 34));
23048	23069	return __ret;
23049	23070	}
23050	23071	#else
23051	23072	__ai __attribute__((target("neon"))) int32x4_t vqrdmulhq_s32(int32x4_t __p0, int32x4_t __p1) {
23052	23073	int32x4_t __ret;
23053		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
23054		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
23055		__ret = (int32x4_t) __builtin_neon_vqrdmulhq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
23056		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	23074	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	23075	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	23076	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqrdmulhq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 34));
	23077	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
23057	23078	return __ret;
23058	23079	}
23059	23080	__ai __attribute__((target("neon"))) int32x4_t __noswap_vqrdmulhq_s32(int32x4_t __p0, int32x4_t __p1) {
23060	23081	int32x4_t __ret;
23061		__ret = (int32x4_t) __builtin_neon_vqrdmulhq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
	23082	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqrdmulhq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 34));
23062	23083	return __ret;
23063	23084	}
23064	23085	#endif
...	...	@@ -23066,21 +23087,21 @@ __ai __attribute__((target("neon"))) int32x4_t __noswap_vqrdmulhq_s32(int32x4_t
23066	23087	#ifdef __LITTLE_ENDIAN__
23067	23088	__ai __attribute__((target("neon"))) int16x8_t vqrdmulhq_s16(int16x8_t __p0, int16x8_t __p1) {
23068	23089	int16x8_t __ret;
23069		__ret = (int16x8_t) __builtin_neon_vqrdmulhq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
	23090	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqrdmulhq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 33));
23070	23091	return __ret;
23071	23092	}
23072	23093	#else
23073	23094	__ai __attribute__((target("neon"))) int16x8_t vqrdmulhq_s16(int16x8_t __p0, int16x8_t __p1) {
23074	23095	int16x8_t __ret;
23075		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
23076		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
23077		__ret = (int16x8_t) __builtin_neon_vqrdmulhq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
23078		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	23096	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	23097	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	23098	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqrdmulhq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 33));
	23099	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
23079	23100	return __ret;
23080	23101	}
23081	23102	__ai __attribute__((target("neon"))) int16x8_t __noswap_vqrdmulhq_s16(int16x8_t __p0, int16x8_t __p1) {
23082	23103	int16x8_t __ret;
23083		__ret = (int16x8_t) __builtin_neon_vqrdmulhq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
	23104	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqrdmulhq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 33));
23084	23105	return __ret;
23085	23106	}
23086	23107	#endif
...	...	@@ -23088,21 +23109,21 @@ __ai __attribute__((target("neon"))) int16x8_t __noswap_vqrdmulhq_s16(int16x8_t
23088	23109	#ifdef __LITTLE_ENDIAN__
23089	23110	__ai __attribute__((target("neon"))) int32x2_t vqrdmulh_s32(int32x2_t __p0, int32x2_t __p1) {
23090	23111	int32x2_t __ret;
23091		__ret = (int32x2_t) __builtin_neon_vqrdmulh_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
	23112	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqrdmulh_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 2));
23092	23113	return __ret;
23093	23114	}
23094	23115	#else
23095	23116	__ai __attribute__((target("neon"))) int32x2_t vqrdmulh_s32(int32x2_t __p0, int32x2_t __p1) {
23096	23117	int32x2_t __ret;
23097		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
23098		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
23099		__ret = (int32x2_t) __builtin_neon_vqrdmulh_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
23100		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	23118	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	23119	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	23120	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqrdmulh_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 2));
	23121	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
23101	23122	return __ret;
23102	23123	}
23103	23124	__ai __attribute__((target("neon"))) int32x2_t __noswap_vqrdmulh_s32(int32x2_t __p0, int32x2_t __p1) {
23104	23125	int32x2_t __ret;
23105		__ret = (int32x2_t) __builtin_neon_vqrdmulh_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
	23126	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqrdmulh_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 2));
23106	23127	return __ret;
23107	23128	}
23108	23129	#endif
...	...	@@ -23110,21 +23131,21 @@ __ai __attribute__((target("neon"))) int32x2_t __noswap_vqrdmulh_s32(int32x2_t _
23110	23131	#ifdef __LITTLE_ENDIAN__
23111	23132	__ai __attribute__((target("neon"))) int16x4_t vqrdmulh_s16(int16x4_t __p0, int16x4_t __p1) {
23112	23133	int16x4_t __ret;
23113		__ret = (int16x4_t) __builtin_neon_vqrdmulh_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
	23134	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqrdmulh_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 1));
23114	23135	return __ret;
23115	23136	}
23116	23137	#else
23117	23138	__ai __attribute__((target("neon"))) int16x4_t vqrdmulh_s16(int16x4_t __p0, int16x4_t __p1) {
23118	23139	int16x4_t __ret;
23119		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
23120		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
23121		__ret = (int16x4_t) __builtin_neon_vqrdmulh_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
23122		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	23140	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	23141	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	23142	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqrdmulh_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 1));
	23143	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
23123	23144	return __ret;
23124	23145	}
23125	23146	__ai __attribute__((target("neon"))) int16x4_t __noswap_vqrdmulh_s16(int16x4_t __p0, int16x4_t __p1) {
23126	23147	int16x4_t __ret;
23127		__ret = (int16x4_t) __builtin_neon_vqrdmulh_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
	23148	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqrdmulh_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 1));
23128	23149	return __ret;
23129	23150	}
23130	23151	#endif
...	...	@@ -23138,9 +23159,9 @@ __ai __attribute__((target("neon"))) int32x4_t vqrdmulhq_n_s32(int32x4_t __p0, i
23138	23159	#else
23139	23160	__ai __attribute__((target("neon"))) int32x4_t vqrdmulhq_n_s32(int32x4_t __p0, int32_t __p1) {
23140	23161	int32x4_t __ret;
23141		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	23162	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
23142	23163	__ret = __noswap_vqrdmulhq_s32(__rev0, (int32x4_t) {__p1, __p1, __p1, __p1});
23143		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	23164	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
23144	23165	return __ret;
23145	23166	}
23146	23167	#endif
...	...	@@ -23154,9 +23175,9 @@ __ai __attribute__((target("neon"))) int16x8_t vqrdmulhq_n_s16(int16x8_t __p0, i
23154	23175	#else
23155	23176	__ai __attribute__((target("neon"))) int16x8_t vqrdmulhq_n_s16(int16x8_t __p0, int16_t __p1) {
23156	23177	int16x8_t __ret;
23157		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	23178	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
23158	23179	__ret = __noswap_vqrdmulhq_s16(__rev0, (int16x8_t) {__p1, __p1, __p1, __p1, __p1, __p1, __p1, __p1});
23159		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	23180	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
23160	23181	return __ret;
23161	23182	}
23162	23183	#endif
...	...	@@ -23170,9 +23191,9 @@ __ai __attribute__((target("neon"))) int32x2_t vqrdmulh_n_s32(int32x2_t __p0, in
23170	23191	#else
23171	23192	__ai __attribute__((target("neon"))) int32x2_t vqrdmulh_n_s32(int32x2_t __p0, int32_t __p1) {
23172	23193	int32x2_t __ret;
23173		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
	23194	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
23174	23195	__ret = __noswap_vqrdmulh_s32(__rev0, (int32x2_t) {__p1, __p1});
23175		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	23196	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
23176	23197	return __ret;
23177	23198	}
23178	23199	#endif
...	...	@@ -23186,9 +23207,9 @@ __ai __attribute__((target("neon"))) int16x4_t vqrdmulh_n_s16(int16x4_t __p0, in
23186	23207	#else
23187	23208	__ai __attribute__((target("neon"))) int16x4_t vqrdmulh_n_s16(int16x4_t __p0, int16_t __p1) {
23188	23209	int16x4_t __ret;
23189		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	23210	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
23190	23211	__ret = __noswap_vqrdmulh_s16(__rev0, (int16x4_t) {__p1, __p1, __p1, __p1});
23191		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	23212	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
23192	23213	return __ret;
23193	23214	}
23194	23215	#endif
...	...	@@ -23196,16 +23217,16 @@ __ai __attribute__((target("neon"))) int16x4_t vqrdmulh_n_s16(int16x4_t __p0, in
23196	23217	#ifdef __LITTLE_ENDIAN__
23197	23218	__ai __attribute__((target("neon"))) uint8x16_t vqrshlq_u8(uint8x16_t __p0, int8x16_t __p1) {
23198	23219	uint8x16_t __ret;
23199		__ret = (uint8x16_t) __builtin_neon_vqrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
	23220	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vqrshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 48));
23200	23221	return __ret;
23201	23222	}
23202	23223	#else
23203	23224	__ai __attribute__((target("neon"))) uint8x16_t vqrshlq_u8(uint8x16_t __p0, int8x16_t __p1) {
23204	23225	uint8x16_t __ret;
23205		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
23206		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
23207		__ret = (uint8x16_t) __builtin_neon_vqrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
23208		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	23226	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	23227	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	23228	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vqrshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 48));
	23229	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
23209	23230	return __ret;
23210	23231	}
23211	23232	#endif
...	...	@@ -23213,16 +23234,16 @@ __ai __attribute__((target("neon"))) uint8x16_t vqrshlq_u8(uint8x16_t __p0, int8
23213	23234	#ifdef __LITTLE_ENDIAN__
23214	23235	__ai __attribute__((target("neon"))) uint32x4_t vqrshlq_u32(uint32x4_t __p0, int32x4_t __p1) {
23215	23236	uint32x4_t __ret;
23216		__ret = (uint32x4_t) __builtin_neon_vqrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
	23237	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vqrshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50));
23217	23238	return __ret;
23218	23239	}
23219	23240	#else
23220	23241	__ai __attribute__((target("neon"))) uint32x4_t vqrshlq_u32(uint32x4_t __p0, int32x4_t __p1) {
23221	23242	uint32x4_t __ret;
23222		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
23223		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
23224		__ret = (uint32x4_t) __builtin_neon_vqrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
23225		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	23243	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	23244	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	23245	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vqrshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 50));
	23246	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
23226	23247	return __ret;
23227	23248	}
23228	23249	#endif
...	...	@@ -23230,16 +23251,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vqrshlq_u32(uint32x4_t __p0, int
23230	23251	#ifdef __LITTLE_ENDIAN__
23231	23252	__ai __attribute__((target("neon"))) uint64x2_t vqrshlq_u64(uint64x2_t __p0, int64x2_t __p1) {
23232	23253	uint64x2_t __ret;
23233		__ret = (uint64x2_t) __builtin_neon_vqrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 51);
	23254	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vqrshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 51));
23234	23255	return __ret;
23235	23256	}
23236	23257	#else
23237	23258	__ai __attribute__((target("neon"))) uint64x2_t vqrshlq_u64(uint64x2_t __p0, int64x2_t __p1) {
23238	23259	uint64x2_t __ret;
23239		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
23240		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
23241		__ret = (uint64x2_t) __builtin_neon_vqrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 51);
23242		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	23260	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	23261	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	23262	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vqrshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 51));
	23263	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
23243	23264	return __ret;
23244	23265	}
23245	23266	#endif
...	...	@@ -23247,16 +23268,16 @@ __ai __attribute__((target("neon"))) uint64x2_t vqrshlq_u64(uint64x2_t __p0, int
23247	23268	#ifdef __LITTLE_ENDIAN__
23248	23269	__ai __attribute__((target("neon"))) uint16x8_t vqrshlq_u16(uint16x8_t __p0, int16x8_t __p1) {
23249	23270	uint16x8_t __ret;
23250		__ret = (uint16x8_t) __builtin_neon_vqrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
	23271	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vqrshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 49));
23251	23272	return __ret;
23252	23273	}
23253	23274	#else
23254	23275	__ai __attribute__((target("neon"))) uint16x8_t vqrshlq_u16(uint16x8_t __p0, int16x8_t __p1) {
23255	23276	uint16x8_t __ret;
23256		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
23257		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
23258		__ret = (uint16x8_t) __builtin_neon_vqrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
23259		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	23277	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	23278	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	23279	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vqrshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 49));
	23280	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
23260	23281	return __ret;
23261	23282	}
23262	23283	#endif
...	...	@@ -23264,16 +23285,16 @@ __ai __attribute__((target("neon"))) uint16x8_t vqrshlq_u16(uint16x8_t __p0, int
23264	23285	#ifdef __LITTLE_ENDIAN__
23265	23286	__ai __attribute__((target("neon"))) int8x16_t vqrshlq_s8(int8x16_t __p0, int8x16_t __p1) {
23266	23287	int8x16_t __ret;
23267		__ret = (int8x16_t) __builtin_neon_vqrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
	23288	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vqrshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 32));
23268	23289	return __ret;
23269	23290	}
23270	23291	#else
23271	23292	__ai __attribute__((target("neon"))) int8x16_t vqrshlq_s8(int8x16_t __p0, int8x16_t __p1) {
23272	23293	int8x16_t __ret;
23273		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
23274		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
23275		__ret = (int8x16_t) __builtin_neon_vqrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 32);
23276		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	23294	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	23295	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	23296	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vqrshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 32));
	23297	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
23277	23298	return __ret;
23278	23299	}
23279	23300	#endif
...	...	@@ -23281,16 +23302,16 @@ __ai __attribute__((target("neon"))) int8x16_t vqrshlq_s8(int8x16_t __p0, int8x1
23281	23302	#ifdef __LITTLE_ENDIAN__
23282	23303	__ai __attribute__((target("neon"))) int32x4_t vqrshlq_s32(int32x4_t __p0, int32x4_t __p1) {
23283	23304	int32x4_t __ret;
23284		__ret = (int32x4_t) __builtin_neon_vqrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
	23305	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqrshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 34));
23285	23306	return __ret;
23286	23307	}
23287	23308	#else
23288	23309	__ai __attribute__((target("neon"))) int32x4_t vqrshlq_s32(int32x4_t __p0, int32x4_t __p1) {
23289	23310	int32x4_t __ret;
23290		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
23291		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
23292		__ret = (int32x4_t) __builtin_neon_vqrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
23293		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	23311	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	23312	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	23313	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqrshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 34));
	23314	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
23294	23315	return __ret;
23295	23316	}
23296	23317	#endif
...	...	@@ -23298,16 +23319,16 @@ __ai __attribute__((target("neon"))) int32x4_t vqrshlq_s32(int32x4_t __p0, int32
23298	23319	#ifdef __LITTLE_ENDIAN__
23299	23320	__ai __attribute__((target("neon"))) int64x2_t vqrshlq_s64(int64x2_t __p0, int64x2_t __p1) {
23300	23321	int64x2_t __ret;
23301		__ret = (int64x2_t) __builtin_neon_vqrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 35);
	23322	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vqrshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 35));
23302	23323	return __ret;
23303	23324	}
23304	23325	#else
23305	23326	__ai __attribute__((target("neon"))) int64x2_t vqrshlq_s64(int64x2_t __p0, int64x2_t __p1) {
23306	23327	int64x2_t __ret;
23307		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
23308		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
23309		__ret = (int64x2_t) __builtin_neon_vqrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 35);
23310		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	23328	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	23329	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	23330	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vqrshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 35));
	23331	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
23311	23332	return __ret;
23312	23333	}
23313	23334	#endif
...	...	@@ -23315,16 +23336,16 @@ __ai __attribute__((target("neon"))) int64x2_t vqrshlq_s64(int64x2_t __p0, int64
23315	23336	#ifdef __LITTLE_ENDIAN__
23316	23337	__ai __attribute__((target("neon"))) int16x8_t vqrshlq_s16(int16x8_t __p0, int16x8_t __p1) {
23317	23338	int16x8_t __ret;
23318		__ret = (int16x8_t) __builtin_neon_vqrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
	23339	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqrshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 33));
23319	23340	return __ret;
23320	23341	}
23321	23342	#else
23322	23343	__ai __attribute__((target("neon"))) int16x8_t vqrshlq_s16(int16x8_t __p0, int16x8_t __p1) {
23323	23344	int16x8_t __ret;
23324		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
23325		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
23326		__ret = (int16x8_t) __builtin_neon_vqrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
23327		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	23345	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	23346	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	23347	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqrshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 33));
	23348	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
23328	23349	return __ret;
23329	23350	}
23330	23351	#endif
...	...	@@ -23332,16 +23353,16 @@ __ai __attribute__((target("neon"))) int16x8_t vqrshlq_s16(int16x8_t __p0, int16
23332	23353	#ifdef __LITTLE_ENDIAN__
23333	23354	__ai __attribute__((target("neon"))) uint8x8_t vqrshl_u8(uint8x8_t __p0, int8x8_t __p1) {
23334	23355	uint8x8_t __ret;
23335		__ret = (uint8x8_t) __builtin_neon_vqrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
	23356	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqrshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 16));
23336	23357	return __ret;
23337	23358	}
23338	23359	#else
23339	23360	__ai __attribute__((target("neon"))) uint8x8_t vqrshl_u8(uint8x8_t __p0, int8x8_t __p1) {
23340	23361	uint8x8_t __ret;
23341		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
23342		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
23343		__ret = (uint8x8_t) __builtin_neon_vqrshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
23344		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	23362	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	23363	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	23364	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqrshl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 16));
	23365	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
23345	23366	return __ret;
23346	23367	}
23347	23368	#endif
...	...	@@ -23349,38 +23370,38 @@ __ai __attribute__((target("neon"))) uint8x8_t vqrshl_u8(uint8x8_t __p0, int8x8_
23349	23370	#ifdef __LITTLE_ENDIAN__
23350	23371	__ai __attribute__((target("neon"))) uint32x2_t vqrshl_u32(uint32x2_t __p0, int32x2_t __p1) {
23351	23372	uint32x2_t __ret;
23352		__ret = (uint32x2_t) __builtin_neon_vqrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
	23373	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vqrshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 18));
23353	23374	return __ret;
23354	23375	}
23355	23376	#else
23356	23377	__ai __attribute__((target("neon"))) uint32x2_t vqrshl_u32(uint32x2_t __p0, int32x2_t __p1) {
23357	23378	uint32x2_t __ret;
23358		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
23359		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
23360		__ret = (uint32x2_t) __builtin_neon_vqrshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
23361		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	23379	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	23380	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	23381	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vqrshl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 18));
	23382	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
23362	23383	return __ret;
23363	23384	}
23364	23385	#endif
23365	23386
23366	23387	__ai __attribute__((target("neon"))) uint64x1_t vqrshl_u64(uint64x1_t __p0, int64x1_t __p1) {
23367	23388	uint64x1_t __ret;
23368		__ret = (uint64x1_t) __builtin_neon_vqrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 19);
	23389	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vqrshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 19));
23369	23390	return __ret;
23370	23391	}
23371	23392	#ifdef __LITTLE_ENDIAN__
23372	23393	__ai __attribute__((target("neon"))) uint16x4_t vqrshl_u16(uint16x4_t __p0, int16x4_t __p1) {
23373	23394	uint16x4_t __ret;
23374		__ret = (uint16x4_t) __builtin_neon_vqrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
	23395	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vqrshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 17));
23375	23396	return __ret;
23376	23397	}
23377	23398	#else
23378	23399	__ai __attribute__((target("neon"))) uint16x4_t vqrshl_u16(uint16x4_t __p0, int16x4_t __p1) {
23379	23400	uint16x4_t __ret;
23380		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
23381		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
23382		__ret = (uint16x4_t) __builtin_neon_vqrshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
23383		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	23401	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	23402	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	23403	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vqrshl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 17));
	23404	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
23384	23405	return __ret;
23385	23406	}
23386	23407	#endif
...	...	@@ -23388,16 +23409,16 @@ __ai __attribute__((target("neon"))) uint16x4_t vqrshl_u16(uint16x4_t __p0, int1
23388	23409	#ifdef __LITTLE_ENDIAN__
23389	23410	__ai __attribute__((target("neon"))) int8x8_t vqrshl_s8(int8x8_t __p0, int8x8_t __p1) {
23390	23411	int8x8_t __ret;
23391		__ret = (int8x8_t) __builtin_neon_vqrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
	23412	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqrshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 0));
23392	23413	return __ret;
23393	23414	}
23394	23415	#else
23395	23416	__ai __attribute__((target("neon"))) int8x8_t vqrshl_s8(int8x8_t __p0, int8x8_t __p1) {
23396	23417	int8x8_t __ret;
23397		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
23398		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
23399		__ret = (int8x8_t) __builtin_neon_vqrshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
23400		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	23418	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	23419	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	23420	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqrshl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 0));
	23421	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
23401	23422	return __ret;
23402	23423	}
23403	23424	#endif
...	...	@@ -23405,38 +23426,38 @@ __ai __attribute__((target("neon"))) int8x8_t vqrshl_s8(int8x8_t __p0, int8x8_t
23405	23426	#ifdef __LITTLE_ENDIAN__
23406	23427	__ai __attribute__((target("neon"))) int32x2_t vqrshl_s32(int32x2_t __p0, int32x2_t __p1) {
23407	23428	int32x2_t __ret;
23408		__ret = (int32x2_t) __builtin_neon_vqrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
	23429	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqrshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 2));
23409	23430	return __ret;
23410	23431	}
23411	23432	#else
23412	23433	__ai __attribute__((target("neon"))) int32x2_t vqrshl_s32(int32x2_t __p0, int32x2_t __p1) {
23413	23434	int32x2_t __ret;
23414		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
23415		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
23416		__ret = (int32x2_t) __builtin_neon_vqrshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
23417		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	23435	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	23436	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	23437	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqrshl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 2));
	23438	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
23418	23439	return __ret;
23419	23440	}
23420	23441	#endif
23421	23442
23422	23443	__ai __attribute__((target("neon"))) int64x1_t vqrshl_s64(int64x1_t __p0, int64x1_t __p1) {
23423	23444	int64x1_t __ret;
23424		__ret = (int64x1_t) __builtin_neon_vqrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 3);
	23445	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vqrshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 3));
23425	23446	return __ret;
23426	23447	}
23427	23448	#ifdef __LITTLE_ENDIAN__
23428	23449	__ai __attribute__((target("neon"))) int16x4_t vqrshl_s16(int16x4_t __p0, int16x4_t __p1) {
23429	23450	int16x4_t __ret;
23430		__ret = (int16x4_t) __builtin_neon_vqrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
	23451	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqrshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 1));
23431	23452	return __ret;
23432	23453	}
23433	23454	#else
23434	23455	__ai __attribute__((target("neon"))) int16x4_t vqrshl_s16(int16x4_t __p0, int16x4_t __p1) {
23435	23456	int16x4_t __ret;
23436		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
23437		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
23438		__ret = (int16x4_t) __builtin_neon_vqrshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
23439		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	23457	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	23458	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	23459	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqrshl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 1));
	23460	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
23440	23461	return __ret;
23441	23462	}
23442	23463	#endif
...	...	@@ -23445,22 +23466,22 @@ __ai __attribute__((target("neon"))) int16x4_t vqrshl_s16(int16x4_t __p0, int16x
23445	23466	#define vqrshrn_n_u32(__p0, __p1) __extension__ ({ \
23446	23467	uint16x4_t __ret; \
23447	23468	uint32x4_t __s0 = __p0; \
23448		__ret = (uint16x4_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__s0, __p1, 17); \
	23469	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vqrshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 17)); \
23449	23470	__ret; \
23450	23471	})
23451	23472	#else
23452	23473	#define vqrshrn_n_u32(__p0, __p1) __extension__ ({ \
23453	23474	uint16x4_t __ret; \
23454	23475	uint32x4_t __s0 = __p0; \
23455		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
23456		__ret = (uint16x4_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__rev0, __p1, 17); \
23457		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	23476	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	23477	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vqrshrn_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 17)); \
	23478	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
23458	23479	__ret; \
23459	23480	})
23460	23481	#define __noswap_vqrshrn_n_u32(__p0, __p1) __extension__ ({ \
23461	23482	uint16x4_t __ret; \
23462	23483	uint32x4_t __s0 = __p0; \
23463		__ret = (uint16x4_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__s0, __p1, 17); \
	23484	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vqrshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 17)); \
23464	23485	__ret; \
23465	23486	})
23466	23487	#endif
...	...	@@ -23469,22 +23490,22 @@ __ai __attribute__((target("neon"))) int16x4_t vqrshl_s16(int16x4_t __p0, int16x
23469	23490	#define vqrshrn_n_u64(__p0, __p1) __extension__ ({ \
23470	23491	uint32x2_t __ret; \
23471	23492	uint64x2_t __s0 = __p0; \
23472		__ret = (uint32x2_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__s0, __p1, 18); \
	23493	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vqrshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 18)); \
23473	23494	__ret; \
23474	23495	})
23475	23496	#else
23476	23497	#define vqrshrn_n_u64(__p0, __p1) __extension__ ({ \
23477	23498	uint32x2_t __ret; \
23478	23499	uint64x2_t __s0 = __p0; \
23479		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
23480		__ret = (uint32x2_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__rev0, __p1, 18); \
23481		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	23500	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	23501	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vqrshrn_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 18)); \
	23502	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
23482	23503	__ret; \
23483	23504	})
23484	23505	#define __noswap_vqrshrn_n_u64(__p0, __p1) __extension__ ({ \
23485	23506	uint32x2_t __ret; \
23486	23507	uint64x2_t __s0 = __p0; \
23487		__ret = (uint32x2_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__s0, __p1, 18); \
	23508	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vqrshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 18)); \
23488	23509	__ret; \
23489	23510	})
23490	23511	#endif
...	...	@@ -23493,22 +23514,22 @@ __ai __attribute__((target("neon"))) int16x4_t vqrshl_s16(int16x4_t __p0, int16x
23493	23514	#define vqrshrn_n_u16(__p0, __p1) __extension__ ({ \
23494	23515	uint8x8_t __ret; \
23495	23516	uint16x8_t __s0 = __p0; \
23496		__ret = (uint8x8_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__s0, __p1, 16); \
	23517	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqrshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 16)); \
23497	23518	__ret; \
23498	23519	})
23499	23520	#else
23500	23521	#define vqrshrn_n_u16(__p0, __p1) __extension__ ({ \
23501	23522	uint8x8_t __ret; \
23502	23523	uint16x8_t __s0 = __p0; \
23503		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
23504		__ret = (uint8x8_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__rev0, __p1, 16); \
23505		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	23524	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	23525	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqrshrn_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 16)); \
	23526	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
23506	23527	__ret; \
23507	23528	})
23508	23529	#define __noswap_vqrshrn_n_u16(__p0, __p1) __extension__ ({ \
23509	23530	uint8x8_t __ret; \
23510	23531	uint16x8_t __s0 = __p0; \
23511		__ret = (uint8x8_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__s0, __p1, 16); \
	23532	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqrshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 16)); \
23512	23533	__ret; \
23513	23534	})
23514	23535	#endif
...	...	@@ -23517,22 +23538,22 @@ __ai __attribute__((target("neon"))) int16x4_t vqrshl_s16(int16x4_t __p0, int16x
23517	23538	#define vqrshrn_n_s32(__p0, __p1) __extension__ ({ \
23518	23539	int16x4_t __ret; \
23519	23540	int32x4_t __s0 = __p0; \
23520		__ret = (int16x4_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__s0, __p1, 1); \
	23541	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqrshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 1)); \
23521	23542	__ret; \
23522	23543	})
23523	23544	#else
23524	23545	#define vqrshrn_n_s32(__p0, __p1) __extension__ ({ \
23525	23546	int16x4_t __ret; \
23526	23547	int32x4_t __s0 = __p0; \
23527		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
23528		__ret = (int16x4_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__rev0, __p1, 1); \
23529		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	23548	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	23549	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqrshrn_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 1)); \
	23550	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
23530	23551	__ret; \
23531	23552	})
23532	23553	#define __noswap_vqrshrn_n_s32(__p0, __p1) __extension__ ({ \
23533	23554	int16x4_t __ret; \
23534	23555	int32x4_t __s0 = __p0; \
23535		__ret = (int16x4_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__s0, __p1, 1); \
	23556	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqrshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 1)); \
23536	23557	__ret; \
23537	23558	})
23538	23559	#endif
...	...	@@ -23541,22 +23562,22 @@ __ai __attribute__((target("neon"))) int16x4_t vqrshl_s16(int16x4_t __p0, int16x
23541	23562	#define vqrshrn_n_s64(__p0, __p1) __extension__ ({ \
23542	23563	int32x2_t __ret; \
23543	23564	int64x2_t __s0 = __p0; \
23544		__ret = (int32x2_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__s0, __p1, 2); \
	23565	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqrshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 2)); \
23545	23566	__ret; \
23546	23567	})
23547	23568	#else
23548	23569	#define vqrshrn_n_s64(__p0, __p1) __extension__ ({ \
23549	23570	int32x2_t __ret; \
23550	23571	int64x2_t __s0 = __p0; \
23551		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
23552		__ret = (int32x2_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__rev0, __p1, 2); \
23553		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	23572	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	23573	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqrshrn_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 2)); \
	23574	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
23554	23575	__ret; \
23555	23576	})
23556	23577	#define __noswap_vqrshrn_n_s64(__p0, __p1) __extension__ ({ \
23557	23578	int32x2_t __ret; \
23558	23579	int64x2_t __s0 = __p0; \
23559		__ret = (int32x2_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__s0, __p1, 2); \
	23580	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqrshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 2)); \
23560	23581	__ret; \
23561	23582	})
23562	23583	#endif
...	...	@@ -23565,22 +23586,22 @@ __ai __attribute__((target("neon"))) int16x4_t vqrshl_s16(int16x4_t __p0, int16x
23565	23586	#define vqrshrn_n_s16(__p0, __p1) __extension__ ({ \
23566	23587	int8x8_t __ret; \
23567	23588	int16x8_t __s0 = __p0; \
23568		__ret = (int8x8_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__s0, __p1, 0); \
	23589	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqrshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 0)); \
23569	23590	__ret; \
23570	23591	})
23571	23592	#else
23572	23593	#define vqrshrn_n_s16(__p0, __p1) __extension__ ({ \
23573	23594	int8x8_t __ret; \
23574	23595	int16x8_t __s0 = __p0; \
23575		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
23576		__ret = (int8x8_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__rev0, __p1, 0); \
23577		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	23596	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	23597	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqrshrn_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 0)); \
	23598	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
23578	23599	__ret; \
23579	23600	})
23580	23601	#define __noswap_vqrshrn_n_s16(__p0, __p1) __extension__ ({ \
23581	23602	int8x8_t __ret; \
23582	23603	int16x8_t __s0 = __p0; \
23583		__ret = (int8x8_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__s0, __p1, 0); \
	23604	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqrshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 0)); \
23584	23605	__ret; \
23585	23606	})
23586	23607	#endif
...	...	@@ -23589,22 +23610,22 @@ __ai __attribute__((target("neon"))) int16x4_t vqrshl_s16(int16x4_t __p0, int16x
23589	23610	#define vqrshrun_n_s32(__p0, __p1) __extension__ ({ \
23590	23611	uint16x4_t __ret; \
23591	23612	int32x4_t __s0 = __p0; \
23592		__ret = (uint16x4_t) __builtin_neon_vqrshrun_n_v((int8x16_t)__s0, __p1, 17); \
	23613	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vqrshrun_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 17)); \
23593	23614	__ret; \
23594	23615	})
23595	23616	#else
23596	23617	#define vqrshrun_n_s32(__p0, __p1) __extension__ ({ \
23597	23618	uint16x4_t __ret; \
23598	23619	int32x4_t __s0 = __p0; \
23599		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
23600		__ret = (uint16x4_t) __builtin_neon_vqrshrun_n_v((int8x16_t)__rev0, __p1, 17); \
23601		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	23620	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	23621	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vqrshrun_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 17)); \
	23622	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
23602	23623	__ret; \
23603	23624	})
23604	23625	#define __noswap_vqrshrun_n_s32(__p0, __p1) __extension__ ({ \
23605	23626	uint16x4_t __ret; \
23606	23627	int32x4_t __s0 = __p0; \
23607		__ret = (uint16x4_t) __builtin_neon_vqrshrun_n_v((int8x16_t)__s0, __p1, 17); \
	23628	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vqrshrun_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 17)); \
23608	23629	__ret; \
23609	23630	})
23610	23631	#endif
...	...	@@ -23613,22 +23634,22 @@ __ai __attribute__((target("neon"))) int16x4_t vqrshl_s16(int16x4_t __p0, int16x
23613	23634	#define vqrshrun_n_s64(__p0, __p1) __extension__ ({ \
23614	23635	uint32x2_t __ret; \
23615	23636	int64x2_t __s0 = __p0; \
23616		__ret = (uint32x2_t) __builtin_neon_vqrshrun_n_v((int8x16_t)__s0, __p1, 18); \
	23637	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vqrshrun_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 18)); \
23617	23638	__ret; \
23618	23639	})
23619	23640	#else
23620	23641	#define vqrshrun_n_s64(__p0, __p1) __extension__ ({ \
23621	23642	uint32x2_t __ret; \
23622	23643	int64x2_t __s0 = __p0; \
23623		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
23624		__ret = (uint32x2_t) __builtin_neon_vqrshrun_n_v((int8x16_t)__rev0, __p1, 18); \
23625		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	23644	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	23645	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vqrshrun_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 18)); \
	23646	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
23626	23647	__ret; \
23627	23648	})
23628	23649	#define __noswap_vqrshrun_n_s64(__p0, __p1) __extension__ ({ \
23629	23650	uint32x2_t __ret; \
23630	23651	int64x2_t __s0 = __p0; \
23631		__ret = (uint32x2_t) __builtin_neon_vqrshrun_n_v((int8x16_t)__s0, __p1, 18); \
	23652	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vqrshrun_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 18)); \
23632	23653	__ret; \
23633	23654	})
23634	23655	#endif
...	...	@@ -23637,22 +23658,22 @@ __ai __attribute__((target("neon"))) int16x4_t vqrshl_s16(int16x4_t __p0, int16x
23637	23658	#define vqrshrun_n_s16(__p0, __p1) __extension__ ({ \
23638	23659	uint8x8_t __ret; \
23639	23660	int16x8_t __s0 = __p0; \
23640		__ret = (uint8x8_t) __builtin_neon_vqrshrun_n_v((int8x16_t)__s0, __p1, 16); \
	23661	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqrshrun_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 16)); \
23641	23662	__ret; \
23642	23663	})
23643	23664	#else
23644	23665	#define vqrshrun_n_s16(__p0, __p1) __extension__ ({ \
23645	23666	uint8x8_t __ret; \
23646	23667	int16x8_t __s0 = __p0; \
23647		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
23648		__ret = (uint8x8_t) __builtin_neon_vqrshrun_n_v((int8x16_t)__rev0, __p1, 16); \
23649		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	23668	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	23669	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqrshrun_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 16)); \
	23670	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
23650	23671	__ret; \
23651	23672	})
23652	23673	#define __noswap_vqrshrun_n_s16(__p0, __p1) __extension__ ({ \
23653	23674	uint8x8_t __ret; \
23654	23675	int16x8_t __s0 = __p0; \
23655		__ret = (uint8x8_t) __builtin_neon_vqrshrun_n_v((int8x16_t)__s0, __p1, 16); \
	23676	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqrshrun_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 16)); \
23656	23677	__ret; \
23657	23678	})
23658	23679	#endif
...	...	@@ -23660,16 +23681,16 @@ __ai __attribute__((target("neon"))) int16x4_t vqrshl_s16(int16x4_t __p0, int16x
23660	23681	#ifdef __LITTLE_ENDIAN__
23661	23682	__ai __attribute__((target("neon"))) uint8x16_t vqshlq_u8(uint8x16_t __p0, int8x16_t __p1) {
23662	23683	uint8x16_t __ret;
23663		__ret = (uint8x16_t) __builtin_neon_vqshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
	23684	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vqshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 48));
23664	23685	return __ret;
23665	23686	}
23666	23687	#else
23667	23688	__ai __attribute__((target("neon"))) uint8x16_t vqshlq_u8(uint8x16_t __p0, int8x16_t __p1) {
23668	23689	uint8x16_t __ret;
23669		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
23670		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
23671		__ret = (uint8x16_t) __builtin_neon_vqshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
23672		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	23690	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	23691	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	23692	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vqshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 48));
	23693	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
23673	23694	return __ret;
23674	23695	}
23675	23696	#endif
...	...	@@ -23677,16 +23698,16 @@ __ai __attribute__((target("neon"))) uint8x16_t vqshlq_u8(uint8x16_t __p0, int8x
23677	23698	#ifdef __LITTLE_ENDIAN__
23678	23699	__ai __attribute__((target("neon"))) uint32x4_t vqshlq_u32(uint32x4_t __p0, int32x4_t __p1) {
23679	23700	uint32x4_t __ret;
23680		__ret = (uint32x4_t) __builtin_neon_vqshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
	23701	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vqshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50));
23681	23702	return __ret;
23682	23703	}
23683	23704	#else
23684	23705	__ai __attribute__((target("neon"))) uint32x4_t vqshlq_u32(uint32x4_t __p0, int32x4_t __p1) {
23685	23706	uint32x4_t __ret;
23686		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
23687		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
23688		__ret = (uint32x4_t) __builtin_neon_vqshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
23689		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	23707	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	23708	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	23709	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vqshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 50));
	23710	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
23690	23711	return __ret;
23691	23712	}
23692	23713	#endif
...	...	@@ -23694,16 +23715,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vqshlq_u32(uint32x4_t __p0, int3
23694	23715	#ifdef __LITTLE_ENDIAN__
23695	23716	__ai __attribute__((target("neon"))) uint64x2_t vqshlq_u64(uint64x2_t __p0, int64x2_t __p1) {
23696	23717	uint64x2_t __ret;
23697		__ret = (uint64x2_t) __builtin_neon_vqshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 51);
	23718	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vqshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 51));
23698	23719	return __ret;
23699	23720	}
23700	23721	#else
23701	23722	__ai __attribute__((target("neon"))) uint64x2_t vqshlq_u64(uint64x2_t __p0, int64x2_t __p1) {
23702	23723	uint64x2_t __ret;
23703		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
23704		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
23705		__ret = (uint64x2_t) __builtin_neon_vqshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 51);
23706		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	23724	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	23725	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	23726	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vqshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 51));
	23727	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
23707	23728	return __ret;
23708	23729	}
23709	23730	#endif
...	...	@@ -23711,16 +23732,16 @@ __ai __attribute__((target("neon"))) uint64x2_t vqshlq_u64(uint64x2_t __p0, int6
23711	23732	#ifdef __LITTLE_ENDIAN__
23712	23733	__ai __attribute__((target("neon"))) uint16x8_t vqshlq_u16(uint16x8_t __p0, int16x8_t __p1) {
23713	23734	uint16x8_t __ret;
23714		__ret = (uint16x8_t) __builtin_neon_vqshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
	23735	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vqshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 49));
23715	23736	return __ret;
23716	23737	}
23717	23738	#else
23718	23739	__ai __attribute__((target("neon"))) uint16x8_t vqshlq_u16(uint16x8_t __p0, int16x8_t __p1) {
23719	23740	uint16x8_t __ret;
23720		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
23721		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
23722		__ret = (uint16x8_t) __builtin_neon_vqshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
23723		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	23741	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	23742	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	23743	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vqshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 49));
	23744	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
23724	23745	return __ret;
23725	23746	}
23726	23747	#endif
...	...	@@ -23728,16 +23749,16 @@ __ai __attribute__((target("neon"))) uint16x8_t vqshlq_u16(uint16x8_t __p0, int1
23728	23749	#ifdef __LITTLE_ENDIAN__
23729	23750	__ai __attribute__((target("neon"))) int8x16_t vqshlq_s8(int8x16_t __p0, int8x16_t __p1) {
23730	23751	int8x16_t __ret;
23731		__ret = (int8x16_t) __builtin_neon_vqshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
	23752	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vqshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 32));
23732	23753	return __ret;
23733	23754	}
23734	23755	#else
23735	23756	__ai __attribute__((target("neon"))) int8x16_t vqshlq_s8(int8x16_t __p0, int8x16_t __p1) {
23736	23757	int8x16_t __ret;
23737		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
23738		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
23739		__ret = (int8x16_t) __builtin_neon_vqshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 32);
23740		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	23758	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	23759	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	23760	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vqshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 32));
	23761	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
23741	23762	return __ret;
23742	23763	}
23743	23764	#endif
...	...	@@ -23745,16 +23766,16 @@ __ai __attribute__((target("neon"))) int8x16_t vqshlq_s8(int8x16_t __p0, int8x16
23745	23766	#ifdef __LITTLE_ENDIAN__
23746	23767	__ai __attribute__((target("neon"))) int32x4_t vqshlq_s32(int32x4_t __p0, int32x4_t __p1) {
23747	23768	int32x4_t __ret;
23748		__ret = (int32x4_t) __builtin_neon_vqshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
	23769	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 34));
23749	23770	return __ret;
23750	23771	}
23751	23772	#else
23752	23773	__ai __attribute__((target("neon"))) int32x4_t vqshlq_s32(int32x4_t __p0, int32x4_t __p1) {
23753	23774	int32x4_t __ret;
23754		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
23755		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
23756		__ret = (int32x4_t) __builtin_neon_vqshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
23757		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	23775	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	23776	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	23777	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 34));
	23778	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
23758	23779	return __ret;
23759	23780	}
23760	23781	#endif
...	...	@@ -23762,16 +23783,16 @@ __ai __attribute__((target("neon"))) int32x4_t vqshlq_s32(int32x4_t __p0, int32x
23762	23783	#ifdef __LITTLE_ENDIAN__
23763	23784	__ai __attribute__((target("neon"))) int64x2_t vqshlq_s64(int64x2_t __p0, int64x2_t __p1) {
23764	23785	int64x2_t __ret;
23765		__ret = (int64x2_t) __builtin_neon_vqshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 35);
	23786	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vqshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 35));
23766	23787	return __ret;
23767	23788	}
23768	23789	#else
23769	23790	__ai __attribute__((target("neon"))) int64x2_t vqshlq_s64(int64x2_t __p0, int64x2_t __p1) {
23770	23791	int64x2_t __ret;
23771		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
23772		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
23773		__ret = (int64x2_t) __builtin_neon_vqshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 35);
23774		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	23792	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	23793	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	23794	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vqshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 35));
	23795	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
23775	23796	return __ret;
23776	23797	}
23777	23798	#endif
...	...	@@ -23779,16 +23800,16 @@ __ai __attribute__((target("neon"))) int64x2_t vqshlq_s64(int64x2_t __p0, int64x
23779	23800	#ifdef __LITTLE_ENDIAN__
23780	23801	__ai __attribute__((target("neon"))) int16x8_t vqshlq_s16(int16x8_t __p0, int16x8_t __p1) {
23781	23802	int16x8_t __ret;
23782		__ret = (int16x8_t) __builtin_neon_vqshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
	23803	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 33));
23783	23804	return __ret;
23784	23805	}
23785	23806	#else
23786	23807	__ai __attribute__((target("neon"))) int16x8_t vqshlq_s16(int16x8_t __p0, int16x8_t __p1) {
23787	23808	int16x8_t __ret;
23788		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
23789		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
23790		__ret = (int16x8_t) __builtin_neon_vqshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
23791		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	23809	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	23810	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	23811	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 33));
	23812	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
23792	23813	return __ret;
23793	23814	}
23794	23815	#endif
...	...	@@ -23796,16 +23817,16 @@ __ai __attribute__((target("neon"))) int16x8_t vqshlq_s16(int16x8_t __p0, int16x
23796	23817	#ifdef __LITTLE_ENDIAN__
23797	23818	__ai __attribute__((target("neon"))) uint8x8_t vqshl_u8(uint8x8_t __p0, int8x8_t __p1) {
23798	23819	uint8x8_t __ret;
23799		__ret = (uint8x8_t) __builtin_neon_vqshl_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
	23820	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 16));
23800	23821	return __ret;
23801	23822	}
23802	23823	#else
23803	23824	__ai __attribute__((target("neon"))) uint8x8_t vqshl_u8(uint8x8_t __p0, int8x8_t __p1) {
23804	23825	uint8x8_t __ret;
23805		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
23806		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
23807		__ret = (uint8x8_t) __builtin_neon_vqshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
23808		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	23826	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	23827	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	23828	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqshl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 16));
	23829	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
23809	23830	return __ret;
23810	23831	}
23811	23832	#endif
...	...	@@ -23813,38 +23834,38 @@ __ai __attribute__((target("neon"))) uint8x8_t vqshl_u8(uint8x8_t __p0, int8x8_t
23813	23834	#ifdef __LITTLE_ENDIAN__
23814	23835	__ai __attribute__((target("neon"))) uint32x2_t vqshl_u32(uint32x2_t __p0, int32x2_t __p1) {
23815	23836	uint32x2_t __ret;
23816		__ret = (uint32x2_t) __builtin_neon_vqshl_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
	23837	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vqshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 18));
23817	23838	return __ret;
23818	23839	}
23819	23840	#else
23820	23841	__ai __attribute__((target("neon"))) uint32x2_t vqshl_u32(uint32x2_t __p0, int32x2_t __p1) {
23821	23842	uint32x2_t __ret;
23822		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
23823		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
23824		__ret = (uint32x2_t) __builtin_neon_vqshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
23825		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	23843	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	23844	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	23845	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vqshl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 18));
	23846	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
23826	23847	return __ret;
23827	23848	}
23828	23849	#endif
23829	23850
23830	23851	__ai __attribute__((target("neon"))) uint64x1_t vqshl_u64(uint64x1_t __p0, int64x1_t __p1) {
23831	23852	uint64x1_t __ret;
23832		__ret = (uint64x1_t) __builtin_neon_vqshl_v((int8x8_t)__p0, (int8x8_t)__p1, 19);
	23853	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vqshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 19));
23833	23854	return __ret;
23834	23855	}
23835	23856	#ifdef __LITTLE_ENDIAN__
23836	23857	__ai __attribute__((target("neon"))) uint16x4_t vqshl_u16(uint16x4_t __p0, int16x4_t __p1) {
23837	23858	uint16x4_t __ret;
23838		__ret = (uint16x4_t) __builtin_neon_vqshl_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
	23859	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vqshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 17));
23839	23860	return __ret;
23840	23861	}
23841	23862	#else
23842	23863	__ai __attribute__((target("neon"))) uint16x4_t vqshl_u16(uint16x4_t __p0, int16x4_t __p1) {
23843	23864	uint16x4_t __ret;
23844		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
23845		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
23846		__ret = (uint16x4_t) __builtin_neon_vqshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
23847		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	23865	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	23866	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	23867	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vqshl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 17));
	23868	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
23848	23869	return __ret;
23849	23870	}
23850	23871	#endif
...	...	@@ -23852,16 +23873,16 @@ __ai __attribute__((target("neon"))) uint16x4_t vqshl_u16(uint16x4_t __p0, int16
23852	23873	#ifdef __LITTLE_ENDIAN__
23853	23874	__ai __attribute__((target("neon"))) int8x8_t vqshl_s8(int8x8_t __p0, int8x8_t __p1) {
23854	23875	int8x8_t __ret;
23855		__ret = (int8x8_t) __builtin_neon_vqshl_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
	23876	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 0));
23856	23877	return __ret;
23857	23878	}
23858	23879	#else
23859	23880	__ai __attribute__((target("neon"))) int8x8_t vqshl_s8(int8x8_t __p0, int8x8_t __p1) {
23860	23881	int8x8_t __ret;
23861		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
23862		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
23863		__ret = (int8x8_t) __builtin_neon_vqshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
23864		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	23882	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	23883	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	23884	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqshl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 0));
	23885	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
23865	23886	return __ret;
23866	23887	}
23867	23888	#endif
...	...	@@ -23869,38 +23890,38 @@ __ai __attribute__((target("neon"))) int8x8_t vqshl_s8(int8x8_t __p0, int8x8_t _
23869	23890	#ifdef __LITTLE_ENDIAN__
23870	23891	__ai __attribute__((target("neon"))) int32x2_t vqshl_s32(int32x2_t __p0, int32x2_t __p1) {
23871	23892	int32x2_t __ret;
23872		__ret = (int32x2_t) __builtin_neon_vqshl_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
	23893	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 2));
23873	23894	return __ret;
23874	23895	}
23875	23896	#else
23876	23897	__ai __attribute__((target("neon"))) int32x2_t vqshl_s32(int32x2_t __p0, int32x2_t __p1) {
23877	23898	int32x2_t __ret;
23878		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
23879		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
23880		__ret = (int32x2_t) __builtin_neon_vqshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
23881		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	23899	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	23900	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	23901	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqshl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 2));
	23902	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
23882	23903	return __ret;
23883	23904	}
23884	23905	#endif
23885	23906
23886	23907	__ai __attribute__((target("neon"))) int64x1_t vqshl_s64(int64x1_t __p0, int64x1_t __p1) {
23887	23908	int64x1_t __ret;
23888		__ret = (int64x1_t) __builtin_neon_vqshl_v((int8x8_t)__p0, (int8x8_t)__p1, 3);
	23909	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vqshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 3));
23889	23910	return __ret;
23890	23911	}
23891	23912	#ifdef __LITTLE_ENDIAN__
23892	23913	__ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4_t __p1) {
23893	23914	int16x4_t __ret;
23894		__ret = (int16x4_t) __builtin_neon_vqshl_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
	23915	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 1));
23895	23916	return __ret;
23896	23917	}
23897	23918	#else
23898	23919	__ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4_t __p1) {
23899	23920	int16x4_t __ret;
23900		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
23901		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
23902		__ret = (int16x4_t) __builtin_neon_vqshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
23903		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	23921	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	23922	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	23923	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqshl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 1));
	23924	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
23904	23925	return __ret;
23905	23926	}
23906	23927	#endif
...	...	@@ -23909,16 +23930,16 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
23909	23930	#define vqshlq_n_u8(__p0, __p1) __extension__ ({ \
23910	23931	uint8x16_t __ret; \
23911	23932	uint8x16_t __s0 = __p0; \
23912		__ret = (uint8x16_t) __builtin_neon_vqshlq_n_v((int8x16_t)__s0, __p1, 48); \
	23933	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vqshlq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 48)); \
23913	23934	__ret; \
23914	23935	})
23915	23936	#else
23916	23937	#define vqshlq_n_u8(__p0, __p1) __extension__ ({ \
23917	23938	uint8x16_t __ret; \
23918	23939	uint8x16_t __s0 = __p0; \
23919		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
23920		__ret = (uint8x16_t) __builtin_neon_vqshlq_n_v((int8x16_t)__rev0, __p1, 48); \
23921		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	23940	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	23941	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vqshlq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 48)); \
	23942	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
23922	23943	__ret; \
23923	23944	})
23924	23945	#endif
...	...	@@ -23927,16 +23948,16 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
23927	23948	#define vqshlq_n_u32(__p0, __p1) __extension__ ({ \
23928	23949	uint32x4_t __ret; \
23929	23950	uint32x4_t __s0 = __p0; \
23930		__ret = (uint32x4_t) __builtin_neon_vqshlq_n_v((int8x16_t)__s0, __p1, 50); \
	23951	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vqshlq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 50)); \
23931	23952	__ret; \
23932	23953	})
23933	23954	#else
23934	23955	#define vqshlq_n_u32(__p0, __p1) __extension__ ({ \
23935	23956	uint32x4_t __ret; \
23936	23957	uint32x4_t __s0 = __p0; \
23937		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
23938		__ret = (uint32x4_t) __builtin_neon_vqshlq_n_v((int8x16_t)__rev0, __p1, 50); \
23939		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	23958	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	23959	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vqshlq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 50)); \
	23960	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
23940	23961	__ret; \
23941	23962	})
23942	23963	#endif
...	...	@@ -23945,16 +23966,16 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
23945	23966	#define vqshlq_n_u64(__p0, __p1) __extension__ ({ \
23946	23967	uint64x2_t __ret; \
23947	23968	uint64x2_t __s0 = __p0; \
23948		__ret = (uint64x2_t) __builtin_neon_vqshlq_n_v((int8x16_t)__s0, __p1, 51); \
	23969	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vqshlq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 51)); \
23949	23970	__ret; \
23950	23971	})
23951	23972	#else
23952	23973	#define vqshlq_n_u64(__p0, __p1) __extension__ ({ \
23953	23974	uint64x2_t __ret; \
23954	23975	uint64x2_t __s0 = __p0; \
23955		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
23956		__ret = (uint64x2_t) __builtin_neon_vqshlq_n_v((int8x16_t)__rev0, __p1, 51); \
23957		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	23976	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	23977	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vqshlq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 51)); \
	23978	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
23958	23979	__ret; \
23959	23980	})
23960	23981	#endif
...	...	@@ -23963,16 +23984,16 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
23963	23984	#define vqshlq_n_u16(__p0, __p1) __extension__ ({ \
23964	23985	uint16x8_t __ret; \
23965	23986	uint16x8_t __s0 = __p0; \
23966		__ret = (uint16x8_t) __builtin_neon_vqshlq_n_v((int8x16_t)__s0, __p1, 49); \
	23987	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vqshlq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 49)); \
23967	23988	__ret; \
23968	23989	})
23969	23990	#else
23970	23991	#define vqshlq_n_u16(__p0, __p1) __extension__ ({ \
23971	23992	uint16x8_t __ret; \
23972	23993	uint16x8_t __s0 = __p0; \
23973		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
23974		__ret = (uint16x8_t) __builtin_neon_vqshlq_n_v((int8x16_t)__rev0, __p1, 49); \
23975		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	23994	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	23995	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vqshlq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 49)); \
	23996	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
23976	23997	__ret; \
23977	23998	})
23978	23999	#endif
...	...	@@ -23981,16 +24002,16 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
23981	24002	#define vqshlq_n_s8(__p0, __p1) __extension__ ({ \
23982	24003	int8x16_t __ret; \
23983	24004	int8x16_t __s0 = __p0; \
23984		__ret = (int8x16_t) __builtin_neon_vqshlq_n_v((int8x16_t)__s0, __p1, 32); \
	24005	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vqshlq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 32)); \
23985	24006	__ret; \
23986	24007	})
23987	24008	#else
23988	24009	#define vqshlq_n_s8(__p0, __p1) __extension__ ({ \
23989	24010	int8x16_t __ret; \
23990	24011	int8x16_t __s0 = __p0; \
23991		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
23992		__ret = (int8x16_t) __builtin_neon_vqshlq_n_v((int8x16_t)__rev0, __p1, 32); \
23993		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	24012	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	24013	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vqshlq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 32)); \
	24014	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
23994	24015	__ret; \
23995	24016	})
23996	24017	#endif
...	...	@@ -23999,16 +24020,16 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
23999	24020	#define vqshlq_n_s32(__p0, __p1) __extension__ ({ \
24000	24021	int32x4_t __ret; \
24001	24022	int32x4_t __s0 = __p0; \
24002		__ret = (int32x4_t) __builtin_neon_vqshlq_n_v((int8x16_t)__s0, __p1, 34); \
	24023	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqshlq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 34)); \
24003	24024	__ret; \
24004	24025	})
24005	24026	#else
24006	24027	#define vqshlq_n_s32(__p0, __p1) __extension__ ({ \
24007	24028	int32x4_t __ret; \
24008	24029	int32x4_t __s0 = __p0; \
24009		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
24010		__ret = (int32x4_t) __builtin_neon_vqshlq_n_v((int8x16_t)__rev0, __p1, 34); \
24011		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	24030	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	24031	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqshlq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 34)); \
	24032	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
24012	24033	__ret; \
24013	24034	})
24014	24035	#endif
...	...	@@ -24017,16 +24038,16 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
24017	24038	#define vqshlq_n_s64(__p0, __p1) __extension__ ({ \
24018	24039	int64x2_t __ret; \
24019	24040	int64x2_t __s0 = __p0; \
24020		__ret = (int64x2_t) __builtin_neon_vqshlq_n_v((int8x16_t)__s0, __p1, 35); \
	24041	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vqshlq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 35)); \
24021	24042	__ret; \
24022	24043	})
24023	24044	#else
24024	24045	#define vqshlq_n_s64(__p0, __p1) __extension__ ({ \
24025	24046	int64x2_t __ret; \
24026	24047	int64x2_t __s0 = __p0; \
24027		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
24028		__ret = (int64x2_t) __builtin_neon_vqshlq_n_v((int8x16_t)__rev0, __p1, 35); \
24029		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	24048	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	24049	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vqshlq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 35)); \
	24050	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
24030	24051	__ret; \
24031	24052	})
24032	24053	#endif
...	...	@@ -24035,16 +24056,16 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
24035	24056	#define vqshlq_n_s16(__p0, __p1) __extension__ ({ \
24036	24057	int16x8_t __ret; \
24037	24058	int16x8_t __s0 = __p0; \
24038		__ret = (int16x8_t) __builtin_neon_vqshlq_n_v((int8x16_t)__s0, __p1, 33); \
	24059	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqshlq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 33)); \
24039	24060	__ret; \
24040	24061	})
24041	24062	#else
24042	24063	#define vqshlq_n_s16(__p0, __p1) __extension__ ({ \
24043	24064	int16x8_t __ret; \
24044	24065	int16x8_t __s0 = __p0; \
24045		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
24046		__ret = (int16x8_t) __builtin_neon_vqshlq_n_v((int8x16_t)__rev0, __p1, 33); \
24047		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	24066	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	24067	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqshlq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 33)); \
	24068	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
24048	24069	__ret; \
24049	24070	})
24050	24071	#endif
...	...	@@ -24053,16 +24074,16 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
24053	24074	#define vqshl_n_u8(__p0, __p1) __extension__ ({ \
24054	24075	uint8x8_t __ret; \
24055	24076	uint8x8_t __s0 = __p0; \
24056		__ret = (uint8x8_t) __builtin_neon_vqshl_n_v((int8x8_t)__s0, __p1, 16); \
	24077	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqshl_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 16)); \
24057	24078	__ret; \
24058	24079	})
24059	24080	#else
24060	24081	#define vqshl_n_u8(__p0, __p1) __extension__ ({ \
24061	24082	uint8x8_t __ret; \
24062	24083	uint8x8_t __s0 = __p0; \
24063		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
24064		__ret = (uint8x8_t) __builtin_neon_vqshl_n_v((int8x8_t)__rev0, __p1, 16); \
24065		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	24084	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	24085	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqshl_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 16)); \
	24086	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
24066	24087	__ret; \
24067	24088	})
24068	24089	#endif
...	...	@@ -24071,16 +24092,16 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
24071	24092	#define vqshl_n_u32(__p0, __p1) __extension__ ({ \
24072	24093	uint32x2_t __ret; \
24073	24094	uint32x2_t __s0 = __p0; \
24074		__ret = (uint32x2_t) __builtin_neon_vqshl_n_v((int8x8_t)__s0, __p1, 18); \
	24095	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vqshl_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 18)); \
24075	24096	__ret; \
24076	24097	})
24077	24098	#else
24078	24099	#define vqshl_n_u32(__p0, __p1) __extension__ ({ \
24079	24100	uint32x2_t __ret; \
24080	24101	uint32x2_t __s0 = __p0; \
24081		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
24082		__ret = (uint32x2_t) __builtin_neon_vqshl_n_v((int8x8_t)__rev0, __p1, 18); \
24083		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	24102	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	24103	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vqshl_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 18)); \
	24104	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
24084	24105	__ret; \
24085	24106	})
24086	24107	#endif
...	...	@@ -24088,23 +24109,23 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
24088	24109	#define vqshl_n_u64(__p0, __p1) __extension__ ({ \
24089	24110	uint64x1_t __ret; \
24090	24111	uint64x1_t __s0 = __p0; \
24091		__ret = (uint64x1_t) __builtin_neon_vqshl_n_v((int8x8_t)__s0, __p1, 19); \
	24112	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vqshl_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 19)); \
24092	24113	__ret; \
24093	24114	})
24094	24115	#ifdef __LITTLE_ENDIAN__
24095	24116	#define vqshl_n_u16(__p0, __p1) __extension__ ({ \
24096	24117	uint16x4_t __ret; \
24097	24118	uint16x4_t __s0 = __p0; \
24098		__ret = (uint16x4_t) __builtin_neon_vqshl_n_v((int8x8_t)__s0, __p1, 17); \
	24119	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vqshl_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 17)); \
24099	24120	__ret; \
24100	24121	})
24101	24122	#else
24102	24123	#define vqshl_n_u16(__p0, __p1) __extension__ ({ \
24103	24124	uint16x4_t __ret; \
24104	24125	uint16x4_t __s0 = __p0; \
24105		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
24106		__ret = (uint16x4_t) __builtin_neon_vqshl_n_v((int8x8_t)__rev0, __p1, 17); \
24107		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	24126	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	24127	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vqshl_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 17)); \
	24128	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
24108	24129	__ret; \
24109	24130	})
24110	24131	#endif
...	...	@@ -24113,16 +24134,16 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
24113	24134	#define vqshl_n_s8(__p0, __p1) __extension__ ({ \
24114	24135	int8x8_t __ret; \
24115	24136	int8x8_t __s0 = __p0; \
24116		__ret = (int8x8_t) __builtin_neon_vqshl_n_v((int8x8_t)__s0, __p1, 0); \
	24137	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqshl_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 0)); \
24117	24138	__ret; \
24118	24139	})
24119	24140	#else
24120	24141	#define vqshl_n_s8(__p0, __p1) __extension__ ({ \
24121	24142	int8x8_t __ret; \
24122	24143	int8x8_t __s0 = __p0; \
24123		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
24124		__ret = (int8x8_t) __builtin_neon_vqshl_n_v((int8x8_t)__rev0, __p1, 0); \
24125		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	24144	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	24145	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqshl_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 0)); \
	24146	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
24126	24147	__ret; \
24127	24148	})
24128	24149	#endif
...	...	@@ -24131,16 +24152,16 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
24131	24152	#define vqshl_n_s32(__p0, __p1) __extension__ ({ \
24132	24153	int32x2_t __ret; \
24133	24154	int32x2_t __s0 = __p0; \
24134		__ret = (int32x2_t) __builtin_neon_vqshl_n_v((int8x8_t)__s0, __p1, 2); \
	24155	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqshl_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 2)); \
24135	24156	__ret; \
24136	24157	})
24137	24158	#else
24138	24159	#define vqshl_n_s32(__p0, __p1) __extension__ ({ \
24139	24160	int32x2_t __ret; \
24140	24161	int32x2_t __s0 = __p0; \
24141		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
24142		__ret = (int32x2_t) __builtin_neon_vqshl_n_v((int8x8_t)__rev0, __p1, 2); \
24143		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	24162	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	24163	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqshl_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 2)); \
	24164	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
24144	24165	__ret; \
24145	24166	})
24146	24167	#endif
...	...	@@ -24148,23 +24169,23 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
24148	24169	#define vqshl_n_s64(__p0, __p1) __extension__ ({ \
24149	24170	int64x1_t __ret; \
24150	24171	int64x1_t __s0 = __p0; \
24151		__ret = (int64x1_t) __builtin_neon_vqshl_n_v((int8x8_t)__s0, __p1, 3); \
	24172	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vqshl_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 3)); \
24152	24173	__ret; \
24153	24174	})
24154	24175	#ifdef __LITTLE_ENDIAN__
24155	24176	#define vqshl_n_s16(__p0, __p1) __extension__ ({ \
24156	24177	int16x4_t __ret; \
24157	24178	int16x4_t __s0 = __p0; \
24158		__ret = (int16x4_t) __builtin_neon_vqshl_n_v((int8x8_t)__s0, __p1, 1); \
	24179	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqshl_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 1)); \
24159	24180	__ret; \
24160	24181	})
24161	24182	#else
24162	24183	#define vqshl_n_s16(__p0, __p1) __extension__ ({ \
24163	24184	int16x4_t __ret; \
24164	24185	int16x4_t __s0 = __p0; \
24165		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
24166		__ret = (int16x4_t) __builtin_neon_vqshl_n_v((int8x8_t)__rev0, __p1, 1); \
24167		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	24186	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	24187	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqshl_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 1)); \
	24188	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
24168	24189	__ret; \
24169	24190	})
24170	24191	#endif
...	...	@@ -24173,16 +24194,16 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
24173	24194	#define vqshluq_n_s8(__p0, __p1) __extension__ ({ \
24174	24195	uint8x16_t __ret; \
24175	24196	int8x16_t __s0 = __p0; \
24176		__ret = (uint8x16_t) __builtin_neon_vqshluq_n_v((int8x16_t)__s0, __p1, 48); \
	24197	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vqshluq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 48)); \
24177	24198	__ret; \
24178	24199	})
24179	24200	#else
24180	24201	#define vqshluq_n_s8(__p0, __p1) __extension__ ({ \
24181	24202	uint8x16_t __ret; \
24182	24203	int8x16_t __s0 = __p0; \
24183		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
24184		__ret = (uint8x16_t) __builtin_neon_vqshluq_n_v((int8x16_t)__rev0, __p1, 48); \
24185		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	24204	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	24205	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vqshluq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 48)); \
	24206	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
24186	24207	__ret; \
24187	24208	})
24188	24209	#endif
...	...	@@ -24191,16 +24212,16 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
24191	24212	#define vqshluq_n_s32(__p0, __p1) __extension__ ({ \
24192	24213	uint32x4_t __ret; \
24193	24214	int32x4_t __s0 = __p0; \
24194		__ret = (uint32x4_t) __builtin_neon_vqshluq_n_v((int8x16_t)__s0, __p1, 50); \
	24215	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vqshluq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 50)); \
24195	24216	__ret; \
24196	24217	})
24197	24218	#else
24198	24219	#define vqshluq_n_s32(__p0, __p1) __extension__ ({ \
24199	24220	uint32x4_t __ret; \
24200	24221	int32x4_t __s0 = __p0; \
24201		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
24202		__ret = (uint32x4_t) __builtin_neon_vqshluq_n_v((int8x16_t)__rev0, __p1, 50); \
24203		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	24222	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	24223	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vqshluq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 50)); \
	24224	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
24204	24225	__ret; \
24205	24226	})
24206	24227	#endif
...	...	@@ -24209,16 +24230,16 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
24209	24230	#define vqshluq_n_s64(__p0, __p1) __extension__ ({ \
24210	24231	uint64x2_t __ret; \
24211	24232	int64x2_t __s0 = __p0; \
24212		__ret = (uint64x2_t) __builtin_neon_vqshluq_n_v((int8x16_t)__s0, __p1, 51); \
	24233	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vqshluq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 51)); \
24213	24234	__ret; \
24214	24235	})
24215	24236	#else
24216	24237	#define vqshluq_n_s64(__p0, __p1) __extension__ ({ \
24217	24238	uint64x2_t __ret; \
24218	24239	int64x2_t __s0 = __p0; \
24219		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
24220		__ret = (uint64x2_t) __builtin_neon_vqshluq_n_v((int8x16_t)__rev0, __p1, 51); \
24221		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	24240	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	24241	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vqshluq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 51)); \
	24242	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
24222	24243	__ret; \
24223	24244	})
24224	24245	#endif
...	...	@@ -24227,16 +24248,16 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
24227	24248	#define vqshluq_n_s16(__p0, __p1) __extension__ ({ \
24228	24249	uint16x8_t __ret; \
24229	24250	int16x8_t __s0 = __p0; \
24230		__ret = (uint16x8_t) __builtin_neon_vqshluq_n_v((int8x16_t)__s0, __p1, 49); \
	24251	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vqshluq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 49)); \
24231	24252	__ret; \
24232	24253	})
24233	24254	#else
24234	24255	#define vqshluq_n_s16(__p0, __p1) __extension__ ({ \
24235	24256	uint16x8_t __ret; \
24236	24257	int16x8_t __s0 = __p0; \
24237		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
24238		__ret = (uint16x8_t) __builtin_neon_vqshluq_n_v((int8x16_t)__rev0, __p1, 49); \
24239		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	24258	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	24259	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vqshluq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 49)); \
	24260	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
24240	24261	__ret; \
24241	24262	})
24242	24263	#endif
...	...	@@ -24245,16 +24266,16 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
24245	24266	#define vqshlu_n_s8(__p0, __p1) __extension__ ({ \
24246	24267	uint8x8_t __ret; \
24247	24268	int8x8_t __s0 = __p0; \
24248		__ret = (uint8x8_t) __builtin_neon_vqshlu_n_v((int8x8_t)__s0, __p1, 16); \
	24269	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqshlu_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 16)); \
24249	24270	__ret; \
24250	24271	})
24251	24272	#else
24252	24273	#define vqshlu_n_s8(__p0, __p1) __extension__ ({ \
24253	24274	uint8x8_t __ret; \
24254	24275	int8x8_t __s0 = __p0; \
24255		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
24256		__ret = (uint8x8_t) __builtin_neon_vqshlu_n_v((int8x8_t)__rev0, __p1, 16); \
24257		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	24276	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	24277	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqshlu_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 16)); \
	24278	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
24258	24279	__ret; \
24259	24280	})
24260	24281	#endif
...	...	@@ -24263,16 +24284,16 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
24263	24284	#define vqshlu_n_s32(__p0, __p1) __extension__ ({ \
24264	24285	uint32x2_t __ret; \
24265	24286	int32x2_t __s0 = __p0; \
24266		__ret = (uint32x2_t) __builtin_neon_vqshlu_n_v((int8x8_t)__s0, __p1, 18); \
	24287	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vqshlu_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 18)); \
24267	24288	__ret; \
24268	24289	})
24269	24290	#else
24270	24291	#define vqshlu_n_s32(__p0, __p1) __extension__ ({ \
24271	24292	uint32x2_t __ret; \
24272	24293	int32x2_t __s0 = __p0; \
24273		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
24274		__ret = (uint32x2_t) __builtin_neon_vqshlu_n_v((int8x8_t)__rev0, __p1, 18); \
24275		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	24294	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	24295	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vqshlu_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 18)); \
	24296	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
24276	24297	__ret; \
24277	24298	})
24278	24299	#endif
...	...	@@ -24280,23 +24301,23 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
24280	24301	#define vqshlu_n_s64(__p0, __p1) __extension__ ({ \
24281	24302	uint64x1_t __ret; \
24282	24303	int64x1_t __s0 = __p0; \
24283		__ret = (uint64x1_t) __builtin_neon_vqshlu_n_v((int8x8_t)__s0, __p1, 19); \
	24304	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vqshlu_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 19)); \
24284	24305	__ret; \
24285	24306	})
24286	24307	#ifdef __LITTLE_ENDIAN__
24287	24308	#define vqshlu_n_s16(__p0, __p1) __extension__ ({ \
24288	24309	uint16x4_t __ret; \
24289	24310	int16x4_t __s0 = __p0; \
24290		__ret = (uint16x4_t) __builtin_neon_vqshlu_n_v((int8x8_t)__s0, __p1, 17); \
	24311	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vqshlu_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 17)); \
24291	24312	__ret; \
24292	24313	})
24293	24314	#else
24294	24315	#define vqshlu_n_s16(__p0, __p1) __extension__ ({ \
24295	24316	uint16x4_t __ret; \
24296	24317	int16x4_t __s0 = __p0; \
24297		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
24298		__ret = (uint16x4_t) __builtin_neon_vqshlu_n_v((int8x8_t)__rev0, __p1, 17); \
24299		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	24318	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	24319	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vqshlu_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 17)); \
	24320	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
24300	24321	__ret; \
24301	24322	})
24302	24323	#endif
...	...	@@ -24305,22 +24326,22 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
24305	24326	#define vqshrn_n_u32(__p0, __p1) __extension__ ({ \
24306	24327	uint16x4_t __ret; \
24307	24328	uint32x4_t __s0 = __p0; \
24308		__ret = (uint16x4_t) __builtin_neon_vqshrn_n_v((int8x16_t)__s0, __p1, 17); \
	24329	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vqshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 17)); \
24309	24330	__ret; \
24310	24331	})
24311	24332	#else
24312	24333	#define vqshrn_n_u32(__p0, __p1) __extension__ ({ \
24313	24334	uint16x4_t __ret; \
24314	24335	uint32x4_t __s0 = __p0; \
24315		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
24316		__ret = (uint16x4_t) __builtin_neon_vqshrn_n_v((int8x16_t)__rev0, __p1, 17); \
24317		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	24336	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	24337	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vqshrn_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 17)); \
	24338	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
24318	24339	__ret; \
24319	24340	})
24320	24341	#define __noswap_vqshrn_n_u32(__p0, __p1) __extension__ ({ \
24321	24342	uint16x4_t __ret; \
24322	24343	uint32x4_t __s0 = __p0; \
24323		__ret = (uint16x4_t) __builtin_neon_vqshrn_n_v((int8x16_t)__s0, __p1, 17); \
	24344	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vqshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 17)); \
24324	24345	__ret; \
24325	24346	})
24326	24347	#endif
...	...	@@ -24329,22 +24350,22 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
24329	24350	#define vqshrn_n_u64(__p0, __p1) __extension__ ({ \
24330	24351	uint32x2_t __ret; \
24331	24352	uint64x2_t __s0 = __p0; \
24332		__ret = (uint32x2_t) __builtin_neon_vqshrn_n_v((int8x16_t)__s0, __p1, 18); \
	24353	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vqshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 18)); \
24333	24354	__ret; \
24334	24355	})
24335	24356	#else
24336	24357	#define vqshrn_n_u64(__p0, __p1) __extension__ ({ \
24337	24358	uint32x2_t __ret; \
24338	24359	uint64x2_t __s0 = __p0; \
24339		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
24340		__ret = (uint32x2_t) __builtin_neon_vqshrn_n_v((int8x16_t)__rev0, __p1, 18); \
24341		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	24360	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	24361	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vqshrn_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 18)); \
	24362	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
24342	24363	__ret; \
24343	24364	})
24344	24365	#define __noswap_vqshrn_n_u64(__p0, __p1) __extension__ ({ \
24345	24366	uint32x2_t __ret; \
24346	24367	uint64x2_t __s0 = __p0; \
24347		__ret = (uint32x2_t) __builtin_neon_vqshrn_n_v((int8x16_t)__s0, __p1, 18); \
	24368	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vqshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 18)); \
24348	24369	__ret; \
24349	24370	})
24350	24371	#endif
...	...	@@ -24353,22 +24374,22 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
24353	24374	#define vqshrn_n_u16(__p0, __p1) __extension__ ({ \
24354	24375	uint8x8_t __ret; \
24355	24376	uint16x8_t __s0 = __p0; \
24356		__ret = (uint8x8_t) __builtin_neon_vqshrn_n_v((int8x16_t)__s0, __p1, 16); \
	24377	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 16)); \
24357	24378	__ret; \
24358	24379	})
24359	24380	#else
24360	24381	#define vqshrn_n_u16(__p0, __p1) __extension__ ({ \
24361	24382	uint8x8_t __ret; \
24362	24383	uint16x8_t __s0 = __p0; \
24363		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
24364		__ret = (uint8x8_t) __builtin_neon_vqshrn_n_v((int8x16_t)__rev0, __p1, 16); \
24365		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	24384	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	24385	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqshrn_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 16)); \
	24386	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
24366	24387	__ret; \
24367	24388	})
24368	24389	#define __noswap_vqshrn_n_u16(__p0, __p1) __extension__ ({ \
24369	24390	uint8x8_t __ret; \
24370	24391	uint16x8_t __s0 = __p0; \
24371		__ret = (uint8x8_t) __builtin_neon_vqshrn_n_v((int8x16_t)__s0, __p1, 16); \
	24392	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 16)); \
24372	24393	__ret; \
24373	24394	})
24374	24395	#endif
...	...	@@ -24377,22 +24398,22 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
24377	24398	#define vqshrn_n_s32(__p0, __p1) __extension__ ({ \
24378	24399	int16x4_t __ret; \
24379	24400	int32x4_t __s0 = __p0; \
24380		__ret = (int16x4_t) __builtin_neon_vqshrn_n_v((int8x16_t)__s0, __p1, 1); \
	24401	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 1)); \
24381	24402	__ret; \
24382	24403	})
24383	24404	#else
24384	24405	#define vqshrn_n_s32(__p0, __p1) __extension__ ({ \
24385	24406	int16x4_t __ret; \
24386	24407	int32x4_t __s0 = __p0; \
24387		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
24388		__ret = (int16x4_t) __builtin_neon_vqshrn_n_v((int8x16_t)__rev0, __p1, 1); \
24389		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	24408	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	24409	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqshrn_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 1)); \
	24410	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
24390	24411	__ret; \
24391	24412	})
24392	24413	#define __noswap_vqshrn_n_s32(__p0, __p1) __extension__ ({ \
24393	24414	int16x4_t __ret; \
24394	24415	int32x4_t __s0 = __p0; \
24395		__ret = (int16x4_t) __builtin_neon_vqshrn_n_v((int8x16_t)__s0, __p1, 1); \
	24416	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 1)); \
24396	24417	__ret; \
24397	24418	})
24398	24419	#endif
...	...	@@ -24401,22 +24422,22 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
24401	24422	#define vqshrn_n_s64(__p0, __p1) __extension__ ({ \
24402	24423	int32x2_t __ret; \
24403	24424	int64x2_t __s0 = __p0; \
24404		__ret = (int32x2_t) __builtin_neon_vqshrn_n_v((int8x16_t)__s0, __p1, 2); \
	24425	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 2)); \
24405	24426	__ret; \
24406	24427	})
24407	24428	#else
24408	24429	#define vqshrn_n_s64(__p0, __p1) __extension__ ({ \
24409	24430	int32x2_t __ret; \
24410	24431	int64x2_t __s0 = __p0; \
24411		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
24412		__ret = (int32x2_t) __builtin_neon_vqshrn_n_v((int8x16_t)__rev0, __p1, 2); \
24413		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	24432	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	24433	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqshrn_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 2)); \
	24434	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
24414	24435	__ret; \
24415	24436	})
24416	24437	#define __noswap_vqshrn_n_s64(__p0, __p1) __extension__ ({ \
24417	24438	int32x2_t __ret; \
24418	24439	int64x2_t __s0 = __p0; \
24419		__ret = (int32x2_t) __builtin_neon_vqshrn_n_v((int8x16_t)__s0, __p1, 2); \
	24440	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 2)); \
24420	24441	__ret; \
24421	24442	})
24422	24443	#endif
...	...	@@ -24425,22 +24446,22 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
24425	24446	#define vqshrn_n_s16(__p0, __p1) __extension__ ({ \
24426	24447	int8x8_t __ret; \
24427	24448	int16x8_t __s0 = __p0; \
24428		__ret = (int8x8_t) __builtin_neon_vqshrn_n_v((int8x16_t)__s0, __p1, 0); \
	24449	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 0)); \
24429	24450	__ret; \
24430	24451	})
24431	24452	#else
24432	24453	#define vqshrn_n_s16(__p0, __p1) __extension__ ({ \
24433	24454	int8x8_t __ret; \
24434	24455	int16x8_t __s0 = __p0; \
24435		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
24436		__ret = (int8x8_t) __builtin_neon_vqshrn_n_v((int8x16_t)__rev0, __p1, 0); \
24437		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	24456	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	24457	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqshrn_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 0)); \
	24458	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
24438	24459	__ret; \
24439	24460	})
24440	24461	#define __noswap_vqshrn_n_s16(__p0, __p1) __extension__ ({ \
24441	24462	int8x8_t __ret; \
24442	24463	int16x8_t __s0 = __p0; \
24443		__ret = (int8x8_t) __builtin_neon_vqshrn_n_v((int8x16_t)__s0, __p1, 0); \
	24464	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 0)); \
24444	24465	__ret; \
24445	24466	})
24446	24467	#endif
...	...	@@ -24449,22 +24470,22 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
24449	24470	#define vqshrun_n_s32(__p0, __p1) __extension__ ({ \
24450	24471	uint16x4_t __ret; \
24451	24472	int32x4_t __s0 = __p0; \
24452		__ret = (uint16x4_t) __builtin_neon_vqshrun_n_v((int8x16_t)__s0, __p1, 17); \
	24473	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vqshrun_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 17)); \
24453	24474	__ret; \
24454	24475	})
24455	24476	#else
24456	24477	#define vqshrun_n_s32(__p0, __p1) __extension__ ({ \
24457	24478	uint16x4_t __ret; \
24458	24479	int32x4_t __s0 = __p0; \
24459		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
24460		__ret = (uint16x4_t) __builtin_neon_vqshrun_n_v((int8x16_t)__rev0, __p1, 17); \
24461		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	24480	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	24481	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vqshrun_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 17)); \
	24482	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
24462	24483	__ret; \
24463	24484	})
24464	24485	#define __noswap_vqshrun_n_s32(__p0, __p1) __extension__ ({ \
24465	24486	uint16x4_t __ret; \
24466	24487	int32x4_t __s0 = __p0; \
24467		__ret = (uint16x4_t) __builtin_neon_vqshrun_n_v((int8x16_t)__s0, __p1, 17); \
	24488	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vqshrun_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 17)); \
24468	24489	__ret; \
24469	24490	})
24470	24491	#endif
...	...	@@ -24473,22 +24494,22 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
24473	24494	#define vqshrun_n_s64(__p0, __p1) __extension__ ({ \
24474	24495	uint32x2_t __ret; \
24475	24496	int64x2_t __s0 = __p0; \
24476		__ret = (uint32x2_t) __builtin_neon_vqshrun_n_v((int8x16_t)__s0, __p1, 18); \
	24497	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vqshrun_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 18)); \
24477	24498	__ret; \
24478	24499	})
24479	24500	#else
24480	24501	#define vqshrun_n_s64(__p0, __p1) __extension__ ({ \
24481	24502	uint32x2_t __ret; \
24482	24503	int64x2_t __s0 = __p0; \
24483		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
24484		__ret = (uint32x2_t) __builtin_neon_vqshrun_n_v((int8x16_t)__rev0, __p1, 18); \
24485		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	24504	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	24505	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vqshrun_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 18)); \
	24506	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
24486	24507	__ret; \
24487	24508	})
24488	24509	#define __noswap_vqshrun_n_s64(__p0, __p1) __extension__ ({ \
24489	24510	uint32x2_t __ret; \
24490	24511	int64x2_t __s0 = __p0; \
24491		__ret = (uint32x2_t) __builtin_neon_vqshrun_n_v((int8x16_t)__s0, __p1, 18); \
	24512	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vqshrun_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 18)); \
24492	24513	__ret; \
24493	24514	})
24494	24515	#endif
...	...	@@ -24497,22 +24518,22 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
24497	24518	#define vqshrun_n_s16(__p0, __p1) __extension__ ({ \
24498	24519	uint8x8_t __ret; \
24499	24520	int16x8_t __s0 = __p0; \
24500		__ret = (uint8x8_t) __builtin_neon_vqshrun_n_v((int8x16_t)__s0, __p1, 16); \
	24521	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqshrun_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 16)); \
24501	24522	__ret; \
24502	24523	})
24503	24524	#else
24504	24525	#define vqshrun_n_s16(__p0, __p1) __extension__ ({ \
24505	24526	uint8x8_t __ret; \
24506	24527	int16x8_t __s0 = __p0; \
24507		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
24508		__ret = (uint8x8_t) __builtin_neon_vqshrun_n_v((int8x16_t)__rev0, __p1, 16); \
24509		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	24528	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	24529	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqshrun_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 16)); \
	24530	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
24510	24531	__ret; \
24511	24532	})
24512	24533	#define __noswap_vqshrun_n_s16(__p0, __p1) __extension__ ({ \
24513	24534	uint8x8_t __ret; \
24514	24535	int16x8_t __s0 = __p0; \
24515		__ret = (uint8x8_t) __builtin_neon_vqshrun_n_v((int8x16_t)__s0, __p1, 16); \
	24536	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqshrun_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 16)); \
24516	24537	__ret; \
24517	24538	})
24518	24539	#endif
...	...	@@ -24520,16 +24541,16 @@ __ai __attribute__((target("neon"))) int16x4_t vqshl_s16(int16x4_t __p0, int16x4
24520	24541	#ifdef __LITTLE_ENDIAN__
24521	24542	__ai __attribute__((target("neon"))) uint8x16_t vqsubq_u8(uint8x16_t __p0, uint8x16_t __p1) {
24522	24543	uint8x16_t __ret;
24523		__ret = (uint8x16_t) __builtin_neon_vqsubq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
	24544	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vqsubq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 48));
24524	24545	return __ret;
24525	24546	}
24526	24547	#else
24527	24548	__ai __attribute__((target("neon"))) uint8x16_t vqsubq_u8(uint8x16_t __p0, uint8x16_t __p1) {
24528	24549	uint8x16_t __ret;
24529		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
24530		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
24531		__ret = (uint8x16_t) __builtin_neon_vqsubq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
24532		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	24550	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	24551	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	24552	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vqsubq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 48));
	24553	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
24533	24554	return __ret;
24534	24555	}
24535	24556	#endif
...	...	@@ -24537,16 +24558,16 @@ __ai __attribute__((target("neon"))) uint8x16_t vqsubq_u8(uint8x16_t __p0, uint8
24537	24558	#ifdef __LITTLE_ENDIAN__
24538	24559	__ai __attribute__((target("neon"))) uint32x4_t vqsubq_u32(uint32x4_t __p0, uint32x4_t __p1) {
24539	24560	uint32x4_t __ret;
24540		__ret = (uint32x4_t) __builtin_neon_vqsubq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
	24561	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vqsubq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50));
24541	24562	return __ret;
24542	24563	}
24543	24564	#else
24544	24565	__ai __attribute__((target("neon"))) uint32x4_t vqsubq_u32(uint32x4_t __p0, uint32x4_t __p1) {
24545	24566	uint32x4_t __ret;
24546		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
24547		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
24548		__ret = (uint32x4_t) __builtin_neon_vqsubq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
24549		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	24567	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	24568	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	24569	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vqsubq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 50));
	24570	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
24550	24571	return __ret;
24551	24572	}
24552	24573	#endif
...	...	@@ -24554,16 +24575,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vqsubq_u32(uint32x4_t __p0, uint
24554	24575	#ifdef __LITTLE_ENDIAN__
24555	24576	__ai __attribute__((target("neon"))) uint64x2_t vqsubq_u64(uint64x2_t __p0, uint64x2_t __p1) {
24556	24577	uint64x2_t __ret;
24557		__ret = (uint64x2_t) __builtin_neon_vqsubq_v((int8x16_t)__p0, (int8x16_t)__p1, 51);
	24578	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vqsubq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 51));
24558	24579	return __ret;
24559	24580	}
24560	24581	#else
24561	24582	__ai __attribute__((target("neon"))) uint64x2_t vqsubq_u64(uint64x2_t __p0, uint64x2_t __p1) {
24562	24583	uint64x2_t __ret;
24563		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
24564		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
24565		__ret = (uint64x2_t) __builtin_neon_vqsubq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 51);
24566		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	24584	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	24585	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	24586	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vqsubq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 51));
	24587	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
24567	24588	return __ret;
24568	24589	}
24569	24590	#endif
...	...	@@ -24571,16 +24592,16 @@ __ai __attribute__((target("neon"))) uint64x2_t vqsubq_u64(uint64x2_t __p0, uint
24571	24592	#ifdef __LITTLE_ENDIAN__
24572	24593	__ai __attribute__((target("neon"))) uint16x8_t vqsubq_u16(uint16x8_t __p0, uint16x8_t __p1) {
24573	24594	uint16x8_t __ret;
24574		__ret = (uint16x8_t) __builtin_neon_vqsubq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
	24595	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vqsubq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 49));
24575	24596	return __ret;
24576	24597	}
24577	24598	#else
24578	24599	__ai __attribute__((target("neon"))) uint16x8_t vqsubq_u16(uint16x8_t __p0, uint16x8_t __p1) {
24579	24600	uint16x8_t __ret;
24580		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
24581		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
24582		__ret = (uint16x8_t) __builtin_neon_vqsubq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
24583		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	24601	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	24602	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	24603	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vqsubq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 49));
	24604	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
24584	24605	return __ret;
24585	24606	}
24586	24607	#endif
...	...	@@ -24588,16 +24609,16 @@ __ai __attribute__((target("neon"))) uint16x8_t vqsubq_u16(uint16x8_t __p0, uint
24588	24609	#ifdef __LITTLE_ENDIAN__
24589	24610	__ai __attribute__((target("neon"))) int8x16_t vqsubq_s8(int8x16_t __p0, int8x16_t __p1) {
24590	24611	int8x16_t __ret;
24591		__ret = (int8x16_t) __builtin_neon_vqsubq_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
	24612	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vqsubq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 32));
24592	24613	return __ret;
24593	24614	}
24594	24615	#else
24595	24616	__ai __attribute__((target("neon"))) int8x16_t vqsubq_s8(int8x16_t __p0, int8x16_t __p1) {
24596	24617	int8x16_t __ret;
24597		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
24598		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
24599		__ret = (int8x16_t) __builtin_neon_vqsubq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 32);
24600		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	24618	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	24619	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	24620	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vqsubq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 32));
	24621	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
24601	24622	return __ret;
24602	24623	}
24603	24624	#endif
...	...	@@ -24605,16 +24626,16 @@ __ai __attribute__((target("neon"))) int8x16_t vqsubq_s8(int8x16_t __p0, int8x16
24605	24626	#ifdef __LITTLE_ENDIAN__
24606	24627	__ai __attribute__((target("neon"))) int32x4_t vqsubq_s32(int32x4_t __p0, int32x4_t __p1) {
24607	24628	int32x4_t __ret;
24608		__ret = (int32x4_t) __builtin_neon_vqsubq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
	24629	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqsubq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 34));
24609	24630	return __ret;
24610	24631	}
24611	24632	#else
24612	24633	__ai __attribute__((target("neon"))) int32x4_t vqsubq_s32(int32x4_t __p0, int32x4_t __p1) {
24613	24634	int32x4_t __ret;
24614		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
24615		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
24616		__ret = (int32x4_t) __builtin_neon_vqsubq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
24617		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	24635	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	24636	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	24637	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqsubq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 34));
	24638	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
24618	24639	return __ret;
24619	24640	}
24620	24641	#endif
...	...	@@ -24622,16 +24643,16 @@ __ai __attribute__((target("neon"))) int32x4_t vqsubq_s32(int32x4_t __p0, int32x
24622	24643	#ifdef __LITTLE_ENDIAN__
24623	24644	__ai __attribute__((target("neon"))) int64x2_t vqsubq_s64(int64x2_t __p0, int64x2_t __p1) {
24624	24645	int64x2_t __ret;
24625		__ret = (int64x2_t) __builtin_neon_vqsubq_v((int8x16_t)__p0, (int8x16_t)__p1, 35);
	24646	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vqsubq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 35));
24626	24647	return __ret;
24627	24648	}
24628	24649	#else
24629	24650	__ai __attribute__((target("neon"))) int64x2_t vqsubq_s64(int64x2_t __p0, int64x2_t __p1) {
24630	24651	int64x2_t __ret;
24631		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
24632		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
24633		__ret = (int64x2_t) __builtin_neon_vqsubq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 35);
24634		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	24652	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	24653	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	24654	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vqsubq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 35));
	24655	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
24635	24656	return __ret;
24636	24657	}
24637	24658	#endif
...	...	@@ -24639,16 +24660,16 @@ __ai __attribute__((target("neon"))) int64x2_t vqsubq_s64(int64x2_t __p0, int64x
24639	24660	#ifdef __LITTLE_ENDIAN__
24640	24661	__ai __attribute__((target("neon"))) int16x8_t vqsubq_s16(int16x8_t __p0, int16x8_t __p1) {
24641	24662	int16x8_t __ret;
24642		__ret = (int16x8_t) __builtin_neon_vqsubq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
	24663	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqsubq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 33));
24643	24664	return __ret;
24644	24665	}
24645	24666	#else
24646	24667	__ai __attribute__((target("neon"))) int16x8_t vqsubq_s16(int16x8_t __p0, int16x8_t __p1) {
24647	24668	int16x8_t __ret;
24648		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
24649		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
24650		__ret = (int16x8_t) __builtin_neon_vqsubq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
24651		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	24669	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	24670	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	24671	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqsubq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 33));
	24672	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
24652	24673	return __ret;
24653	24674	}
24654	24675	#endif
...	...	@@ -24656,16 +24677,16 @@ __ai __attribute__((target("neon"))) int16x8_t vqsubq_s16(int16x8_t __p0, int16x
24656	24677	#ifdef __LITTLE_ENDIAN__
24657	24678	__ai __attribute__((target("neon"))) uint8x8_t vqsub_u8(uint8x8_t __p0, uint8x8_t __p1) {
24658	24679	uint8x8_t __ret;
24659		__ret = (uint8x8_t) __builtin_neon_vqsub_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
	24680	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqsub_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 16));
24660	24681	return __ret;
24661	24682	}
24662	24683	#else
24663	24684	__ai __attribute__((target("neon"))) uint8x8_t vqsub_u8(uint8x8_t __p0, uint8x8_t __p1) {
24664	24685	uint8x8_t __ret;
24665		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
24666		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
24667		__ret = (uint8x8_t) __builtin_neon_vqsub_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
24668		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	24686	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	24687	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	24688	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqsub_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 16));
	24689	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
24669	24690	return __ret;
24670	24691	}
24671	24692	#endif
...	...	@@ -24673,38 +24694,38 @@ __ai __attribute__((target("neon"))) uint8x8_t vqsub_u8(uint8x8_t __p0, uint8x8_
24673	24694	#ifdef __LITTLE_ENDIAN__
24674	24695	__ai __attribute__((target("neon"))) uint32x2_t vqsub_u32(uint32x2_t __p0, uint32x2_t __p1) {
24675	24696	uint32x2_t __ret;
24676		__ret = (uint32x2_t) __builtin_neon_vqsub_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
	24697	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vqsub_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 18));
24677	24698	return __ret;
24678	24699	}
24679	24700	#else
24680	24701	__ai __attribute__((target("neon"))) uint32x2_t vqsub_u32(uint32x2_t __p0, uint32x2_t __p1) {
24681	24702	uint32x2_t __ret;
24682		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
24683		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
24684		__ret = (uint32x2_t) __builtin_neon_vqsub_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
24685		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	24703	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	24704	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	24705	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vqsub_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 18));
	24706	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
24686	24707	return __ret;
24687	24708	}
24688	24709	#endif
24689	24710
24690	24711	__ai __attribute__((target("neon"))) uint64x1_t vqsub_u64(uint64x1_t __p0, uint64x1_t __p1) {
24691	24712	uint64x1_t __ret;
24692		__ret = (uint64x1_t) __builtin_neon_vqsub_v((int8x8_t)__p0, (int8x8_t)__p1, 19);
	24713	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vqsub_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 19));
24693	24714	return __ret;
24694	24715	}
24695	24716	#ifdef __LITTLE_ENDIAN__
24696	24717	__ai __attribute__((target("neon"))) uint16x4_t vqsub_u16(uint16x4_t __p0, uint16x4_t __p1) {
24697	24718	uint16x4_t __ret;
24698		__ret = (uint16x4_t) __builtin_neon_vqsub_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
	24719	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vqsub_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 17));
24699	24720	return __ret;
24700	24721	}
24701	24722	#else
24702	24723	__ai __attribute__((target("neon"))) uint16x4_t vqsub_u16(uint16x4_t __p0, uint16x4_t __p1) {
24703	24724	uint16x4_t __ret;
24704		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
24705		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
24706		__ret = (uint16x4_t) __builtin_neon_vqsub_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
24707		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	24725	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	24726	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	24727	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vqsub_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 17));
	24728	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
24708	24729	return __ret;
24709	24730	}
24710	24731	#endif
...	...	@@ -24712,16 +24733,16 @@ __ai __attribute__((target("neon"))) uint16x4_t vqsub_u16(uint16x4_t __p0, uint1
24712	24733	#ifdef __LITTLE_ENDIAN__
24713	24734	__ai __attribute__((target("neon"))) int8x8_t vqsub_s8(int8x8_t __p0, int8x8_t __p1) {
24714	24735	int8x8_t __ret;
24715		__ret = (int8x8_t) __builtin_neon_vqsub_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
	24736	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqsub_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 0));
24716	24737	return __ret;
24717	24738	}
24718	24739	#else
24719	24740	__ai __attribute__((target("neon"))) int8x8_t vqsub_s8(int8x8_t __p0, int8x8_t __p1) {
24720	24741	int8x8_t __ret;
24721		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
24722		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
24723		__ret = (int8x8_t) __builtin_neon_vqsub_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
24724		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	24742	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	24743	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	24744	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqsub_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 0));
	24745	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
24725	24746	return __ret;
24726	24747	}
24727	24748	#endif
...	...	@@ -24729,38 +24750,38 @@ __ai __attribute__((target("neon"))) int8x8_t vqsub_s8(int8x8_t __p0, int8x8_t _
24729	24750	#ifdef __LITTLE_ENDIAN__
24730	24751	__ai __attribute__((target("neon"))) int32x2_t vqsub_s32(int32x2_t __p0, int32x2_t __p1) {
24731	24752	int32x2_t __ret;
24732		__ret = (int32x2_t) __builtin_neon_vqsub_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
	24753	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqsub_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 2));
24733	24754	return __ret;
24734	24755	}
24735	24756	#else
24736	24757	__ai __attribute__((target("neon"))) int32x2_t vqsub_s32(int32x2_t __p0, int32x2_t __p1) {
24737	24758	int32x2_t __ret;
24738		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
24739		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
24740		__ret = (int32x2_t) __builtin_neon_vqsub_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
24741		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	24759	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	24760	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	24761	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqsub_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 2));
	24762	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
24742	24763	return __ret;
24743	24764	}
24744	24765	#endif
24745	24766
24746	24767	__ai __attribute__((target("neon"))) int64x1_t vqsub_s64(int64x1_t __p0, int64x1_t __p1) {
24747	24768	int64x1_t __ret;
24748		__ret = (int64x1_t) __builtin_neon_vqsub_v((int8x8_t)__p0, (int8x8_t)__p1, 3);
	24769	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vqsub_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 3));
24749	24770	return __ret;
24750	24771	}
24751	24772	#ifdef __LITTLE_ENDIAN__
24752	24773	__ai __attribute__((target("neon"))) int16x4_t vqsub_s16(int16x4_t __p0, int16x4_t __p1) {
24753	24774	int16x4_t __ret;
24754		__ret = (int16x4_t) __builtin_neon_vqsub_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
	24775	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqsub_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 1));
24755	24776	return __ret;
24756	24777	}
24757	24778	#else
24758	24779	__ai __attribute__((target("neon"))) int16x4_t vqsub_s16(int16x4_t __p0, int16x4_t __p1) {
24759	24780	int16x4_t __ret;
24760		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
24761		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
24762		__ret = (int16x4_t) __builtin_neon_vqsub_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
24763		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	24781	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	24782	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	24783	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqsub_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 1));
	24784	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
24764	24785	return __ret;
24765	24786	}
24766	24787	#endif
...	...	@@ -24768,21 +24789,21 @@ __ai __attribute__((target("neon"))) int16x4_t vqsub_s16(int16x4_t __p0, int16x4
24768	24789	#ifdef __LITTLE_ENDIAN__
24769	24790	__ai __attribute__((target("neon"))) uint16x4_t vraddhn_u32(uint32x4_t __p0, uint32x4_t __p1) {
24770	24791	uint16x4_t __ret;
24771		__ret = (uint16x4_t) __builtin_neon_vraddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 17);
	24792	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vraddhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 17));
24772	24793	return __ret;
24773	24794	}
24774	24795	#else
24775	24796	__ai __attribute__((target("neon"))) uint16x4_t vraddhn_u32(uint32x4_t __p0, uint32x4_t __p1) {
24776	24797	uint16x4_t __ret;
24777		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
24778		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
24779		__ret = (uint16x4_t) __builtin_neon_vraddhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 17);
24780		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	24798	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	24799	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	24800	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vraddhn_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 17));
	24801	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
24781	24802	return __ret;
24782	24803	}
24783	24804	__ai __attribute__((target("neon"))) uint16x4_t __noswap_vraddhn_u32(uint32x4_t __p0, uint32x4_t __p1) {
24784	24805	uint16x4_t __ret;
24785		__ret = (uint16x4_t) __builtin_neon_vraddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 17);
	24806	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vraddhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 17));
24786	24807	return __ret;
24787	24808	}
24788	24809	#endif
...	...	@@ -24790,21 +24811,21 @@ __ai __attribute__((target("neon"))) uint16x4_t __noswap_vraddhn_u32(uint32x4_t
24790	24811	#ifdef __LITTLE_ENDIAN__
24791	24812	__ai __attribute__((target("neon"))) uint32x2_t vraddhn_u64(uint64x2_t __p0, uint64x2_t __p1) {
24792	24813	uint32x2_t __ret;
24793		__ret = (uint32x2_t) __builtin_neon_vraddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 18);
	24814	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vraddhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 18));
24794	24815	return __ret;
24795	24816	}
24796	24817	#else
24797	24818	__ai __attribute__((target("neon"))) uint32x2_t vraddhn_u64(uint64x2_t __p0, uint64x2_t __p1) {
24798	24819	uint32x2_t __ret;
24799		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
24800		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
24801		__ret = (uint32x2_t) __builtin_neon_vraddhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 18);
24802		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	24820	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	24821	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	24822	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vraddhn_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 18));
	24823	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
24803	24824	return __ret;
24804	24825	}
24805	24826	__ai __attribute__((target("neon"))) uint32x2_t __noswap_vraddhn_u64(uint64x2_t __p0, uint64x2_t __p1) {
24806	24827	uint32x2_t __ret;
24807		__ret = (uint32x2_t) __builtin_neon_vraddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 18);
	24828	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vraddhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 18));
24808	24829	return __ret;
24809	24830	}
24810	24831	#endif
...	...	@@ -24812,21 +24833,21 @@ __ai __attribute__((target("neon"))) uint32x2_t __noswap_vraddhn_u64(uint64x2_t
24812	24833	#ifdef __LITTLE_ENDIAN__
24813	24834	__ai __attribute__((target("neon"))) uint8x8_t vraddhn_u16(uint16x8_t __p0, uint16x8_t __p1) {
24814	24835	uint8x8_t __ret;
24815		__ret = (uint8x8_t) __builtin_neon_vraddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 16);
	24836	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vraddhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 16));
24816	24837	return __ret;
24817	24838	}
24818	24839	#else
24819	24840	__ai __attribute__((target("neon"))) uint8x8_t vraddhn_u16(uint16x8_t __p0, uint16x8_t __p1) {
24820	24841	uint8x8_t __ret;
24821		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
24822		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
24823		__ret = (uint8x8_t) __builtin_neon_vraddhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 16);
24824		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	24842	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	24843	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	24844	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vraddhn_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 16));
	24845	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
24825	24846	return __ret;
24826	24847	}
24827	24848	__ai __attribute__((target("neon"))) uint8x8_t __noswap_vraddhn_u16(uint16x8_t __p0, uint16x8_t __p1) {
24828	24849	uint8x8_t __ret;
24829		__ret = (uint8x8_t) __builtin_neon_vraddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 16);
	24850	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vraddhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 16));
24830	24851	return __ret;
24831	24852	}
24832	24853	#endif
...	...	@@ -24834,21 +24855,21 @@ __ai __attribute__((target("neon"))) uint8x8_t __noswap_vraddhn_u16(uint16x8_t _
24834	24855	#ifdef __LITTLE_ENDIAN__
24835	24856	__ai __attribute__((target("neon"))) int16x4_t vraddhn_s32(int32x4_t __p0, int32x4_t __p1) {
24836	24857	int16x4_t __ret;
24837		__ret = (int16x4_t) __builtin_neon_vraddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 1);
	24858	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vraddhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 1));
24838	24859	return __ret;
24839	24860	}
24840	24861	#else
24841	24862	__ai __attribute__((target("neon"))) int16x4_t vraddhn_s32(int32x4_t __p0, int32x4_t __p1) {
24842	24863	int16x4_t __ret;
24843		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
24844		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
24845		__ret = (int16x4_t) __builtin_neon_vraddhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 1);
24846		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	24864	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	24865	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	24866	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vraddhn_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 1));
	24867	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
24847	24868	return __ret;
24848	24869	}
24849	24870	__ai __attribute__((target("neon"))) int16x4_t __noswap_vraddhn_s32(int32x4_t __p0, int32x4_t __p1) {
24850	24871	int16x4_t __ret;
24851		__ret = (int16x4_t) __builtin_neon_vraddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 1);
	24872	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vraddhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 1));
24852	24873	return __ret;
24853	24874	}
24854	24875	#endif
...	...	@@ -24856,21 +24877,21 @@ __ai __attribute__((target("neon"))) int16x4_t __noswap_vraddhn_s32(int32x4_t __
24856	24877	#ifdef __LITTLE_ENDIAN__
24857	24878	__ai __attribute__((target("neon"))) int32x2_t vraddhn_s64(int64x2_t __p0, int64x2_t __p1) {
24858	24879	int32x2_t __ret;
24859		__ret = (int32x2_t) __builtin_neon_vraddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 2);
	24880	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vraddhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 2));
24860	24881	return __ret;
24861	24882	}
24862	24883	#else
24863	24884	__ai __attribute__((target("neon"))) int32x2_t vraddhn_s64(int64x2_t __p0, int64x2_t __p1) {
24864	24885	int32x2_t __ret;
24865		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
24866		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
24867		__ret = (int32x2_t) __builtin_neon_vraddhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 2);
24868		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	24886	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	24887	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	24888	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vraddhn_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 2));
	24889	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
24869	24890	return __ret;
24870	24891	}
24871	24892	__ai __attribute__((target("neon"))) int32x2_t __noswap_vraddhn_s64(int64x2_t __p0, int64x2_t __p1) {
24872	24893	int32x2_t __ret;
24873		__ret = (int32x2_t) __builtin_neon_vraddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 2);
	24894	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vraddhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 2));
24874	24895	return __ret;
24875	24896	}
24876	24897	#endif
...	...	@@ -24878,21 +24899,21 @@ __ai __attribute__((target("neon"))) int32x2_t __noswap_vraddhn_s64(int64x2_t __
24878	24899	#ifdef __LITTLE_ENDIAN__
24879	24900	__ai __attribute__((target("neon"))) int8x8_t vraddhn_s16(int16x8_t __p0, int16x8_t __p1) {
24880	24901	int8x8_t __ret;
24881		__ret = (int8x8_t) __builtin_neon_vraddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 0);
	24902	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vraddhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 0));
24882	24903	return __ret;
24883	24904	}
24884	24905	#else
24885	24906	__ai __attribute__((target("neon"))) int8x8_t vraddhn_s16(int16x8_t __p0, int16x8_t __p1) {
24886	24907	int8x8_t __ret;
24887		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
24888		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
24889		__ret = (int8x8_t) __builtin_neon_vraddhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 0);
24890		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	24908	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	24909	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	24910	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vraddhn_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 0));
	24911	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
24891	24912	return __ret;
24892	24913	}
24893	24914	__ai __attribute__((target("neon"))) int8x8_t __noswap_vraddhn_s16(int16x8_t __p0, int16x8_t __p1) {
24894	24915	int8x8_t __ret;
24895		__ret = (int8x8_t) __builtin_neon_vraddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 0);
	24916	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vraddhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 0));
24896	24917	return __ret;
24897	24918	}
24898	24919	#endif
...	...	@@ -24900,15 +24921,15 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vraddhn_s16(int16x8_t __p
24900	24921	#ifdef __LITTLE_ENDIAN__
24901	24922	__ai __attribute__((target("neon"))) uint32x4_t vrecpeq_u32(uint32x4_t __p0) {
24902	24923	uint32x4_t __ret;
24903		__ret = (uint32x4_t) __builtin_neon_vrecpeq_v((int8x16_t)__p0, 50);
	24924	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vrecpeq_v(__builtin_bit_cast(int8x16_t, __p0), 50));
24904	24925	return __ret;
24905	24926	}
24906	24927	#else
24907	24928	__ai __attribute__((target("neon"))) uint32x4_t vrecpeq_u32(uint32x4_t __p0) {
24908	24929	uint32x4_t __ret;
24909		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
24910		__ret = (uint32x4_t) __builtin_neon_vrecpeq_v((int8x16_t)__rev0, 50);
24911		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	24930	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	24931	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vrecpeq_v(__builtin_bit_cast(int8x16_t, __rev0), 50));
	24932	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
24912	24933	return __ret;
24913	24934	}
24914	24935	#endif
...	...	@@ -24916,15 +24937,15 @@ __ai __attribute__((target("neon"))) uint32x4_t vrecpeq_u32(uint32x4_t __p0) {
24916	24937	#ifdef __LITTLE_ENDIAN__
24917	24938	__ai __attribute__((target("neon"))) float32x4_t vrecpeq_f32(float32x4_t __p0) {
24918	24939	float32x4_t __ret;
24919		__ret = (float32x4_t) __builtin_neon_vrecpeq_v((int8x16_t)__p0, 41);
	24940	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vrecpeq_v(__builtin_bit_cast(int8x16_t, __p0), 41));
24920	24941	return __ret;
24921	24942	}
24922	24943	#else
24923	24944	__ai __attribute__((target("neon"))) float32x4_t vrecpeq_f32(float32x4_t __p0) {
24924	24945	float32x4_t __ret;
24925		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
24926		__ret = (float32x4_t) __builtin_neon_vrecpeq_v((int8x16_t)__rev0, 41);
24927		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	24946	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	24947	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vrecpeq_v(__builtin_bit_cast(int8x16_t, __rev0), 41));
	24948	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
24928	24949	return __ret;
24929	24950	}
24930	24951	#endif
...	...	@@ -24932,15 +24953,15 @@ __ai __attribute__((target("neon"))) float32x4_t vrecpeq_f32(float32x4_t __p0) {
24932	24953	#ifdef __LITTLE_ENDIAN__
24933	24954	__ai __attribute__((target("neon"))) uint32x2_t vrecpe_u32(uint32x2_t __p0) {
24934	24955	uint32x2_t __ret;
24935		__ret = (uint32x2_t) __builtin_neon_vrecpe_v((int8x8_t)__p0, 18);
	24956	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vrecpe_v(__builtin_bit_cast(int8x8_t, __p0), 18));
24936	24957	return __ret;
24937	24958	}
24938	24959	#else
24939	24960	__ai __attribute__((target("neon"))) uint32x2_t vrecpe_u32(uint32x2_t __p0) {
24940	24961	uint32x2_t __ret;
24941		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
24942		__ret = (uint32x2_t) __builtin_neon_vrecpe_v((int8x8_t)__rev0, 18);
24943		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	24962	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	24963	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vrecpe_v(__builtin_bit_cast(int8x8_t, __rev0), 18));
	24964	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
24944	24965	return __ret;
24945	24966	}
24946	24967	#endif
...	...	@@ -24948,15 +24969,15 @@ __ai __attribute__((target("neon"))) uint32x2_t vrecpe_u32(uint32x2_t __p0) {
24948	24969	#ifdef __LITTLE_ENDIAN__
24949	24970	__ai __attribute__((target("neon"))) float32x2_t vrecpe_f32(float32x2_t __p0) {
24950	24971	float32x2_t __ret;
24951		__ret = (float32x2_t) __builtin_neon_vrecpe_v((int8x8_t)__p0, 9);
	24972	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vrecpe_v(__builtin_bit_cast(int8x8_t, __p0), 9));
24952	24973	return __ret;
24953	24974	}
24954	24975	#else
24955	24976	__ai __attribute__((target("neon"))) float32x2_t vrecpe_f32(float32x2_t __p0) {
24956	24977	float32x2_t __ret;
24957		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
24958		__ret = (float32x2_t) __builtin_neon_vrecpe_v((int8x8_t)__rev0, 9);
24959		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	24978	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	24979	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vrecpe_v(__builtin_bit_cast(int8x8_t, __rev0), 9));
	24980	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
24960	24981	return __ret;
24961	24982	}
24962	24983	#endif
...	...	@@ -24964,16 +24985,16 @@ __ai __attribute__((target("neon"))) float32x2_t vrecpe_f32(float32x2_t __p0) {
24964	24985	#ifdef __LITTLE_ENDIAN__
24965	24986	__ai __attribute__((target("neon"))) float32x4_t vrecpsq_f32(float32x4_t __p0, float32x4_t __p1) {
24966	24987	float32x4_t __ret;
24967		__ret = (float32x4_t) __builtin_neon_vrecpsq_v((int8x16_t)__p0, (int8x16_t)__p1, 41);
	24988	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vrecpsq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 41));
24968	24989	return __ret;
24969	24990	}
24970	24991	#else
24971	24992	__ai __attribute__((target("neon"))) float32x4_t vrecpsq_f32(float32x4_t __p0, float32x4_t __p1) {
24972	24993	float32x4_t __ret;
24973		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
24974		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
24975		__ret = (float32x4_t) __builtin_neon_vrecpsq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 41);
24976		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	24994	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	24995	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	24996	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vrecpsq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 41));
	24997	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
24977	24998	return __ret;
24978	24999	}
24979	25000	#endif
...	...	@@ -24981,16 +25002,16 @@ __ai __attribute__((target("neon"))) float32x4_t vrecpsq_f32(float32x4_t __p0, f
24981	25002	#ifdef __LITTLE_ENDIAN__
24982	25003	__ai __attribute__((target("neon"))) float32x2_t vrecps_f32(float32x2_t __p0, float32x2_t __p1) {
24983	25004	float32x2_t __ret;
24984		__ret = (float32x2_t) __builtin_neon_vrecps_v((int8x8_t)__p0, (int8x8_t)__p1, 9);
	25005	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vrecps_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 9));
24985	25006	return __ret;
24986	25007	}
24987	25008	#else
24988	25009	__ai __attribute__((target("neon"))) float32x2_t vrecps_f32(float32x2_t __p0, float32x2_t __p1) {
24989	25010	float32x2_t __ret;
24990		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
24991		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
24992		__ret = (float32x2_t) __builtin_neon_vrecps_v((int8x8_t)__rev0, (int8x8_t)__rev1, 9);
24993		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	25011	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	25012	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	25013	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vrecps_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 9));
	25014	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
24994	25015	return __ret;
24995	25016	}
24996	25017	#endif
...	...	@@ -25004,9 +25025,9 @@ __ai __attribute__((target("neon"))) poly8x8_t vrev16_p8(poly8x8_t __p0) {
25004	25025	#else
25005	25026	__ai __attribute__((target("neon"))) poly8x8_t vrev16_p8(poly8x8_t __p0) {
25006	25027	poly8x8_t __ret;
25007		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	25028	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
25008	25029	__ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2, 5, 4, 7, 6);
25009		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	25030	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
25010	25031	return __ret;
25011	25032	}
25012	25033	#endif
...	...	@@ -25020,9 +25041,9 @@ __ai __attribute__((target("neon"))) poly8x16_t vrev16q_p8(poly8x16_t __p0) {
25020	25041	#else
25021	25042	__ai __attribute__((target("neon"))) poly8x16_t vrev16q_p8(poly8x16_t __p0) {
25022	25043	poly8x16_t __ret;
25023		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	25044	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
25024	25045	__ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2, 5, 4, 7, 6, 9, 8, 11, 10, 13, 12, 15, 14);
25025		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	25046	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
25026	25047	return __ret;
25027	25048	}
25028	25049	#endif
...	...	@@ -25036,9 +25057,9 @@ __ai __attribute__((target("neon"))) uint8x16_t vrev16q_u8(uint8x16_t __p0) {
25036	25057	#else
25037	25058	__ai __attribute__((target("neon"))) uint8x16_t vrev16q_u8(uint8x16_t __p0) {
25038	25059	uint8x16_t __ret;
25039		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	25060	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
25040	25061	__ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2, 5, 4, 7, 6, 9, 8, 11, 10, 13, 12, 15, 14);
25041		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	25062	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
25042	25063	return __ret;
25043	25064	}
25044	25065	#endif
...	...	@@ -25052,9 +25073,9 @@ __ai __attribute__((target("neon"))) int8x16_t vrev16q_s8(int8x16_t __p0) {
25052	25073	#else
25053	25074	__ai __attribute__((target("neon"))) int8x16_t vrev16q_s8(int8x16_t __p0) {
25054	25075	int8x16_t __ret;
25055		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	25076	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
25056	25077	__ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2, 5, 4, 7, 6, 9, 8, 11, 10, 13, 12, 15, 14);
25057		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	25078	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
25058	25079	return __ret;
25059	25080	}
25060	25081	#endif
...	...	@@ -25068,9 +25089,9 @@ __ai __attribute__((target("neon"))) uint8x8_t vrev16_u8(uint8x8_t __p0) {
25068	25089	#else
25069	25090	__ai __attribute__((target("neon"))) uint8x8_t vrev16_u8(uint8x8_t __p0) {
25070	25091	uint8x8_t __ret;
25071		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	25092	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
25072	25093	__ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2, 5, 4, 7, 6);
25073		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	25094	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
25074	25095	return __ret;
25075	25096	}
25076	25097	#endif
...	...	@@ -25084,9 +25105,9 @@ __ai __attribute__((target("neon"))) int8x8_t vrev16_s8(int8x8_t __p0) {
25084	25105	#else
25085	25106	__ai __attribute__((target("neon"))) int8x8_t vrev16_s8(int8x8_t __p0) {
25086	25107	int8x8_t __ret;
25087		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	25108	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
25088	25109	__ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2, 5, 4, 7, 6);
25089		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	25110	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
25090	25111	return __ret;
25091	25112	}
25092	25113	#endif
...	...	@@ -25100,9 +25121,9 @@ __ai __attribute__((target("neon"))) poly8x8_t vrev32_p8(poly8x8_t __p0) {
25100	25121	#else
25101	25122	__ai __attribute__((target("neon"))) poly8x8_t vrev32_p8(poly8x8_t __p0) {
25102	25123	poly8x8_t __ret;
25103		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	25124	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
25104	25125	__ret = __builtin_shufflevector(__rev0, __rev0, 3, 2, 1, 0, 7, 6, 5, 4);
25105		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	25126	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
25106	25127	return __ret;
25107	25128	}
25108	25129	#endif
...	...	@@ -25116,9 +25137,9 @@ __ai __attribute__((target("neon"))) poly16x4_t vrev32_p16(poly16x4_t __p0) {
25116	25137	#else
25117	25138	__ai __attribute__((target("neon"))) poly16x4_t vrev32_p16(poly16x4_t __p0) {
25118	25139	poly16x4_t __ret;
25119		poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	25140	poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
25120	25141	__ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2);
25121		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	25142	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
25122	25143	return __ret;
25123	25144	}
25124	25145	#endif
...	...	@@ -25132,9 +25153,9 @@ __ai __attribute__((target("neon"))) poly8x16_t vrev32q_p8(poly8x16_t __p0) {
25132	25153	#else
25133	25154	__ai __attribute__((target("neon"))) poly8x16_t vrev32q_p8(poly8x16_t __p0) {
25134	25155	poly8x16_t __ret;
25135		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	25156	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
25136	25157	__ret = __builtin_shufflevector(__rev0, __rev0, 3, 2, 1, 0, 7, 6, 5, 4, 11, 10, 9, 8, 15, 14, 13, 12);
25137		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	25158	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
25138	25159	return __ret;
25139	25160	}
25140	25161	#endif
...	...	@@ -25148,9 +25169,9 @@ __ai __attribute__((target("neon"))) poly16x8_t vrev32q_p16(poly16x8_t __p0) {
25148	25169	#else
25149	25170	__ai __attribute__((target("neon"))) poly16x8_t vrev32q_p16(poly16x8_t __p0) {
25150	25171	poly16x8_t __ret;
25151		poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	25172	poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
25152	25173	__ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2, 5, 4, 7, 6);
25153		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	25174	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
25154	25175	return __ret;
25155	25176	}
25156	25177	#endif
...	...	@@ -25164,9 +25185,9 @@ __ai __attribute__((target("neon"))) uint8x16_t vrev32q_u8(uint8x16_t __p0) {
25164	25185	#else
25165	25186	__ai __attribute__((target("neon"))) uint8x16_t vrev32q_u8(uint8x16_t __p0) {
25166	25187	uint8x16_t __ret;
25167		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	25188	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
25168	25189	__ret = __builtin_shufflevector(__rev0, __rev0, 3, 2, 1, 0, 7, 6, 5, 4, 11, 10, 9, 8, 15, 14, 13, 12);
25169		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	25190	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
25170	25191	return __ret;
25171	25192	}
25172	25193	#endif
...	...	@@ -25180,9 +25201,9 @@ __ai __attribute__((target("neon"))) uint16x8_t vrev32q_u16(uint16x8_t __p0) {
25180	25201	#else
25181	25202	__ai __attribute__((target("neon"))) uint16x8_t vrev32q_u16(uint16x8_t __p0) {
25182	25203	uint16x8_t __ret;
25183		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	25204	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
25184	25205	__ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2, 5, 4, 7, 6);
25185		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	25206	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
25186	25207	return __ret;
25187	25208	}
25188	25209	#endif
...	...	@@ -25196,9 +25217,9 @@ __ai __attribute__((target("neon"))) int8x16_t vrev32q_s8(int8x16_t __p0) {
25196	25217	#else
25197	25218	__ai __attribute__((target("neon"))) int8x16_t vrev32q_s8(int8x16_t __p0) {
25198	25219	int8x16_t __ret;
25199		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	25220	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
25200	25221	__ret = __builtin_shufflevector(__rev0, __rev0, 3, 2, 1, 0, 7, 6, 5, 4, 11, 10, 9, 8, 15, 14, 13, 12);
25201		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	25222	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
25202	25223	return __ret;
25203	25224	}
25204	25225	#endif
...	...	@@ -25212,9 +25233,9 @@ __ai __attribute__((target("neon"))) int16x8_t vrev32q_s16(int16x8_t __p0) {
25212	25233	#else
25213	25234	__ai __attribute__((target("neon"))) int16x8_t vrev32q_s16(int16x8_t __p0) {
25214	25235	int16x8_t __ret;
25215		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	25236	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
25216	25237	__ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2, 5, 4, 7, 6);
25217		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	25238	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
25218	25239	return __ret;
25219	25240	}
25220	25241	#endif
...	...	@@ -25228,9 +25249,9 @@ __ai __attribute__((target("neon"))) uint8x8_t vrev32_u8(uint8x8_t __p0) {
25228	25249	#else
25229	25250	__ai __attribute__((target("neon"))) uint8x8_t vrev32_u8(uint8x8_t __p0) {
25230	25251	uint8x8_t __ret;
25231		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	25252	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
25232	25253	__ret = __builtin_shufflevector(__rev0, __rev0, 3, 2, 1, 0, 7, 6, 5, 4);
25233		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	25254	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
25234	25255	return __ret;
25235	25256	}
25236	25257	#endif
...	...	@@ -25244,9 +25265,9 @@ __ai __attribute__((target("neon"))) uint16x4_t vrev32_u16(uint16x4_t __p0) {
25244	25265	#else
25245	25266	__ai __attribute__((target("neon"))) uint16x4_t vrev32_u16(uint16x4_t __p0) {
25246	25267	uint16x4_t __ret;
25247		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	25268	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
25248	25269	__ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2);
25249		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	25270	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
25250	25271	return __ret;
25251	25272	}
25252	25273	#endif
...	...	@@ -25260,9 +25281,9 @@ __ai __attribute__((target("neon"))) int8x8_t vrev32_s8(int8x8_t __p0) {
25260	25281	#else
25261	25282	__ai __attribute__((target("neon"))) int8x8_t vrev32_s8(int8x8_t __p0) {
25262	25283	int8x8_t __ret;
25263		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	25284	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
25264	25285	__ret = __builtin_shufflevector(__rev0, __rev0, 3, 2, 1, 0, 7, 6, 5, 4);
25265		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	25286	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
25266	25287	return __ret;
25267	25288	}
25268	25289	#endif
...	...	@@ -25276,9 +25297,9 @@ __ai __attribute__((target("neon"))) int16x4_t vrev32_s16(int16x4_t __p0) {
25276	25297	#else
25277	25298	__ai __attribute__((target("neon"))) int16x4_t vrev32_s16(int16x4_t __p0) {
25278	25299	int16x4_t __ret;
25279		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	25300	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
25280	25301	__ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2);
25281		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	25302	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
25282	25303	return __ret;
25283	25304	}
25284	25305	#endif
...	...	@@ -25292,9 +25313,9 @@ __ai __attribute__((target("neon"))) poly8x8_t vrev64_p8(poly8x8_t __p0) {
25292	25313	#else
25293	25314	__ai __attribute__((target("neon"))) poly8x8_t vrev64_p8(poly8x8_t __p0) {
25294	25315	poly8x8_t __ret;
25295		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	25316	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
25296	25317	__ret = __builtin_shufflevector(__rev0, __rev0, 7, 6, 5, 4, 3, 2, 1, 0);
25297		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	25318	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
25298	25319	return __ret;
25299	25320	}
25300	25321	#endif
...	...	@@ -25308,9 +25329,9 @@ __ai __attribute__((target("neon"))) poly16x4_t vrev64_p16(poly16x4_t __p0) {
25308	25329	#else
25309	25330	__ai __attribute__((target("neon"))) poly16x4_t vrev64_p16(poly16x4_t __p0) {
25310	25331	poly16x4_t __ret;
25311		poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	25332	poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
25312	25333	__ret = __builtin_shufflevector(__rev0, __rev0, 3, 2, 1, 0);
25313		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	25334	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
25314	25335	return __ret;
25315	25336	}
25316	25337	#endif
...	...	@@ -25324,9 +25345,9 @@ __ai __attribute__((target("neon"))) poly8x16_t vrev64q_p8(poly8x16_t __p0) {
25324	25345	#else
25325	25346	__ai __attribute__((target("neon"))) poly8x16_t vrev64q_p8(poly8x16_t __p0) {
25326	25347	poly8x16_t __ret;
25327		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	25348	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
25328	25349	__ret = __builtin_shufflevector(__rev0, __rev0, 7, 6, 5, 4, 3, 2, 1, 0, 15, 14, 13, 12, 11, 10, 9, 8);
25329		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	25350	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
25330	25351	return __ret;
25331	25352	}
25332	25353	#endif
...	...	@@ -25340,9 +25361,9 @@ __ai __attribute__((target("neon"))) poly16x8_t vrev64q_p16(poly16x8_t __p0) {
25340	25361	#else
25341	25362	__ai __attribute__((target("neon"))) poly16x8_t vrev64q_p16(poly16x8_t __p0) {
25342	25363	poly16x8_t __ret;
25343		poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	25364	poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
25344	25365	__ret = __builtin_shufflevector(__rev0, __rev0, 3, 2, 1, 0, 7, 6, 5, 4);
25345		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	25366	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
25346	25367	return __ret;
25347	25368	}
25348	25369	#endif
...	...	@@ -25356,9 +25377,9 @@ __ai __attribute__((target("neon"))) uint8x16_t vrev64q_u8(uint8x16_t __p0) {
25356	25377	#else
25357	25378	__ai __attribute__((target("neon"))) uint8x16_t vrev64q_u8(uint8x16_t __p0) {
25358	25379	uint8x16_t __ret;
25359		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	25380	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
25360	25381	__ret = __builtin_shufflevector(__rev0, __rev0, 7, 6, 5, 4, 3, 2, 1, 0, 15, 14, 13, 12, 11, 10, 9, 8);
25361		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	25382	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
25362	25383	return __ret;
25363	25384	}
25364	25385	#endif
...	...	@@ -25372,9 +25393,9 @@ __ai __attribute__((target("neon"))) uint32x4_t vrev64q_u32(uint32x4_t __p0) {
25372	25393	#else
25373	25394	__ai __attribute__((target("neon"))) uint32x4_t vrev64q_u32(uint32x4_t __p0) {
25374	25395	uint32x4_t __ret;
25375		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	25396	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
25376	25397	__ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2);
25377		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	25398	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
25378	25399	return __ret;
25379	25400	}
25380	25401	#endif
...	...	@@ -25388,9 +25409,9 @@ __ai __attribute__((target("neon"))) uint16x8_t vrev64q_u16(uint16x8_t __p0) {
25388	25409	#else
25389	25410	__ai __attribute__((target("neon"))) uint16x8_t vrev64q_u16(uint16x8_t __p0) {
25390	25411	uint16x8_t __ret;
25391		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	25412	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
25392	25413	__ret = __builtin_shufflevector(__rev0, __rev0, 3, 2, 1, 0, 7, 6, 5, 4);
25393		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	25414	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
25394	25415	return __ret;
25395	25416	}
25396	25417	#endif
...	...	@@ -25404,9 +25425,9 @@ __ai __attribute__((target("neon"))) int8x16_t vrev64q_s8(int8x16_t __p0) {
25404	25425	#else
25405	25426	__ai __attribute__((target("neon"))) int8x16_t vrev64q_s8(int8x16_t __p0) {
25406	25427	int8x16_t __ret;
25407		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	25428	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
25408	25429	__ret = __builtin_shufflevector(__rev0, __rev0, 7, 6, 5, 4, 3, 2, 1, 0, 15, 14, 13, 12, 11, 10, 9, 8);
25409		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	25430	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
25410	25431	return __ret;
25411	25432	}
25412	25433	#endif
...	...	@@ -25420,9 +25441,9 @@ __ai __attribute__((target("neon"))) float32x4_t vrev64q_f32(float32x4_t __p0) {
25420	25441	#else
25421	25442	__ai __attribute__((target("neon"))) float32x4_t vrev64q_f32(float32x4_t __p0) {
25422	25443	float32x4_t __ret;
25423		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	25444	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
25424	25445	__ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2);
25425		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	25446	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
25426	25447	return __ret;
25427	25448	}
25428	25449	#endif
...	...	@@ -25436,9 +25457,9 @@ __ai __attribute__((target("neon"))) int32x4_t vrev64q_s32(int32x4_t __p0) {
25436	25457	#else
25437	25458	__ai __attribute__((target("neon"))) int32x4_t vrev64q_s32(int32x4_t __p0) {
25438	25459	int32x4_t __ret;
25439		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	25460	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
25440	25461	__ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2);
25441		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	25462	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
25442	25463	return __ret;
25443	25464	}
25444	25465	#endif
...	...	@@ -25452,9 +25473,9 @@ __ai __attribute__((target("neon"))) int16x8_t vrev64q_s16(int16x8_t __p0) {
25452	25473	#else
25453	25474	__ai __attribute__((target("neon"))) int16x8_t vrev64q_s16(int16x8_t __p0) {
25454	25475	int16x8_t __ret;
25455		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	25476	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
25456	25477	__ret = __builtin_shufflevector(__rev0, __rev0, 3, 2, 1, 0, 7, 6, 5, 4);
25457		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	25478	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
25458	25479	return __ret;
25459	25480	}
25460	25481	#endif
...	...	@@ -25468,9 +25489,9 @@ __ai __attribute__((target("neon"))) uint8x8_t vrev64_u8(uint8x8_t __p0) {
25468	25489	#else
25469	25490	__ai __attribute__((target("neon"))) uint8x8_t vrev64_u8(uint8x8_t __p0) {
25470	25491	uint8x8_t __ret;
25471		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	25492	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
25472	25493	__ret = __builtin_shufflevector(__rev0, __rev0, 7, 6, 5, 4, 3, 2, 1, 0);
25473		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	25494	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
25474	25495	return __ret;
25475	25496	}
25476	25497	#endif
...	...	@@ -25484,9 +25505,9 @@ __ai __attribute__((target("neon"))) uint32x2_t vrev64_u32(uint32x2_t __p0) {
25484	25505	#else
25485	25506	__ai __attribute__((target("neon"))) uint32x2_t vrev64_u32(uint32x2_t __p0) {
25486	25507	uint32x2_t __ret;
25487		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
	25508	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
25488	25509	__ret = __builtin_shufflevector(__rev0, __rev0, 1, 0);
25489		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	25510	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
25490	25511	return __ret;
25491	25512	}
25492	25513	#endif
...	...	@@ -25500,9 +25521,9 @@ __ai __attribute__((target("neon"))) uint16x4_t vrev64_u16(uint16x4_t __p0) {
25500	25521	#else
25501	25522	__ai __attribute__((target("neon"))) uint16x4_t vrev64_u16(uint16x4_t __p0) {
25502	25523	uint16x4_t __ret;
25503		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	25524	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
25504	25525	__ret = __builtin_shufflevector(__rev0, __rev0, 3, 2, 1, 0);
25505		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	25526	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
25506	25527	return __ret;
25507	25528	}
25508	25529	#endif
...	...	@@ -25516,9 +25537,9 @@ __ai __attribute__((target("neon"))) int8x8_t vrev64_s8(int8x8_t __p0) {
25516	25537	#else
25517	25538	__ai __attribute__((target("neon"))) int8x8_t vrev64_s8(int8x8_t __p0) {
25518	25539	int8x8_t __ret;
25519		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	25540	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
25520	25541	__ret = __builtin_shufflevector(__rev0, __rev0, 7, 6, 5, 4, 3, 2, 1, 0);
25521		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	25542	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
25522	25543	return __ret;
25523	25544	}
25524	25545	#endif
...	...	@@ -25532,9 +25553,9 @@ __ai __attribute__((target("neon"))) float32x2_t vrev64_f32(float32x2_t __p0) {
25532	25553	#else
25533	25554	__ai __attribute__((target("neon"))) float32x2_t vrev64_f32(float32x2_t __p0) {
25534	25555	float32x2_t __ret;
25535		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
	25556	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
25536	25557	__ret = __builtin_shufflevector(__rev0, __rev0, 1, 0);
25537		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	25558	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
25538	25559	return __ret;
25539	25560	}
25540	25561	#endif
...	...	@@ -25548,9 +25569,9 @@ __ai __attribute__((target("neon"))) int32x2_t vrev64_s32(int32x2_t __p0) {
25548	25569	#else
25549	25570	__ai __attribute__((target("neon"))) int32x2_t vrev64_s32(int32x2_t __p0) {
25550	25571	int32x2_t __ret;
25551		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
	25572	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
25552	25573	__ret = __builtin_shufflevector(__rev0, __rev0, 1, 0);
25553		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	25574	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
25554	25575	return __ret;
25555	25576	}
25556	25577	#endif
...	...	@@ -25564,9 +25585,9 @@ __ai __attribute__((target("neon"))) int16x4_t vrev64_s16(int16x4_t __p0) {
25564	25585	#else
25565	25586	__ai __attribute__((target("neon"))) int16x4_t vrev64_s16(int16x4_t __p0) {
25566	25587	int16x4_t __ret;
25567		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	25588	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
25568	25589	__ret = __builtin_shufflevector(__rev0, __rev0, 3, 2, 1, 0);
25569		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	25590	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
25570	25591	return __ret;
25571	25592	}
25572	25593	#endif
...	...	@@ -25580,9 +25601,9 @@ __ai __attribute__((target("neon"))) float16x8_t vrev64q_f16(float16x8_t __p0) {
25580	25601	#else
25581	25602	__ai __attribute__((target("neon"))) float16x8_t vrev64q_f16(float16x8_t __p0) {
25582	25603	float16x8_t __ret;
25583		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	25604	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
25584	25605	__ret = __builtin_shufflevector(__rev0, __rev0, 3, 2, 1, 0, 7, 6, 5, 4);
25585		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	25606	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
25586	25607	return __ret;
25587	25608	}
25588	25609	#endif
...	...	@@ -25596,9 +25617,9 @@ __ai __attribute__((target("neon"))) float16x4_t vrev64_f16(float16x4_t __p0) {
25596	25617	#else
25597	25618	__ai __attribute__((target("neon"))) float16x4_t vrev64_f16(float16x4_t __p0) {
25598	25619	float16x4_t __ret;
25599		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	25620	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
25600	25621	__ret = __builtin_shufflevector(__rev0, __rev0, 3, 2, 1, 0);
25601		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	25622	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
25602	25623	return __ret;
25603	25624	}
25604	25625	#endif
...	...	@@ -25606,16 +25627,16 @@ __ai __attribute__((target("neon"))) float16x4_t vrev64_f16(float16x4_t __p0) {
25606	25627	#ifdef __LITTLE_ENDIAN__
25607	25628	__ai __attribute__((target("neon"))) uint8x16_t vrhaddq_u8(uint8x16_t __p0, uint8x16_t __p1) {
25608	25629	uint8x16_t __ret;
25609		__ret = (uint8x16_t) __builtin_neon_vrhaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
	25630	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vrhaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 48));
25610	25631	return __ret;
25611	25632	}
25612	25633	#else
25613	25634	__ai __attribute__((target("neon"))) uint8x16_t vrhaddq_u8(uint8x16_t __p0, uint8x16_t __p1) {
25614	25635	uint8x16_t __ret;
25615		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
25616		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
25617		__ret = (uint8x16_t) __builtin_neon_vrhaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
25618		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	25636	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	25637	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	25638	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vrhaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 48));
	25639	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
25619	25640	return __ret;
25620	25641	}
25621	25642	#endif
...	...	@@ -25623,16 +25644,16 @@ __ai __attribute__((target("neon"))) uint8x16_t vrhaddq_u8(uint8x16_t __p0, uint
25623	25644	#ifdef __LITTLE_ENDIAN__
25624	25645	__ai __attribute__((target("neon"))) uint32x4_t vrhaddq_u32(uint32x4_t __p0, uint32x4_t __p1) {
25625	25646	uint32x4_t __ret;
25626		__ret = (uint32x4_t) __builtin_neon_vrhaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
	25647	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vrhaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50));
25627	25648	return __ret;
25628	25649	}
25629	25650	#else
25630	25651	__ai __attribute__((target("neon"))) uint32x4_t vrhaddq_u32(uint32x4_t __p0, uint32x4_t __p1) {
25631	25652	uint32x4_t __ret;
25632		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
25633		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
25634		__ret = (uint32x4_t) __builtin_neon_vrhaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
25635		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	25653	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	25654	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	25655	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vrhaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 50));
	25656	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
25636	25657	return __ret;
25637	25658	}
25638	25659	#endif
...	...	@@ -25640,16 +25661,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vrhaddq_u32(uint32x4_t __p0, uin
25640	25661	#ifdef __LITTLE_ENDIAN__
25641	25662	__ai __attribute__((target("neon"))) uint16x8_t vrhaddq_u16(uint16x8_t __p0, uint16x8_t __p1) {
25642	25663	uint16x8_t __ret;
25643		__ret = (uint16x8_t) __builtin_neon_vrhaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
	25664	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vrhaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 49));
25644	25665	return __ret;
25645	25666	}
25646	25667	#else
25647	25668	__ai __attribute__((target("neon"))) uint16x8_t vrhaddq_u16(uint16x8_t __p0, uint16x8_t __p1) {
25648	25669	uint16x8_t __ret;
25649		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
25650		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
25651		__ret = (uint16x8_t) __builtin_neon_vrhaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
25652		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	25670	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	25671	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	25672	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vrhaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 49));
	25673	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
25653	25674	return __ret;
25654	25675	}
25655	25676	#endif
...	...	@@ -25657,16 +25678,16 @@ __ai __attribute__((target("neon"))) uint16x8_t vrhaddq_u16(uint16x8_t __p0, uin
25657	25678	#ifdef __LITTLE_ENDIAN__
25658	25679	__ai __attribute__((target("neon"))) int8x16_t vrhaddq_s8(int8x16_t __p0, int8x16_t __p1) {
25659	25680	int8x16_t __ret;
25660		__ret = (int8x16_t) __builtin_neon_vrhaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
	25681	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vrhaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 32));
25661	25682	return __ret;
25662	25683	}
25663	25684	#else
25664	25685	__ai __attribute__((target("neon"))) int8x16_t vrhaddq_s8(int8x16_t __p0, int8x16_t __p1) {
25665	25686	int8x16_t __ret;
25666		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
25667		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
25668		__ret = (int8x16_t) __builtin_neon_vrhaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 32);
25669		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	25687	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	25688	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	25689	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vrhaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 32));
	25690	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
25670	25691	return __ret;
25671	25692	}
25672	25693	#endif
...	...	@@ -25674,16 +25695,16 @@ __ai __attribute__((target("neon"))) int8x16_t vrhaddq_s8(int8x16_t __p0, int8x1
25674	25695	#ifdef __LITTLE_ENDIAN__
25675	25696	__ai __attribute__((target("neon"))) int32x4_t vrhaddq_s32(int32x4_t __p0, int32x4_t __p1) {
25676	25697	int32x4_t __ret;
25677		__ret = (int32x4_t) __builtin_neon_vrhaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
	25698	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vrhaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 34));
25678	25699	return __ret;
25679	25700	}
25680	25701	#else
25681	25702	__ai __attribute__((target("neon"))) int32x4_t vrhaddq_s32(int32x4_t __p0, int32x4_t __p1) {
25682	25703	int32x4_t __ret;
25683		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
25684		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
25685		__ret = (int32x4_t) __builtin_neon_vrhaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
25686		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	25704	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	25705	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	25706	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vrhaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 34));
	25707	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
25687	25708	return __ret;
25688	25709	}
25689	25710	#endif
...	...	@@ -25691,16 +25712,16 @@ __ai __attribute__((target("neon"))) int32x4_t vrhaddq_s32(int32x4_t __p0, int32
25691	25712	#ifdef __LITTLE_ENDIAN__
25692	25713	__ai __attribute__((target("neon"))) int16x8_t vrhaddq_s16(int16x8_t __p0, int16x8_t __p1) {
25693	25714	int16x8_t __ret;
25694		__ret = (int16x8_t) __builtin_neon_vrhaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
	25715	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vrhaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 33));
25695	25716	return __ret;
25696	25717	}
25697	25718	#else
25698	25719	__ai __attribute__((target("neon"))) int16x8_t vrhaddq_s16(int16x8_t __p0, int16x8_t __p1) {
25699	25720	int16x8_t __ret;
25700		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
25701		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
25702		__ret = (int16x8_t) __builtin_neon_vrhaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
25703		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	25721	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	25722	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	25723	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vrhaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 33));
	25724	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
25704	25725	return __ret;
25705	25726	}
25706	25727	#endif
...	...	@@ -25708,16 +25729,16 @@ __ai __attribute__((target("neon"))) int16x8_t vrhaddq_s16(int16x8_t __p0, int16
25708	25729	#ifdef __LITTLE_ENDIAN__
25709	25730	__ai __attribute__((target("neon"))) uint8x8_t vrhadd_u8(uint8x8_t __p0, uint8x8_t __p1) {
25710	25731	uint8x8_t __ret;
25711		__ret = (uint8x8_t) __builtin_neon_vrhadd_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
	25732	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vrhadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 16));
25712	25733	return __ret;
25713	25734	}
25714	25735	#else
25715	25736	__ai __attribute__((target("neon"))) uint8x8_t vrhadd_u8(uint8x8_t __p0, uint8x8_t __p1) {
25716	25737	uint8x8_t __ret;
25717		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
25718		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
25719		__ret = (uint8x8_t) __builtin_neon_vrhadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
25720		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	25738	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	25739	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	25740	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vrhadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 16));
	25741	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
25721	25742	return __ret;
25722	25743	}
25723	25744	#endif
...	...	@@ -25725,16 +25746,16 @@ __ai __attribute__((target("neon"))) uint8x8_t vrhadd_u8(uint8x8_t __p0, uint8x8
25725	25746	#ifdef __LITTLE_ENDIAN__
25726	25747	__ai __attribute__((target("neon"))) uint32x2_t vrhadd_u32(uint32x2_t __p0, uint32x2_t __p1) {
25727	25748	uint32x2_t __ret;
25728		__ret = (uint32x2_t) __builtin_neon_vrhadd_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
	25749	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vrhadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 18));
25729	25750	return __ret;
25730	25751	}
25731	25752	#else
25732	25753	__ai __attribute__((target("neon"))) uint32x2_t vrhadd_u32(uint32x2_t __p0, uint32x2_t __p1) {
25733	25754	uint32x2_t __ret;
25734		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
25735		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
25736		__ret = (uint32x2_t) __builtin_neon_vrhadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
25737		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	25755	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	25756	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	25757	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vrhadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 18));
	25758	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
25738	25759	return __ret;
25739	25760	}
25740	25761	#endif
...	...	@@ -25742,16 +25763,16 @@ __ai __attribute__((target("neon"))) uint32x2_t vrhadd_u32(uint32x2_t __p0, uint
25742	25763	#ifdef __LITTLE_ENDIAN__
25743	25764	__ai __attribute__((target("neon"))) uint16x4_t vrhadd_u16(uint16x4_t __p0, uint16x4_t __p1) {
25744	25765	uint16x4_t __ret;
25745		__ret = (uint16x4_t) __builtin_neon_vrhadd_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
	25766	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vrhadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 17));
25746	25767	return __ret;
25747	25768	}
25748	25769	#else
25749	25770	__ai __attribute__((target("neon"))) uint16x4_t vrhadd_u16(uint16x4_t __p0, uint16x4_t __p1) {
25750	25771	uint16x4_t __ret;
25751		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
25752		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
25753		__ret = (uint16x4_t) __builtin_neon_vrhadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
25754		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	25772	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	25773	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	25774	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vrhadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 17));
	25775	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
25755	25776	return __ret;
25756	25777	}
25757	25778	#endif
...	...	@@ -25759,16 +25780,16 @@ __ai __attribute__((target("neon"))) uint16x4_t vrhadd_u16(uint16x4_t __p0, uint
25759	25780	#ifdef __LITTLE_ENDIAN__
25760	25781	__ai __attribute__((target("neon"))) int8x8_t vrhadd_s8(int8x8_t __p0, int8x8_t __p1) {
25761	25782	int8x8_t __ret;
25762		__ret = (int8x8_t) __builtin_neon_vrhadd_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
	25783	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vrhadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 0));
25763	25784	return __ret;
25764	25785	}
25765	25786	#else
25766	25787	__ai __attribute__((target("neon"))) int8x8_t vrhadd_s8(int8x8_t __p0, int8x8_t __p1) {
25767	25788	int8x8_t __ret;
25768		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
25769		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
25770		__ret = (int8x8_t) __builtin_neon_vrhadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
25771		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	25789	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	25790	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	25791	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vrhadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 0));
	25792	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
25772	25793	return __ret;
25773	25794	}
25774	25795	#endif
...	...	@@ -25776,16 +25797,16 @@ __ai __attribute__((target("neon"))) int8x8_t vrhadd_s8(int8x8_t __p0, int8x8_t
25776	25797	#ifdef __LITTLE_ENDIAN__
25777	25798	__ai __attribute__((target("neon"))) int32x2_t vrhadd_s32(int32x2_t __p0, int32x2_t __p1) {
25778	25799	int32x2_t __ret;
25779		__ret = (int32x2_t) __builtin_neon_vrhadd_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
	25800	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vrhadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 2));
25780	25801	return __ret;
25781	25802	}
25782	25803	#else
25783	25804	__ai __attribute__((target("neon"))) int32x2_t vrhadd_s32(int32x2_t __p0, int32x2_t __p1) {
25784	25805	int32x2_t __ret;
25785		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
25786		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
25787		__ret = (int32x2_t) __builtin_neon_vrhadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
25788		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	25806	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	25807	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	25808	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vrhadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 2));
	25809	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
25789	25810	return __ret;
25790	25811	}
25791	25812	#endif
...	...	@@ -25793,16 +25814,16 @@ __ai __attribute__((target("neon"))) int32x2_t vrhadd_s32(int32x2_t __p0, int32x
25793	25814	#ifdef __LITTLE_ENDIAN__
25794	25815	__ai __attribute__((target("neon"))) int16x4_t vrhadd_s16(int16x4_t __p0, int16x4_t __p1) {
25795	25816	int16x4_t __ret;
25796		__ret = (int16x4_t) __builtin_neon_vrhadd_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
	25817	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vrhadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 1));
25797	25818	return __ret;
25798	25819	}
25799	25820	#else
25800	25821	__ai __attribute__((target("neon"))) int16x4_t vrhadd_s16(int16x4_t __p0, int16x4_t __p1) {
25801	25822	int16x4_t __ret;
25802		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
25803		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
25804		__ret = (int16x4_t) __builtin_neon_vrhadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
25805		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	25823	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	25824	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	25825	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vrhadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 1));
	25826	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
25806	25827	return __ret;
25807	25828	}
25808	25829	#endif
...	...	@@ -25810,16 +25831,16 @@ __ai __attribute__((target("neon"))) int16x4_t vrhadd_s16(int16x4_t __p0, int16x
25810	25831	#ifdef __LITTLE_ENDIAN__
25811	25832	__ai __attribute__((target("neon"))) uint8x16_t vrshlq_u8(uint8x16_t __p0, int8x16_t __p1) {
25812	25833	uint8x16_t __ret;
25813		__ret = (uint8x16_t) __builtin_neon_vrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
	25834	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vrshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 48));
25814	25835	return __ret;
25815	25836	}
25816	25837	#else
25817	25838	__ai __attribute__((target("neon"))) uint8x16_t vrshlq_u8(uint8x16_t __p0, int8x16_t __p1) {
25818	25839	uint8x16_t __ret;
25819		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
25820		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
25821		__ret = (uint8x16_t) __builtin_neon_vrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
25822		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	25840	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	25841	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	25842	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vrshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 48));
	25843	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
25823	25844	return __ret;
25824	25845	}
25825	25846	#endif
...	...	@@ -25827,16 +25848,16 @@ __ai __attribute__((target("neon"))) uint8x16_t vrshlq_u8(uint8x16_t __p0, int8x
25827	25848	#ifdef __LITTLE_ENDIAN__
25828	25849	__ai __attribute__((target("neon"))) uint32x4_t vrshlq_u32(uint32x4_t __p0, int32x4_t __p1) {
25829	25850	uint32x4_t __ret;
25830		__ret = (uint32x4_t) __builtin_neon_vrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
	25851	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vrshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50));
25831	25852	return __ret;
25832	25853	}
25833	25854	#else
25834	25855	__ai __attribute__((target("neon"))) uint32x4_t vrshlq_u32(uint32x4_t __p0, int32x4_t __p1) {
25835	25856	uint32x4_t __ret;
25836		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
25837		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
25838		__ret = (uint32x4_t) __builtin_neon_vrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
25839		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	25857	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	25858	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	25859	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vrshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 50));
	25860	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
25840	25861	return __ret;
25841	25862	}
25842	25863	#endif
...	...	@@ -25844,16 +25865,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vrshlq_u32(uint32x4_t __p0, int3
25844	25865	#ifdef __LITTLE_ENDIAN__
25845	25866	__ai __attribute__((target("neon"))) uint64x2_t vrshlq_u64(uint64x2_t __p0, int64x2_t __p1) {
25846	25867	uint64x2_t __ret;
25847		__ret = (uint64x2_t) __builtin_neon_vrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 51);
	25868	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vrshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 51));
25848	25869	return __ret;
25849	25870	}
25850	25871	#else
25851	25872	__ai __attribute__((target("neon"))) uint64x2_t vrshlq_u64(uint64x2_t __p0, int64x2_t __p1) {
25852	25873	uint64x2_t __ret;
25853		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
25854		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
25855		__ret = (uint64x2_t) __builtin_neon_vrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 51);
25856		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	25874	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	25875	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	25876	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vrshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 51));
	25877	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
25857	25878	return __ret;
25858	25879	}
25859	25880	#endif
...	...	@@ -25861,16 +25882,16 @@ __ai __attribute__((target("neon"))) uint64x2_t vrshlq_u64(uint64x2_t __p0, int6
25861	25882	#ifdef __LITTLE_ENDIAN__
25862	25883	__ai __attribute__((target("neon"))) uint16x8_t vrshlq_u16(uint16x8_t __p0, int16x8_t __p1) {
25863	25884	uint16x8_t __ret;
25864		__ret = (uint16x8_t) __builtin_neon_vrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
	25885	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vrshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 49));
25865	25886	return __ret;
25866	25887	}
25867	25888	#else
25868	25889	__ai __attribute__((target("neon"))) uint16x8_t vrshlq_u16(uint16x8_t __p0, int16x8_t __p1) {
25869	25890	uint16x8_t __ret;
25870		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
25871		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
25872		__ret = (uint16x8_t) __builtin_neon_vrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
25873		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	25891	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	25892	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	25893	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vrshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 49));
	25894	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
25874	25895	return __ret;
25875	25896	}
25876	25897	#endif
...	...	@@ -25878,16 +25899,16 @@ __ai __attribute__((target("neon"))) uint16x8_t vrshlq_u16(uint16x8_t __p0, int1
25878	25899	#ifdef __LITTLE_ENDIAN__
25879	25900	__ai __attribute__((target("neon"))) int8x16_t vrshlq_s8(int8x16_t __p0, int8x16_t __p1) {
25880	25901	int8x16_t __ret;
25881		__ret = (int8x16_t) __builtin_neon_vrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
	25902	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vrshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 32));
25882	25903	return __ret;
25883	25904	}
25884	25905	#else
25885	25906	__ai __attribute__((target("neon"))) int8x16_t vrshlq_s8(int8x16_t __p0, int8x16_t __p1) {
25886	25907	int8x16_t __ret;
25887		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
25888		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
25889		__ret = (int8x16_t) __builtin_neon_vrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 32);
25890		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	25908	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	25909	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	25910	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vrshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 32));
	25911	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
25891	25912	return __ret;
25892	25913	}
25893	25914	#endif
...	...	@@ -25895,16 +25916,16 @@ __ai __attribute__((target("neon"))) int8x16_t vrshlq_s8(int8x16_t __p0, int8x16
25895	25916	#ifdef __LITTLE_ENDIAN__
25896	25917	__ai __attribute__((target("neon"))) int32x4_t vrshlq_s32(int32x4_t __p0, int32x4_t __p1) {
25897	25918	int32x4_t __ret;
25898		__ret = (int32x4_t) __builtin_neon_vrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
	25919	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vrshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 34));
25899	25920	return __ret;
25900	25921	}
25901	25922	#else
25902	25923	__ai __attribute__((target("neon"))) int32x4_t vrshlq_s32(int32x4_t __p0, int32x4_t __p1) {
25903	25924	int32x4_t __ret;
25904		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
25905		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
25906		__ret = (int32x4_t) __builtin_neon_vrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
25907		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	25925	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	25926	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	25927	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vrshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 34));
	25928	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
25908	25929	return __ret;
25909	25930	}
25910	25931	#endif
...	...	@@ -25912,16 +25933,16 @@ __ai __attribute__((target("neon"))) int32x4_t vrshlq_s32(int32x4_t __p0, int32x
25912	25933	#ifdef __LITTLE_ENDIAN__
25913	25934	__ai __attribute__((target("neon"))) int64x2_t vrshlq_s64(int64x2_t __p0, int64x2_t __p1) {
25914	25935	int64x2_t __ret;
25915		__ret = (int64x2_t) __builtin_neon_vrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 35);
	25936	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vrshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 35));
25916	25937	return __ret;
25917	25938	}
25918	25939	#else
25919	25940	__ai __attribute__((target("neon"))) int64x2_t vrshlq_s64(int64x2_t __p0, int64x2_t __p1) {
25920	25941	int64x2_t __ret;
25921		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
25922		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
25923		__ret = (int64x2_t) __builtin_neon_vrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 35);
25924		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	25942	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	25943	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	25944	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vrshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 35));
	25945	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
25925	25946	return __ret;
25926	25947	}
25927	25948	#endif
...	...	@@ -25929,16 +25950,16 @@ __ai __attribute__((target("neon"))) int64x2_t vrshlq_s64(int64x2_t __p0, int64x
25929	25950	#ifdef __LITTLE_ENDIAN__
25930	25951	__ai __attribute__((target("neon"))) int16x8_t vrshlq_s16(int16x8_t __p0, int16x8_t __p1) {
25931	25952	int16x8_t __ret;
25932		__ret = (int16x8_t) __builtin_neon_vrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
	25953	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vrshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 33));
25933	25954	return __ret;
25934	25955	}
25935	25956	#else
25936	25957	__ai __attribute__((target("neon"))) int16x8_t vrshlq_s16(int16x8_t __p0, int16x8_t __p1) {
25937	25958	int16x8_t __ret;
25938		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
25939		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
25940		__ret = (int16x8_t) __builtin_neon_vrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
25941		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	25959	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	25960	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	25961	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vrshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 33));
	25962	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
25942	25963	return __ret;
25943	25964	}
25944	25965	#endif
...	...	@@ -25946,16 +25967,16 @@ __ai __attribute__((target("neon"))) int16x8_t vrshlq_s16(int16x8_t __p0, int16x
25946	25967	#ifdef __LITTLE_ENDIAN__
25947	25968	__ai __attribute__((target("neon"))) uint8x8_t vrshl_u8(uint8x8_t __p0, int8x8_t __p1) {
25948	25969	uint8x8_t __ret;
25949		__ret = (uint8x8_t) __builtin_neon_vrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
	25970	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vrshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 16));
25950	25971	return __ret;
25951	25972	}
25952	25973	#else
25953	25974	__ai __attribute__((target("neon"))) uint8x8_t vrshl_u8(uint8x8_t __p0, int8x8_t __p1) {
25954	25975	uint8x8_t __ret;
25955		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
25956		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
25957		__ret = (uint8x8_t) __builtin_neon_vrshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
25958		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	25976	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	25977	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	25978	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vrshl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 16));
	25979	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
25959	25980	return __ret;
25960	25981	}
25961	25982	#endif
...	...	@@ -25963,38 +25984,38 @@ __ai __attribute__((target("neon"))) uint8x8_t vrshl_u8(uint8x8_t __p0, int8x8_t
25963	25984	#ifdef __LITTLE_ENDIAN__
25964	25985	__ai __attribute__((target("neon"))) uint32x2_t vrshl_u32(uint32x2_t __p0, int32x2_t __p1) {
25965	25986	uint32x2_t __ret;
25966		__ret = (uint32x2_t) __builtin_neon_vrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
	25987	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vrshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 18));
25967	25988	return __ret;
25968	25989	}
25969	25990	#else
25970	25991	__ai __attribute__((target("neon"))) uint32x2_t vrshl_u32(uint32x2_t __p0, int32x2_t __p1) {
25971	25992	uint32x2_t __ret;
25972		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
25973		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
25974		__ret = (uint32x2_t) __builtin_neon_vrshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
25975		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	25993	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	25994	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	25995	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vrshl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 18));
	25996	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
25976	25997	return __ret;
25977	25998	}
25978	25999	#endif
25979	26000
25980	26001	__ai __attribute__((target("neon"))) uint64x1_t vrshl_u64(uint64x1_t __p0, int64x1_t __p1) {
25981	26002	uint64x1_t __ret;
25982		__ret = (uint64x1_t) __builtin_neon_vrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 19);
	26003	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vrshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 19));
25983	26004	return __ret;
25984	26005	}
25985	26006	#ifdef __LITTLE_ENDIAN__
25986	26007	__ai __attribute__((target("neon"))) uint16x4_t vrshl_u16(uint16x4_t __p0, int16x4_t __p1) {
25987	26008	uint16x4_t __ret;
25988		__ret = (uint16x4_t) __builtin_neon_vrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
	26009	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vrshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 17));
25989	26010	return __ret;
25990	26011	}
25991	26012	#else
25992	26013	__ai __attribute__((target("neon"))) uint16x4_t vrshl_u16(uint16x4_t __p0, int16x4_t __p1) {
25993	26014	uint16x4_t __ret;
25994		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
25995		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
25996		__ret = (uint16x4_t) __builtin_neon_vrshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
25997		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	26015	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	26016	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	26017	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vrshl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 17));
	26018	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
25998	26019	return __ret;
25999	26020	}
26000	26021	#endif
...	...	@@ -26002,16 +26023,16 @@ __ai __attribute__((target("neon"))) uint16x4_t vrshl_u16(uint16x4_t __p0, int16
26002	26023	#ifdef __LITTLE_ENDIAN__
26003	26024	__ai __attribute__((target("neon"))) int8x8_t vrshl_s8(int8x8_t __p0, int8x8_t __p1) {
26004	26025	int8x8_t __ret;
26005		__ret = (int8x8_t) __builtin_neon_vrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
	26026	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vrshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 0));
26006	26027	return __ret;
26007	26028	}
26008	26029	#else
26009	26030	__ai __attribute__((target("neon"))) int8x8_t vrshl_s8(int8x8_t __p0, int8x8_t __p1) {
26010	26031	int8x8_t __ret;
26011		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
26012		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
26013		__ret = (int8x8_t) __builtin_neon_vrshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
26014		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	26032	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	26033	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	26034	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vrshl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 0));
	26035	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
26015	26036	return __ret;
26016	26037	}
26017	26038	#endif
...	...	@@ -26019,38 +26040,38 @@ __ai __attribute__((target("neon"))) int8x8_t vrshl_s8(int8x8_t __p0, int8x8_t _
26019	26040	#ifdef __LITTLE_ENDIAN__
26020	26041	__ai __attribute__((target("neon"))) int32x2_t vrshl_s32(int32x2_t __p0, int32x2_t __p1) {
26021	26042	int32x2_t __ret;
26022		__ret = (int32x2_t) __builtin_neon_vrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
	26043	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vrshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 2));
26023	26044	return __ret;
26024	26045	}
26025	26046	#else
26026	26047	__ai __attribute__((target("neon"))) int32x2_t vrshl_s32(int32x2_t __p0, int32x2_t __p1) {
26027	26048	int32x2_t __ret;
26028		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
26029		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
26030		__ret = (int32x2_t) __builtin_neon_vrshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
26031		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	26049	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	26050	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	26051	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vrshl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 2));
	26052	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
26032	26053	return __ret;
26033	26054	}
26034	26055	#endif
26035	26056
26036	26057	__ai __attribute__((target("neon"))) int64x1_t vrshl_s64(int64x1_t __p0, int64x1_t __p1) {
26037	26058	int64x1_t __ret;
26038		__ret = (int64x1_t) __builtin_neon_vrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 3);
	26059	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vrshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 3));
26039	26060	return __ret;
26040	26061	}
26041	26062	#ifdef __LITTLE_ENDIAN__
26042	26063	__ai __attribute__((target("neon"))) int16x4_t vrshl_s16(int16x4_t __p0, int16x4_t __p1) {
26043	26064	int16x4_t __ret;
26044		__ret = (int16x4_t) __builtin_neon_vrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
	26065	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vrshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 1));
26045	26066	return __ret;
26046	26067	}
26047	26068	#else
26048	26069	__ai __attribute__((target("neon"))) int16x4_t vrshl_s16(int16x4_t __p0, int16x4_t __p1) {
26049	26070	int16x4_t __ret;
26050		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
26051		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
26052		__ret = (int16x4_t) __builtin_neon_vrshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
26053		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	26071	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	26072	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	26073	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vrshl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 1));
	26074	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
26054	26075	return __ret;
26055	26076	}
26056	26077	#endif
...	...	@@ -26059,16 +26080,16 @@ __ai __attribute__((target("neon"))) int16x4_t vrshl_s16(int16x4_t __p0, int16x4
26059	26080	#define vrshrq_n_u8(__p0, __p1) __extension__ ({ \
26060	26081	uint8x16_t __ret; \
26061	26082	uint8x16_t __s0 = __p0; \
26062		__ret = (uint8x16_t) __builtin_neon_vrshrq_n_v((int8x16_t)__s0, __p1, 48); \
	26083	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vrshrq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 48)); \
26063	26084	__ret; \
26064	26085	})
26065	26086	#else
26066	26087	#define vrshrq_n_u8(__p0, __p1) __extension__ ({ \
26067	26088	uint8x16_t __ret; \
26068	26089	uint8x16_t __s0 = __p0; \
26069		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
26070		__ret = (uint8x16_t) __builtin_neon_vrshrq_n_v((int8x16_t)__rev0, __p1, 48); \
26071		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	26090	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	26091	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vrshrq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 48)); \
	26092	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
26072	26093	__ret; \
26073	26094	})
26074	26095	#endif
...	...	@@ -26077,16 +26098,16 @@ __ai __attribute__((target("neon"))) int16x4_t vrshl_s16(int16x4_t __p0, int16x4
26077	26098	#define vrshrq_n_u32(__p0, __p1) __extension__ ({ \
26078	26099	uint32x4_t __ret; \
26079	26100	uint32x4_t __s0 = __p0; \
26080		__ret = (uint32x4_t) __builtin_neon_vrshrq_n_v((int8x16_t)__s0, __p1, 50); \
	26101	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vrshrq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 50)); \
26081	26102	__ret; \
26082	26103	})
26083	26104	#else
26084	26105	#define vrshrq_n_u32(__p0, __p1) __extension__ ({ \
26085	26106	uint32x4_t __ret; \
26086	26107	uint32x4_t __s0 = __p0; \
26087		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
26088		__ret = (uint32x4_t) __builtin_neon_vrshrq_n_v((int8x16_t)__rev0, __p1, 50); \
26089		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	26108	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	26109	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vrshrq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 50)); \
	26110	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
26090	26111	__ret; \
26091	26112	})
26092	26113	#endif
...	...	@@ -26095,16 +26116,16 @@ __ai __attribute__((target("neon"))) int16x4_t vrshl_s16(int16x4_t __p0, int16x4
26095	26116	#define vrshrq_n_u64(__p0, __p1) __extension__ ({ \
26096	26117	uint64x2_t __ret; \
26097	26118	uint64x2_t __s0 = __p0; \
26098		__ret = (uint64x2_t) __builtin_neon_vrshrq_n_v((int8x16_t)__s0, __p1, 51); \
	26119	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vrshrq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 51)); \
26099	26120	__ret; \
26100	26121	})
26101	26122	#else
26102	26123	#define vrshrq_n_u64(__p0, __p1) __extension__ ({ \
26103	26124	uint64x2_t __ret; \
26104	26125	uint64x2_t __s0 = __p0; \
26105		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
26106		__ret = (uint64x2_t) __builtin_neon_vrshrq_n_v((int8x16_t)__rev0, __p1, 51); \
26107		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	26126	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	26127	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vrshrq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 51)); \
	26128	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
26108	26129	__ret; \
26109	26130	})
26110	26131	#endif
...	...	@@ -26113,16 +26134,16 @@ __ai __attribute__((target("neon"))) int16x4_t vrshl_s16(int16x4_t __p0, int16x4
26113	26134	#define vrshrq_n_u16(__p0, __p1) __extension__ ({ \
26114	26135	uint16x8_t __ret; \
26115	26136	uint16x8_t __s0 = __p0; \
26116		__ret = (uint16x8_t) __builtin_neon_vrshrq_n_v((int8x16_t)__s0, __p1, 49); \
	26137	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vrshrq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 49)); \
26117	26138	__ret; \
26118	26139	})
26119	26140	#else
26120	26141	#define vrshrq_n_u16(__p0, __p1) __extension__ ({ \
26121	26142	uint16x8_t __ret; \
26122	26143	uint16x8_t __s0 = __p0; \
26123		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
26124		__ret = (uint16x8_t) __builtin_neon_vrshrq_n_v((int8x16_t)__rev0, __p1, 49); \
26125		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	26144	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	26145	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vrshrq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 49)); \
	26146	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
26126	26147	__ret; \
26127	26148	})
26128	26149	#endif
...	...	@@ -26131,16 +26152,16 @@ __ai __attribute__((target("neon"))) int16x4_t vrshl_s16(int16x4_t __p0, int16x4
26131	26152	#define vrshrq_n_s8(__p0, __p1) __extension__ ({ \
26132	26153	int8x16_t __ret; \
26133	26154	int8x16_t __s0 = __p0; \
26134		__ret = (int8x16_t) __builtin_neon_vrshrq_n_v((int8x16_t)__s0, __p1, 32); \
	26155	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vrshrq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 32)); \
26135	26156	__ret; \
26136	26157	})
26137	26158	#else
26138	26159	#define vrshrq_n_s8(__p0, __p1) __extension__ ({ \
26139	26160	int8x16_t __ret; \
26140	26161	int8x16_t __s0 = __p0; \
26141		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
26142		__ret = (int8x16_t) __builtin_neon_vrshrq_n_v((int8x16_t)__rev0, __p1, 32); \
26143		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	26162	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	26163	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vrshrq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 32)); \
	26164	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
26144	26165	__ret; \
26145	26166	})
26146	26167	#endif
...	...	@@ -26149,16 +26170,16 @@ __ai __attribute__((target("neon"))) int16x4_t vrshl_s16(int16x4_t __p0, int16x4
26149	26170	#define vrshrq_n_s32(__p0, __p1) __extension__ ({ \
26150	26171	int32x4_t __ret; \
26151	26172	int32x4_t __s0 = __p0; \
26152		__ret = (int32x4_t) __builtin_neon_vrshrq_n_v((int8x16_t)__s0, __p1, 34); \
	26173	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vrshrq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 34)); \
26153	26174	__ret; \
26154	26175	})
26155	26176	#else
26156	26177	#define vrshrq_n_s32(__p0, __p1) __extension__ ({ \
26157	26178	int32x4_t __ret; \
26158	26179	int32x4_t __s0 = __p0; \
26159		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
26160		__ret = (int32x4_t) __builtin_neon_vrshrq_n_v((int8x16_t)__rev0, __p1, 34); \
26161		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	26180	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	26181	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vrshrq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 34)); \
	26182	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
26162	26183	__ret; \
26163	26184	})
26164	26185	#endif
...	...	@@ -26167,16 +26188,16 @@ __ai __attribute__((target("neon"))) int16x4_t vrshl_s16(int16x4_t __p0, int16x4
26167	26188	#define vrshrq_n_s64(__p0, __p1) __extension__ ({ \
26168	26189	int64x2_t __ret; \
26169	26190	int64x2_t __s0 = __p0; \
26170		__ret = (int64x2_t) __builtin_neon_vrshrq_n_v((int8x16_t)__s0, __p1, 35); \
	26191	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vrshrq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 35)); \
26171	26192	__ret; \
26172	26193	})
26173	26194	#else
26174	26195	#define vrshrq_n_s64(__p0, __p1) __extension__ ({ \
26175	26196	int64x2_t __ret; \
26176	26197	int64x2_t __s0 = __p0; \
26177		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
26178		__ret = (int64x2_t) __builtin_neon_vrshrq_n_v((int8x16_t)__rev0, __p1, 35); \
26179		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	26198	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	26199	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vrshrq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 35)); \
	26200	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
26180	26201	__ret; \
26181	26202	})
26182	26203	#endif
...	...	@@ -26185,16 +26206,16 @@ __ai __attribute__((target("neon"))) int16x4_t vrshl_s16(int16x4_t __p0, int16x4
26185	26206	#define vrshrq_n_s16(__p0, __p1) __extension__ ({ \
26186	26207	int16x8_t __ret; \
26187	26208	int16x8_t __s0 = __p0; \
26188		__ret = (int16x8_t) __builtin_neon_vrshrq_n_v((int8x16_t)__s0, __p1, 33); \
	26209	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vrshrq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 33)); \
26189	26210	__ret; \
26190	26211	})
26191	26212	#else
26192	26213	#define vrshrq_n_s16(__p0, __p1) __extension__ ({ \
26193	26214	int16x8_t __ret; \
26194	26215	int16x8_t __s0 = __p0; \
26195		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
26196		__ret = (int16x8_t) __builtin_neon_vrshrq_n_v((int8x16_t)__rev0, __p1, 33); \
26197		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	26216	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	26217	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vrshrq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 33)); \
	26218	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
26198	26219	__ret; \
26199	26220	})
26200	26221	#endif
...	...	@@ -26203,16 +26224,16 @@ __ai __attribute__((target("neon"))) int16x4_t vrshl_s16(int16x4_t __p0, int16x4
26203	26224	#define vrshr_n_u8(__p0, __p1) __extension__ ({ \
26204	26225	uint8x8_t __ret; \
26205	26226	uint8x8_t __s0 = __p0; \
26206		__ret = (uint8x8_t) __builtin_neon_vrshr_n_v((int8x8_t)__s0, __p1, 16); \
	26227	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vrshr_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 16)); \
26207	26228	__ret; \
26208	26229	})
26209	26230	#else
26210	26231	#define vrshr_n_u8(__p0, __p1) __extension__ ({ \
26211	26232	uint8x8_t __ret; \
26212	26233	uint8x8_t __s0 = __p0; \
26213		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
26214		__ret = (uint8x8_t) __builtin_neon_vrshr_n_v((int8x8_t)__rev0, __p1, 16); \
26215		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	26234	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	26235	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vrshr_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 16)); \
	26236	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
26216	26237	__ret; \
26217	26238	})
26218	26239	#endif
...	...	@@ -26221,16 +26242,16 @@ __ai __attribute__((target("neon"))) int16x4_t vrshl_s16(int16x4_t __p0, int16x4
26221	26242	#define vrshr_n_u32(__p0, __p1) __extension__ ({ \
26222	26243	uint32x2_t __ret; \
26223	26244	uint32x2_t __s0 = __p0; \
26224		__ret = (uint32x2_t) __builtin_neon_vrshr_n_v((int8x8_t)__s0, __p1, 18); \
	26245	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vrshr_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 18)); \
26225	26246	__ret; \
26226	26247	})
26227	26248	#else
26228	26249	#define vrshr_n_u32(__p0, __p1) __extension__ ({ \
26229	26250	uint32x2_t __ret; \
26230	26251	uint32x2_t __s0 = __p0; \
26231		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
26232		__ret = (uint32x2_t) __builtin_neon_vrshr_n_v((int8x8_t)__rev0, __p1, 18); \
26233		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	26252	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	26253	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vrshr_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 18)); \
	26254	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
26234	26255	__ret; \
26235	26256	})
26236	26257	#endif
...	...	@@ -26238,23 +26259,23 @@ __ai __attribute__((target("neon"))) int16x4_t vrshl_s16(int16x4_t __p0, int16x4
26238	26259	#define vrshr_n_u64(__p0, __p1) __extension__ ({ \
26239	26260	uint64x1_t __ret; \
26240	26261	uint64x1_t __s0 = __p0; \
26241		__ret = (uint64x1_t) __builtin_neon_vrshr_n_v((int8x8_t)__s0, __p1, 19); \
	26262	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vrshr_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 19)); \
26242	26263	__ret; \
26243	26264	})
26244	26265	#ifdef __LITTLE_ENDIAN__
26245	26266	#define vrshr_n_u16(__p0, __p1) __extension__ ({ \
26246	26267	uint16x4_t __ret; \
26247	26268	uint16x4_t __s0 = __p0; \
26248		__ret = (uint16x4_t) __builtin_neon_vrshr_n_v((int8x8_t)__s0, __p1, 17); \
	26269	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vrshr_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 17)); \
26249	26270	__ret; \
26250	26271	})
26251	26272	#else
26252	26273	#define vrshr_n_u16(__p0, __p1) __extension__ ({ \
26253	26274	uint16x4_t __ret; \
26254	26275	uint16x4_t __s0 = __p0; \
26255		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
26256		__ret = (uint16x4_t) __builtin_neon_vrshr_n_v((int8x8_t)__rev0, __p1, 17); \
26257		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	26276	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	26277	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vrshr_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 17)); \
	26278	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
26258	26279	__ret; \
26259	26280	})
26260	26281	#endif
...	...	@@ -26263,16 +26284,16 @@ __ai __attribute__((target("neon"))) int16x4_t vrshl_s16(int16x4_t __p0, int16x4
26263	26284	#define vrshr_n_s8(__p0, __p1) __extension__ ({ \
26264	26285	int8x8_t __ret; \
26265	26286	int8x8_t __s0 = __p0; \
26266		__ret = (int8x8_t) __builtin_neon_vrshr_n_v((int8x8_t)__s0, __p1, 0); \
	26287	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vrshr_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 0)); \
26267	26288	__ret; \
26268	26289	})
26269	26290	#else
26270	26291	#define vrshr_n_s8(__p0, __p1) __extension__ ({ \
26271	26292	int8x8_t __ret; \
26272	26293	int8x8_t __s0 = __p0; \
26273		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
26274		__ret = (int8x8_t) __builtin_neon_vrshr_n_v((int8x8_t)__rev0, __p1, 0); \
26275		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	26294	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	26295	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vrshr_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 0)); \
	26296	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
26276	26297	__ret; \
26277	26298	})
26278	26299	#endif
...	...	@@ -26281,16 +26302,16 @@ __ai __attribute__((target("neon"))) int16x4_t vrshl_s16(int16x4_t __p0, int16x4
26281	26302	#define vrshr_n_s32(__p0, __p1) __extension__ ({ \
26282	26303	int32x2_t __ret; \
26283	26304	int32x2_t __s0 = __p0; \
26284		__ret = (int32x2_t) __builtin_neon_vrshr_n_v((int8x8_t)__s0, __p1, 2); \
	26305	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vrshr_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 2)); \
26285	26306	__ret; \
26286	26307	})
26287	26308	#else
26288	26309	#define vrshr_n_s32(__p0, __p1) __extension__ ({ \
26289	26310	int32x2_t __ret; \
26290	26311	int32x2_t __s0 = __p0; \
26291		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
26292		__ret = (int32x2_t) __builtin_neon_vrshr_n_v((int8x8_t)__rev0, __p1, 2); \
26293		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	26312	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	26313	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vrshr_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 2)); \
	26314	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
26294	26315	__ret; \
26295	26316	})
26296	26317	#endif
...	...	@@ -26298,23 +26319,23 @@ __ai __attribute__((target("neon"))) int16x4_t vrshl_s16(int16x4_t __p0, int16x4
26298	26319	#define vrshr_n_s64(__p0, __p1) __extension__ ({ \
26299	26320	int64x1_t __ret; \
26300	26321	int64x1_t __s0 = __p0; \
26301		__ret = (int64x1_t) __builtin_neon_vrshr_n_v((int8x8_t)__s0, __p1, 3); \
	26322	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vrshr_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 3)); \
26302	26323	__ret; \
26303	26324	})
26304	26325	#ifdef __LITTLE_ENDIAN__
26305	26326	#define vrshr_n_s16(__p0, __p1) __extension__ ({ \
26306	26327	int16x4_t __ret; \
26307	26328	int16x4_t __s0 = __p0; \
26308		__ret = (int16x4_t) __builtin_neon_vrshr_n_v((int8x8_t)__s0, __p1, 1); \
	26329	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vrshr_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 1)); \
26309	26330	__ret; \
26310	26331	})
26311	26332	#else
26312	26333	#define vrshr_n_s16(__p0, __p1) __extension__ ({ \
26313	26334	int16x4_t __ret; \
26314	26335	int16x4_t __s0 = __p0; \
26315		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
26316		__ret = (int16x4_t) __builtin_neon_vrshr_n_v((int8x8_t)__rev0, __p1, 1); \
26317		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	26336	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	26337	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vrshr_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 1)); \
	26338	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
26318	26339	__ret; \
26319	26340	})
26320	26341	#endif
...	...	@@ -26323,22 +26344,22 @@ __ai __attribute__((target("neon"))) int16x4_t vrshl_s16(int16x4_t __p0, int16x4
26323	26344	#define vrshrn_n_u32(__p0, __p1) __extension__ ({ \
26324	26345	uint16x4_t __ret; \
26325	26346	uint32x4_t __s0 = __p0; \
26326		__ret = (uint16x4_t) __builtin_neon_vrshrn_n_v((int8x16_t)__s0, __p1, 17); \
	26347	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vrshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 17)); \
26327	26348	__ret; \
26328	26349	})
26329	26350	#else
26330	26351	#define vrshrn_n_u32(__p0, __p1) __extension__ ({ \
26331	26352	uint16x4_t __ret; \
26332	26353	uint32x4_t __s0 = __p0; \
26333		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
26334		__ret = (uint16x4_t) __builtin_neon_vrshrn_n_v((int8x16_t)__rev0, __p1, 17); \
26335		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	26354	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	26355	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vrshrn_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 17)); \
	26356	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
26336	26357	__ret; \
26337	26358	})
26338	26359	#define __noswap_vrshrn_n_u32(__p0, __p1) __extension__ ({ \
26339	26360	uint16x4_t __ret; \
26340	26361	uint32x4_t __s0 = __p0; \
26341		__ret = (uint16x4_t) __builtin_neon_vrshrn_n_v((int8x16_t)__s0, __p1, 17); \
	26362	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vrshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 17)); \
26342	26363	__ret; \
26343	26364	})
26344	26365	#endif
...	...	@@ -26347,22 +26368,22 @@ __ai __attribute__((target("neon"))) int16x4_t vrshl_s16(int16x4_t __p0, int16x4
26347	26368	#define vrshrn_n_u64(__p0, __p1) __extension__ ({ \
26348	26369	uint32x2_t __ret; \
26349	26370	uint64x2_t __s0 = __p0; \
26350		__ret = (uint32x2_t) __builtin_neon_vrshrn_n_v((int8x16_t)__s0, __p1, 18); \
	26371	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vrshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 18)); \
26351	26372	__ret; \
26352	26373	})
26353	26374	#else
26354	26375	#define vrshrn_n_u64(__p0, __p1) __extension__ ({ \
26355	26376	uint32x2_t __ret; \
26356	26377	uint64x2_t __s0 = __p0; \
26357		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
26358		__ret = (uint32x2_t) __builtin_neon_vrshrn_n_v((int8x16_t)__rev0, __p1, 18); \
26359		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	26378	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	26379	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vrshrn_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 18)); \
	26380	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
26360	26381	__ret; \
26361	26382	})
26362	26383	#define __noswap_vrshrn_n_u64(__p0, __p1) __extension__ ({ \
26363	26384	uint32x2_t __ret; \
26364	26385	uint64x2_t __s0 = __p0; \
26365		__ret = (uint32x2_t) __builtin_neon_vrshrn_n_v((int8x16_t)__s0, __p1, 18); \
	26386	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vrshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 18)); \
26366	26387	__ret; \
26367	26388	})
26368	26389	#endif
...	...	@@ -26371,22 +26392,22 @@ __ai __attribute__((target("neon"))) int16x4_t vrshl_s16(int16x4_t __p0, int16x4
26371	26392	#define vrshrn_n_u16(__p0, __p1) __extension__ ({ \
26372	26393	uint8x8_t __ret; \
26373	26394	uint16x8_t __s0 = __p0; \
26374		__ret = (uint8x8_t) __builtin_neon_vrshrn_n_v((int8x16_t)__s0, __p1, 16); \
	26395	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vrshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 16)); \
26375	26396	__ret; \
26376	26397	})
26377	26398	#else
26378	26399	#define vrshrn_n_u16(__p0, __p1) __extension__ ({ \
26379	26400	uint8x8_t __ret; \
26380	26401	uint16x8_t __s0 = __p0; \
26381		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
26382		__ret = (uint8x8_t) __builtin_neon_vrshrn_n_v((int8x16_t)__rev0, __p1, 16); \
26383		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	26402	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	26403	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vrshrn_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 16)); \
	26404	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
26384	26405	__ret; \
26385	26406	})
26386	26407	#define __noswap_vrshrn_n_u16(__p0, __p1) __extension__ ({ \
26387	26408	uint8x8_t __ret; \
26388	26409	uint16x8_t __s0 = __p0; \
26389		__ret = (uint8x8_t) __builtin_neon_vrshrn_n_v((int8x16_t)__s0, __p1, 16); \
	26410	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vrshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 16)); \
26390	26411	__ret; \
26391	26412	})
26392	26413	#endif
...	...	@@ -26395,22 +26416,22 @@ __ai __attribute__((target("neon"))) int16x4_t vrshl_s16(int16x4_t __p0, int16x4
26395	26416	#define vrshrn_n_s32(__p0, __p1) __extension__ ({ \
26396	26417	int16x4_t __ret; \
26397	26418	int32x4_t __s0 = __p0; \
26398		__ret = (int16x4_t) __builtin_neon_vrshrn_n_v((int8x16_t)__s0, __p1, 1); \
	26419	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vrshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 1)); \
26399	26420	__ret; \
26400	26421	})
26401	26422	#else
26402	26423	#define vrshrn_n_s32(__p0, __p1) __extension__ ({ \
26403	26424	int16x4_t __ret; \
26404	26425	int32x4_t __s0 = __p0; \
26405		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
26406		__ret = (int16x4_t) __builtin_neon_vrshrn_n_v((int8x16_t)__rev0, __p1, 1); \
26407		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	26426	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	26427	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vrshrn_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 1)); \
	26428	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
26408	26429	__ret; \
26409	26430	})
26410	26431	#define __noswap_vrshrn_n_s32(__p0, __p1) __extension__ ({ \
26411	26432	int16x4_t __ret; \
26412	26433	int32x4_t __s0 = __p0; \
26413		__ret = (int16x4_t) __builtin_neon_vrshrn_n_v((int8x16_t)__s0, __p1, 1); \
	26434	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vrshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 1)); \
26414	26435	__ret; \
26415	26436	})
26416	26437	#endif
...	...	@@ -26419,22 +26440,22 @@ __ai __attribute__((target("neon"))) int16x4_t vrshl_s16(int16x4_t __p0, int16x4
26419	26440	#define vrshrn_n_s64(__p0, __p1) __extension__ ({ \
26420	26441	int32x2_t __ret; \
26421	26442	int64x2_t __s0 = __p0; \
26422		__ret = (int32x2_t) __builtin_neon_vrshrn_n_v((int8x16_t)__s0, __p1, 2); \
	26443	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vrshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 2)); \
26423	26444	__ret; \
26424	26445	})
26425	26446	#else
26426	26447	#define vrshrn_n_s64(__p0, __p1) __extension__ ({ \
26427	26448	int32x2_t __ret; \
26428	26449	int64x2_t __s0 = __p0; \
26429		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
26430		__ret = (int32x2_t) __builtin_neon_vrshrn_n_v((int8x16_t)__rev0, __p1, 2); \
26431		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	26450	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	26451	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vrshrn_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 2)); \
	26452	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
26432	26453	__ret; \
26433	26454	})
26434	26455	#define __noswap_vrshrn_n_s64(__p0, __p1) __extension__ ({ \
26435	26456	int32x2_t __ret; \
26436	26457	int64x2_t __s0 = __p0; \
26437		__ret = (int32x2_t) __builtin_neon_vrshrn_n_v((int8x16_t)__s0, __p1, 2); \
	26458	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vrshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 2)); \
26438	26459	__ret; \
26439	26460	})
26440	26461	#endif
...	...	@@ -26443,22 +26464,22 @@ __ai __attribute__((target("neon"))) int16x4_t vrshl_s16(int16x4_t __p0, int16x4
26443	26464	#define vrshrn_n_s16(__p0, __p1) __extension__ ({ \
26444	26465	int8x8_t __ret; \
26445	26466	int16x8_t __s0 = __p0; \
26446		__ret = (int8x8_t) __builtin_neon_vrshrn_n_v((int8x16_t)__s0, __p1, 0); \
	26467	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vrshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 0)); \
26447	26468	__ret; \
26448	26469	})
26449	26470	#else
26450	26471	#define vrshrn_n_s16(__p0, __p1) __extension__ ({ \
26451	26472	int8x8_t __ret; \
26452	26473	int16x8_t __s0 = __p0; \
26453		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
26454		__ret = (int8x8_t) __builtin_neon_vrshrn_n_v((int8x16_t)__rev0, __p1, 0); \
26455		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	26474	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	26475	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vrshrn_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 0)); \
	26476	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
26456	26477	__ret; \
26457	26478	})
26458	26479	#define __noswap_vrshrn_n_s16(__p0, __p1) __extension__ ({ \
26459	26480	int8x8_t __ret; \
26460	26481	int16x8_t __s0 = __p0; \
26461		__ret = (int8x8_t) __builtin_neon_vrshrn_n_v((int8x16_t)__s0, __p1, 0); \
	26482	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vrshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 0)); \
26462	26483	__ret; \
26463	26484	})
26464	26485	#endif
...	...	@@ -26466,15 +26487,15 @@ __ai __attribute__((target("neon"))) int16x4_t vrshl_s16(int16x4_t __p0, int16x4
26466	26487	#ifdef __LITTLE_ENDIAN__
26467	26488	__ai __attribute__((target("neon"))) uint32x4_t vrsqrteq_u32(uint32x4_t __p0) {
26468	26489	uint32x4_t __ret;
26469		__ret = (uint32x4_t) __builtin_neon_vrsqrteq_v((int8x16_t)__p0, 50);
	26490	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vrsqrteq_v(__builtin_bit_cast(int8x16_t, __p0), 50));
26470	26491	return __ret;
26471	26492	}
26472	26493	#else
26473	26494	__ai __attribute__((target("neon"))) uint32x4_t vrsqrteq_u32(uint32x4_t __p0) {
26474	26495	uint32x4_t __ret;
26475		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
26476		__ret = (uint32x4_t) __builtin_neon_vrsqrteq_v((int8x16_t)__rev0, 50);
26477		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	26496	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	26497	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vrsqrteq_v(__builtin_bit_cast(int8x16_t, __rev0), 50));
	26498	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
26478	26499	return __ret;
26479	26500	}
26480	26501	#endif
...	...	@@ -26482,15 +26503,15 @@ __ai __attribute__((target("neon"))) uint32x4_t vrsqrteq_u32(uint32x4_t __p0) {
26482	26503	#ifdef __LITTLE_ENDIAN__
26483	26504	__ai __attribute__((target("neon"))) float32x4_t vrsqrteq_f32(float32x4_t __p0) {
26484	26505	float32x4_t __ret;
26485		__ret = (float32x4_t) __builtin_neon_vrsqrteq_v((int8x16_t)__p0, 41);
	26506	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vrsqrteq_v(__builtin_bit_cast(int8x16_t, __p0), 41));
26486	26507	return __ret;
26487	26508	}
26488	26509	#else
26489	26510	__ai __attribute__((target("neon"))) float32x4_t vrsqrteq_f32(float32x4_t __p0) {
26490	26511	float32x4_t __ret;
26491		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
26492		__ret = (float32x4_t) __builtin_neon_vrsqrteq_v((int8x16_t)__rev0, 41);
26493		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	26512	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	26513	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vrsqrteq_v(__builtin_bit_cast(int8x16_t, __rev0), 41));
	26514	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
26494	26515	return __ret;
26495	26516	}
26496	26517	#endif
...	...	@@ -26498,15 +26519,15 @@ __ai __attribute__((target("neon"))) float32x4_t vrsqrteq_f32(float32x4_t __p0)
26498	26519	#ifdef __LITTLE_ENDIAN__
26499	26520	__ai __attribute__((target("neon"))) uint32x2_t vrsqrte_u32(uint32x2_t __p0) {
26500	26521	uint32x2_t __ret;
26501		__ret = (uint32x2_t) __builtin_neon_vrsqrte_v((int8x8_t)__p0, 18);
	26522	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vrsqrte_v(__builtin_bit_cast(int8x8_t, __p0), 18));
26502	26523	return __ret;
26503	26524	}
26504	26525	#else
26505	26526	__ai __attribute__((target("neon"))) uint32x2_t vrsqrte_u32(uint32x2_t __p0) {
26506	26527	uint32x2_t __ret;
26507		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
26508		__ret = (uint32x2_t) __builtin_neon_vrsqrte_v((int8x8_t)__rev0, 18);
26509		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	26528	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	26529	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vrsqrte_v(__builtin_bit_cast(int8x8_t, __rev0), 18));
	26530	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
26510	26531	return __ret;
26511	26532	}
26512	26533	#endif
...	...	@@ -26514,15 +26535,15 @@ __ai __attribute__((target("neon"))) uint32x2_t vrsqrte_u32(uint32x2_t __p0) {
26514	26535	#ifdef __LITTLE_ENDIAN__
26515	26536	__ai __attribute__((target("neon"))) float32x2_t vrsqrte_f32(float32x2_t __p0) {
26516	26537	float32x2_t __ret;
26517		__ret = (float32x2_t) __builtin_neon_vrsqrte_v((int8x8_t)__p0, 9);
	26538	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vrsqrte_v(__builtin_bit_cast(int8x8_t, __p0), 9));
26518	26539	return __ret;
26519	26540	}
26520	26541	#else
26521	26542	__ai __attribute__((target("neon"))) float32x2_t vrsqrte_f32(float32x2_t __p0) {
26522	26543	float32x2_t __ret;
26523		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
26524		__ret = (float32x2_t) __builtin_neon_vrsqrte_v((int8x8_t)__rev0, 9);
26525		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	26544	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	26545	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vrsqrte_v(__builtin_bit_cast(int8x8_t, __rev0), 9));
	26546	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
26526	26547	return __ret;
26527	26548	}
26528	26549	#endif
...	...	@@ -26530,16 +26551,16 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrte_f32(float32x2_t __p0) {
26530	26551	#ifdef __LITTLE_ENDIAN__
26531	26552	__ai __attribute__((target("neon"))) float32x4_t vrsqrtsq_f32(float32x4_t __p0, float32x4_t __p1) {
26532	26553	float32x4_t __ret;
26533		__ret = (float32x4_t) __builtin_neon_vrsqrtsq_v((int8x16_t)__p0, (int8x16_t)__p1, 41);
	26554	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vrsqrtsq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 41));
26534	26555	return __ret;
26535	26556	}
26536	26557	#else
26537	26558	__ai __attribute__((target("neon"))) float32x4_t vrsqrtsq_f32(float32x4_t __p0, float32x4_t __p1) {
26538	26559	float32x4_t __ret;
26539		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
26540		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
26541		__ret = (float32x4_t) __builtin_neon_vrsqrtsq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 41);
26542		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	26560	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	26561	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	26562	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vrsqrtsq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 41));
	26563	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
26543	26564	return __ret;
26544	26565	}
26545	26566	#endif
...	...	@@ -26547,16 +26568,16 @@ __ai __attribute__((target("neon"))) float32x4_t vrsqrtsq_f32(float32x4_t __p0,
26547	26568	#ifdef __LITTLE_ENDIAN__
26548	26569	__ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, float32x2_t __p1) {
26549	26570	float32x2_t __ret;
26550		__ret = (float32x2_t) __builtin_neon_vrsqrts_v((int8x8_t)__p0, (int8x8_t)__p1, 9);
	26571	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vrsqrts_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 9));
26551	26572	return __ret;
26552	26573	}
26553	26574	#else
26554	26575	__ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, float32x2_t __p1) {
26555	26576	float32x2_t __ret;
26556		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
26557		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
26558		__ret = (float32x2_t) __builtin_neon_vrsqrts_v((int8x8_t)__rev0, (int8x8_t)__rev1, 9);
26559		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	26577	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	26578	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	26579	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vrsqrts_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 9));
	26580	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
26560	26581	return __ret;
26561	26582	}
26562	26583	#endif
...	...	@@ -26566,7 +26587,7 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26566	26587	uint8x16_t __ret; \
26567	26588	uint8x16_t __s0 = __p0; \
26568	26589	uint8x16_t __s1 = __p1; \
26569		__ret = (uint8x16_t) __builtin_neon_vrsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 48); \
	26590	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vrsraq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 48)); \
26570	26591	__ret; \
26571	26592	})
26572	26593	#else
...	...	@@ -26574,10 +26595,10 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26574	26595	uint8x16_t __ret; \
26575	26596	uint8x16_t __s0 = __p0; \
26576	26597	uint8x16_t __s1 = __p1; \
26577		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
26578		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
26579		__ret = (uint8x16_t) __builtin_neon_vrsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 48); \
26580		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	26598	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	26599	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	26600	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vrsraq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 48)); \
	26601	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
26581	26602	__ret; \
26582	26603	})
26583	26604	#endif
...	...	@@ -26587,7 +26608,7 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26587	26608	uint32x4_t __ret; \
26588	26609	uint32x4_t __s0 = __p0; \
26589	26610	uint32x4_t __s1 = __p1; \
26590		__ret = (uint32x4_t) __builtin_neon_vrsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 50); \
	26611	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vrsraq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 50)); \
26591	26612	__ret; \
26592	26613	})
26593	26614	#else
...	...	@@ -26595,10 +26616,10 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26595	26616	uint32x4_t __ret; \
26596	26617	uint32x4_t __s0 = __p0; \
26597	26618	uint32x4_t __s1 = __p1; \
26598		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
26599		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
26600		__ret = (uint32x4_t) __builtin_neon_vrsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 50); \
26601		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	26619	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	26620	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	26621	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vrsraq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 50)); \
	26622	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
26602	26623	__ret; \
26603	26624	})
26604	26625	#endif
...	...	@@ -26608,7 +26629,7 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26608	26629	uint64x2_t __ret; \
26609	26630	uint64x2_t __s0 = __p0; \
26610	26631	uint64x2_t __s1 = __p1; \
26611		__ret = (uint64x2_t) __builtin_neon_vrsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 51); \
	26632	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vrsraq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 51)); \
26612	26633	__ret; \
26613	26634	})
26614	26635	#else
...	...	@@ -26616,10 +26637,10 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26616	26637	uint64x2_t __ret; \
26617	26638	uint64x2_t __s0 = __p0; \
26618	26639	uint64x2_t __s1 = __p1; \
26619		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
26620		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
26621		__ret = (uint64x2_t) __builtin_neon_vrsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 51); \
26622		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	26640	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	26641	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	26642	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vrsraq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 51)); \
	26643	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
26623	26644	__ret; \
26624	26645	})
26625	26646	#endif
...	...	@@ -26629,7 +26650,7 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26629	26650	uint16x8_t __ret; \
26630	26651	uint16x8_t __s0 = __p0; \
26631	26652	uint16x8_t __s1 = __p1; \
26632		__ret = (uint16x8_t) __builtin_neon_vrsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 49); \
	26653	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vrsraq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 49)); \
26633	26654	__ret; \
26634	26655	})
26635	26656	#else
...	...	@@ -26637,10 +26658,10 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26637	26658	uint16x8_t __ret; \
26638	26659	uint16x8_t __s0 = __p0; \
26639	26660	uint16x8_t __s1 = __p1; \
26640		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
26641		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
26642		__ret = (uint16x8_t) __builtin_neon_vrsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 49); \
26643		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	26661	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	26662	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	26663	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vrsraq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 49)); \
	26664	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
26644	26665	__ret; \
26645	26666	})
26646	26667	#endif
...	...	@@ -26650,7 +26671,7 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26650	26671	int8x16_t __ret; \
26651	26672	int8x16_t __s0 = __p0; \
26652	26673	int8x16_t __s1 = __p1; \
26653		__ret = (int8x16_t) __builtin_neon_vrsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 32); \
	26674	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vrsraq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 32)); \
26654	26675	__ret; \
26655	26676	})
26656	26677	#else
...	...	@@ -26658,10 +26679,10 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26658	26679	int8x16_t __ret; \
26659	26680	int8x16_t __s0 = __p0; \
26660	26681	int8x16_t __s1 = __p1; \
26661		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
26662		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
26663		__ret = (int8x16_t) __builtin_neon_vrsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 32); \
26664		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	26682	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	26683	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	26684	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vrsraq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 32)); \
	26685	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
26665	26686	__ret; \
26666	26687	})
26667	26688	#endif
...	...	@@ -26671,7 +26692,7 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26671	26692	int32x4_t __ret; \
26672	26693	int32x4_t __s0 = __p0; \
26673	26694	int32x4_t __s1 = __p1; \
26674		__ret = (int32x4_t) __builtin_neon_vrsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 34); \
	26695	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vrsraq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 34)); \
26675	26696	__ret; \
26676	26697	})
26677	26698	#else
...	...	@@ -26679,10 +26700,10 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26679	26700	int32x4_t __ret; \
26680	26701	int32x4_t __s0 = __p0; \
26681	26702	int32x4_t __s1 = __p1; \
26682		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
26683		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
26684		__ret = (int32x4_t) __builtin_neon_vrsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 34); \
26685		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	26703	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	26704	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	26705	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vrsraq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 34)); \
	26706	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
26686	26707	__ret; \
26687	26708	})
26688	26709	#endif
...	...	@@ -26692,7 +26713,7 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26692	26713	int64x2_t __ret; \
26693	26714	int64x2_t __s0 = __p0; \
26694	26715	int64x2_t __s1 = __p1; \
26695		__ret = (int64x2_t) __builtin_neon_vrsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 35); \
	26716	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vrsraq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 35)); \
26696	26717	__ret; \
26697	26718	})
26698	26719	#else
...	...	@@ -26700,10 +26721,10 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26700	26721	int64x2_t __ret; \
26701	26722	int64x2_t __s0 = __p0; \
26702	26723	int64x2_t __s1 = __p1; \
26703		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
26704		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
26705		__ret = (int64x2_t) __builtin_neon_vrsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 35); \
26706		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	26724	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	26725	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	26726	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vrsraq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 35)); \
	26727	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
26707	26728	__ret; \
26708	26729	})
26709	26730	#endif
...	...	@@ -26713,7 +26734,7 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26713	26734	int16x8_t __ret; \
26714	26735	int16x8_t __s0 = __p0; \
26715	26736	int16x8_t __s1 = __p1; \
26716		__ret = (int16x8_t) __builtin_neon_vrsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 33); \
	26737	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vrsraq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 33)); \
26717	26738	__ret; \
26718	26739	})
26719	26740	#else
...	...	@@ -26721,10 +26742,10 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26721	26742	int16x8_t __ret; \
26722	26743	int16x8_t __s0 = __p0; \
26723	26744	int16x8_t __s1 = __p1; \
26724		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
26725		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
26726		__ret = (int16x8_t) __builtin_neon_vrsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 33); \
26727		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	26745	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	26746	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	26747	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vrsraq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 33)); \
	26748	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
26728	26749	__ret; \
26729	26750	})
26730	26751	#endif
...	...	@@ -26734,7 +26755,7 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26734	26755	uint8x8_t __ret; \
26735	26756	uint8x8_t __s0 = __p0; \
26736	26757	uint8x8_t __s1 = __p1; \
26737		__ret = (uint8x8_t) __builtin_neon_vrsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 16); \
	26758	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vrsra_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 16)); \
26738	26759	__ret; \
26739	26760	})
26740	26761	#else
...	...	@@ -26742,10 +26763,10 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26742	26763	uint8x8_t __ret; \
26743	26764	uint8x8_t __s0 = __p0; \
26744	26765	uint8x8_t __s1 = __p1; \
26745		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
26746		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
26747		__ret = (uint8x8_t) __builtin_neon_vrsra_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 16); \
26748		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	26766	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	26767	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	26768	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vrsra_n_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 16)); \
	26769	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
26749	26770	__ret; \
26750	26771	})
26751	26772	#endif
...	...	@@ -26755,7 +26776,7 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26755	26776	uint32x2_t __ret; \
26756	26777	uint32x2_t __s0 = __p0; \
26757	26778	uint32x2_t __s1 = __p1; \
26758		__ret = (uint32x2_t) __builtin_neon_vrsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 18); \
	26779	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vrsra_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 18)); \
26759	26780	__ret; \
26760	26781	})
26761	26782	#else
...	...	@@ -26763,10 +26784,10 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26763	26784	uint32x2_t __ret; \
26764	26785	uint32x2_t __s0 = __p0; \
26765	26786	uint32x2_t __s1 = __p1; \
26766		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
26767		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
26768		__ret = (uint32x2_t) __builtin_neon_vrsra_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 18); \
26769		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	26787	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	26788	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_32); \
	26789	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vrsra_n_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 18)); \
	26790	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
26770	26791	__ret; \
26771	26792	})
26772	26793	#endif
...	...	@@ -26775,7 +26796,7 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26775	26796	uint64x1_t __ret; \
26776	26797	uint64x1_t __s0 = __p0; \
26777	26798	uint64x1_t __s1 = __p1; \
26778		__ret = (uint64x1_t) __builtin_neon_vrsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 19); \
	26799	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vrsra_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 19)); \
26779	26800	__ret; \
26780	26801	})
26781	26802	#ifdef __LITTLE_ENDIAN__
...	...	@@ -26783,7 +26804,7 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26783	26804	uint16x4_t __ret; \
26784	26805	uint16x4_t __s0 = __p0; \
26785	26806	uint16x4_t __s1 = __p1; \
26786		__ret = (uint16x4_t) __builtin_neon_vrsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 17); \
	26807	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vrsra_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 17)); \
26787	26808	__ret; \
26788	26809	})
26789	26810	#else
...	...	@@ -26791,10 +26812,10 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26791	26812	uint16x4_t __ret; \
26792	26813	uint16x4_t __s0 = __p0; \
26793	26814	uint16x4_t __s1 = __p1; \
26794		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
26795		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
26796		__ret = (uint16x4_t) __builtin_neon_vrsra_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 17); \
26797		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	26815	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	26816	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	26817	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vrsra_n_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 17)); \
	26818	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
26798	26819	__ret; \
26799	26820	})
26800	26821	#endif
...	...	@@ -26804,7 +26825,7 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26804	26825	int8x8_t __ret; \
26805	26826	int8x8_t __s0 = __p0; \
26806	26827	int8x8_t __s1 = __p1; \
26807		__ret = (int8x8_t) __builtin_neon_vrsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 0); \
	26828	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vrsra_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 0)); \
26808	26829	__ret; \
26809	26830	})
26810	26831	#else
...	...	@@ -26812,10 +26833,10 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26812	26833	int8x8_t __ret; \
26813	26834	int8x8_t __s0 = __p0; \
26814	26835	int8x8_t __s1 = __p1; \
26815		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
26816		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
26817		__ret = (int8x8_t) __builtin_neon_vrsra_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 0); \
26818		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	26836	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	26837	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	26838	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vrsra_n_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 0)); \
	26839	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
26819	26840	__ret; \
26820	26841	})
26821	26842	#endif
...	...	@@ -26825,7 +26846,7 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26825	26846	int32x2_t __ret; \
26826	26847	int32x2_t __s0 = __p0; \
26827	26848	int32x2_t __s1 = __p1; \
26828		__ret = (int32x2_t) __builtin_neon_vrsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 2); \
	26849	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vrsra_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 2)); \
26829	26850	__ret; \
26830	26851	})
26831	26852	#else
...	...	@@ -26833,10 +26854,10 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26833	26854	int32x2_t __ret; \
26834	26855	int32x2_t __s0 = __p0; \
26835	26856	int32x2_t __s1 = __p1; \
26836		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
26837		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
26838		__ret = (int32x2_t) __builtin_neon_vrsra_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 2); \
26839		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	26857	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	26858	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_32); \
	26859	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vrsra_n_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 2)); \
	26860	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
26840	26861	__ret; \
26841	26862	})
26842	26863	#endif
...	...	@@ -26845,7 +26866,7 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26845	26866	int64x1_t __ret; \
26846	26867	int64x1_t __s0 = __p0; \
26847	26868	int64x1_t __s1 = __p1; \
26848		__ret = (int64x1_t) __builtin_neon_vrsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 3); \
	26869	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vrsra_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 3)); \
26849	26870	__ret; \
26850	26871	})
26851	26872	#ifdef __LITTLE_ENDIAN__
...	...	@@ -26853,7 +26874,7 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26853	26874	int16x4_t __ret; \
26854	26875	int16x4_t __s0 = __p0; \
26855	26876	int16x4_t __s1 = __p1; \
26856		__ret = (int16x4_t) __builtin_neon_vrsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 1); \
	26877	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vrsra_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 1)); \
26857	26878	__ret; \
26858	26879	})
26859	26880	#else
...	...	@@ -26861,10 +26882,10 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26861	26882	int16x4_t __ret; \
26862	26883	int16x4_t __s0 = __p0; \
26863	26884	int16x4_t __s1 = __p1; \
26864		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
26865		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
26866		__ret = (int16x4_t) __builtin_neon_vrsra_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 1); \
26867		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	26885	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	26886	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	26887	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vrsra_n_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 1)); \
	26888	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
26868	26889	__ret; \
26869	26890	})
26870	26891	#endif
...	...	@@ -26872,21 +26893,21 @@ __ai __attribute__((target("neon"))) float32x2_t vrsqrts_f32(float32x2_t __p0, f
26872	26893	#ifdef __LITTLE_ENDIAN__
26873	26894	__ai __attribute__((target("neon"))) uint16x4_t vrsubhn_u32(uint32x4_t __p0, uint32x4_t __p1) {
26874	26895	uint16x4_t __ret;
26875		__ret = (uint16x4_t) __builtin_neon_vrsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 17);
	26896	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vrsubhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 17));
26876	26897	return __ret;
26877	26898	}
26878	26899	#else
26879	26900	__ai __attribute__((target("neon"))) uint16x4_t vrsubhn_u32(uint32x4_t __p0, uint32x4_t __p1) {
26880	26901	uint16x4_t __ret;
26881		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
26882		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
26883		__ret = (uint16x4_t) __builtin_neon_vrsubhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 17);
26884		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	26902	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	26903	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	26904	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vrsubhn_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 17));
	26905	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
26885	26906	return __ret;
26886	26907	}
26887	26908	__ai __attribute__((target("neon"))) uint16x4_t __noswap_vrsubhn_u32(uint32x4_t __p0, uint32x4_t __p1) {
26888	26909	uint16x4_t __ret;
26889		__ret = (uint16x4_t) __builtin_neon_vrsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 17);
	26910	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vrsubhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 17));
26890	26911	return __ret;
26891	26912	}
26892	26913	#endif
...	...	@@ -26894,21 +26915,21 @@ __ai __attribute__((target("neon"))) uint16x4_t __noswap_vrsubhn_u32(uint32x4_t
26894	26915	#ifdef __LITTLE_ENDIAN__
26895	26916	__ai __attribute__((target("neon"))) uint32x2_t vrsubhn_u64(uint64x2_t __p0, uint64x2_t __p1) {
26896	26917	uint32x2_t __ret;
26897		__ret = (uint32x2_t) __builtin_neon_vrsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 18);
	26918	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vrsubhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 18));
26898	26919	return __ret;
26899	26920	}
26900	26921	#else
26901	26922	__ai __attribute__((target("neon"))) uint32x2_t vrsubhn_u64(uint64x2_t __p0, uint64x2_t __p1) {
26902	26923	uint32x2_t __ret;
26903		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
26904		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
26905		__ret = (uint32x2_t) __builtin_neon_vrsubhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 18);
26906		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	26924	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	26925	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	26926	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vrsubhn_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 18));
	26927	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
26907	26928	return __ret;
26908	26929	}
26909	26930	__ai __attribute__((target("neon"))) uint32x2_t __noswap_vrsubhn_u64(uint64x2_t __p0, uint64x2_t __p1) {
26910	26931	uint32x2_t __ret;
26911		__ret = (uint32x2_t) __builtin_neon_vrsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 18);
	26932	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vrsubhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 18));
26912	26933	return __ret;
26913	26934	}
26914	26935	#endif
...	...	@@ -26916,21 +26937,21 @@ __ai __attribute__((target("neon"))) uint32x2_t __noswap_vrsubhn_u64(uint64x2_t
26916	26937	#ifdef __LITTLE_ENDIAN__
26917	26938	__ai __attribute__((target("neon"))) uint8x8_t vrsubhn_u16(uint16x8_t __p0, uint16x8_t __p1) {
26918	26939	uint8x8_t __ret;
26919		__ret = (uint8x8_t) __builtin_neon_vrsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 16);
	26940	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vrsubhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 16));
26920	26941	return __ret;
26921	26942	}
26922	26943	#else
26923	26944	__ai __attribute__((target("neon"))) uint8x8_t vrsubhn_u16(uint16x8_t __p0, uint16x8_t __p1) {
26924	26945	uint8x8_t __ret;
26925		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
26926		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
26927		__ret = (uint8x8_t) __builtin_neon_vrsubhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 16);
26928		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	26946	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	26947	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	26948	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vrsubhn_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 16));
	26949	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
26929	26950	return __ret;
26930	26951	}
26931	26952	__ai __attribute__((target("neon"))) uint8x8_t __noswap_vrsubhn_u16(uint16x8_t __p0, uint16x8_t __p1) {
26932	26953	uint8x8_t __ret;
26933		__ret = (uint8x8_t) __builtin_neon_vrsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 16);
	26954	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vrsubhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 16));
26934	26955	return __ret;
26935	26956	}
26936	26957	#endif
...	...	@@ -26938,21 +26959,21 @@ __ai __attribute__((target("neon"))) uint8x8_t __noswap_vrsubhn_u16(uint16x8_t _
26938	26959	#ifdef __LITTLE_ENDIAN__
26939	26960	__ai __attribute__((target("neon"))) int16x4_t vrsubhn_s32(int32x4_t __p0, int32x4_t __p1) {
26940	26961	int16x4_t __ret;
26941		__ret = (int16x4_t) __builtin_neon_vrsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 1);
	26962	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vrsubhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 1));
26942	26963	return __ret;
26943	26964	}
26944	26965	#else
26945	26966	__ai __attribute__((target("neon"))) int16x4_t vrsubhn_s32(int32x4_t __p0, int32x4_t __p1) {
26946	26967	int16x4_t __ret;
26947		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
26948		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
26949		__ret = (int16x4_t) __builtin_neon_vrsubhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 1);
26950		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	26968	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	26969	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	26970	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vrsubhn_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 1));
	26971	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
26951	26972	return __ret;
26952	26973	}
26953	26974	__ai __attribute__((target("neon"))) int16x4_t __noswap_vrsubhn_s32(int32x4_t __p0, int32x4_t __p1) {
26954	26975	int16x4_t __ret;
26955		__ret = (int16x4_t) __builtin_neon_vrsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 1);
	26976	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vrsubhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 1));
26956	26977	return __ret;
26957	26978	}
26958	26979	#endif
...	...	@@ -26960,21 +26981,21 @@ __ai __attribute__((target("neon"))) int16x4_t __noswap_vrsubhn_s32(int32x4_t __
26960	26981	#ifdef __LITTLE_ENDIAN__
26961	26982	__ai __attribute__((target("neon"))) int32x2_t vrsubhn_s64(int64x2_t __p0, int64x2_t __p1) {
26962	26983	int32x2_t __ret;
26963		__ret = (int32x2_t) __builtin_neon_vrsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 2);
	26984	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vrsubhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 2));
26964	26985	return __ret;
26965	26986	}
26966	26987	#else
26967	26988	__ai __attribute__((target("neon"))) int32x2_t vrsubhn_s64(int64x2_t __p0, int64x2_t __p1) {
26968	26989	int32x2_t __ret;
26969		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
26970		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
26971		__ret = (int32x2_t) __builtin_neon_vrsubhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 2);
26972		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	26990	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	26991	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	26992	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vrsubhn_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 2));
	26993	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
26973	26994	return __ret;
26974	26995	}
26975	26996	__ai __attribute__((target("neon"))) int32x2_t __noswap_vrsubhn_s64(int64x2_t __p0, int64x2_t __p1) {
26976	26997	int32x2_t __ret;
26977		__ret = (int32x2_t) __builtin_neon_vrsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 2);
	26998	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vrsubhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 2));
26978	26999	return __ret;
26979	27000	}
26980	27001	#endif
...	...	@@ -26982,21 +27003,21 @@ __ai __attribute__((target("neon"))) int32x2_t __noswap_vrsubhn_s64(int64x2_t __
26982	27003	#ifdef __LITTLE_ENDIAN__
26983	27004	__ai __attribute__((target("neon"))) int8x8_t vrsubhn_s16(int16x8_t __p0, int16x8_t __p1) {
26984	27005	int8x8_t __ret;
26985		__ret = (int8x8_t) __builtin_neon_vrsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 0);
	27006	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vrsubhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 0));
26986	27007	return __ret;
26987	27008	}
26988	27009	#else
26989	27010	__ai __attribute__((target("neon"))) int8x8_t vrsubhn_s16(int16x8_t __p0, int16x8_t __p1) {
26990	27011	int8x8_t __ret;
26991		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
26992		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
26993		__ret = (int8x8_t) __builtin_neon_vrsubhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 0);
26994		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	27012	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	27013	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	27014	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vrsubhn_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 0));
	27015	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
26995	27016	return __ret;
26996	27017	}
26997	27018	__ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p0, int16x8_t __p1) {
26998	27019	int8x8_t __ret;
26999		__ret = (int8x8_t) __builtin_neon_vrsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 0);
	27020	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vrsubhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 0));
27000	27021	return __ret;
27001	27022	}
27002	27023	#endif
...	...	@@ -27006,7 +27027,7 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27006	27027	poly8x8_t __ret; \
27007	27028	poly8_t __s0 = __p0; \
27008	27029	poly8x8_t __s1 = __p1; \
27009		__ret = (poly8x8_t) __builtin_neon_vset_lane_i8(__s0, (poly8x8_t)__s1, __p2); \
	27030	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vset_lane_i8(__s0, __s1, __p2)); \
27010	27031	__ret; \
27011	27032	})
27012	27033	#else
...	...	@@ -27014,16 +27035,16 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27014	27035	poly8x8_t __ret; \
27015	27036	poly8_t __s0 = __p0; \
27016	27037	poly8x8_t __s1 = __p1; \
27017		poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
27018		__ret = (poly8x8_t) __builtin_neon_vset_lane_i8(__s0, (poly8x8_t)__rev1, __p2); \
27019		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	27038	poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	27039	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vset_lane_i8(__s0, __rev1, __p2)); \
	27040	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
27020	27041	__ret; \
27021	27042	})
27022	27043	#define __noswap_vset_lane_p8(__p0, __p1, __p2) __extension__ ({ \
27023	27044	poly8x8_t __ret; \
27024	27045	poly8_t __s0 = __p0; \
27025	27046	poly8x8_t __s1 = __p1; \
27026		__ret = (poly8x8_t) __builtin_neon_vset_lane_i8(__s0, (poly8x8_t)__s1, __p2); \
	27047	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vset_lane_i8(__s0, __s1, __p2)); \
27027	27048	__ret; \
27028	27049	})
27029	27050	#endif
...	...	@@ -27033,7 +27054,7 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27033	27054	poly16x4_t __ret; \
27034	27055	poly16_t __s0 = __p0; \
27035	27056	poly16x4_t __s1 = __p1; \
27036		__ret = (poly16x4_t) __builtin_neon_vset_lane_i16(__s0, (poly16x4_t)__s1, __p2); \
	27057	__ret = __builtin_bit_cast(poly16x4_t, __builtin_neon_vset_lane_i16(__s0, __s1, __p2)); \
27037	27058	__ret; \
27038	27059	})
27039	27060	#else
...	...	@@ -27041,16 +27062,16 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27041	27062	poly16x4_t __ret; \
27042	27063	poly16_t __s0 = __p0; \
27043	27064	poly16x4_t __s1 = __p1; \
27044		poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
27045		__ret = (poly16x4_t) __builtin_neon_vset_lane_i16(__s0, (poly16x4_t)__rev1, __p2); \
27046		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	27065	poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	27066	__ret = __builtin_bit_cast(poly16x4_t, __builtin_neon_vset_lane_i16(__s0, __rev1, __p2)); \
	27067	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
27047	27068	__ret; \
27048	27069	})
27049	27070	#define __noswap_vset_lane_p16(__p0, __p1, __p2) __extension__ ({ \
27050	27071	poly16x4_t __ret; \
27051	27072	poly16_t __s0 = __p0; \
27052	27073	poly16x4_t __s1 = __p1; \
27053		__ret = (poly16x4_t) __builtin_neon_vset_lane_i16(__s0, (poly16x4_t)__s1, __p2); \
	27074	__ret = __builtin_bit_cast(poly16x4_t, __builtin_neon_vset_lane_i16(__s0, __s1, __p2)); \
27054	27075	__ret; \
27055	27076	})
27056	27077	#endif
...	...	@@ -27060,7 +27081,7 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27060	27081	poly8x16_t __ret; \
27061	27082	poly8_t __s0 = __p0; \
27062	27083	poly8x16_t __s1 = __p1; \
27063		__ret = (poly8x16_t) __builtin_neon_vsetq_lane_i8(__s0, (poly8x16_t)__s1, __p2); \
	27084	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vsetq_lane_i8(__s0, __s1, __p2)); \
27064	27085	__ret; \
27065	27086	})
27066	27087	#else
...	...	@@ -27068,16 +27089,16 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27068	27089	poly8x16_t __ret; \
27069	27090	poly8_t __s0 = __p0; \
27070	27091	poly8x16_t __s1 = __p1; \
27071		poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27072		__ret = (poly8x16_t) __builtin_neon_vsetq_lane_i8(__s0, (poly8x16_t)__rev1, __p2); \
27073		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	27092	poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	27093	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vsetq_lane_i8(__s0, __rev1, __p2)); \
	27094	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
27074	27095	__ret; \
27075	27096	})
27076	27097	#define __noswap_vsetq_lane_p8(__p0, __p1, __p2) __extension__ ({ \
27077	27098	poly8x16_t __ret; \
27078	27099	poly8_t __s0 = __p0; \
27079	27100	poly8x16_t __s1 = __p1; \
27080		__ret = (poly8x16_t) __builtin_neon_vsetq_lane_i8(__s0, (poly8x16_t)__s1, __p2); \
	27101	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vsetq_lane_i8(__s0, __s1, __p2)); \
27081	27102	__ret; \
27082	27103	})
27083	27104	#endif
...	...	@@ -27087,7 +27108,7 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27087	27108	poly16x8_t __ret; \
27088	27109	poly16_t __s0 = __p0; \
27089	27110	poly16x8_t __s1 = __p1; \
27090		__ret = (poly16x8_t) __builtin_neon_vsetq_lane_i16(__s0, (poly16x8_t)__s1, __p2); \
	27111	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vsetq_lane_i16(__s0, __s1, __p2)); \
27091	27112	__ret; \
27092	27113	})
27093	27114	#else
...	...	@@ -27095,16 +27116,16 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27095	27116	poly16x8_t __ret; \
27096	27117	poly16_t __s0 = __p0; \
27097	27118	poly16x8_t __s1 = __p1; \
27098		poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
27099		__ret = (poly16x8_t) __builtin_neon_vsetq_lane_i16(__s0, (poly16x8_t)__rev1, __p2); \
27100		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	27119	poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	27120	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vsetq_lane_i16(__s0, __rev1, __p2)); \
	27121	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
27101	27122	__ret; \
27102	27123	})
27103	27124	#define __noswap_vsetq_lane_p16(__p0, __p1, __p2) __extension__ ({ \
27104	27125	poly16x8_t __ret; \
27105	27126	poly16_t __s0 = __p0; \
27106	27127	poly16x8_t __s1 = __p1; \
27107		__ret = (poly16x8_t) __builtin_neon_vsetq_lane_i16(__s0, (poly16x8_t)__s1, __p2); \
	27128	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vsetq_lane_i16(__s0, __s1, __p2)); \
27108	27129	__ret; \
27109	27130	})
27110	27131	#endif
...	...	@@ -27114,7 +27135,7 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27114	27135	uint8x16_t __ret; \
27115	27136	uint8_t __s0 = __p0; \
27116	27137	uint8x16_t __s1 = __p1; \
27117		__ret = (uint8x16_t) __builtin_neon_vsetq_lane_i8(__s0, (int8x16_t)__s1, __p2); \
	27138	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vsetq_lane_i8(__s0, __builtin_bit_cast(int8x16_t, __s1), __p2)); \
27118	27139	__ret; \
27119	27140	})
27120	27141	#else
...	...	@@ -27122,16 +27143,16 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27122	27143	uint8x16_t __ret; \
27123	27144	uint8_t __s0 = __p0; \
27124	27145	uint8x16_t __s1 = __p1; \
27125		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27126		__ret = (uint8x16_t) __builtin_neon_vsetq_lane_i8(__s0, (int8x16_t)__rev1, __p2); \
27127		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	27146	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	27147	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vsetq_lane_i8(__s0, __builtin_bit_cast(int8x16_t, __rev1), __p2)); \
	27148	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
27128	27149	__ret; \
27129	27150	})
27130	27151	#define __noswap_vsetq_lane_u8(__p0, __p1, __p2) __extension__ ({ \
27131	27152	uint8x16_t __ret; \
27132	27153	uint8_t __s0 = __p0; \
27133	27154	uint8x16_t __s1 = __p1; \
27134		__ret = (uint8x16_t) __builtin_neon_vsetq_lane_i8(__s0, (int8x16_t)__s1, __p2); \
	27155	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vsetq_lane_i8(__s0, __builtin_bit_cast(int8x16_t, __s1), __p2)); \
27135	27156	__ret; \
27136	27157	})
27137	27158	#endif
...	...	@@ -27141,7 +27162,7 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27141	27162	uint32x4_t __ret; \
27142	27163	uint32_t __s0 = __p0; \
27143	27164	uint32x4_t __s1 = __p1; \
27144		__ret = (uint32x4_t) __builtin_neon_vsetq_lane_i32(__s0, (int32x4_t)__s1, __p2); \
	27165	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsetq_lane_i32(__s0, __builtin_bit_cast(int32x4_t, __s1), __p2)); \
27145	27166	__ret; \
27146	27167	})
27147	27168	#else
...	...	@@ -27149,16 +27170,16 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27149	27170	uint32x4_t __ret; \
27150	27171	uint32_t __s0 = __p0; \
27151	27172	uint32x4_t __s1 = __p1; \
27152		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
27153		__ret = (uint32x4_t) __builtin_neon_vsetq_lane_i32(__s0, (int32x4_t)__rev1, __p2); \
27154		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	27173	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	27174	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsetq_lane_i32(__s0, __builtin_bit_cast(int32x4_t, __rev1), __p2)); \
	27175	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
27155	27176	__ret; \
27156	27177	})
27157	27178	#define __noswap_vsetq_lane_u32(__p0, __p1, __p2) __extension__ ({ \
27158	27179	uint32x4_t __ret; \
27159	27180	uint32_t __s0 = __p0; \
27160	27181	uint32x4_t __s1 = __p1; \
27161		__ret = (uint32x4_t) __builtin_neon_vsetq_lane_i32(__s0, (int32x4_t)__s1, __p2); \
	27182	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsetq_lane_i32(__s0, __builtin_bit_cast(int32x4_t, __s1), __p2)); \
27162	27183	__ret; \
27163	27184	})
27164	27185	#endif
...	...	@@ -27168,7 +27189,7 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27168	27189	uint64x2_t __ret; \
27169	27190	uint64_t __s0 = __p0; \
27170	27191	uint64x2_t __s1 = __p1; \
27171		__ret = (uint64x2_t) __builtin_neon_vsetq_lane_i64(__s0, (int64x2_t)__s1, __p2); \
	27192	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vsetq_lane_i64(__s0, __builtin_bit_cast(int64x2_t, __s1), __p2)); \
27172	27193	__ret; \
27173	27194	})
27174	27195	#else
...	...	@@ -27176,16 +27197,16 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27176	27197	uint64x2_t __ret; \
27177	27198	uint64_t __s0 = __p0; \
27178	27199	uint64x2_t __s1 = __p1; \
27179		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
27180		__ret = (uint64x2_t) __builtin_neon_vsetq_lane_i64(__s0, (int64x2_t)__rev1, __p2); \
27181		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	27200	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	27201	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vsetq_lane_i64(__s0, __builtin_bit_cast(int64x2_t, __rev1), __p2)); \
	27202	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
27182	27203	__ret; \
27183	27204	})
27184	27205	#define __noswap_vsetq_lane_u64(__p0, __p1, __p2) __extension__ ({ \
27185	27206	uint64x2_t __ret; \
27186	27207	uint64_t __s0 = __p0; \
27187	27208	uint64x2_t __s1 = __p1; \
27188		__ret = (uint64x2_t) __builtin_neon_vsetq_lane_i64(__s0, (int64x2_t)__s1, __p2); \
	27209	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vsetq_lane_i64(__s0, __builtin_bit_cast(int64x2_t, __s1), __p2)); \
27189	27210	__ret; \
27190	27211	})
27191	27212	#endif
...	...	@@ -27195,7 +27216,7 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27195	27216	uint16x8_t __ret; \
27196	27217	uint16_t __s0 = __p0; \
27197	27218	uint16x8_t __s1 = __p1; \
27198		__ret = (uint16x8_t) __builtin_neon_vsetq_lane_i16(__s0, (int16x8_t)__s1, __p2); \
	27219	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vsetq_lane_i16(__s0, __builtin_bit_cast(int16x8_t, __s1), __p2)); \
27199	27220	__ret; \
27200	27221	})
27201	27222	#else
...	...	@@ -27203,16 +27224,16 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27203	27224	uint16x8_t __ret; \
27204	27225	uint16_t __s0 = __p0; \
27205	27226	uint16x8_t __s1 = __p1; \
27206		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
27207		__ret = (uint16x8_t) __builtin_neon_vsetq_lane_i16(__s0, (int16x8_t)__rev1, __p2); \
27208		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	27227	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	27228	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vsetq_lane_i16(__s0, __builtin_bit_cast(int16x8_t, __rev1), __p2)); \
	27229	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
27209	27230	__ret; \
27210	27231	})
27211	27232	#define __noswap_vsetq_lane_u16(__p0, __p1, __p2) __extension__ ({ \
27212	27233	uint16x8_t __ret; \
27213	27234	uint16_t __s0 = __p0; \
27214	27235	uint16x8_t __s1 = __p1; \
27215		__ret = (uint16x8_t) __builtin_neon_vsetq_lane_i16(__s0, (int16x8_t)__s1, __p2); \
	27236	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vsetq_lane_i16(__s0, __builtin_bit_cast(int16x8_t, __s1), __p2)); \
27216	27237	__ret; \
27217	27238	})
27218	27239	#endif
...	...	@@ -27222,7 +27243,7 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27222	27243	int8x16_t __ret; \
27223	27244	int8_t __s0 = __p0; \
27224	27245	int8x16_t __s1 = __p1; \
27225		__ret = (int8x16_t) __builtin_neon_vsetq_lane_i8(__s0, (int8x16_t)__s1, __p2); \
	27246	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vsetq_lane_i8(__s0, __builtin_bit_cast(int8x16_t, __s1), __p2)); \
27226	27247	__ret; \
27227	27248	})
27228	27249	#else
...	...	@@ -27230,16 +27251,16 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27230	27251	int8x16_t __ret; \
27231	27252	int8_t __s0 = __p0; \
27232	27253	int8x16_t __s1 = __p1; \
27233		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27234		__ret = (int8x16_t) __builtin_neon_vsetq_lane_i8(__s0, (int8x16_t)__rev1, __p2); \
27235		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	27254	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	27255	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vsetq_lane_i8(__s0, __builtin_bit_cast(int8x16_t, __rev1), __p2)); \
	27256	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
27236	27257	__ret; \
27237	27258	})
27238	27259	#define __noswap_vsetq_lane_s8(__p0, __p1, __p2) __extension__ ({ \
27239	27260	int8x16_t __ret; \
27240	27261	int8_t __s0 = __p0; \
27241	27262	int8x16_t __s1 = __p1; \
27242		__ret = (int8x16_t) __builtin_neon_vsetq_lane_i8(__s0, (int8x16_t)__s1, __p2); \
	27263	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vsetq_lane_i8(__s0, __builtin_bit_cast(int8x16_t, __s1), __p2)); \
27243	27264	__ret; \
27244	27265	})
27245	27266	#endif
...	...	@@ -27249,7 +27270,7 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27249	27270	float32x4_t __ret; \
27250	27271	float32_t __s0 = __p0; \
27251	27272	float32x4_t __s1 = __p1; \
27252		__ret = (float32x4_t) __builtin_neon_vsetq_lane_f32(__s0, (float32x4_t)__s1, __p2); \
	27273	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vsetq_lane_f32(__s0, __s1, __p2)); \
27253	27274	__ret; \
27254	27275	})
27255	27276	#else
...	...	@@ -27257,16 +27278,16 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27257	27278	float32x4_t __ret; \
27258	27279	float32_t __s0 = __p0; \
27259	27280	float32x4_t __s1 = __p1; \
27260		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
27261		__ret = (float32x4_t) __builtin_neon_vsetq_lane_f32(__s0, (float32x4_t)__rev1, __p2); \
27262		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	27281	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	27282	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vsetq_lane_f32(__s0, __rev1, __p2)); \
	27283	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
27263	27284	__ret; \
27264	27285	})
27265	27286	#define __noswap_vsetq_lane_f32(__p0, __p1, __p2) __extension__ ({ \
27266	27287	float32x4_t __ret; \
27267	27288	float32_t __s0 = __p0; \
27268	27289	float32x4_t __s1 = __p1; \
27269		__ret = (float32x4_t) __builtin_neon_vsetq_lane_f32(__s0, (float32x4_t)__s1, __p2); \
	27290	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vsetq_lane_f32(__s0, __s1, __p2)); \
27270	27291	__ret; \
27271	27292	})
27272	27293	#endif
...	...	@@ -27276,7 +27297,7 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27276	27297	int32x4_t __ret; \
27277	27298	int32_t __s0 = __p0; \
27278	27299	int32x4_t __s1 = __p1; \
27279		__ret = (int32x4_t) __builtin_neon_vsetq_lane_i32(__s0, (int32x4_t)__s1, __p2); \
	27300	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vsetq_lane_i32(__s0, __builtin_bit_cast(int32x4_t, __s1), __p2)); \
27280	27301	__ret; \
27281	27302	})
27282	27303	#else
...	...	@@ -27284,16 +27305,16 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27284	27305	int32x4_t __ret; \
27285	27306	int32_t __s0 = __p0; \
27286	27307	int32x4_t __s1 = __p1; \
27287		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
27288		__ret = (int32x4_t) __builtin_neon_vsetq_lane_i32(__s0, (int32x4_t)__rev1, __p2); \
27289		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	27308	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	27309	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vsetq_lane_i32(__s0, __builtin_bit_cast(int32x4_t, __rev1), __p2)); \
	27310	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
27290	27311	__ret; \
27291	27312	})
27292	27313	#define __noswap_vsetq_lane_s32(__p0, __p1, __p2) __extension__ ({ \
27293	27314	int32x4_t __ret; \
27294	27315	int32_t __s0 = __p0; \
27295	27316	int32x4_t __s1 = __p1; \
27296		__ret = (int32x4_t) __builtin_neon_vsetq_lane_i32(__s0, (int32x4_t)__s1, __p2); \
	27317	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vsetq_lane_i32(__s0, __builtin_bit_cast(int32x4_t, __s1), __p2)); \
27297	27318	__ret; \
27298	27319	})
27299	27320	#endif
...	...	@@ -27303,7 +27324,7 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27303	27324	int64x2_t __ret; \
27304	27325	int64_t __s0 = __p0; \
27305	27326	int64x2_t __s1 = __p1; \
27306		__ret = (int64x2_t) __builtin_neon_vsetq_lane_i64(__s0, (int64x2_t)__s1, __p2); \
	27327	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vsetq_lane_i64(__s0, __builtin_bit_cast(int64x2_t, __s1), __p2)); \
27307	27328	__ret; \
27308	27329	})
27309	27330	#else
...	...	@@ -27311,16 +27332,16 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27311	27332	int64x2_t __ret; \
27312	27333	int64_t __s0 = __p0; \
27313	27334	int64x2_t __s1 = __p1; \
27314		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
27315		__ret = (int64x2_t) __builtin_neon_vsetq_lane_i64(__s0, (int64x2_t)__rev1, __p2); \
27316		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	27335	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	27336	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vsetq_lane_i64(__s0, __builtin_bit_cast(int64x2_t, __rev1), __p2)); \
	27337	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
27317	27338	__ret; \
27318	27339	})
27319	27340	#define __noswap_vsetq_lane_s64(__p0, __p1, __p2) __extension__ ({ \
27320	27341	int64x2_t __ret; \
27321	27342	int64_t __s0 = __p0; \
27322	27343	int64x2_t __s1 = __p1; \
27323		__ret = (int64x2_t) __builtin_neon_vsetq_lane_i64(__s0, (int64x2_t)__s1, __p2); \
	27344	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vsetq_lane_i64(__s0, __builtin_bit_cast(int64x2_t, __s1), __p2)); \
27324	27345	__ret; \
27325	27346	})
27326	27347	#endif
...	...	@@ -27330,7 +27351,7 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27330	27351	int16x8_t __ret; \
27331	27352	int16_t __s0 = __p0; \
27332	27353	int16x8_t __s1 = __p1; \
27333		__ret = (int16x8_t) __builtin_neon_vsetq_lane_i16(__s0, (int16x8_t)__s1, __p2); \
	27354	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vsetq_lane_i16(__s0, __builtin_bit_cast(int16x8_t, __s1), __p2)); \
27334	27355	__ret; \
27335	27356	})
27336	27357	#else
...	...	@@ -27338,16 +27359,16 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27338	27359	int16x8_t __ret; \
27339	27360	int16_t __s0 = __p0; \
27340	27361	int16x8_t __s1 = __p1; \
27341		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
27342		__ret = (int16x8_t) __builtin_neon_vsetq_lane_i16(__s0, (int16x8_t)__rev1, __p2); \
27343		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	27362	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	27363	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vsetq_lane_i16(__s0, __builtin_bit_cast(int16x8_t, __rev1), __p2)); \
	27364	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
27344	27365	__ret; \
27345	27366	})
27346	27367	#define __noswap_vsetq_lane_s16(__p0, __p1, __p2) __extension__ ({ \
27347	27368	int16x8_t __ret; \
27348	27369	int16_t __s0 = __p0; \
27349	27370	int16x8_t __s1 = __p1; \
27350		__ret = (int16x8_t) __builtin_neon_vsetq_lane_i16(__s0, (int16x8_t)__s1, __p2); \
	27371	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vsetq_lane_i16(__s0, __builtin_bit_cast(int16x8_t, __s1), __p2)); \
27351	27372	__ret; \
27352	27373	})
27353	27374	#endif
...	...	@@ -27357,7 +27378,7 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27357	27378	uint8x8_t __ret; \
27358	27379	uint8_t __s0 = __p0; \
27359	27380	uint8x8_t __s1 = __p1; \
27360		__ret = (uint8x8_t) __builtin_neon_vset_lane_i8(__s0, (int8x8_t)__s1, __p2); \
	27381	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vset_lane_i8(__s0, __builtin_bit_cast(int8x8_t, __s1), __p2)); \
27361	27382	__ret; \
27362	27383	})
27363	27384	#else
...	...	@@ -27365,16 +27386,16 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27365	27386	uint8x8_t __ret; \
27366	27387	uint8_t __s0 = __p0; \
27367	27388	uint8x8_t __s1 = __p1; \
27368		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
27369		__ret = (uint8x8_t) __builtin_neon_vset_lane_i8(__s0, (int8x8_t)__rev1, __p2); \
27370		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	27389	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	27390	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vset_lane_i8(__s0, __builtin_bit_cast(int8x8_t, __rev1), __p2)); \
	27391	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
27371	27392	__ret; \
27372	27393	})
27373	27394	#define __noswap_vset_lane_u8(__p0, __p1, __p2) __extension__ ({ \
27374	27395	uint8x8_t __ret; \
27375	27396	uint8_t __s0 = __p0; \
27376	27397	uint8x8_t __s1 = __p1; \
27377		__ret = (uint8x8_t) __builtin_neon_vset_lane_i8(__s0, (int8x8_t)__s1, __p2); \
	27398	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vset_lane_i8(__s0, __builtin_bit_cast(int8x8_t, __s1), __p2)); \
27378	27399	__ret; \
27379	27400	})
27380	27401	#endif
...	...	@@ -27384,7 +27405,7 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27384	27405	uint32x2_t __ret; \
27385	27406	uint32_t __s0 = __p0; \
27386	27407	uint32x2_t __s1 = __p1; \
27387		__ret = (uint32x2_t) __builtin_neon_vset_lane_i32(__s0, (int32x2_t)__s1, __p2); \
	27408	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vset_lane_i32(__s0, __builtin_bit_cast(int32x2_t, __s1), __p2)); \
27388	27409	__ret; \
27389	27410	})
27390	27411	#else
...	...	@@ -27392,16 +27413,16 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27392	27413	uint32x2_t __ret; \
27393	27414	uint32_t __s0 = __p0; \
27394	27415	uint32x2_t __s1 = __p1; \
27395		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
27396		__ret = (uint32x2_t) __builtin_neon_vset_lane_i32(__s0, (int32x2_t)__rev1, __p2); \
27397		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	27416	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_32); \
	27417	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vset_lane_i32(__s0, __builtin_bit_cast(int32x2_t, __rev1), __p2)); \
	27418	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
27398	27419	__ret; \
27399	27420	})
27400	27421	#define __noswap_vset_lane_u32(__p0, __p1, __p2) __extension__ ({ \
27401	27422	uint32x2_t __ret; \
27402	27423	uint32_t __s0 = __p0; \
27403	27424	uint32x2_t __s1 = __p1; \
27404		__ret = (uint32x2_t) __builtin_neon_vset_lane_i32(__s0, (int32x2_t)__s1, __p2); \
	27425	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vset_lane_i32(__s0, __builtin_bit_cast(int32x2_t, __s1), __p2)); \
27405	27426	__ret; \
27406	27427	})
27407	27428	#endif
...	...	@@ -27410,7 +27431,7 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27410	27431	uint64x1_t __ret; \
27411	27432	uint64_t __s0 = __p0; \
27412	27433	uint64x1_t __s1 = __p1; \
27413		__ret = (uint64x1_t) __builtin_neon_vset_lane_i64(__s0, (int64x1_t)__s1, __p2); \
	27434	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vset_lane_i64(__s0, __builtin_bit_cast(int64x1_t, __s1), __p2)); \
27414	27435	__ret; \
27415	27436	})
27416	27437	#ifdef __LITTLE_ENDIAN__
...	...	@@ -27418,7 +27439,7 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27418	27439	uint16x4_t __ret; \
27419	27440	uint16_t __s0 = __p0; \
27420	27441	uint16x4_t __s1 = __p1; \
27421		__ret = (uint16x4_t) __builtin_neon_vset_lane_i16(__s0, (int16x4_t)__s1, __p2); \
	27442	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vset_lane_i16(__s0, __builtin_bit_cast(int16x4_t, __s1), __p2)); \
27422	27443	__ret; \
27423	27444	})
27424	27445	#else
...	...	@@ -27426,16 +27447,16 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27426	27447	uint16x4_t __ret; \
27427	27448	uint16_t __s0 = __p0; \
27428	27449	uint16x4_t __s1 = __p1; \
27429		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
27430		__ret = (uint16x4_t) __builtin_neon_vset_lane_i16(__s0, (int16x4_t)__rev1, __p2); \
27431		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	27450	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	27451	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vset_lane_i16(__s0, __builtin_bit_cast(int16x4_t, __rev1), __p2)); \
	27452	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
27432	27453	__ret; \
27433	27454	})
27434	27455	#define __noswap_vset_lane_u16(__p0, __p1, __p2) __extension__ ({ \
27435	27456	uint16x4_t __ret; \
27436	27457	uint16_t __s0 = __p0; \
27437	27458	uint16x4_t __s1 = __p1; \
27438		__ret = (uint16x4_t) __builtin_neon_vset_lane_i16(__s0, (int16x4_t)__s1, __p2); \
	27459	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vset_lane_i16(__s0, __builtin_bit_cast(int16x4_t, __s1), __p2)); \
27439	27460	__ret; \
27440	27461	})
27441	27462	#endif
...	...	@@ -27445,7 +27466,7 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27445	27466	int8x8_t __ret; \
27446	27467	int8_t __s0 = __p0; \
27447	27468	int8x8_t __s1 = __p1; \
27448		__ret = (int8x8_t) __builtin_neon_vset_lane_i8(__s0, (int8x8_t)__s1, __p2); \
	27469	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vset_lane_i8(__s0, __builtin_bit_cast(int8x8_t, __s1), __p2)); \
27449	27470	__ret; \
27450	27471	})
27451	27472	#else
...	...	@@ -27453,16 +27474,16 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27453	27474	int8x8_t __ret; \
27454	27475	int8_t __s0 = __p0; \
27455	27476	int8x8_t __s1 = __p1; \
27456		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
27457		__ret = (int8x8_t) __builtin_neon_vset_lane_i8(__s0, (int8x8_t)__rev1, __p2); \
27458		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	27477	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	27478	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vset_lane_i8(__s0, __builtin_bit_cast(int8x8_t, __rev1), __p2)); \
	27479	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
27459	27480	__ret; \
27460	27481	})
27461	27482	#define __noswap_vset_lane_s8(__p0, __p1, __p2) __extension__ ({ \
27462	27483	int8x8_t __ret; \
27463	27484	int8_t __s0 = __p0; \
27464	27485	int8x8_t __s1 = __p1; \
27465		__ret = (int8x8_t) __builtin_neon_vset_lane_i8(__s0, (int8x8_t)__s1, __p2); \
	27486	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vset_lane_i8(__s0, __builtin_bit_cast(int8x8_t, __s1), __p2)); \
27466	27487	__ret; \
27467	27488	})
27468	27489	#endif
...	...	@@ -27472,7 +27493,7 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27472	27493	float32x2_t __ret; \
27473	27494	float32_t __s0 = __p0; \
27474	27495	float32x2_t __s1 = __p1; \
27475		__ret = (float32x2_t) __builtin_neon_vset_lane_f32(__s0, (float32x2_t)__s1, __p2); \
	27496	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vset_lane_f32(__s0, __s1, __p2)); \
27476	27497	__ret; \
27477	27498	})
27478	27499	#else
...	...	@@ -27480,16 +27501,16 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27480	27501	float32x2_t __ret; \
27481	27502	float32_t __s0 = __p0; \
27482	27503	float32x2_t __s1 = __p1; \
27483		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
27484		__ret = (float32x2_t) __builtin_neon_vset_lane_f32(__s0, (float32x2_t)__rev1, __p2); \
27485		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	27504	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_32); \
	27505	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vset_lane_f32(__s0, __rev1, __p2)); \
	27506	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
27486	27507	__ret; \
27487	27508	})
27488	27509	#define __noswap_vset_lane_f32(__p0, __p1, __p2) __extension__ ({ \
27489	27510	float32x2_t __ret; \
27490	27511	float32_t __s0 = __p0; \
27491	27512	float32x2_t __s1 = __p1; \
27492		__ret = (float32x2_t) __builtin_neon_vset_lane_f32(__s0, (float32x2_t)__s1, __p2); \
	27513	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vset_lane_f32(__s0, __s1, __p2)); \
27493	27514	__ret; \
27494	27515	})
27495	27516	#endif
...	...	@@ -27499,7 +27520,7 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27499	27520	int32x2_t __ret; \
27500	27521	int32_t __s0 = __p0; \
27501	27522	int32x2_t __s1 = __p1; \
27502		__ret = (int32x2_t) __builtin_neon_vset_lane_i32(__s0, (int32x2_t)__s1, __p2); \
	27523	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vset_lane_i32(__s0, __builtin_bit_cast(int32x2_t, __s1), __p2)); \
27503	27524	__ret; \
27504	27525	})
27505	27526	#else
...	...	@@ -27507,16 +27528,16 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27507	27528	int32x2_t __ret; \
27508	27529	int32_t __s0 = __p0; \
27509	27530	int32x2_t __s1 = __p1; \
27510		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
27511		__ret = (int32x2_t) __builtin_neon_vset_lane_i32(__s0, (int32x2_t)__rev1, __p2); \
27512		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	27531	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_32); \
	27532	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vset_lane_i32(__s0, __builtin_bit_cast(int32x2_t, __rev1), __p2)); \
	27533	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
27513	27534	__ret; \
27514	27535	})
27515	27536	#define __noswap_vset_lane_s32(__p0, __p1, __p2) __extension__ ({ \
27516	27537	int32x2_t __ret; \
27517	27538	int32_t __s0 = __p0; \
27518	27539	int32x2_t __s1 = __p1; \
27519		__ret = (int32x2_t) __builtin_neon_vset_lane_i32(__s0, (int32x2_t)__s1, __p2); \
	27540	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vset_lane_i32(__s0, __builtin_bit_cast(int32x2_t, __s1), __p2)); \
27520	27541	__ret; \
27521	27542	})
27522	27543	#endif
...	...	@@ -27525,7 +27546,7 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27525	27546	int64x1_t __ret; \
27526	27547	int64_t __s0 = __p0; \
27527	27548	int64x1_t __s1 = __p1; \
27528		__ret = (int64x1_t) __builtin_neon_vset_lane_i64(__s0, (int64x1_t)__s1, __p2); \
	27549	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vset_lane_i64(__s0, __builtin_bit_cast(int64x1_t, __s1), __p2)); \
27529	27550	__ret; \
27530	27551	})
27531	27552	#ifdef __LITTLE_ENDIAN__
...	...	@@ -27533,7 +27554,7 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27533	27554	int16x4_t __ret; \
27534	27555	int16_t __s0 = __p0; \
27535	27556	int16x4_t __s1 = __p1; \
27536		__ret = (int16x4_t) __builtin_neon_vset_lane_i16(__s0, (int16x4_t)__s1, __p2); \
	27557	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vset_lane_i16(__s0, __builtin_bit_cast(int16x4_t, __s1), __p2)); \
27537	27558	__ret; \
27538	27559	})
27539	27560	#else
...	...	@@ -27541,16 +27562,16 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27541	27562	int16x4_t __ret; \
27542	27563	int16_t __s0 = __p0; \
27543	27564	int16x4_t __s1 = __p1; \
27544		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
27545		__ret = (int16x4_t) __builtin_neon_vset_lane_i16(__s0, (int16x4_t)__rev1, __p2); \
27546		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	27565	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	27566	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vset_lane_i16(__s0, __builtin_bit_cast(int16x4_t, __rev1), __p2)); \
	27567	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
27547	27568	__ret; \
27548	27569	})
27549	27570	#define __noswap_vset_lane_s16(__p0, __p1, __p2) __extension__ ({ \
27550	27571	int16x4_t __ret; \
27551	27572	int16_t __s0 = __p0; \
27552	27573	int16x4_t __s1 = __p1; \
27553		__ret = (int16x4_t) __builtin_neon_vset_lane_i16(__s0, (int16x4_t)__s1, __p2); \
	27574	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vset_lane_i16(__s0, __builtin_bit_cast(int16x4_t, __s1), __p2)); \
27554	27575	__ret; \
27555	27576	})
27556	27577	#endif
...	...	@@ -27558,16 +27579,16 @@ __ai __attribute__((target("neon"))) int8x8_t __noswap_vrsubhn_s16(int16x8_t __p
27558	27579	#ifdef __LITTLE_ENDIAN__
27559	27580	__ai __attribute__((target("neon"))) uint8x16_t vshlq_u8(uint8x16_t __p0, int8x16_t __p1) {
27560	27581	uint8x16_t __ret;
27561		__ret = (uint8x16_t) __builtin_neon_vshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
	27582	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 48));
27562	27583	return __ret;
27563	27584	}
27564	27585	#else
27565	27586	__ai __attribute__((target("neon"))) uint8x16_t vshlq_u8(uint8x16_t __p0, int8x16_t __p1) {
27566	27587	uint8x16_t __ret;
27567		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
27568		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
27569		__ret = (uint8x16_t) __builtin_neon_vshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
27570		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	27588	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	27589	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	27590	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 48));
	27591	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
27571	27592	return __ret;
27572	27593	}
27573	27594	#endif
...	...	@@ -27575,16 +27596,16 @@ __ai __attribute__((target("neon"))) uint8x16_t vshlq_u8(uint8x16_t __p0, int8x1
27575	27596	#ifdef __LITTLE_ENDIAN__
27576	27597	__ai __attribute__((target("neon"))) uint32x4_t vshlq_u32(uint32x4_t __p0, int32x4_t __p1) {
27577	27598	uint32x4_t __ret;
27578		__ret = (uint32x4_t) __builtin_neon_vshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
	27599	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50));
27579	27600	return __ret;
27580	27601	}
27581	27602	#else
27582	27603	__ai __attribute__((target("neon"))) uint32x4_t vshlq_u32(uint32x4_t __p0, int32x4_t __p1) {
27583	27604	uint32x4_t __ret;
27584		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
27585		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
27586		__ret = (uint32x4_t) __builtin_neon_vshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
27587		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	27605	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	27606	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	27607	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 50));
	27608	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
27588	27609	return __ret;
27589	27610	}
27590	27611	#endif
...	...	@@ -27592,16 +27613,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vshlq_u32(uint32x4_t __p0, int32
27592	27613	#ifdef __LITTLE_ENDIAN__
27593	27614	__ai __attribute__((target("neon"))) uint64x2_t vshlq_u64(uint64x2_t __p0, int64x2_t __p1) {
27594	27615	uint64x2_t __ret;
27595		__ret = (uint64x2_t) __builtin_neon_vshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 51);
	27616	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 51));
27596	27617	return __ret;
27597	27618	}
27598	27619	#else
27599	27620	__ai __attribute__((target("neon"))) uint64x2_t vshlq_u64(uint64x2_t __p0, int64x2_t __p1) {
27600	27621	uint64x2_t __ret;
27601		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
27602		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
27603		__ret = (uint64x2_t) __builtin_neon_vshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 51);
27604		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	27622	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	27623	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	27624	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 51));
	27625	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
27605	27626	return __ret;
27606	27627	}
27607	27628	#endif
...	...	@@ -27609,16 +27630,16 @@ __ai __attribute__((target("neon"))) uint64x2_t vshlq_u64(uint64x2_t __p0, int64
27609	27630	#ifdef __LITTLE_ENDIAN__
27610	27631	__ai __attribute__((target("neon"))) uint16x8_t vshlq_u16(uint16x8_t __p0, int16x8_t __p1) {
27611	27632	uint16x8_t __ret;
27612		__ret = (uint16x8_t) __builtin_neon_vshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
	27633	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 49));
27613	27634	return __ret;
27614	27635	}
27615	27636	#else
27616	27637	__ai __attribute__((target("neon"))) uint16x8_t vshlq_u16(uint16x8_t __p0, int16x8_t __p1) {
27617	27638	uint16x8_t __ret;
27618		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
27619		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
27620		__ret = (uint16x8_t) __builtin_neon_vshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
27621		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	27639	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	27640	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	27641	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 49));
	27642	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
27622	27643	return __ret;
27623	27644	}
27624	27645	#endif
...	...	@@ -27626,16 +27647,16 @@ __ai __attribute__((target("neon"))) uint16x8_t vshlq_u16(uint16x8_t __p0, int16
27626	27647	#ifdef __LITTLE_ENDIAN__
27627	27648	__ai __attribute__((target("neon"))) int8x16_t vshlq_s8(int8x16_t __p0, int8x16_t __p1) {
27628	27649	int8x16_t __ret;
27629		__ret = (int8x16_t) __builtin_neon_vshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
	27650	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 32));
27630	27651	return __ret;
27631	27652	}
27632	27653	#else
27633	27654	__ai __attribute__((target("neon"))) int8x16_t vshlq_s8(int8x16_t __p0, int8x16_t __p1) {
27634	27655	int8x16_t __ret;
27635		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
27636		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
27637		__ret = (int8x16_t) __builtin_neon_vshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 32);
27638		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	27656	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	27657	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	27658	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 32));
	27659	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
27639	27660	return __ret;
27640	27661	}
27641	27662	#endif
...	...	@@ -27643,16 +27664,16 @@ __ai __attribute__((target("neon"))) int8x16_t vshlq_s8(int8x16_t __p0, int8x16_
27643	27664	#ifdef __LITTLE_ENDIAN__
27644	27665	__ai __attribute__((target("neon"))) int32x4_t vshlq_s32(int32x4_t __p0, int32x4_t __p1) {
27645	27666	int32x4_t __ret;
27646		__ret = (int32x4_t) __builtin_neon_vshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
	27667	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 34));
27647	27668	return __ret;
27648	27669	}
27649	27670	#else
27650	27671	__ai __attribute__((target("neon"))) int32x4_t vshlq_s32(int32x4_t __p0, int32x4_t __p1) {
27651	27672	int32x4_t __ret;
27652		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
27653		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
27654		__ret = (int32x4_t) __builtin_neon_vshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
27655		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	27673	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	27674	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	27675	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 34));
	27676	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
27656	27677	return __ret;
27657	27678	}
27658	27679	#endif
...	...	@@ -27660,16 +27681,16 @@ __ai __attribute__((target("neon"))) int32x4_t vshlq_s32(int32x4_t __p0, int32x4
27660	27681	#ifdef __LITTLE_ENDIAN__
27661	27682	__ai __attribute__((target("neon"))) int64x2_t vshlq_s64(int64x2_t __p0, int64x2_t __p1) {
27662	27683	int64x2_t __ret;
27663		__ret = (int64x2_t) __builtin_neon_vshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 35);
	27684	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 35));
27664	27685	return __ret;
27665	27686	}
27666	27687	#else
27667	27688	__ai __attribute__((target("neon"))) int64x2_t vshlq_s64(int64x2_t __p0, int64x2_t __p1) {
27668	27689	int64x2_t __ret;
27669		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
27670		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
27671		__ret = (int64x2_t) __builtin_neon_vshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 35);
27672		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	27690	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	27691	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	27692	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 35));
	27693	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
27673	27694	return __ret;
27674	27695	}
27675	27696	#endif
...	...	@@ -27677,16 +27698,16 @@ __ai __attribute__((target("neon"))) int64x2_t vshlq_s64(int64x2_t __p0, int64x2
27677	27698	#ifdef __LITTLE_ENDIAN__
27678	27699	__ai __attribute__((target("neon"))) int16x8_t vshlq_s16(int16x8_t __p0, int16x8_t __p1) {
27679	27700	int16x8_t __ret;
27680		__ret = (int16x8_t) __builtin_neon_vshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
	27701	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vshlq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 33));
27681	27702	return __ret;
27682	27703	}
27683	27704	#else
27684	27705	__ai __attribute__((target("neon"))) int16x8_t vshlq_s16(int16x8_t __p0, int16x8_t __p1) {
27685	27706	int16x8_t __ret;
27686		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
27687		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
27688		__ret = (int16x8_t) __builtin_neon_vshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
27689		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	27707	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	27708	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	27709	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vshlq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 33));
	27710	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
27690	27711	return __ret;
27691	27712	}
27692	27713	#endif
...	...	@@ -27694,16 +27715,16 @@ __ai __attribute__((target("neon"))) int16x8_t vshlq_s16(int16x8_t __p0, int16x8
27694	27715	#ifdef __LITTLE_ENDIAN__
27695	27716	__ai __attribute__((target("neon"))) uint8x8_t vshl_u8(uint8x8_t __p0, int8x8_t __p1) {
27696	27717	uint8x8_t __ret;
27697		__ret = (uint8x8_t) __builtin_neon_vshl_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
	27718	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 16));
27698	27719	return __ret;
27699	27720	}
27700	27721	#else
27701	27722	__ai __attribute__((target("neon"))) uint8x8_t vshl_u8(uint8x8_t __p0, int8x8_t __p1) {
27702	27723	uint8x8_t __ret;
27703		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
27704		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
27705		__ret = (uint8x8_t) __builtin_neon_vshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
27706		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	27724	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	27725	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	27726	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vshl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 16));
	27727	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
27707	27728	return __ret;
27708	27729	}
27709	27730	#endif
...	...	@@ -27711,38 +27732,38 @@ __ai __attribute__((target("neon"))) uint8x8_t vshl_u8(uint8x8_t __p0, int8x8_t
27711	27732	#ifdef __LITTLE_ENDIAN__
27712	27733	__ai __attribute__((target("neon"))) uint32x2_t vshl_u32(uint32x2_t __p0, int32x2_t __p1) {
27713	27734	uint32x2_t __ret;
27714		__ret = (uint32x2_t) __builtin_neon_vshl_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
	27735	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 18));
27715	27736	return __ret;
27716	27737	}
27717	27738	#else
27718	27739	__ai __attribute__((target("neon"))) uint32x2_t vshl_u32(uint32x2_t __p0, int32x2_t __p1) {
27719	27740	uint32x2_t __ret;
27720		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
27721		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
27722		__ret = (uint32x2_t) __builtin_neon_vshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
27723		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	27741	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	27742	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	27743	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vshl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 18));
	27744	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
27724	27745	return __ret;
27725	27746	}
27726	27747	#endif
27727	27748
27728	27749	__ai __attribute__((target("neon"))) uint64x1_t vshl_u64(uint64x1_t __p0, int64x1_t __p1) {
27729	27750	uint64x1_t __ret;
27730		__ret = (uint64x1_t) __builtin_neon_vshl_v((int8x8_t)__p0, (int8x8_t)__p1, 19);
	27751	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 19));
27731	27752	return __ret;
27732	27753	}
27733	27754	#ifdef __LITTLE_ENDIAN__
27734	27755	__ai __attribute__((target("neon"))) uint16x4_t vshl_u16(uint16x4_t __p0, int16x4_t __p1) {
27735	27756	uint16x4_t __ret;
27736		__ret = (uint16x4_t) __builtin_neon_vshl_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
	27757	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 17));
27737	27758	return __ret;
27738	27759	}
27739	27760	#else
27740	27761	__ai __attribute__((target("neon"))) uint16x4_t vshl_u16(uint16x4_t __p0, int16x4_t __p1) {
27741	27762	uint16x4_t __ret;
27742		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
27743		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
27744		__ret = (uint16x4_t) __builtin_neon_vshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
27745		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	27763	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	27764	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	27765	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vshl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 17));
	27766	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
27746	27767	return __ret;
27747	27768	}
27748	27769	#endif
...	...	@@ -27750,16 +27771,16 @@ __ai __attribute__((target("neon"))) uint16x4_t vshl_u16(uint16x4_t __p0, int16x
27750	27771	#ifdef __LITTLE_ENDIAN__
27751	27772	__ai __attribute__((target("neon"))) int8x8_t vshl_s8(int8x8_t __p0, int8x8_t __p1) {
27752	27773	int8x8_t __ret;
27753		__ret = (int8x8_t) __builtin_neon_vshl_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
	27774	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 0));
27754	27775	return __ret;
27755	27776	}
27756	27777	#else
27757	27778	__ai __attribute__((target("neon"))) int8x8_t vshl_s8(int8x8_t __p0, int8x8_t __p1) {
27758	27779	int8x8_t __ret;
27759		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
27760		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
27761		__ret = (int8x8_t) __builtin_neon_vshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
27762		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	27780	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	27781	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	27782	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vshl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 0));
	27783	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
27763	27784	return __ret;
27764	27785	}
27765	27786	#endif
...	...	@@ -27767,38 +27788,38 @@ __ai __attribute__((target("neon"))) int8x8_t vshl_s8(int8x8_t __p0, int8x8_t __
27767	27788	#ifdef __LITTLE_ENDIAN__
27768	27789	__ai __attribute__((target("neon"))) int32x2_t vshl_s32(int32x2_t __p0, int32x2_t __p1) {
27769	27790	int32x2_t __ret;
27770		__ret = (int32x2_t) __builtin_neon_vshl_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
	27791	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 2));
27771	27792	return __ret;
27772	27793	}
27773	27794	#else
27774	27795	__ai __attribute__((target("neon"))) int32x2_t vshl_s32(int32x2_t __p0, int32x2_t __p1) {
27775	27796	int32x2_t __ret;
27776		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
27777		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
27778		__ret = (int32x2_t) __builtin_neon_vshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
27779		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	27797	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	27798	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	27799	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vshl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 2));
	27800	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
27780	27801	return __ret;
27781	27802	}
27782	27803	#endif
27783	27804
27784	27805	__ai __attribute__((target("neon"))) int64x1_t vshl_s64(int64x1_t __p0, int64x1_t __p1) {
27785	27806	int64x1_t __ret;
27786		__ret = (int64x1_t) __builtin_neon_vshl_v((int8x8_t)__p0, (int8x8_t)__p1, 3);
	27807	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 3));
27787	27808	return __ret;
27788	27809	}
27789	27810	#ifdef __LITTLE_ENDIAN__
27790	27811	__ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_t __p1) {
27791	27812	int16x4_t __ret;
27792		__ret = (int16x4_t) __builtin_neon_vshl_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
	27813	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vshl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 1));
27793	27814	return __ret;
27794	27815	}
27795	27816	#else
27796	27817	__ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_t __p1) {
27797	27818	int16x4_t __ret;
27798		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
27799		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
27800		__ret = (int16x4_t) __builtin_neon_vshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
27801		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	27819	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	27820	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	27821	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vshl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 1));
	27822	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
27802	27823	return __ret;
27803	27824	}
27804	27825	#endif
...	...	@@ -27807,16 +27828,16 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
27807	27828	#define vshlq_n_u8(__p0, __p1) __extension__ ({ \
27808	27829	uint8x16_t __ret; \
27809	27830	uint8x16_t __s0 = __p0; \
27810		__ret = (uint8x16_t) __builtin_neon_vshlq_n_v((int8x16_t)__s0, __p1, 48); \
	27831	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vshlq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 48)); \
27811	27832	__ret; \
27812	27833	})
27813	27834	#else
27814	27835	#define vshlq_n_u8(__p0, __p1) __extension__ ({ \
27815	27836	uint8x16_t __ret; \
27816	27837	uint8x16_t __s0 = __p0; \
27817		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27818		__ret = (uint8x16_t) __builtin_neon_vshlq_n_v((int8x16_t)__rev0, __p1, 48); \
27819		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	27838	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	27839	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vshlq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 48)); \
	27840	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
27820	27841	__ret; \
27821	27842	})
27822	27843	#endif
...	...	@@ -27825,16 +27846,16 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
27825	27846	#define vshlq_n_u32(__p0, __p1) __extension__ ({ \
27826	27847	uint32x4_t __ret; \
27827	27848	uint32x4_t __s0 = __p0; \
27828		__ret = (uint32x4_t) __builtin_neon_vshlq_n_v((int8x16_t)__s0, __p1, 50); \
	27849	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vshlq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 50)); \
27829	27850	__ret; \
27830	27851	})
27831	27852	#else
27832	27853	#define vshlq_n_u32(__p0, __p1) __extension__ ({ \
27833	27854	uint32x4_t __ret; \
27834	27855	uint32x4_t __s0 = __p0; \
27835		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
27836		__ret = (uint32x4_t) __builtin_neon_vshlq_n_v((int8x16_t)__rev0, __p1, 50); \
27837		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	27856	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	27857	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vshlq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 50)); \
	27858	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
27838	27859	__ret; \
27839	27860	})
27840	27861	#endif
...	...	@@ -27843,16 +27864,16 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
27843	27864	#define vshlq_n_u64(__p0, __p1) __extension__ ({ \
27844	27865	uint64x2_t __ret; \
27845	27866	uint64x2_t __s0 = __p0; \
27846		__ret = (uint64x2_t) __builtin_neon_vshlq_n_v((int8x16_t)__s0, __p1, 51); \
	27867	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vshlq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 51)); \
27847	27868	__ret; \
27848	27869	})
27849	27870	#else
27850	27871	#define vshlq_n_u64(__p0, __p1) __extension__ ({ \
27851	27872	uint64x2_t __ret; \
27852	27873	uint64x2_t __s0 = __p0; \
27853		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
27854		__ret = (uint64x2_t) __builtin_neon_vshlq_n_v((int8x16_t)__rev0, __p1, 51); \
27855		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	27874	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	27875	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vshlq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 51)); \
	27876	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
27856	27877	__ret; \
27857	27878	})
27858	27879	#endif
...	...	@@ -27861,16 +27882,16 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
27861	27882	#define vshlq_n_u16(__p0, __p1) __extension__ ({ \
27862	27883	uint16x8_t __ret; \
27863	27884	uint16x8_t __s0 = __p0; \
27864		__ret = (uint16x8_t) __builtin_neon_vshlq_n_v((int8x16_t)__s0, __p1, 49); \
	27885	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vshlq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 49)); \
27865	27886	__ret; \
27866	27887	})
27867	27888	#else
27868	27889	#define vshlq_n_u16(__p0, __p1) __extension__ ({ \
27869	27890	uint16x8_t __ret; \
27870	27891	uint16x8_t __s0 = __p0; \
27871		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
27872		__ret = (uint16x8_t) __builtin_neon_vshlq_n_v((int8x16_t)__rev0, __p1, 49); \
27873		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	27892	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	27893	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vshlq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 49)); \
	27894	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
27874	27895	__ret; \
27875	27896	})
27876	27897	#endif
...	...	@@ -27879,16 +27900,16 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
27879	27900	#define vshlq_n_s8(__p0, __p1) __extension__ ({ \
27880	27901	int8x16_t __ret; \
27881	27902	int8x16_t __s0 = __p0; \
27882		__ret = (int8x16_t) __builtin_neon_vshlq_n_v((int8x16_t)__s0, __p1, 32); \
	27903	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vshlq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 32)); \
27883	27904	__ret; \
27884	27905	})
27885	27906	#else
27886	27907	#define vshlq_n_s8(__p0, __p1) __extension__ ({ \
27887	27908	int8x16_t __ret; \
27888	27909	int8x16_t __s0 = __p0; \
27889		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27890		__ret = (int8x16_t) __builtin_neon_vshlq_n_v((int8x16_t)__rev0, __p1, 32); \
27891		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	27910	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	27911	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vshlq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 32)); \
	27912	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
27892	27913	__ret; \
27893	27914	})
27894	27915	#endif
...	...	@@ -27897,16 +27918,16 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
27897	27918	#define vshlq_n_s32(__p0, __p1) __extension__ ({ \
27898	27919	int32x4_t __ret; \
27899	27920	int32x4_t __s0 = __p0; \
27900		__ret = (int32x4_t) __builtin_neon_vshlq_n_v((int8x16_t)__s0, __p1, 34); \
	27921	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vshlq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 34)); \
27901	27922	__ret; \
27902	27923	})
27903	27924	#else
27904	27925	#define vshlq_n_s32(__p0, __p1) __extension__ ({ \
27905	27926	int32x4_t __ret; \
27906	27927	int32x4_t __s0 = __p0; \
27907		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
27908		__ret = (int32x4_t) __builtin_neon_vshlq_n_v((int8x16_t)__rev0, __p1, 34); \
27909		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	27928	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	27929	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vshlq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 34)); \
	27930	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
27910	27931	__ret; \
27911	27932	})
27912	27933	#endif
...	...	@@ -27915,16 +27936,16 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
27915	27936	#define vshlq_n_s64(__p0, __p1) __extension__ ({ \
27916	27937	int64x2_t __ret; \
27917	27938	int64x2_t __s0 = __p0; \
27918		__ret = (int64x2_t) __builtin_neon_vshlq_n_v((int8x16_t)__s0, __p1, 35); \
	27939	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vshlq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 35)); \
27919	27940	__ret; \
27920	27941	})
27921	27942	#else
27922	27943	#define vshlq_n_s64(__p0, __p1) __extension__ ({ \
27923	27944	int64x2_t __ret; \
27924	27945	int64x2_t __s0 = __p0; \
27925		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
27926		__ret = (int64x2_t) __builtin_neon_vshlq_n_v((int8x16_t)__rev0, __p1, 35); \
27927		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	27946	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	27947	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vshlq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 35)); \
	27948	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
27928	27949	__ret; \
27929	27950	})
27930	27951	#endif
...	...	@@ -27933,16 +27954,16 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
27933	27954	#define vshlq_n_s16(__p0, __p1) __extension__ ({ \
27934	27955	int16x8_t __ret; \
27935	27956	int16x8_t __s0 = __p0; \
27936		__ret = (int16x8_t) __builtin_neon_vshlq_n_v((int8x16_t)__s0, __p1, 33); \
	27957	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vshlq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 33)); \
27937	27958	__ret; \
27938	27959	})
27939	27960	#else
27940	27961	#define vshlq_n_s16(__p0, __p1) __extension__ ({ \
27941	27962	int16x8_t __ret; \
27942	27963	int16x8_t __s0 = __p0; \
27943		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
27944		__ret = (int16x8_t) __builtin_neon_vshlq_n_v((int8x16_t)__rev0, __p1, 33); \
27945		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	27964	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	27965	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vshlq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 33)); \
	27966	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
27946	27967	__ret; \
27947	27968	})
27948	27969	#endif
...	...	@@ -27951,16 +27972,16 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
27951	27972	#define vshl_n_u8(__p0, __p1) __extension__ ({ \
27952	27973	uint8x8_t __ret; \
27953	27974	uint8x8_t __s0 = __p0; \
27954		__ret = (uint8x8_t) __builtin_neon_vshl_n_v((int8x8_t)__s0, __p1, 16); \
	27975	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vshl_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 16)); \
27955	27976	__ret; \
27956	27977	})
27957	27978	#else
27958	27979	#define vshl_n_u8(__p0, __p1) __extension__ ({ \
27959	27980	uint8x8_t __ret; \
27960	27981	uint8x8_t __s0 = __p0; \
27961		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
27962		__ret = (uint8x8_t) __builtin_neon_vshl_n_v((int8x8_t)__rev0, __p1, 16); \
27963		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	27982	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	27983	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vshl_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 16)); \
	27984	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
27964	27985	__ret; \
27965	27986	})
27966	27987	#endif
...	...	@@ -27969,16 +27990,16 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
27969	27990	#define vshl_n_u32(__p0, __p1) __extension__ ({ \
27970	27991	uint32x2_t __ret; \
27971	27992	uint32x2_t __s0 = __p0; \
27972		__ret = (uint32x2_t) __builtin_neon_vshl_n_v((int8x8_t)__s0, __p1, 18); \
	27993	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vshl_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 18)); \
27973	27994	__ret; \
27974	27995	})
27975	27996	#else
27976	27997	#define vshl_n_u32(__p0, __p1) __extension__ ({ \
27977	27998	uint32x2_t __ret; \
27978	27999	uint32x2_t __s0 = __p0; \
27979		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
27980		__ret = (uint32x2_t) __builtin_neon_vshl_n_v((int8x8_t)__rev0, __p1, 18); \
27981		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	28000	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	28001	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vshl_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 18)); \
	28002	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
27982	28003	__ret; \
27983	28004	})
27984	28005	#endif
...	...	@@ -27986,23 +28007,23 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
27986	28007	#define vshl_n_u64(__p0, __p1) __extension__ ({ \
27987	28008	uint64x1_t __ret; \
27988	28009	uint64x1_t __s0 = __p0; \
27989		__ret = (uint64x1_t) __builtin_neon_vshl_n_v((int8x8_t)__s0, __p1, 19); \
	28010	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vshl_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 19)); \
27990	28011	__ret; \
27991	28012	})
27992	28013	#ifdef __LITTLE_ENDIAN__
27993	28014	#define vshl_n_u16(__p0, __p1) __extension__ ({ \
27994	28015	uint16x4_t __ret; \
27995	28016	uint16x4_t __s0 = __p0; \
27996		__ret = (uint16x4_t) __builtin_neon_vshl_n_v((int8x8_t)__s0, __p1, 17); \
	28017	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vshl_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 17)); \
27997	28018	__ret; \
27998	28019	})
27999	28020	#else
28000	28021	#define vshl_n_u16(__p0, __p1) __extension__ ({ \
28001	28022	uint16x4_t __ret; \
28002	28023	uint16x4_t __s0 = __p0; \
28003		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
28004		__ret = (uint16x4_t) __builtin_neon_vshl_n_v((int8x8_t)__rev0, __p1, 17); \
28005		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	28024	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	28025	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vshl_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 17)); \
	28026	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
28006	28027	__ret; \
28007	28028	})
28008	28029	#endif
...	...	@@ -28011,16 +28032,16 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28011	28032	#define vshl_n_s8(__p0, __p1) __extension__ ({ \
28012	28033	int8x8_t __ret; \
28013	28034	int8x8_t __s0 = __p0; \
28014		__ret = (int8x8_t) __builtin_neon_vshl_n_v((int8x8_t)__s0, __p1, 0); \
	28035	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vshl_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 0)); \
28015	28036	__ret; \
28016	28037	})
28017	28038	#else
28018	28039	#define vshl_n_s8(__p0, __p1) __extension__ ({ \
28019	28040	int8x8_t __ret; \
28020	28041	int8x8_t __s0 = __p0; \
28021		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
28022		__ret = (int8x8_t) __builtin_neon_vshl_n_v((int8x8_t)__rev0, __p1, 0); \
28023		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	28042	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	28043	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vshl_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 0)); \
	28044	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
28024	28045	__ret; \
28025	28046	})
28026	28047	#endif
...	...	@@ -28029,16 +28050,16 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28029	28050	#define vshl_n_s32(__p0, __p1) __extension__ ({ \
28030	28051	int32x2_t __ret; \
28031	28052	int32x2_t __s0 = __p0; \
28032		__ret = (int32x2_t) __builtin_neon_vshl_n_v((int8x8_t)__s0, __p1, 2); \
	28053	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vshl_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 2)); \
28033	28054	__ret; \
28034	28055	})
28035	28056	#else
28036	28057	#define vshl_n_s32(__p0, __p1) __extension__ ({ \
28037	28058	int32x2_t __ret; \
28038	28059	int32x2_t __s0 = __p0; \
28039		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
28040		__ret = (int32x2_t) __builtin_neon_vshl_n_v((int8x8_t)__rev0, __p1, 2); \
28041		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	28060	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	28061	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vshl_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 2)); \
	28062	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
28042	28063	__ret; \
28043	28064	})
28044	28065	#endif
...	...	@@ -28046,23 +28067,23 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28046	28067	#define vshl_n_s64(__p0, __p1) __extension__ ({ \
28047	28068	int64x1_t __ret; \
28048	28069	int64x1_t __s0 = __p0; \
28049		__ret = (int64x1_t) __builtin_neon_vshl_n_v((int8x8_t)__s0, __p1, 3); \
	28070	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vshl_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 3)); \
28050	28071	__ret; \
28051	28072	})
28052	28073	#ifdef __LITTLE_ENDIAN__
28053	28074	#define vshl_n_s16(__p0, __p1) __extension__ ({ \
28054	28075	int16x4_t __ret; \
28055	28076	int16x4_t __s0 = __p0; \
28056		__ret = (int16x4_t) __builtin_neon_vshl_n_v((int8x8_t)__s0, __p1, 1); \
	28077	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vshl_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 1)); \
28057	28078	__ret; \
28058	28079	})
28059	28080	#else
28060	28081	#define vshl_n_s16(__p0, __p1) __extension__ ({ \
28061	28082	int16x4_t __ret; \
28062	28083	int16x4_t __s0 = __p0; \
28063		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
28064		__ret = (int16x4_t) __builtin_neon_vshl_n_v((int8x8_t)__rev0, __p1, 1); \
28065		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	28084	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	28085	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vshl_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 1)); \
	28086	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
28066	28087	__ret; \
28067	28088	})
28068	28089	#endif
...	...	@@ -28071,22 +28092,22 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28071	28092	#define vshll_n_u8(__p0, __p1) __extension__ ({ \
28072	28093	uint16x8_t __ret; \
28073	28094	uint8x8_t __s0 = __p0; \
28074		__ret = (uint16x8_t) __builtin_neon_vshll_n_v((int8x8_t)__s0, __p1, 49); \
	28095	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vshll_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 49)); \
28075	28096	__ret; \
28076	28097	})
28077	28098	#else
28078	28099	#define vshll_n_u8(__p0, __p1) __extension__ ({ \
28079	28100	uint16x8_t __ret; \
28080	28101	uint8x8_t __s0 = __p0; \
28081		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
28082		__ret = (uint16x8_t) __builtin_neon_vshll_n_v((int8x8_t)__rev0, __p1, 49); \
28083		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	28102	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	28103	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vshll_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 49)); \
	28104	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
28084	28105	__ret; \
28085	28106	})
28086	28107	#define __noswap_vshll_n_u8(__p0, __p1) __extension__ ({ \
28087	28108	uint16x8_t __ret; \
28088	28109	uint8x8_t __s0 = __p0; \
28089		__ret = (uint16x8_t) __builtin_neon_vshll_n_v((int8x8_t)__s0, __p1, 49); \
	28110	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vshll_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 49)); \
28090	28111	__ret; \
28091	28112	})
28092	28113	#endif
...	...	@@ -28095,22 +28116,22 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28095	28116	#define vshll_n_u32(__p0, __p1) __extension__ ({ \
28096	28117	uint64x2_t __ret; \
28097	28118	uint32x2_t __s0 = __p0; \
28098		__ret = (uint64x2_t) __builtin_neon_vshll_n_v((int8x8_t)__s0, __p1, 51); \
	28119	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vshll_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 51)); \
28099	28120	__ret; \
28100	28121	})
28101	28122	#else
28102	28123	#define vshll_n_u32(__p0, __p1) __extension__ ({ \
28103	28124	uint64x2_t __ret; \
28104	28125	uint32x2_t __s0 = __p0; \
28105		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
28106		__ret = (uint64x2_t) __builtin_neon_vshll_n_v((int8x8_t)__rev0, __p1, 51); \
28107		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	28126	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	28127	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vshll_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 51)); \
	28128	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
28108	28129	__ret; \
28109	28130	})
28110	28131	#define __noswap_vshll_n_u32(__p0, __p1) __extension__ ({ \
28111	28132	uint64x2_t __ret; \
28112	28133	uint32x2_t __s0 = __p0; \
28113		__ret = (uint64x2_t) __builtin_neon_vshll_n_v((int8x8_t)__s0, __p1, 51); \
	28134	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vshll_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 51)); \
28114	28135	__ret; \
28115	28136	})
28116	28137	#endif
...	...	@@ -28119,22 +28140,22 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28119	28140	#define vshll_n_u16(__p0, __p1) __extension__ ({ \
28120	28141	uint32x4_t __ret; \
28121	28142	uint16x4_t __s0 = __p0; \
28122		__ret = (uint32x4_t) __builtin_neon_vshll_n_v((int8x8_t)__s0, __p1, 50); \
	28143	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vshll_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 50)); \
28123	28144	__ret; \
28124	28145	})
28125	28146	#else
28126	28147	#define vshll_n_u16(__p0, __p1) __extension__ ({ \
28127	28148	uint32x4_t __ret; \
28128	28149	uint16x4_t __s0 = __p0; \
28129		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
28130		__ret = (uint32x4_t) __builtin_neon_vshll_n_v((int8x8_t)__rev0, __p1, 50); \
28131		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	28150	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	28151	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vshll_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 50)); \
	28152	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
28132	28153	__ret; \
28133	28154	})
28134	28155	#define __noswap_vshll_n_u16(__p0, __p1) __extension__ ({ \
28135	28156	uint32x4_t __ret; \
28136	28157	uint16x4_t __s0 = __p0; \
28137		__ret = (uint32x4_t) __builtin_neon_vshll_n_v((int8x8_t)__s0, __p1, 50); \
	28158	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vshll_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 50)); \
28138	28159	__ret; \
28139	28160	})
28140	28161	#endif
...	...	@@ -28143,22 +28164,22 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28143	28164	#define vshll_n_s8(__p0, __p1) __extension__ ({ \
28144	28165	int16x8_t __ret; \
28145	28166	int8x8_t __s0 = __p0; \
28146		__ret = (int16x8_t) __builtin_neon_vshll_n_v((int8x8_t)__s0, __p1, 33); \
	28167	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vshll_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 33)); \
28147	28168	__ret; \
28148	28169	})
28149	28170	#else
28150	28171	#define vshll_n_s8(__p0, __p1) __extension__ ({ \
28151	28172	int16x8_t __ret; \
28152	28173	int8x8_t __s0 = __p0; \
28153		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
28154		__ret = (int16x8_t) __builtin_neon_vshll_n_v((int8x8_t)__rev0, __p1, 33); \
28155		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	28174	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	28175	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vshll_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 33)); \
	28176	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
28156	28177	__ret; \
28157	28178	})
28158	28179	#define __noswap_vshll_n_s8(__p0, __p1) __extension__ ({ \
28159	28180	int16x8_t __ret; \
28160	28181	int8x8_t __s0 = __p0; \
28161		__ret = (int16x8_t) __builtin_neon_vshll_n_v((int8x8_t)__s0, __p1, 33); \
	28182	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vshll_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 33)); \
28162	28183	__ret; \
28163	28184	})
28164	28185	#endif
...	...	@@ -28167,22 +28188,22 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28167	28188	#define vshll_n_s32(__p0, __p1) __extension__ ({ \
28168	28189	int64x2_t __ret; \
28169	28190	int32x2_t __s0 = __p0; \
28170		__ret = (int64x2_t) __builtin_neon_vshll_n_v((int8x8_t)__s0, __p1, 35); \
	28191	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vshll_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 35)); \
28171	28192	__ret; \
28172	28193	})
28173	28194	#else
28174	28195	#define vshll_n_s32(__p0, __p1) __extension__ ({ \
28175	28196	int64x2_t __ret; \
28176	28197	int32x2_t __s0 = __p0; \
28177		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
28178		__ret = (int64x2_t) __builtin_neon_vshll_n_v((int8x8_t)__rev0, __p1, 35); \
28179		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	28198	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	28199	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vshll_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 35)); \
	28200	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
28180	28201	__ret; \
28181	28202	})
28182	28203	#define __noswap_vshll_n_s32(__p0, __p1) __extension__ ({ \
28183	28204	int64x2_t __ret; \
28184	28205	int32x2_t __s0 = __p0; \
28185		__ret = (int64x2_t) __builtin_neon_vshll_n_v((int8x8_t)__s0, __p1, 35); \
	28206	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vshll_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 35)); \
28186	28207	__ret; \
28187	28208	})
28188	28209	#endif
...	...	@@ -28191,22 +28212,22 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28191	28212	#define vshll_n_s16(__p0, __p1) __extension__ ({ \
28192	28213	int32x4_t __ret; \
28193	28214	int16x4_t __s0 = __p0; \
28194		__ret = (int32x4_t) __builtin_neon_vshll_n_v((int8x8_t)__s0, __p1, 34); \
	28215	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vshll_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 34)); \
28195	28216	__ret; \
28196	28217	})
28197	28218	#else
28198	28219	#define vshll_n_s16(__p0, __p1) __extension__ ({ \
28199	28220	int32x4_t __ret; \
28200	28221	int16x4_t __s0 = __p0; \
28201		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
28202		__ret = (int32x4_t) __builtin_neon_vshll_n_v((int8x8_t)__rev0, __p1, 34); \
28203		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	28222	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	28223	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vshll_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 34)); \
	28224	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
28204	28225	__ret; \
28205	28226	})
28206	28227	#define __noswap_vshll_n_s16(__p0, __p1) __extension__ ({ \
28207	28228	int32x4_t __ret; \
28208	28229	int16x4_t __s0 = __p0; \
28209		__ret = (int32x4_t) __builtin_neon_vshll_n_v((int8x8_t)__s0, __p1, 34); \
	28230	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vshll_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 34)); \
28210	28231	__ret; \
28211	28232	})
28212	28233	#endif
...	...	@@ -28215,16 +28236,16 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28215	28236	#define vshrq_n_u8(__p0, __p1) __extension__ ({ \
28216	28237	uint8x16_t __ret; \
28217	28238	uint8x16_t __s0 = __p0; \
28218		__ret = (uint8x16_t) __builtin_neon_vshrq_n_v((int8x16_t)__s0, __p1, 48); \
	28239	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vshrq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 48)); \
28219	28240	__ret; \
28220	28241	})
28221	28242	#else
28222	28243	#define vshrq_n_u8(__p0, __p1) __extension__ ({ \
28223	28244	uint8x16_t __ret; \
28224	28245	uint8x16_t __s0 = __p0; \
28225		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
28226		__ret = (uint8x16_t) __builtin_neon_vshrq_n_v((int8x16_t)__rev0, __p1, 48); \
28227		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	28246	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	28247	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vshrq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 48)); \
	28248	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
28228	28249	__ret; \
28229	28250	})
28230	28251	#endif
...	...	@@ -28233,16 +28254,16 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28233	28254	#define vshrq_n_u32(__p0, __p1) __extension__ ({ \
28234	28255	uint32x4_t __ret; \
28235	28256	uint32x4_t __s0 = __p0; \
28236		__ret = (uint32x4_t) __builtin_neon_vshrq_n_v((int8x16_t)__s0, __p1, 50); \
	28257	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vshrq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 50)); \
28237	28258	__ret; \
28238	28259	})
28239	28260	#else
28240	28261	#define vshrq_n_u32(__p0, __p1) __extension__ ({ \
28241	28262	uint32x4_t __ret; \
28242	28263	uint32x4_t __s0 = __p0; \
28243		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
28244		__ret = (uint32x4_t) __builtin_neon_vshrq_n_v((int8x16_t)__rev0, __p1, 50); \
28245		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	28264	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	28265	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vshrq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 50)); \
	28266	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
28246	28267	__ret; \
28247	28268	})
28248	28269	#endif
...	...	@@ -28251,16 +28272,16 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28251	28272	#define vshrq_n_u64(__p0, __p1) __extension__ ({ \
28252	28273	uint64x2_t __ret; \
28253	28274	uint64x2_t __s0 = __p0; \
28254		__ret = (uint64x2_t) __builtin_neon_vshrq_n_v((int8x16_t)__s0, __p1, 51); \
	28275	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vshrq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 51)); \
28255	28276	__ret; \
28256	28277	})
28257	28278	#else
28258	28279	#define vshrq_n_u64(__p0, __p1) __extension__ ({ \
28259	28280	uint64x2_t __ret; \
28260	28281	uint64x2_t __s0 = __p0; \
28261		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
28262		__ret = (uint64x2_t) __builtin_neon_vshrq_n_v((int8x16_t)__rev0, __p1, 51); \
28263		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	28282	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	28283	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vshrq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 51)); \
	28284	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
28264	28285	__ret; \
28265	28286	})
28266	28287	#endif
...	...	@@ -28269,16 +28290,16 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28269	28290	#define vshrq_n_u16(__p0, __p1) __extension__ ({ \
28270	28291	uint16x8_t __ret; \
28271	28292	uint16x8_t __s0 = __p0; \
28272		__ret = (uint16x8_t) __builtin_neon_vshrq_n_v((int8x16_t)__s0, __p1, 49); \
	28293	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vshrq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 49)); \
28273	28294	__ret; \
28274	28295	})
28275	28296	#else
28276	28297	#define vshrq_n_u16(__p0, __p1) __extension__ ({ \
28277	28298	uint16x8_t __ret; \
28278	28299	uint16x8_t __s0 = __p0; \
28279		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
28280		__ret = (uint16x8_t) __builtin_neon_vshrq_n_v((int8x16_t)__rev0, __p1, 49); \
28281		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	28300	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	28301	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vshrq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 49)); \
	28302	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
28282	28303	__ret; \
28283	28304	})
28284	28305	#endif
...	...	@@ -28287,16 +28308,16 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28287	28308	#define vshrq_n_s8(__p0, __p1) __extension__ ({ \
28288	28309	int8x16_t __ret; \
28289	28310	int8x16_t __s0 = __p0; \
28290		__ret = (int8x16_t) __builtin_neon_vshrq_n_v((int8x16_t)__s0, __p1, 32); \
	28311	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vshrq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 32)); \
28291	28312	__ret; \
28292	28313	})
28293	28314	#else
28294	28315	#define vshrq_n_s8(__p0, __p1) __extension__ ({ \
28295	28316	int8x16_t __ret; \
28296	28317	int8x16_t __s0 = __p0; \
28297		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
28298		__ret = (int8x16_t) __builtin_neon_vshrq_n_v((int8x16_t)__rev0, __p1, 32); \
28299		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	28318	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	28319	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vshrq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 32)); \
	28320	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
28300	28321	__ret; \
28301	28322	})
28302	28323	#endif
...	...	@@ -28305,16 +28326,16 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28305	28326	#define vshrq_n_s32(__p0, __p1) __extension__ ({ \
28306	28327	int32x4_t __ret; \
28307	28328	int32x4_t __s0 = __p0; \
28308		__ret = (int32x4_t) __builtin_neon_vshrq_n_v((int8x16_t)__s0, __p1, 34); \
	28329	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vshrq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 34)); \
28309	28330	__ret; \
28310	28331	})
28311	28332	#else
28312	28333	#define vshrq_n_s32(__p0, __p1) __extension__ ({ \
28313	28334	int32x4_t __ret; \
28314	28335	int32x4_t __s0 = __p0; \
28315		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
28316		__ret = (int32x4_t) __builtin_neon_vshrq_n_v((int8x16_t)__rev0, __p1, 34); \
28317		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	28336	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	28337	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vshrq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 34)); \
	28338	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
28318	28339	__ret; \
28319	28340	})
28320	28341	#endif
...	...	@@ -28323,16 +28344,16 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28323	28344	#define vshrq_n_s64(__p0, __p1) __extension__ ({ \
28324	28345	int64x2_t __ret; \
28325	28346	int64x2_t __s0 = __p0; \
28326		__ret = (int64x2_t) __builtin_neon_vshrq_n_v((int8x16_t)__s0, __p1, 35); \
	28347	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vshrq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 35)); \
28327	28348	__ret; \
28328	28349	})
28329	28350	#else
28330	28351	#define vshrq_n_s64(__p0, __p1) __extension__ ({ \
28331	28352	int64x2_t __ret; \
28332	28353	int64x2_t __s0 = __p0; \
28333		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
28334		__ret = (int64x2_t) __builtin_neon_vshrq_n_v((int8x16_t)__rev0, __p1, 35); \
28335		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	28354	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	28355	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vshrq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 35)); \
	28356	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
28336	28357	__ret; \
28337	28358	})
28338	28359	#endif
...	...	@@ -28341,16 +28362,16 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28341	28362	#define vshrq_n_s16(__p0, __p1) __extension__ ({ \
28342	28363	int16x8_t __ret; \
28343	28364	int16x8_t __s0 = __p0; \
28344		__ret = (int16x8_t) __builtin_neon_vshrq_n_v((int8x16_t)__s0, __p1, 33); \
	28365	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vshrq_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 33)); \
28345	28366	__ret; \
28346	28367	})
28347	28368	#else
28348	28369	#define vshrq_n_s16(__p0, __p1) __extension__ ({ \
28349	28370	int16x8_t __ret; \
28350	28371	int16x8_t __s0 = __p0; \
28351		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
28352		__ret = (int16x8_t) __builtin_neon_vshrq_n_v((int8x16_t)__rev0, __p1, 33); \
28353		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	28372	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	28373	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vshrq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 33)); \
	28374	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
28354	28375	__ret; \
28355	28376	})
28356	28377	#endif
...	...	@@ -28359,16 +28380,16 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28359	28380	#define vshr_n_u8(__p0, __p1) __extension__ ({ \
28360	28381	uint8x8_t __ret; \
28361	28382	uint8x8_t __s0 = __p0; \
28362		__ret = (uint8x8_t) __builtin_neon_vshr_n_v((int8x8_t)__s0, __p1, 16); \
	28383	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vshr_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 16)); \
28363	28384	__ret; \
28364	28385	})
28365	28386	#else
28366	28387	#define vshr_n_u8(__p0, __p1) __extension__ ({ \
28367	28388	uint8x8_t __ret; \
28368	28389	uint8x8_t __s0 = __p0; \
28369		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
28370		__ret = (uint8x8_t) __builtin_neon_vshr_n_v((int8x8_t)__rev0, __p1, 16); \
28371		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	28390	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	28391	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vshr_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 16)); \
	28392	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
28372	28393	__ret; \
28373	28394	})
28374	28395	#endif
...	...	@@ -28377,16 +28398,16 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28377	28398	#define vshr_n_u32(__p0, __p1) __extension__ ({ \
28378	28399	uint32x2_t __ret; \
28379	28400	uint32x2_t __s0 = __p0; \
28380		__ret = (uint32x2_t) __builtin_neon_vshr_n_v((int8x8_t)__s0, __p1, 18); \
	28401	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vshr_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 18)); \
28381	28402	__ret; \
28382	28403	})
28383	28404	#else
28384	28405	#define vshr_n_u32(__p0, __p1) __extension__ ({ \
28385	28406	uint32x2_t __ret; \
28386	28407	uint32x2_t __s0 = __p0; \
28387		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
28388		__ret = (uint32x2_t) __builtin_neon_vshr_n_v((int8x8_t)__rev0, __p1, 18); \
28389		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	28408	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	28409	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vshr_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 18)); \
	28410	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
28390	28411	__ret; \
28391	28412	})
28392	28413	#endif
...	...	@@ -28394,23 +28415,23 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28394	28415	#define vshr_n_u64(__p0, __p1) __extension__ ({ \
28395	28416	uint64x1_t __ret; \
28396	28417	uint64x1_t __s0 = __p0; \
28397		__ret = (uint64x1_t) __builtin_neon_vshr_n_v((int8x8_t)__s0, __p1, 19); \
	28418	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vshr_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 19)); \
28398	28419	__ret; \
28399	28420	})
28400	28421	#ifdef __LITTLE_ENDIAN__
28401	28422	#define vshr_n_u16(__p0, __p1) __extension__ ({ \
28402	28423	uint16x4_t __ret; \
28403	28424	uint16x4_t __s0 = __p0; \
28404		__ret = (uint16x4_t) __builtin_neon_vshr_n_v((int8x8_t)__s0, __p1, 17); \
	28425	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vshr_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 17)); \
28405	28426	__ret; \
28406	28427	})
28407	28428	#else
28408	28429	#define vshr_n_u16(__p0, __p1) __extension__ ({ \
28409	28430	uint16x4_t __ret; \
28410	28431	uint16x4_t __s0 = __p0; \
28411		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
28412		__ret = (uint16x4_t) __builtin_neon_vshr_n_v((int8x8_t)__rev0, __p1, 17); \
28413		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	28432	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	28433	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vshr_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 17)); \
	28434	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
28414	28435	__ret; \
28415	28436	})
28416	28437	#endif
...	...	@@ -28419,16 +28440,16 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28419	28440	#define vshr_n_s8(__p0, __p1) __extension__ ({ \
28420	28441	int8x8_t __ret; \
28421	28442	int8x8_t __s0 = __p0; \
28422		__ret = (int8x8_t) __builtin_neon_vshr_n_v((int8x8_t)__s0, __p1, 0); \
	28443	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vshr_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 0)); \
28423	28444	__ret; \
28424	28445	})
28425	28446	#else
28426	28447	#define vshr_n_s8(__p0, __p1) __extension__ ({ \
28427	28448	int8x8_t __ret; \
28428	28449	int8x8_t __s0 = __p0; \
28429		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
28430		__ret = (int8x8_t) __builtin_neon_vshr_n_v((int8x8_t)__rev0, __p1, 0); \
28431		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	28450	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	28451	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vshr_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 0)); \
	28452	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
28432	28453	__ret; \
28433	28454	})
28434	28455	#endif
...	...	@@ -28437,16 +28458,16 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28437	28458	#define vshr_n_s32(__p0, __p1) __extension__ ({ \
28438	28459	int32x2_t __ret; \
28439	28460	int32x2_t __s0 = __p0; \
28440		__ret = (int32x2_t) __builtin_neon_vshr_n_v((int8x8_t)__s0, __p1, 2); \
	28461	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vshr_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 2)); \
28441	28462	__ret; \
28442	28463	})
28443	28464	#else
28444	28465	#define vshr_n_s32(__p0, __p1) __extension__ ({ \
28445	28466	int32x2_t __ret; \
28446	28467	int32x2_t __s0 = __p0; \
28447		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
28448		__ret = (int32x2_t) __builtin_neon_vshr_n_v((int8x8_t)__rev0, __p1, 2); \
28449		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	28468	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	28469	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vshr_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 2)); \
	28470	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
28450	28471	__ret; \
28451	28472	})
28452	28473	#endif
...	...	@@ -28454,23 +28475,23 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28454	28475	#define vshr_n_s64(__p0, __p1) __extension__ ({ \
28455	28476	int64x1_t __ret; \
28456	28477	int64x1_t __s0 = __p0; \
28457		__ret = (int64x1_t) __builtin_neon_vshr_n_v((int8x8_t)__s0, __p1, 3); \
	28478	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vshr_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 3)); \
28458	28479	__ret; \
28459	28480	})
28460	28481	#ifdef __LITTLE_ENDIAN__
28461	28482	#define vshr_n_s16(__p0, __p1) __extension__ ({ \
28462	28483	int16x4_t __ret; \
28463	28484	int16x4_t __s0 = __p0; \
28464		__ret = (int16x4_t) __builtin_neon_vshr_n_v((int8x8_t)__s0, __p1, 1); \
	28485	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vshr_n_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 1)); \
28465	28486	__ret; \
28466	28487	})
28467	28488	#else
28468	28489	#define vshr_n_s16(__p0, __p1) __extension__ ({ \
28469	28490	int16x4_t __ret; \
28470	28491	int16x4_t __s0 = __p0; \
28471		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
28472		__ret = (int16x4_t) __builtin_neon_vshr_n_v((int8x8_t)__rev0, __p1, 1); \
28473		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	28492	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	28493	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vshr_n_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 1)); \
	28494	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
28474	28495	__ret; \
28475	28496	})
28476	28497	#endif
...	...	@@ -28479,22 +28500,22 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28479	28500	#define vshrn_n_u32(__p0, __p1) __extension__ ({ \
28480	28501	uint16x4_t __ret; \
28481	28502	uint32x4_t __s0 = __p0; \
28482		__ret = (uint16x4_t) __builtin_neon_vshrn_n_v((int8x16_t)__s0, __p1, 17); \
	28503	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 17)); \
28483	28504	__ret; \
28484	28505	})
28485	28506	#else
28486	28507	#define vshrn_n_u32(__p0, __p1) __extension__ ({ \
28487	28508	uint16x4_t __ret; \
28488	28509	uint32x4_t __s0 = __p0; \
28489		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
28490		__ret = (uint16x4_t) __builtin_neon_vshrn_n_v((int8x16_t)__rev0, __p1, 17); \
28491		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	28510	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	28511	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vshrn_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 17)); \
	28512	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
28492	28513	__ret; \
28493	28514	})
28494	28515	#define __noswap_vshrn_n_u32(__p0, __p1) __extension__ ({ \
28495	28516	uint16x4_t __ret; \
28496	28517	uint32x4_t __s0 = __p0; \
28497		__ret = (uint16x4_t) __builtin_neon_vshrn_n_v((int8x16_t)__s0, __p1, 17); \
	28518	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 17)); \
28498	28519	__ret; \
28499	28520	})
28500	28521	#endif
...	...	@@ -28503,22 +28524,22 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28503	28524	#define vshrn_n_u64(__p0, __p1) __extension__ ({ \
28504	28525	uint32x2_t __ret; \
28505	28526	uint64x2_t __s0 = __p0; \
28506		__ret = (uint32x2_t) __builtin_neon_vshrn_n_v((int8x16_t)__s0, __p1, 18); \
	28527	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 18)); \
28507	28528	__ret; \
28508	28529	})
28509	28530	#else
28510	28531	#define vshrn_n_u64(__p0, __p1) __extension__ ({ \
28511	28532	uint32x2_t __ret; \
28512	28533	uint64x2_t __s0 = __p0; \
28513		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
28514		__ret = (uint32x2_t) __builtin_neon_vshrn_n_v((int8x16_t)__rev0, __p1, 18); \
28515		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	28534	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	28535	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vshrn_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 18)); \
	28536	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
28516	28537	__ret; \
28517	28538	})
28518	28539	#define __noswap_vshrn_n_u64(__p0, __p1) __extension__ ({ \
28519	28540	uint32x2_t __ret; \
28520	28541	uint64x2_t __s0 = __p0; \
28521		__ret = (uint32x2_t) __builtin_neon_vshrn_n_v((int8x16_t)__s0, __p1, 18); \
	28542	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 18)); \
28522	28543	__ret; \
28523	28544	})
28524	28545	#endif
...	...	@@ -28527,22 +28548,22 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28527	28548	#define vshrn_n_u16(__p0, __p1) __extension__ ({ \
28528	28549	uint8x8_t __ret; \
28529	28550	uint16x8_t __s0 = __p0; \
28530		__ret = (uint8x8_t) __builtin_neon_vshrn_n_v((int8x16_t)__s0, __p1, 16); \
	28551	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 16)); \
28531	28552	__ret; \
28532	28553	})
28533	28554	#else
28534	28555	#define vshrn_n_u16(__p0, __p1) __extension__ ({ \
28535	28556	uint8x8_t __ret; \
28536	28557	uint16x8_t __s0 = __p0; \
28537		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
28538		__ret = (uint8x8_t) __builtin_neon_vshrn_n_v((int8x16_t)__rev0, __p1, 16); \
28539		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	28558	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	28559	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vshrn_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 16)); \
	28560	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
28540	28561	__ret; \
28541	28562	})
28542	28563	#define __noswap_vshrn_n_u16(__p0, __p1) __extension__ ({ \
28543	28564	uint8x8_t __ret; \
28544	28565	uint16x8_t __s0 = __p0; \
28545		__ret = (uint8x8_t) __builtin_neon_vshrn_n_v((int8x16_t)__s0, __p1, 16); \
	28566	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 16)); \
28546	28567	__ret; \
28547	28568	})
28548	28569	#endif
...	...	@@ -28551,22 +28572,22 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28551	28572	#define vshrn_n_s32(__p0, __p1) __extension__ ({ \
28552	28573	int16x4_t __ret; \
28553	28574	int32x4_t __s0 = __p0; \
28554		__ret = (int16x4_t) __builtin_neon_vshrn_n_v((int8x16_t)__s0, __p1, 1); \
	28575	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 1)); \
28555	28576	__ret; \
28556	28577	})
28557	28578	#else
28558	28579	#define vshrn_n_s32(__p0, __p1) __extension__ ({ \
28559	28580	int16x4_t __ret; \
28560	28581	int32x4_t __s0 = __p0; \
28561		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
28562		__ret = (int16x4_t) __builtin_neon_vshrn_n_v((int8x16_t)__rev0, __p1, 1); \
28563		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	28582	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	28583	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vshrn_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 1)); \
	28584	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
28564	28585	__ret; \
28565	28586	})
28566	28587	#define __noswap_vshrn_n_s32(__p0, __p1) __extension__ ({ \
28567	28588	int16x4_t __ret; \
28568	28589	int32x4_t __s0 = __p0; \
28569		__ret = (int16x4_t) __builtin_neon_vshrn_n_v((int8x16_t)__s0, __p1, 1); \
	28590	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 1)); \
28570	28591	__ret; \
28571	28592	})
28572	28593	#endif
...	...	@@ -28575,22 +28596,22 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28575	28596	#define vshrn_n_s64(__p0, __p1) __extension__ ({ \
28576	28597	int32x2_t __ret; \
28577	28598	int64x2_t __s0 = __p0; \
28578		__ret = (int32x2_t) __builtin_neon_vshrn_n_v((int8x16_t)__s0, __p1, 2); \
	28599	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 2)); \
28579	28600	__ret; \
28580	28601	})
28581	28602	#else
28582	28603	#define vshrn_n_s64(__p0, __p1) __extension__ ({ \
28583	28604	int32x2_t __ret; \
28584	28605	int64x2_t __s0 = __p0; \
28585		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
28586		__ret = (int32x2_t) __builtin_neon_vshrn_n_v((int8x16_t)__rev0, __p1, 2); \
28587		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	28606	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	28607	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vshrn_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 2)); \
	28608	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
28588	28609	__ret; \
28589	28610	})
28590	28611	#define __noswap_vshrn_n_s64(__p0, __p1) __extension__ ({ \
28591	28612	int32x2_t __ret; \
28592	28613	int64x2_t __s0 = __p0; \
28593		__ret = (int32x2_t) __builtin_neon_vshrn_n_v((int8x16_t)__s0, __p1, 2); \
	28614	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 2)); \
28594	28615	__ret; \
28595	28616	})
28596	28617	#endif
...	...	@@ -28599,22 +28620,22 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28599	28620	#define vshrn_n_s16(__p0, __p1) __extension__ ({ \
28600	28621	int8x8_t __ret; \
28601	28622	int16x8_t __s0 = __p0; \
28602		__ret = (int8x8_t) __builtin_neon_vshrn_n_v((int8x16_t)__s0, __p1, 0); \
	28623	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 0)); \
28603	28624	__ret; \
28604	28625	})
28605	28626	#else
28606	28627	#define vshrn_n_s16(__p0, __p1) __extension__ ({ \
28607	28628	int8x8_t __ret; \
28608	28629	int16x8_t __s0 = __p0; \
28609		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
28610		__ret = (int8x8_t) __builtin_neon_vshrn_n_v((int8x16_t)__rev0, __p1, 0); \
28611		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	28630	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	28631	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vshrn_n_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 0)); \
	28632	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
28612	28633	__ret; \
28613	28634	})
28614	28635	#define __noswap_vshrn_n_s16(__p0, __p1) __extension__ ({ \
28615	28636	int8x8_t __ret; \
28616	28637	int16x8_t __s0 = __p0; \
28617		__ret = (int8x8_t) __builtin_neon_vshrn_n_v((int8x16_t)__s0, __p1, 0); \
	28638	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vshrn_n_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 0)); \
28618	28639	__ret; \
28619	28640	})
28620	28641	#endif
...	...	@@ -28624,7 +28645,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28624	28645	poly8x8_t __ret; \
28625	28646	poly8x8_t __s0 = __p0; \
28626	28647	poly8x8_t __s1 = __p1; \
28627		__ret = (poly8x8_t) __builtin_neon_vsli_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 4); \
	28648	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vsli_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 4)); \
28628	28649	__ret; \
28629	28650	})
28630	28651	#else
...	...	@@ -28632,10 +28653,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28632	28653	poly8x8_t __ret; \
28633	28654	poly8x8_t __s0 = __p0; \
28634	28655	poly8x8_t __s1 = __p1; \
28635		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
28636		poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
28637		__ret = (poly8x8_t) __builtin_neon_vsli_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 4); \
28638		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	28656	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	28657	poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	28658	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vsli_n_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 4)); \
	28659	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
28639	28660	__ret; \
28640	28661	})
28641	28662	#endif
...	...	@@ -28645,7 +28666,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28645	28666	poly16x4_t __ret; \
28646	28667	poly16x4_t __s0 = __p0; \
28647	28668	poly16x4_t __s1 = __p1; \
28648		__ret = (poly16x4_t) __builtin_neon_vsli_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 5); \
	28669	__ret = __builtin_bit_cast(poly16x4_t, __builtin_neon_vsli_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 5)); \
28649	28670	__ret; \
28650	28671	})
28651	28672	#else
...	...	@@ -28653,10 +28674,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28653	28674	poly16x4_t __ret; \
28654	28675	poly16x4_t __s0 = __p0; \
28655	28676	poly16x4_t __s1 = __p1; \
28656		poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
28657		poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
28658		__ret = (poly16x4_t) __builtin_neon_vsli_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 5); \
28659		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	28677	poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	28678	poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	28679	__ret = __builtin_bit_cast(poly16x4_t, __builtin_neon_vsli_n_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 5)); \
	28680	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
28660	28681	__ret; \
28661	28682	})
28662	28683	#endif
...	...	@@ -28666,7 +28687,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28666	28687	poly8x16_t __ret; \
28667	28688	poly8x16_t __s0 = __p0; \
28668	28689	poly8x16_t __s1 = __p1; \
28669		__ret = (poly8x16_t) __builtin_neon_vsliq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 36); \
	28690	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vsliq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 36)); \
28670	28691	__ret; \
28671	28692	})
28672	28693	#else
...	...	@@ -28674,10 +28695,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28674	28695	poly8x16_t __ret; \
28675	28696	poly8x16_t __s0 = __p0; \
28676	28697	poly8x16_t __s1 = __p1; \
28677		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
28678		poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
28679		__ret = (poly8x16_t) __builtin_neon_vsliq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 36); \
28680		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	28698	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	28699	poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	28700	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vsliq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 36)); \
	28701	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
28681	28702	__ret; \
28682	28703	})
28683	28704	#endif
...	...	@@ -28687,7 +28708,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28687	28708	poly16x8_t __ret; \
28688	28709	poly16x8_t __s0 = __p0; \
28689	28710	poly16x8_t __s1 = __p1; \
28690		__ret = (poly16x8_t) __builtin_neon_vsliq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 37); \
	28711	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vsliq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 37)); \
28691	28712	__ret; \
28692	28713	})
28693	28714	#else
...	...	@@ -28695,10 +28716,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28695	28716	poly16x8_t __ret; \
28696	28717	poly16x8_t __s0 = __p0; \
28697	28718	poly16x8_t __s1 = __p1; \
28698		poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
28699		poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
28700		__ret = (poly16x8_t) __builtin_neon_vsliq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 37); \
28701		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	28719	poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	28720	poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	28721	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vsliq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 37)); \
	28722	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
28702	28723	__ret; \
28703	28724	})
28704	28725	#endif
...	...	@@ -28708,7 +28729,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28708	28729	uint8x16_t __ret; \
28709	28730	uint8x16_t __s0 = __p0; \
28710	28731	uint8x16_t __s1 = __p1; \
28711		__ret = (uint8x16_t) __builtin_neon_vsliq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 48); \
	28732	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vsliq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 48)); \
28712	28733	__ret; \
28713	28734	})
28714	28735	#else
...	...	@@ -28716,10 +28737,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28716	28737	uint8x16_t __ret; \
28717	28738	uint8x16_t __s0 = __p0; \
28718	28739	uint8x16_t __s1 = __p1; \
28719		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
28720		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
28721		__ret = (uint8x16_t) __builtin_neon_vsliq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 48); \
28722		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	28740	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	28741	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	28742	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vsliq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 48)); \
	28743	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
28723	28744	__ret; \
28724	28745	})
28725	28746	#endif
...	...	@@ -28729,7 +28750,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28729	28750	uint32x4_t __ret; \
28730	28751	uint32x4_t __s0 = __p0; \
28731	28752	uint32x4_t __s1 = __p1; \
28732		__ret = (uint32x4_t) __builtin_neon_vsliq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 50); \
	28753	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsliq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 50)); \
28733	28754	__ret; \
28734	28755	})
28735	28756	#else
...	...	@@ -28737,10 +28758,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28737	28758	uint32x4_t __ret; \
28738	28759	uint32x4_t __s0 = __p0; \
28739	28760	uint32x4_t __s1 = __p1; \
28740		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
28741		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
28742		__ret = (uint32x4_t) __builtin_neon_vsliq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 50); \
28743		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	28761	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	28762	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	28763	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsliq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 50)); \
	28764	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
28744	28765	__ret; \
28745	28766	})
28746	28767	#endif
...	...	@@ -28750,7 +28771,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28750	28771	uint64x2_t __ret; \
28751	28772	uint64x2_t __s0 = __p0; \
28752	28773	uint64x2_t __s1 = __p1; \
28753		__ret = (uint64x2_t) __builtin_neon_vsliq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 51); \
	28774	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vsliq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 51)); \
28754	28775	__ret; \
28755	28776	})
28756	28777	#else
...	...	@@ -28758,10 +28779,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28758	28779	uint64x2_t __ret; \
28759	28780	uint64x2_t __s0 = __p0; \
28760	28781	uint64x2_t __s1 = __p1; \
28761		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
28762		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
28763		__ret = (uint64x2_t) __builtin_neon_vsliq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 51); \
28764		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	28782	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	28783	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	28784	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vsliq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 51)); \
	28785	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
28765	28786	__ret; \
28766	28787	})
28767	28788	#endif
...	...	@@ -28771,7 +28792,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28771	28792	uint16x8_t __ret; \
28772	28793	uint16x8_t __s0 = __p0; \
28773	28794	uint16x8_t __s1 = __p1; \
28774		__ret = (uint16x8_t) __builtin_neon_vsliq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 49); \
	28795	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vsliq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 49)); \
28775	28796	__ret; \
28776	28797	})
28777	28798	#else
...	...	@@ -28779,10 +28800,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28779	28800	uint16x8_t __ret; \
28780	28801	uint16x8_t __s0 = __p0; \
28781	28802	uint16x8_t __s1 = __p1; \
28782		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
28783		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
28784		__ret = (uint16x8_t) __builtin_neon_vsliq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 49); \
28785		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	28803	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	28804	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	28805	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vsliq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 49)); \
	28806	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
28786	28807	__ret; \
28787	28808	})
28788	28809	#endif
...	...	@@ -28792,7 +28813,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28792	28813	int8x16_t __ret; \
28793	28814	int8x16_t __s0 = __p0; \
28794	28815	int8x16_t __s1 = __p1; \
28795		__ret = (int8x16_t) __builtin_neon_vsliq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 32); \
	28816	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vsliq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 32)); \
28796	28817	__ret; \
28797	28818	})
28798	28819	#else
...	...	@@ -28800,10 +28821,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28800	28821	int8x16_t __ret; \
28801	28822	int8x16_t __s0 = __p0; \
28802	28823	int8x16_t __s1 = __p1; \
28803		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
28804		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
28805		__ret = (int8x16_t) __builtin_neon_vsliq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 32); \
28806		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	28824	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	28825	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	28826	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vsliq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 32)); \
	28827	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
28807	28828	__ret; \
28808	28829	})
28809	28830	#endif
...	...	@@ -28813,7 +28834,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28813	28834	int32x4_t __ret; \
28814	28835	int32x4_t __s0 = __p0; \
28815	28836	int32x4_t __s1 = __p1; \
28816		__ret = (int32x4_t) __builtin_neon_vsliq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 34); \
	28837	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vsliq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 34)); \
28817	28838	__ret; \
28818	28839	})
28819	28840	#else
...	...	@@ -28821,10 +28842,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28821	28842	int32x4_t __ret; \
28822	28843	int32x4_t __s0 = __p0; \
28823	28844	int32x4_t __s1 = __p1; \
28824		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
28825		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
28826		__ret = (int32x4_t) __builtin_neon_vsliq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 34); \
28827		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	28845	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	28846	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	28847	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vsliq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 34)); \
	28848	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
28828	28849	__ret; \
28829	28850	})
28830	28851	#endif
...	...	@@ -28834,7 +28855,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28834	28855	int64x2_t __ret; \
28835	28856	int64x2_t __s0 = __p0; \
28836	28857	int64x2_t __s1 = __p1; \
28837		__ret = (int64x2_t) __builtin_neon_vsliq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 35); \
	28858	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vsliq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 35)); \
28838	28859	__ret; \
28839	28860	})
28840	28861	#else
...	...	@@ -28842,10 +28863,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28842	28863	int64x2_t __ret; \
28843	28864	int64x2_t __s0 = __p0; \
28844	28865	int64x2_t __s1 = __p1; \
28845		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
28846		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
28847		__ret = (int64x2_t) __builtin_neon_vsliq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 35); \
28848		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	28866	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	28867	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	28868	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vsliq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 35)); \
	28869	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
28849	28870	__ret; \
28850	28871	})
28851	28872	#endif
...	...	@@ -28855,7 +28876,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28855	28876	int16x8_t __ret; \
28856	28877	int16x8_t __s0 = __p0; \
28857	28878	int16x8_t __s1 = __p1; \
28858		__ret = (int16x8_t) __builtin_neon_vsliq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 33); \
	28879	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vsliq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 33)); \
28859	28880	__ret; \
28860	28881	})
28861	28882	#else
...	...	@@ -28863,10 +28884,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28863	28884	int16x8_t __ret; \
28864	28885	int16x8_t __s0 = __p0; \
28865	28886	int16x8_t __s1 = __p1; \
28866		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
28867		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
28868		__ret = (int16x8_t) __builtin_neon_vsliq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 33); \
28869		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	28887	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	28888	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	28889	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vsliq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 33)); \
	28890	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
28870	28891	__ret; \
28871	28892	})
28872	28893	#endif
...	...	@@ -28876,7 +28897,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28876	28897	uint8x8_t __ret; \
28877	28898	uint8x8_t __s0 = __p0; \
28878	28899	uint8x8_t __s1 = __p1; \
28879		__ret = (uint8x8_t) __builtin_neon_vsli_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 16); \
	28900	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vsli_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 16)); \
28880	28901	__ret; \
28881	28902	})
28882	28903	#else
...	...	@@ -28884,10 +28905,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28884	28905	uint8x8_t __ret; \
28885	28906	uint8x8_t __s0 = __p0; \
28886	28907	uint8x8_t __s1 = __p1; \
28887		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
28888		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
28889		__ret = (uint8x8_t) __builtin_neon_vsli_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 16); \
28890		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	28908	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	28909	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	28910	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vsli_n_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 16)); \
	28911	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
28891	28912	__ret; \
28892	28913	})
28893	28914	#endif
...	...	@@ -28897,7 +28918,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28897	28918	uint32x2_t __ret; \
28898	28919	uint32x2_t __s0 = __p0; \
28899	28920	uint32x2_t __s1 = __p1; \
28900		__ret = (uint32x2_t) __builtin_neon_vsli_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 18); \
	28921	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vsli_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 18)); \
28901	28922	__ret; \
28902	28923	})
28903	28924	#else
...	...	@@ -28905,10 +28926,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28905	28926	uint32x2_t __ret; \
28906	28927	uint32x2_t __s0 = __p0; \
28907	28928	uint32x2_t __s1 = __p1; \
28908		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
28909		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
28910		__ret = (uint32x2_t) __builtin_neon_vsli_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 18); \
28911		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	28929	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	28930	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_32); \
	28931	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vsli_n_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 18)); \
	28932	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
28912	28933	__ret; \
28913	28934	})
28914	28935	#endif
...	...	@@ -28917,7 +28938,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28917	28938	uint64x1_t __ret; \
28918	28939	uint64x1_t __s0 = __p0; \
28919	28940	uint64x1_t __s1 = __p1; \
28920		__ret = (uint64x1_t) __builtin_neon_vsli_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 19); \
	28941	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vsli_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 19)); \
28921	28942	__ret; \
28922	28943	})
28923	28944	#ifdef __LITTLE_ENDIAN__
...	...	@@ -28925,7 +28946,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28925	28946	uint16x4_t __ret; \
28926	28947	uint16x4_t __s0 = __p0; \
28927	28948	uint16x4_t __s1 = __p1; \
28928		__ret = (uint16x4_t) __builtin_neon_vsli_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 17); \
	28949	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vsli_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 17)); \
28929	28950	__ret; \
28930	28951	})
28931	28952	#else
...	...	@@ -28933,10 +28954,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28933	28954	uint16x4_t __ret; \
28934	28955	uint16x4_t __s0 = __p0; \
28935	28956	uint16x4_t __s1 = __p1; \
28936		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
28937		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
28938		__ret = (uint16x4_t) __builtin_neon_vsli_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 17); \
28939		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	28957	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	28958	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	28959	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vsli_n_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 17)); \
	28960	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
28940	28961	__ret; \
28941	28962	})
28942	28963	#endif
...	...	@@ -28946,7 +28967,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28946	28967	int8x8_t __ret; \
28947	28968	int8x8_t __s0 = __p0; \
28948	28969	int8x8_t __s1 = __p1; \
28949		__ret = (int8x8_t) __builtin_neon_vsli_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 0); \
	28970	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vsli_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 0)); \
28950	28971	__ret; \
28951	28972	})
28952	28973	#else
...	...	@@ -28954,10 +28975,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28954	28975	int8x8_t __ret; \
28955	28976	int8x8_t __s0 = __p0; \
28956	28977	int8x8_t __s1 = __p1; \
28957		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
28958		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
28959		__ret = (int8x8_t) __builtin_neon_vsli_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 0); \
28960		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	28978	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	28979	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	28980	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vsli_n_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 0)); \
	28981	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
28961	28982	__ret; \
28962	28983	})
28963	28984	#endif
...	...	@@ -28967,7 +28988,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28967	28988	int32x2_t __ret; \
28968	28989	int32x2_t __s0 = __p0; \
28969	28990	int32x2_t __s1 = __p1; \
28970		__ret = (int32x2_t) __builtin_neon_vsli_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 2); \
	28991	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vsli_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 2)); \
28971	28992	__ret; \
28972	28993	})
28973	28994	#else
...	...	@@ -28975,10 +28996,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28975	28996	int32x2_t __ret; \
28976	28997	int32x2_t __s0 = __p0; \
28977	28998	int32x2_t __s1 = __p1; \
28978		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
28979		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
28980		__ret = (int32x2_t) __builtin_neon_vsli_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 2); \
28981		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	28999	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	29000	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_32); \
	29001	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vsli_n_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 2)); \
	29002	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
28982	29003	__ret; \
28983	29004	})
28984	29005	#endif
...	...	@@ -28987,7 +29008,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28987	29008	int64x1_t __ret; \
28988	29009	int64x1_t __s0 = __p0; \
28989	29010	int64x1_t __s1 = __p1; \
28990		__ret = (int64x1_t) __builtin_neon_vsli_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 3); \
	29011	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vsli_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 3)); \
28991	29012	__ret; \
28992	29013	})
28993	29014	#ifdef __LITTLE_ENDIAN__
...	...	@@ -28995,7 +29016,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
28995	29016	int16x4_t __ret; \
28996	29017	int16x4_t __s0 = __p0; \
28997	29018	int16x4_t __s1 = __p1; \
28998		__ret = (int16x4_t) __builtin_neon_vsli_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 1); \
	29019	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vsli_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 1)); \
28999	29020	__ret; \
29000	29021	})
29001	29022	#else
...	...	@@ -29003,10 +29024,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29003	29024	int16x4_t __ret; \
29004	29025	int16x4_t __s0 = __p0; \
29005	29026	int16x4_t __s1 = __p1; \
29006		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
29007		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
29008		__ret = (int16x4_t) __builtin_neon_vsli_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 1); \
29009		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	29027	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	29028	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	29029	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vsli_n_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 1)); \
	29030	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
29010	29031	__ret; \
29011	29032	})
29012	29033	#endif
...	...	@@ -29016,7 +29037,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29016	29037	uint8x16_t __ret; \
29017	29038	uint8x16_t __s0 = __p0; \
29018	29039	uint8x16_t __s1 = __p1; \
29019		__ret = (uint8x16_t) __builtin_neon_vsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 48); \
	29040	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vsraq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 48)); \
29020	29041	__ret; \
29021	29042	})
29022	29043	#else
...	...	@@ -29024,10 +29045,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29024	29045	uint8x16_t __ret; \
29025	29046	uint8x16_t __s0 = __p0; \
29026	29047	uint8x16_t __s1 = __p1; \
29027		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
29028		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
29029		__ret = (uint8x16_t) __builtin_neon_vsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 48); \
29030		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	29048	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	29049	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	29050	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vsraq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 48)); \
	29051	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
29031	29052	__ret; \
29032	29053	})
29033	29054	#endif
...	...	@@ -29037,7 +29058,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29037	29058	uint32x4_t __ret; \
29038	29059	uint32x4_t __s0 = __p0; \
29039	29060	uint32x4_t __s1 = __p1; \
29040		__ret = (uint32x4_t) __builtin_neon_vsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 50); \
	29061	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsraq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 50)); \
29041	29062	__ret; \
29042	29063	})
29043	29064	#else
...	...	@@ -29045,10 +29066,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29045	29066	uint32x4_t __ret; \
29046	29067	uint32x4_t __s0 = __p0; \
29047	29068	uint32x4_t __s1 = __p1; \
29048		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
29049		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
29050		__ret = (uint32x4_t) __builtin_neon_vsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 50); \
29051		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	29069	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	29070	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	29071	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsraq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 50)); \
	29072	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
29052	29073	__ret; \
29053	29074	})
29054	29075	#endif
...	...	@@ -29058,7 +29079,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29058	29079	uint64x2_t __ret; \
29059	29080	uint64x2_t __s0 = __p0; \
29060	29081	uint64x2_t __s1 = __p1; \
29061		__ret = (uint64x2_t) __builtin_neon_vsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 51); \
	29082	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vsraq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 51)); \
29062	29083	__ret; \
29063	29084	})
29064	29085	#else
...	...	@@ -29066,10 +29087,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29066	29087	uint64x2_t __ret; \
29067	29088	uint64x2_t __s0 = __p0; \
29068	29089	uint64x2_t __s1 = __p1; \
29069		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
29070		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
29071		__ret = (uint64x2_t) __builtin_neon_vsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 51); \
29072		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	29090	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	29091	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	29092	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vsraq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 51)); \
	29093	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
29073	29094	__ret; \
29074	29095	})
29075	29096	#endif
...	...	@@ -29079,7 +29100,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29079	29100	uint16x8_t __ret; \
29080	29101	uint16x8_t __s0 = __p0; \
29081	29102	uint16x8_t __s1 = __p1; \
29082		__ret = (uint16x8_t) __builtin_neon_vsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 49); \
	29103	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vsraq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 49)); \
29083	29104	__ret; \
29084	29105	})
29085	29106	#else
...	...	@@ -29087,10 +29108,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29087	29108	uint16x8_t __ret; \
29088	29109	uint16x8_t __s0 = __p0; \
29089	29110	uint16x8_t __s1 = __p1; \
29090		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
29091		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
29092		__ret = (uint16x8_t) __builtin_neon_vsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 49); \
29093		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	29111	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	29112	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	29113	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vsraq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 49)); \
	29114	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
29094	29115	__ret; \
29095	29116	})
29096	29117	#endif
...	...	@@ -29100,7 +29121,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29100	29121	int8x16_t __ret; \
29101	29122	int8x16_t __s0 = __p0; \
29102	29123	int8x16_t __s1 = __p1; \
29103		__ret = (int8x16_t) __builtin_neon_vsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 32); \
	29124	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vsraq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 32)); \
29104	29125	__ret; \
29105	29126	})
29106	29127	#else
...	...	@@ -29108,10 +29129,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29108	29129	int8x16_t __ret; \
29109	29130	int8x16_t __s0 = __p0; \
29110	29131	int8x16_t __s1 = __p1; \
29111		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
29112		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
29113		__ret = (int8x16_t) __builtin_neon_vsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 32); \
29114		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	29132	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	29133	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	29134	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vsraq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 32)); \
	29135	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
29115	29136	__ret; \
29116	29137	})
29117	29138	#endif
...	...	@@ -29121,7 +29142,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29121	29142	int32x4_t __ret; \
29122	29143	int32x4_t __s0 = __p0; \
29123	29144	int32x4_t __s1 = __p1; \
29124		__ret = (int32x4_t) __builtin_neon_vsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 34); \
	29145	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vsraq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 34)); \
29125	29146	__ret; \
29126	29147	})
29127	29148	#else
...	...	@@ -29129,10 +29150,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29129	29150	int32x4_t __ret; \
29130	29151	int32x4_t __s0 = __p0; \
29131	29152	int32x4_t __s1 = __p1; \
29132		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
29133		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
29134		__ret = (int32x4_t) __builtin_neon_vsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 34); \
29135		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	29153	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	29154	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	29155	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vsraq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 34)); \
	29156	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
29136	29157	__ret; \
29137	29158	})
29138	29159	#endif
...	...	@@ -29142,7 +29163,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29142	29163	int64x2_t __ret; \
29143	29164	int64x2_t __s0 = __p0; \
29144	29165	int64x2_t __s1 = __p1; \
29145		__ret = (int64x2_t) __builtin_neon_vsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 35); \
	29166	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vsraq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 35)); \
29146	29167	__ret; \
29147	29168	})
29148	29169	#else
...	...	@@ -29150,10 +29171,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29150	29171	int64x2_t __ret; \
29151	29172	int64x2_t __s0 = __p0; \
29152	29173	int64x2_t __s1 = __p1; \
29153		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
29154		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
29155		__ret = (int64x2_t) __builtin_neon_vsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 35); \
29156		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	29174	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	29175	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	29176	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vsraq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 35)); \
	29177	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
29157	29178	__ret; \
29158	29179	})
29159	29180	#endif
...	...	@@ -29163,7 +29184,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29163	29184	int16x8_t __ret; \
29164	29185	int16x8_t __s0 = __p0; \
29165	29186	int16x8_t __s1 = __p1; \
29166		__ret = (int16x8_t) __builtin_neon_vsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 33); \
	29187	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vsraq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 33)); \
29167	29188	__ret; \
29168	29189	})
29169	29190	#else
...	...	@@ -29171,10 +29192,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29171	29192	int16x8_t __ret; \
29172	29193	int16x8_t __s0 = __p0; \
29173	29194	int16x8_t __s1 = __p1; \
29174		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
29175		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
29176		__ret = (int16x8_t) __builtin_neon_vsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 33); \
29177		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	29195	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	29196	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	29197	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vsraq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 33)); \
	29198	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
29178	29199	__ret; \
29179	29200	})
29180	29201	#endif
...	...	@@ -29184,7 +29205,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29184	29205	uint8x8_t __ret; \
29185	29206	uint8x8_t __s0 = __p0; \
29186	29207	uint8x8_t __s1 = __p1; \
29187		__ret = (uint8x8_t) __builtin_neon_vsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 16); \
	29208	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vsra_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 16)); \
29188	29209	__ret; \
29189	29210	})
29190	29211	#else
...	...	@@ -29192,10 +29213,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29192	29213	uint8x8_t __ret; \
29193	29214	uint8x8_t __s0 = __p0; \
29194	29215	uint8x8_t __s1 = __p1; \
29195		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
29196		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
29197		__ret = (uint8x8_t) __builtin_neon_vsra_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 16); \
29198		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	29216	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	29217	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	29218	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vsra_n_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 16)); \
	29219	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
29199	29220	__ret; \
29200	29221	})
29201	29222	#endif
...	...	@@ -29205,7 +29226,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29205	29226	uint32x2_t __ret; \
29206	29227	uint32x2_t __s0 = __p0; \
29207	29228	uint32x2_t __s1 = __p1; \
29208		__ret = (uint32x2_t) __builtin_neon_vsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 18); \
	29229	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vsra_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 18)); \
29209	29230	__ret; \
29210	29231	})
29211	29232	#else
...	...	@@ -29213,10 +29234,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29213	29234	uint32x2_t __ret; \
29214	29235	uint32x2_t __s0 = __p0; \
29215	29236	uint32x2_t __s1 = __p1; \
29216		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
29217		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
29218		__ret = (uint32x2_t) __builtin_neon_vsra_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 18); \
29219		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	29237	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	29238	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_32); \
	29239	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vsra_n_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 18)); \
	29240	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
29220	29241	__ret; \
29221	29242	})
29222	29243	#endif
...	...	@@ -29225,7 +29246,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29225	29246	uint64x1_t __ret; \
29226	29247	uint64x1_t __s0 = __p0; \
29227	29248	uint64x1_t __s1 = __p1; \
29228		__ret = (uint64x1_t) __builtin_neon_vsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 19); \
	29249	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vsra_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 19)); \
29229	29250	__ret; \
29230	29251	})
29231	29252	#ifdef __LITTLE_ENDIAN__
...	...	@@ -29233,7 +29254,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29233	29254	uint16x4_t __ret; \
29234	29255	uint16x4_t __s0 = __p0; \
29235	29256	uint16x4_t __s1 = __p1; \
29236		__ret = (uint16x4_t) __builtin_neon_vsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 17); \
	29257	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vsra_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 17)); \
29237	29258	__ret; \
29238	29259	})
29239	29260	#else
...	...	@@ -29241,10 +29262,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29241	29262	uint16x4_t __ret; \
29242	29263	uint16x4_t __s0 = __p0; \
29243	29264	uint16x4_t __s1 = __p1; \
29244		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
29245		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
29246		__ret = (uint16x4_t) __builtin_neon_vsra_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 17); \
29247		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	29265	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	29266	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	29267	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vsra_n_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 17)); \
	29268	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
29248	29269	__ret; \
29249	29270	})
29250	29271	#endif
...	...	@@ -29254,7 +29275,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29254	29275	int8x8_t __ret; \
29255	29276	int8x8_t __s0 = __p0; \
29256	29277	int8x8_t __s1 = __p1; \
29257		__ret = (int8x8_t) __builtin_neon_vsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 0); \
	29278	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vsra_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 0)); \
29258	29279	__ret; \
29259	29280	})
29260	29281	#else
...	...	@@ -29262,10 +29283,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29262	29283	int8x8_t __ret; \
29263	29284	int8x8_t __s0 = __p0; \
29264	29285	int8x8_t __s1 = __p1; \
29265		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
29266		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
29267		__ret = (int8x8_t) __builtin_neon_vsra_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 0); \
29268		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	29286	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	29287	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	29288	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vsra_n_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 0)); \
	29289	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
29269	29290	__ret; \
29270	29291	})
29271	29292	#endif
...	...	@@ -29275,7 +29296,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29275	29296	int32x2_t __ret; \
29276	29297	int32x2_t __s0 = __p0; \
29277	29298	int32x2_t __s1 = __p1; \
29278		__ret = (int32x2_t) __builtin_neon_vsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 2); \
	29299	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vsra_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 2)); \
29279	29300	__ret; \
29280	29301	})
29281	29302	#else
...	...	@@ -29283,10 +29304,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29283	29304	int32x2_t __ret; \
29284	29305	int32x2_t __s0 = __p0; \
29285	29306	int32x2_t __s1 = __p1; \
29286		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
29287		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
29288		__ret = (int32x2_t) __builtin_neon_vsra_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 2); \
29289		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	29307	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	29308	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_32); \
	29309	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vsra_n_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 2)); \
	29310	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
29290	29311	__ret; \
29291	29312	})
29292	29313	#endif
...	...	@@ -29295,7 +29316,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29295	29316	int64x1_t __ret; \
29296	29317	int64x1_t __s0 = __p0; \
29297	29318	int64x1_t __s1 = __p1; \
29298		__ret = (int64x1_t) __builtin_neon_vsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 3); \
	29319	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vsra_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 3)); \
29299	29320	__ret; \
29300	29321	})
29301	29322	#ifdef __LITTLE_ENDIAN__
...	...	@@ -29303,7 +29324,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29303	29324	int16x4_t __ret; \
29304	29325	int16x4_t __s0 = __p0; \
29305	29326	int16x4_t __s1 = __p1; \
29306		__ret = (int16x4_t) __builtin_neon_vsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 1); \
	29327	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vsra_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 1)); \
29307	29328	__ret; \
29308	29329	})
29309	29330	#else
...	...	@@ -29311,10 +29332,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29311	29332	int16x4_t __ret; \
29312	29333	int16x4_t __s0 = __p0; \
29313	29334	int16x4_t __s1 = __p1; \
29314		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
29315		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
29316		__ret = (int16x4_t) __builtin_neon_vsra_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 1); \
29317		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	29335	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	29336	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	29337	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vsra_n_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 1)); \
	29338	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
29318	29339	__ret; \
29319	29340	})
29320	29341	#endif
...	...	@@ -29324,7 +29345,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29324	29345	poly8x8_t __ret; \
29325	29346	poly8x8_t __s0 = __p0; \
29326	29347	poly8x8_t __s1 = __p1; \
29327		__ret = (poly8x8_t) __builtin_neon_vsri_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 4); \
	29348	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vsri_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 4)); \
29328	29349	__ret; \
29329	29350	})
29330	29351	#else
...	...	@@ -29332,10 +29353,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29332	29353	poly8x8_t __ret; \
29333	29354	poly8x8_t __s0 = __p0; \
29334	29355	poly8x8_t __s1 = __p1; \
29335		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
29336		poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
29337		__ret = (poly8x8_t) __builtin_neon_vsri_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 4); \
29338		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	29356	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	29357	poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	29358	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vsri_n_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 4)); \
	29359	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
29339	29360	__ret; \
29340	29361	})
29341	29362	#endif
...	...	@@ -29345,7 +29366,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29345	29366	poly16x4_t __ret; \
29346	29367	poly16x4_t __s0 = __p0; \
29347	29368	poly16x4_t __s1 = __p1; \
29348		__ret = (poly16x4_t) __builtin_neon_vsri_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 5); \
	29369	__ret = __builtin_bit_cast(poly16x4_t, __builtin_neon_vsri_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 5)); \
29349	29370	__ret; \
29350	29371	})
29351	29372	#else
...	...	@@ -29353,10 +29374,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29353	29374	poly16x4_t __ret; \
29354	29375	poly16x4_t __s0 = __p0; \
29355	29376	poly16x4_t __s1 = __p1; \
29356		poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
29357		poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
29358		__ret = (poly16x4_t) __builtin_neon_vsri_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 5); \
29359		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	29377	poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	29378	poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	29379	__ret = __builtin_bit_cast(poly16x4_t, __builtin_neon_vsri_n_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 5)); \
	29380	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
29360	29381	__ret; \
29361	29382	})
29362	29383	#endif
...	...	@@ -29366,7 +29387,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29366	29387	poly8x16_t __ret; \
29367	29388	poly8x16_t __s0 = __p0; \
29368	29389	poly8x16_t __s1 = __p1; \
29369		__ret = (poly8x16_t) __builtin_neon_vsriq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 36); \
	29390	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vsriq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 36)); \
29370	29391	__ret; \
29371	29392	})
29372	29393	#else
...	...	@@ -29374,10 +29395,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29374	29395	poly8x16_t __ret; \
29375	29396	poly8x16_t __s0 = __p0; \
29376	29397	poly8x16_t __s1 = __p1; \
29377		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
29378		poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
29379		__ret = (poly8x16_t) __builtin_neon_vsriq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 36); \
29380		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	29398	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	29399	poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	29400	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vsriq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 36)); \
	29401	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
29381	29402	__ret; \
29382	29403	})
29383	29404	#endif
...	...	@@ -29387,7 +29408,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29387	29408	poly16x8_t __ret; \
29388	29409	poly16x8_t __s0 = __p0; \
29389	29410	poly16x8_t __s1 = __p1; \
29390		__ret = (poly16x8_t) __builtin_neon_vsriq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 37); \
	29411	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vsriq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 37)); \
29391	29412	__ret; \
29392	29413	})
29393	29414	#else
...	...	@@ -29395,10 +29416,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29395	29416	poly16x8_t __ret; \
29396	29417	poly16x8_t __s0 = __p0; \
29397	29418	poly16x8_t __s1 = __p1; \
29398		poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
29399		poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
29400		__ret = (poly16x8_t) __builtin_neon_vsriq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 37); \
29401		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	29419	poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	29420	poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	29421	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vsriq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 37)); \
	29422	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
29402	29423	__ret; \
29403	29424	})
29404	29425	#endif
...	...	@@ -29408,7 +29429,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29408	29429	uint8x16_t __ret; \
29409	29430	uint8x16_t __s0 = __p0; \
29410	29431	uint8x16_t __s1 = __p1; \
29411		__ret = (uint8x16_t) __builtin_neon_vsriq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 48); \
	29432	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vsriq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 48)); \
29412	29433	__ret; \
29413	29434	})
29414	29435	#else
...	...	@@ -29416,10 +29437,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29416	29437	uint8x16_t __ret; \
29417	29438	uint8x16_t __s0 = __p0; \
29418	29439	uint8x16_t __s1 = __p1; \
29419		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
29420		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
29421		__ret = (uint8x16_t) __builtin_neon_vsriq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 48); \
29422		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	29440	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	29441	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	29442	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vsriq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 48)); \
	29443	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
29423	29444	__ret; \
29424	29445	})
29425	29446	#endif
...	...	@@ -29429,7 +29450,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29429	29450	uint32x4_t __ret; \
29430	29451	uint32x4_t __s0 = __p0; \
29431	29452	uint32x4_t __s1 = __p1; \
29432		__ret = (uint32x4_t) __builtin_neon_vsriq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 50); \
	29453	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsriq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 50)); \
29433	29454	__ret; \
29434	29455	})
29435	29456	#else
...	...	@@ -29437,10 +29458,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29437	29458	uint32x4_t __ret; \
29438	29459	uint32x4_t __s0 = __p0; \
29439	29460	uint32x4_t __s1 = __p1; \
29440		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
29441		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
29442		__ret = (uint32x4_t) __builtin_neon_vsriq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 50); \
29443		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	29461	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	29462	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	29463	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsriq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 50)); \
	29464	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
29444	29465	__ret; \
29445	29466	})
29446	29467	#endif
...	...	@@ -29450,7 +29471,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29450	29471	uint64x2_t __ret; \
29451	29472	uint64x2_t __s0 = __p0; \
29452	29473	uint64x2_t __s1 = __p1; \
29453		__ret = (uint64x2_t) __builtin_neon_vsriq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 51); \
	29474	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vsriq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 51)); \
29454	29475	__ret; \
29455	29476	})
29456	29477	#else
...	...	@@ -29458,10 +29479,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29458	29479	uint64x2_t __ret; \
29459	29480	uint64x2_t __s0 = __p0; \
29460	29481	uint64x2_t __s1 = __p1; \
29461		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
29462		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
29463		__ret = (uint64x2_t) __builtin_neon_vsriq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 51); \
29464		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	29482	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	29483	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	29484	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vsriq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 51)); \
	29485	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
29465	29486	__ret; \
29466	29487	})
29467	29488	#endif
...	...	@@ -29471,7 +29492,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29471	29492	uint16x8_t __ret; \
29472	29493	uint16x8_t __s0 = __p0; \
29473	29494	uint16x8_t __s1 = __p1; \
29474		__ret = (uint16x8_t) __builtin_neon_vsriq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 49); \
	29495	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vsriq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 49)); \
29475	29496	__ret; \
29476	29497	})
29477	29498	#else
...	...	@@ -29479,10 +29500,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29479	29500	uint16x8_t __ret; \
29480	29501	uint16x8_t __s0 = __p0; \
29481	29502	uint16x8_t __s1 = __p1; \
29482		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
29483		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
29484		__ret = (uint16x8_t) __builtin_neon_vsriq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 49); \
29485		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	29503	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	29504	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	29505	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vsriq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 49)); \
	29506	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
29486	29507	__ret; \
29487	29508	})
29488	29509	#endif
...	...	@@ -29492,7 +29513,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29492	29513	int8x16_t __ret; \
29493	29514	int8x16_t __s0 = __p0; \
29494	29515	int8x16_t __s1 = __p1; \
29495		__ret = (int8x16_t) __builtin_neon_vsriq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 32); \
	29516	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vsriq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 32)); \
29496	29517	__ret; \
29497	29518	})
29498	29519	#else
...	...	@@ -29500,10 +29521,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29500	29521	int8x16_t __ret; \
29501	29522	int8x16_t __s0 = __p0; \
29502	29523	int8x16_t __s1 = __p1; \
29503		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
29504		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
29505		__ret = (int8x16_t) __builtin_neon_vsriq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 32); \
29506		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	29524	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	29525	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	29526	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vsriq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 32)); \
	29527	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
29507	29528	__ret; \
29508	29529	})
29509	29530	#endif
...	...	@@ -29513,7 +29534,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29513	29534	int32x4_t __ret; \
29514	29535	int32x4_t __s0 = __p0; \
29515	29536	int32x4_t __s1 = __p1; \
29516		__ret = (int32x4_t) __builtin_neon_vsriq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 34); \
	29537	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vsriq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 34)); \
29517	29538	__ret; \
29518	29539	})
29519	29540	#else
...	...	@@ -29521,10 +29542,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29521	29542	int32x4_t __ret; \
29522	29543	int32x4_t __s0 = __p0; \
29523	29544	int32x4_t __s1 = __p1; \
29524		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
29525		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
29526		__ret = (int32x4_t) __builtin_neon_vsriq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 34); \
29527		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	29545	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	29546	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	29547	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vsriq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 34)); \
	29548	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
29528	29549	__ret; \
29529	29550	})
29530	29551	#endif
...	...	@@ -29534,7 +29555,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29534	29555	int64x2_t __ret; \
29535	29556	int64x2_t __s0 = __p0; \
29536	29557	int64x2_t __s1 = __p1; \
29537		__ret = (int64x2_t) __builtin_neon_vsriq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 35); \
	29558	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vsriq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 35)); \
29538	29559	__ret; \
29539	29560	})
29540	29561	#else
...	...	@@ -29542,10 +29563,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29542	29563	int64x2_t __ret; \
29543	29564	int64x2_t __s0 = __p0; \
29544	29565	int64x2_t __s1 = __p1; \
29545		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
29546		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
29547		__ret = (int64x2_t) __builtin_neon_vsriq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 35); \
29548		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	29566	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	29567	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	29568	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vsriq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 35)); \
	29569	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
29549	29570	__ret; \
29550	29571	})
29551	29572	#endif
...	...	@@ -29555,7 +29576,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29555	29576	int16x8_t __ret; \
29556	29577	int16x8_t __s0 = __p0; \
29557	29578	int16x8_t __s1 = __p1; \
29558		__ret = (int16x8_t) __builtin_neon_vsriq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 33); \
	29579	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vsriq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 33)); \
29559	29580	__ret; \
29560	29581	})
29561	29582	#else
...	...	@@ -29563,10 +29584,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29563	29584	int16x8_t __ret; \
29564	29585	int16x8_t __s0 = __p0; \
29565	29586	int16x8_t __s1 = __p1; \
29566		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
29567		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
29568		__ret = (int16x8_t) __builtin_neon_vsriq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 33); \
29569		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	29587	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	29588	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	29589	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vsriq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 33)); \
	29590	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
29570	29591	__ret; \
29571	29592	})
29572	29593	#endif
...	...	@@ -29576,7 +29597,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29576	29597	uint8x8_t __ret; \
29577	29598	uint8x8_t __s0 = __p0; \
29578	29599	uint8x8_t __s1 = __p1; \
29579		__ret = (uint8x8_t) __builtin_neon_vsri_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 16); \
	29600	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vsri_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 16)); \
29580	29601	__ret; \
29581	29602	})
29582	29603	#else
...	...	@@ -29584,10 +29605,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29584	29605	uint8x8_t __ret; \
29585	29606	uint8x8_t __s0 = __p0; \
29586	29607	uint8x8_t __s1 = __p1; \
29587		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
29588		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
29589		__ret = (uint8x8_t) __builtin_neon_vsri_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 16); \
29590		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	29608	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	29609	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	29610	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vsri_n_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 16)); \
	29611	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
29591	29612	__ret; \
29592	29613	})
29593	29614	#endif
...	...	@@ -29597,7 +29618,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29597	29618	uint32x2_t __ret; \
29598	29619	uint32x2_t __s0 = __p0; \
29599	29620	uint32x2_t __s1 = __p1; \
29600		__ret = (uint32x2_t) __builtin_neon_vsri_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 18); \
	29621	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vsri_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 18)); \
29601	29622	__ret; \
29602	29623	})
29603	29624	#else
...	...	@@ -29605,10 +29626,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29605	29626	uint32x2_t __ret; \
29606	29627	uint32x2_t __s0 = __p0; \
29607	29628	uint32x2_t __s1 = __p1; \
29608		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
29609		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
29610		__ret = (uint32x2_t) __builtin_neon_vsri_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 18); \
29611		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	29629	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	29630	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_32); \
	29631	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vsri_n_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 18)); \
	29632	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
29612	29633	__ret; \
29613	29634	})
29614	29635	#endif
...	...	@@ -29617,7 +29638,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29617	29638	uint64x1_t __ret; \
29618	29639	uint64x1_t __s0 = __p0; \
29619	29640	uint64x1_t __s1 = __p1; \
29620		__ret = (uint64x1_t) __builtin_neon_vsri_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 19); \
	29641	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vsri_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 19)); \
29621	29642	__ret; \
29622	29643	})
29623	29644	#ifdef __LITTLE_ENDIAN__
...	...	@@ -29625,7 +29646,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29625	29646	uint16x4_t __ret; \
29626	29647	uint16x4_t __s0 = __p0; \
29627	29648	uint16x4_t __s1 = __p1; \
29628		__ret = (uint16x4_t) __builtin_neon_vsri_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 17); \
	29649	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vsri_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 17)); \
29629	29650	__ret; \
29630	29651	})
29631	29652	#else
...	...	@@ -29633,10 +29654,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29633	29654	uint16x4_t __ret; \
29634	29655	uint16x4_t __s0 = __p0; \
29635	29656	uint16x4_t __s1 = __p1; \
29636		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
29637		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
29638		__ret = (uint16x4_t) __builtin_neon_vsri_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 17); \
29639		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	29657	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	29658	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	29659	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vsri_n_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 17)); \
	29660	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
29640	29661	__ret; \
29641	29662	})
29642	29663	#endif
...	...	@@ -29646,7 +29667,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29646	29667	int8x8_t __ret; \
29647	29668	int8x8_t __s0 = __p0; \
29648	29669	int8x8_t __s1 = __p1; \
29649		__ret = (int8x8_t) __builtin_neon_vsri_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 0); \
	29670	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vsri_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 0)); \
29650	29671	__ret; \
29651	29672	})
29652	29673	#else
...	...	@@ -29654,10 +29675,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29654	29675	int8x8_t __ret; \
29655	29676	int8x8_t __s0 = __p0; \
29656	29677	int8x8_t __s1 = __p1; \
29657		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
29658		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
29659		__ret = (int8x8_t) __builtin_neon_vsri_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 0); \
29660		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	29678	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	29679	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	29680	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vsri_n_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 0)); \
	29681	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
29661	29682	__ret; \
29662	29683	})
29663	29684	#endif
...	...	@@ -29667,7 +29688,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29667	29688	int32x2_t __ret; \
29668	29689	int32x2_t __s0 = __p0; \
29669	29690	int32x2_t __s1 = __p1; \
29670		__ret = (int32x2_t) __builtin_neon_vsri_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 2); \
	29691	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vsri_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 2)); \
29671	29692	__ret; \
29672	29693	})
29673	29694	#else
...	...	@@ -29675,10 +29696,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29675	29696	int32x2_t __ret; \
29676	29697	int32x2_t __s0 = __p0; \
29677	29698	int32x2_t __s1 = __p1; \
29678		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
29679		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
29680		__ret = (int32x2_t) __builtin_neon_vsri_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 2); \
29681		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	29699	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	29700	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_32); \
	29701	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vsri_n_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 2)); \
	29702	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
29682	29703	__ret; \
29683	29704	})
29684	29705	#endif
...	...	@@ -29687,7 +29708,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29687	29708	int64x1_t __ret; \
29688	29709	int64x1_t __s0 = __p0; \
29689	29710	int64x1_t __s1 = __p1; \
29690		__ret = (int64x1_t) __builtin_neon_vsri_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 3); \
	29711	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vsri_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 3)); \
29691	29712	__ret; \
29692	29713	})
29693	29714	#ifdef __LITTLE_ENDIAN__
...	...	@@ -29695,7 +29716,7 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29695	29716	int16x4_t __ret; \
29696	29717	int16x4_t __s0 = __p0; \
29697	29718	int16x4_t __s1 = __p1; \
29698		__ret = (int16x4_t) __builtin_neon_vsri_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 1); \
	29719	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vsri_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 1)); \
29699	29720	__ret; \
29700	29721	})
29701	29722	#else
...	...	@@ -29703,10 +29724,10 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29703	29724	int16x4_t __ret; \
29704	29725	int16x4_t __s0 = __p0; \
29705	29726	int16x4_t __s1 = __p1; \
29706		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
29707		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
29708		__ret = (int16x4_t) __builtin_neon_vsri_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 1); \
29709		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	29727	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	29728	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	29729	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vsri_n_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 1)); \
	29730	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
29710	29731	__ret; \
29711	29732	})
29712	29733	#endif
...	...	@@ -29714,3128 +29735,3128 @@ __ai __attribute__((target("neon"))) int16x4_t vshl_s16(int16x4_t __p0, int16x4_
29714	29735	#ifdef __LITTLE_ENDIAN__
29715	29736	#define vst1_p8(__p0, __p1) __extension__ ({ \
29716	29737	poly8x8_t __s1 = __p1; \
29717		__builtin_neon_vst1_v(__p0, (int8x8_t)__s1, 4); \
	29738	__builtin_neon_vst1_v(__p0, __builtin_bit_cast(int8x8_t, __s1), 4); \
29718	29739	})
29719	29740	#else
29720	29741	#define vst1_p8(__p0, __p1) __extension__ ({ \
29721	29742	poly8x8_t __s1 = __p1; \
29722		poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
29723		__builtin_neon_vst1_v(__p0, (int8x8_t)__rev1, 4); \
	29743	poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	29744	__builtin_neon_vst1_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), 4); \
29724	29745	})
29725	29746	#endif
29726	29747
29727	29748	#ifdef __LITTLE_ENDIAN__
29728	29749	#define vst1_p16(__p0, __p1) __extension__ ({ \
29729	29750	poly16x4_t __s1 = __p1; \
29730		__builtin_neon_vst1_v(__p0, (int8x8_t)__s1, 5); \
	29751	__builtin_neon_vst1_v(__p0, __builtin_bit_cast(int8x8_t, __s1), 5); \
29731	29752	})
29732	29753	#else
29733	29754	#define vst1_p16(__p0, __p1) __extension__ ({ \
29734	29755	poly16x4_t __s1 = __p1; \
29735		poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
29736		__builtin_neon_vst1_v(__p0, (int8x8_t)__rev1, 5); \
	29756	poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	29757	__builtin_neon_vst1_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), 5); \
29737	29758	})
29738	29759	#endif
29739	29760
29740	29761	#ifdef __LITTLE_ENDIAN__
29741	29762	#define vst1q_p8(__p0, __p1) __extension__ ({ \
29742	29763	poly8x16_t __s1 = __p1; \
29743		__builtin_neon_vst1q_v(__p0, (int8x16_t)__s1, 36); \
	29764	__builtin_neon_vst1q_v(__p0, __builtin_bit_cast(int8x16_t, __s1), 36); \
29744	29765	})
29745	29766	#else
29746	29767	#define vst1q_p8(__p0, __p1) __extension__ ({ \
29747	29768	poly8x16_t __s1 = __p1; \
29748		poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
29749		__builtin_neon_vst1q_v(__p0, (int8x16_t)__rev1, 36); \
	29769	poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	29770	__builtin_neon_vst1q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), 36); \
29750	29771	})
29751	29772	#endif
29752	29773
29753	29774	#ifdef __LITTLE_ENDIAN__
29754	29775	#define vst1q_p16(__p0, __p1) __extension__ ({ \
29755	29776	poly16x8_t __s1 = __p1; \
29756		__builtin_neon_vst1q_v(__p0, (int8x16_t)__s1, 37); \
	29777	__builtin_neon_vst1q_v(__p0, __builtin_bit_cast(int8x16_t, __s1), 37); \
29757	29778	})
29758	29779	#else
29759	29780	#define vst1q_p16(__p0, __p1) __extension__ ({ \
29760	29781	poly16x8_t __s1 = __p1; \
29761		poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
29762		__builtin_neon_vst1q_v(__p0, (int8x16_t)__rev1, 37); \
	29782	poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	29783	__builtin_neon_vst1q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), 37); \
29763	29784	})
29764	29785	#endif
29765	29786
29766	29787	#ifdef __LITTLE_ENDIAN__
29767	29788	#define vst1q_u8(__p0, __p1) __extension__ ({ \
29768	29789	uint8x16_t __s1 = __p1; \
29769		__builtin_neon_vst1q_v(__p0, (int8x16_t)__s1, 48); \
	29790	__builtin_neon_vst1q_v(__p0, __builtin_bit_cast(int8x16_t, __s1), 48); \
29770	29791	})
29771	29792	#else
29772	29793	#define vst1q_u8(__p0, __p1) __extension__ ({ \
29773	29794	uint8x16_t __s1 = __p1; \
29774		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
29775		__builtin_neon_vst1q_v(__p0, (int8x16_t)__rev1, 48); \
	29795	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	29796	__builtin_neon_vst1q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), 48); \
29776	29797	})
29777	29798	#endif
29778	29799
29779	29800	#ifdef __LITTLE_ENDIAN__
29780	29801	#define vst1q_u32(__p0, __p1) __extension__ ({ \
29781	29802	uint32x4_t __s1 = __p1; \
29782		__builtin_neon_vst1q_v(__p0, (int8x16_t)__s1, 50); \
	29803	__builtin_neon_vst1q_v(__p0, __builtin_bit_cast(int8x16_t, __s1), 50); \
29783	29804	})
29784	29805	#else
29785	29806	#define vst1q_u32(__p0, __p1) __extension__ ({ \
29786	29807	uint32x4_t __s1 = __p1; \
29787		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
29788		__builtin_neon_vst1q_v(__p0, (int8x16_t)__rev1, 50); \
	29808	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	29809	__builtin_neon_vst1q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), 50); \
29789	29810	})
29790	29811	#endif
29791	29812
29792	29813	#ifdef __LITTLE_ENDIAN__
29793	29814	#define vst1q_u64(__p0, __p1) __extension__ ({ \
29794	29815	uint64x2_t __s1 = __p1; \
29795		__builtin_neon_vst1q_v(__p0, (int8x16_t)__s1, 51); \
	29816	__builtin_neon_vst1q_v(__p0, __builtin_bit_cast(int8x16_t, __s1), 51); \
29796	29817	})
29797	29818	#else
29798	29819	#define vst1q_u64(__p0, __p1) __extension__ ({ \
29799	29820	uint64x2_t __s1 = __p1; \
29800		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
29801		__builtin_neon_vst1q_v(__p0, (int8x16_t)__rev1, 51); \
	29821	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	29822	__builtin_neon_vst1q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), 51); \
29802	29823	})
29803	29824	#endif
29804	29825
29805	29826	#ifdef __LITTLE_ENDIAN__
29806	29827	#define vst1q_u16(__p0, __p1) __extension__ ({ \
29807	29828	uint16x8_t __s1 = __p1; \
29808		__builtin_neon_vst1q_v(__p0, (int8x16_t)__s1, 49); \
	29829	__builtin_neon_vst1q_v(__p0, __builtin_bit_cast(int8x16_t, __s1), 49); \
29809	29830	})
29810	29831	#else
29811	29832	#define vst1q_u16(__p0, __p1) __extension__ ({ \
29812	29833	uint16x8_t __s1 = __p1; \
29813		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
29814		__builtin_neon_vst1q_v(__p0, (int8x16_t)__rev1, 49); \
	29834	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	29835	__builtin_neon_vst1q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), 49); \
29815	29836	})
29816	29837	#endif
29817	29838
29818	29839	#ifdef __LITTLE_ENDIAN__
29819	29840	#define vst1q_s8(__p0, __p1) __extension__ ({ \
29820	29841	int8x16_t __s1 = __p1; \
29821		__builtin_neon_vst1q_v(__p0, (int8x16_t)__s1, 32); \
	29842	__builtin_neon_vst1q_v(__p0, __builtin_bit_cast(int8x16_t, __s1), 32); \
29822	29843	})
29823	29844	#else
29824	29845	#define vst1q_s8(__p0, __p1) __extension__ ({ \
29825	29846	int8x16_t __s1 = __p1; \
29826		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
29827		__builtin_neon_vst1q_v(__p0, (int8x16_t)__rev1, 32); \
	29847	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	29848	__builtin_neon_vst1q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), 32); \
29828	29849	})
29829	29850	#endif
29830	29851
29831	29852	#ifdef __LITTLE_ENDIAN__
29832	29853	#define vst1q_f32(__p0, __p1) __extension__ ({ \
29833	29854	float32x4_t __s1 = __p1; \
29834		__builtin_neon_vst1q_v(__p0, (int8x16_t)__s1, 41); \
	29855	__builtin_neon_vst1q_v(__p0, __builtin_bit_cast(int8x16_t, __s1), 41); \
29835	29856	})
29836	29857	#else
29837	29858	#define vst1q_f32(__p0, __p1) __extension__ ({ \
29838	29859	float32x4_t __s1 = __p1; \
29839		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
29840		__builtin_neon_vst1q_v(__p0, (int8x16_t)__rev1, 41); \
	29860	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	29861	__builtin_neon_vst1q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), 41); \
29841	29862	})
29842	29863	#endif
29843	29864
29844	29865	#ifdef __LITTLE_ENDIAN__
29845	29866	#define vst1q_s32(__p0, __p1) __extension__ ({ \
29846	29867	int32x4_t __s1 = __p1; \
29847		__builtin_neon_vst1q_v(__p0, (int8x16_t)__s1, 34); \
	29868	__builtin_neon_vst1q_v(__p0, __builtin_bit_cast(int8x16_t, __s1), 34); \
29848	29869	})
29849	29870	#else
29850	29871	#define vst1q_s32(__p0, __p1) __extension__ ({ \
29851	29872	int32x4_t __s1 = __p1; \
29852		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
29853		__builtin_neon_vst1q_v(__p0, (int8x16_t)__rev1, 34); \
	29873	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	29874	__builtin_neon_vst1q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), 34); \
29854	29875	})
29855	29876	#endif
29856	29877
29857	29878	#ifdef __LITTLE_ENDIAN__
29858	29879	#define vst1q_s64(__p0, __p1) __extension__ ({ \
29859	29880	int64x2_t __s1 = __p1; \
29860		__builtin_neon_vst1q_v(__p0, (int8x16_t)__s1, 35); \
	29881	__builtin_neon_vst1q_v(__p0, __builtin_bit_cast(int8x16_t, __s1), 35); \
29861	29882	})
29862	29883	#else
29863	29884	#define vst1q_s64(__p0, __p1) __extension__ ({ \
29864	29885	int64x2_t __s1 = __p1; \
29865		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
29866		__builtin_neon_vst1q_v(__p0, (int8x16_t)__rev1, 35); \
	29886	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	29887	__builtin_neon_vst1q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), 35); \
29867	29888	})
29868	29889	#endif
29869	29890
29870	29891	#ifdef __LITTLE_ENDIAN__
29871	29892	#define vst1q_s16(__p0, __p1) __extension__ ({ \
29872	29893	int16x8_t __s1 = __p1; \
29873		__builtin_neon_vst1q_v(__p0, (int8x16_t)__s1, 33); \
	29894	__builtin_neon_vst1q_v(__p0, __builtin_bit_cast(int8x16_t, __s1), 33); \
29874	29895	})
29875	29896	#else
29876	29897	#define vst1q_s16(__p0, __p1) __extension__ ({ \
29877	29898	int16x8_t __s1 = __p1; \
29878		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
29879		__builtin_neon_vst1q_v(__p0, (int8x16_t)__rev1, 33); \
	29899	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	29900	__builtin_neon_vst1q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), 33); \
29880	29901	})
29881	29902	#endif
29882	29903
29883	29904	#ifdef __LITTLE_ENDIAN__
29884	29905	#define vst1_u8(__p0, __p1) __extension__ ({ \
29885	29906	uint8x8_t __s1 = __p1; \
29886		__builtin_neon_vst1_v(__p0, (int8x8_t)__s1, 16); \
	29907	__builtin_neon_vst1_v(__p0, __builtin_bit_cast(int8x8_t, __s1), 16); \
29887	29908	})
29888	29909	#else
29889	29910	#define vst1_u8(__p0, __p1) __extension__ ({ \
29890	29911	uint8x8_t __s1 = __p1; \
29891		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
29892		__builtin_neon_vst1_v(__p0, (int8x8_t)__rev1, 16); \
	29912	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	29913	__builtin_neon_vst1_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), 16); \
29893	29914	})
29894	29915	#endif
29895	29916
29896	29917	#ifdef __LITTLE_ENDIAN__
29897	29918	#define vst1_u32(__p0, __p1) __extension__ ({ \
29898	29919	uint32x2_t __s1 = __p1; \
29899		__builtin_neon_vst1_v(__p0, (int8x8_t)__s1, 18); \
	29920	__builtin_neon_vst1_v(__p0, __builtin_bit_cast(int8x8_t, __s1), 18); \
29900	29921	})
29901	29922	#else
29902	29923	#define vst1_u32(__p0, __p1) __extension__ ({ \
29903	29924	uint32x2_t __s1 = __p1; \
29904		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
29905		__builtin_neon_vst1_v(__p0, (int8x8_t)__rev1, 18); \
	29925	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_32); \
	29926	__builtin_neon_vst1_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), 18); \
29906	29927	})
29907	29928	#endif
29908	29929
29909	29930	#define vst1_u64(__p0, __p1) __extension__ ({ \
29910	29931	uint64x1_t __s1 = __p1; \
29911		__builtin_neon_vst1_v(__p0, (int8x8_t)__s1, 19); \
	29932	__builtin_neon_vst1_v(__p0, __builtin_bit_cast(int8x8_t, __s1), 19); \
29912	29933	})
29913	29934	#ifdef __LITTLE_ENDIAN__
29914	29935	#define vst1_u16(__p0, __p1) __extension__ ({ \
29915	29936	uint16x4_t __s1 = __p1; \
29916		__builtin_neon_vst1_v(__p0, (int8x8_t)__s1, 17); \
	29937	__builtin_neon_vst1_v(__p0, __builtin_bit_cast(int8x8_t, __s1), 17); \
29917	29938	})
29918	29939	#else
29919	29940	#define vst1_u16(__p0, __p1) __extension__ ({ \
29920	29941	uint16x4_t __s1 = __p1; \
29921		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
29922		__builtin_neon_vst1_v(__p0, (int8x8_t)__rev1, 17); \
	29942	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	29943	__builtin_neon_vst1_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), 17); \
29923	29944	})
29924	29945	#endif
29925	29946
29926	29947	#ifdef __LITTLE_ENDIAN__
29927	29948	#define vst1_s8(__p0, __p1) __extension__ ({ \
29928	29949	int8x8_t __s1 = __p1; \
29929		__builtin_neon_vst1_v(__p0, (int8x8_t)__s1, 0); \
	29950	__builtin_neon_vst1_v(__p0, __builtin_bit_cast(int8x8_t, __s1), 0); \
29930	29951	})
29931	29952	#else
29932	29953	#define vst1_s8(__p0, __p1) __extension__ ({ \
29933	29954	int8x8_t __s1 = __p1; \
29934		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
29935		__builtin_neon_vst1_v(__p0, (int8x8_t)__rev1, 0); \
	29955	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	29956	__builtin_neon_vst1_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), 0); \
29936	29957	})
29937	29958	#endif
29938	29959
29939	29960	#ifdef __LITTLE_ENDIAN__
29940	29961	#define vst1_f32(__p0, __p1) __extension__ ({ \
29941	29962	float32x2_t __s1 = __p1; \
29942		__builtin_neon_vst1_v(__p0, (int8x8_t)__s1, 9); \
	29963	__builtin_neon_vst1_v(__p0, __builtin_bit_cast(int8x8_t, __s1), 9); \
29943	29964	})
29944	29965	#else
29945	29966	#define vst1_f32(__p0, __p1) __extension__ ({ \
29946	29967	float32x2_t __s1 = __p1; \
29947		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
29948		__builtin_neon_vst1_v(__p0, (int8x8_t)__rev1, 9); \
	29968	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_32); \
	29969	__builtin_neon_vst1_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), 9); \
29949	29970	})
29950	29971	#endif
29951	29972
29952	29973	#ifdef __LITTLE_ENDIAN__
29953	29974	#define vst1_s32(__p0, __p1) __extension__ ({ \
29954	29975	int32x2_t __s1 = __p1; \
29955		__builtin_neon_vst1_v(__p0, (int8x8_t)__s1, 2); \
	29976	__builtin_neon_vst1_v(__p0, __builtin_bit_cast(int8x8_t, __s1), 2); \
29956	29977	})
29957	29978	#else
29958	29979	#define vst1_s32(__p0, __p1) __extension__ ({ \
29959	29980	int32x2_t __s1 = __p1; \
29960		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
29961		__builtin_neon_vst1_v(__p0, (int8x8_t)__rev1, 2); \
	29981	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_32); \
	29982	__builtin_neon_vst1_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), 2); \
29962	29983	})
29963	29984	#endif
29964	29985
29965	29986	#define vst1_s64(__p0, __p1) __extension__ ({ \
29966	29987	int64x1_t __s1 = __p1; \
29967		__builtin_neon_vst1_v(__p0, (int8x8_t)__s1, 3); \
	29988	__builtin_neon_vst1_v(__p0, __builtin_bit_cast(int8x8_t, __s1), 3); \
29968	29989	})
29969	29990	#ifdef __LITTLE_ENDIAN__
29970	29991	#define vst1_s16(__p0, __p1) __extension__ ({ \
29971	29992	int16x4_t __s1 = __p1; \
29972		__builtin_neon_vst1_v(__p0, (int8x8_t)__s1, 1); \
	29993	__builtin_neon_vst1_v(__p0, __builtin_bit_cast(int8x8_t, __s1), 1); \
29973	29994	})
29974	29995	#else
29975	29996	#define vst1_s16(__p0, __p1) __extension__ ({ \
29976	29997	int16x4_t __s1 = __p1; \
29977		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
29978		__builtin_neon_vst1_v(__p0, (int8x8_t)__rev1, 1); \
	29998	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	29999	__builtin_neon_vst1_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), 1); \
29979	30000	})
29980	30001	#endif
29981	30002
29982	30003	#ifdef __LITTLE_ENDIAN__
29983	30004	#define vst1_lane_p8(__p0, __p1, __p2) __extension__ ({ \
29984	30005	poly8x8_t __s1 = __p1; \
29985		__builtin_neon_vst1_lane_v(__p0, (int8x8_t)__s1, __p2, 4); \
	30006	__builtin_neon_vst1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 4); \
29986	30007	})
29987	30008	#else
29988	30009	#define vst1_lane_p8(__p0, __p1, __p2) __extension__ ({ \
29989	30010	poly8x8_t __s1 = __p1; \
29990		poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
29991		__builtin_neon_vst1_lane_v(__p0, (int8x8_t)__rev1, __p2, 4); \
	30011	poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	30012	__builtin_neon_vst1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), __p2, 4); \
29992	30013	})
29993	30014	#endif
29994	30015
29995	30016	#ifdef __LITTLE_ENDIAN__
29996	30017	#define vst1_lane_p16(__p0, __p1, __p2) __extension__ ({ \
29997	30018	poly16x4_t __s1 = __p1; \
29998		__builtin_neon_vst1_lane_v(__p0, (int8x8_t)__s1, __p2, 5); \
	30019	__builtin_neon_vst1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 5); \
29999	30020	})
30000	30021	#else
30001	30022	#define vst1_lane_p16(__p0, __p1, __p2) __extension__ ({ \
30002	30023	poly16x4_t __s1 = __p1; \
30003		poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
30004		__builtin_neon_vst1_lane_v(__p0, (int8x8_t)__rev1, __p2, 5); \
	30024	poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	30025	__builtin_neon_vst1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), __p2, 5); \
30005	30026	})
30006	30027	#endif
30007	30028
30008	30029	#ifdef __LITTLE_ENDIAN__
30009	30030	#define vst1q_lane_p8(__p0, __p1, __p2) __extension__ ({ \
30010	30031	poly8x16_t __s1 = __p1; \
30011		__builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__s1, __p2, 36); \
	30032	__builtin_neon_vst1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 36); \
30012	30033	})
30013	30034	#else
30014	30035	#define vst1q_lane_p8(__p0, __p1, __p2) __extension__ ({ \
30015	30036	poly8x16_t __s1 = __p1; \
30016		poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
30017		__builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 36); \
	30037	poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	30038	__builtin_neon_vst1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 36); \
30018	30039	})
30019	30040	#endif
30020	30041
30021	30042	#ifdef __LITTLE_ENDIAN__
30022	30043	#define vst1q_lane_p16(__p0, __p1, __p2) __extension__ ({ \
30023	30044	poly16x8_t __s1 = __p1; \
30024		__builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__s1, __p2, 37); \
	30045	__builtin_neon_vst1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 37); \
30025	30046	})
30026	30047	#else
30027	30048	#define vst1q_lane_p16(__p0, __p1, __p2) __extension__ ({ \
30028	30049	poly16x8_t __s1 = __p1; \
30029		poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
30030		__builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 37); \
	30050	poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	30051	__builtin_neon_vst1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 37); \
30031	30052	})
30032	30053	#endif
30033	30054
30034	30055	#ifdef __LITTLE_ENDIAN__
30035	30056	#define vst1q_lane_u8(__p0, __p1, __p2) __extension__ ({ \
30036	30057	uint8x16_t __s1 = __p1; \
30037		__builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__s1, __p2, 48); \
	30058	__builtin_neon_vst1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 48); \
30038	30059	})
30039	30060	#else
30040	30061	#define vst1q_lane_u8(__p0, __p1, __p2) __extension__ ({ \
30041	30062	uint8x16_t __s1 = __p1; \
30042		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
30043		__builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 48); \
	30063	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	30064	__builtin_neon_vst1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 48); \
30044	30065	})
30045	30066	#endif
30046	30067
30047	30068	#ifdef __LITTLE_ENDIAN__
30048	30069	#define vst1q_lane_u32(__p0, __p1, __p2) __extension__ ({ \
30049	30070	uint32x4_t __s1 = __p1; \
30050		__builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__s1, __p2, 50); \
	30071	__builtin_neon_vst1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 50); \
30051	30072	})
30052	30073	#else
30053	30074	#define vst1q_lane_u32(__p0, __p1, __p2) __extension__ ({ \
30054	30075	uint32x4_t __s1 = __p1; \
30055		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
30056		__builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 50); \
	30076	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	30077	__builtin_neon_vst1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 50); \
30057	30078	})
30058	30079	#endif
30059	30080
30060	30081	#ifdef __LITTLE_ENDIAN__
30061	30082	#define vst1q_lane_u64(__p0, __p1, __p2) __extension__ ({ \
30062	30083	uint64x2_t __s1 = __p1; \
30063		__builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__s1, __p2, 51); \
	30084	__builtin_neon_vst1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 51); \
30064	30085	})
30065	30086	#else
30066	30087	#define vst1q_lane_u64(__p0, __p1, __p2) __extension__ ({ \
30067	30088	uint64x2_t __s1 = __p1; \
30068		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
30069		__builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 51); \
	30089	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	30090	__builtin_neon_vst1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 51); \
30070	30091	})
30071	30092	#endif
30072	30093
30073	30094	#ifdef __LITTLE_ENDIAN__
30074	30095	#define vst1q_lane_u16(__p0, __p1, __p2) __extension__ ({ \
30075	30096	uint16x8_t __s1 = __p1; \
30076		__builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__s1, __p2, 49); \
	30097	__builtin_neon_vst1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 49); \
30077	30098	})
30078	30099	#else
30079	30100	#define vst1q_lane_u16(__p0, __p1, __p2) __extension__ ({ \
30080	30101	uint16x8_t __s1 = __p1; \
30081		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
30082		__builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 49); \
	30102	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	30103	__builtin_neon_vst1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 49); \
30083	30104	})
30084	30105	#endif
30085	30106
30086	30107	#ifdef __LITTLE_ENDIAN__
30087	30108	#define vst1q_lane_s8(__p0, __p1, __p2) __extension__ ({ \
30088	30109	int8x16_t __s1 = __p1; \
30089		__builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__s1, __p2, 32); \
	30110	__builtin_neon_vst1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 32); \
30090	30111	})
30091	30112	#else
30092	30113	#define vst1q_lane_s8(__p0, __p1, __p2) __extension__ ({ \
30093	30114	int8x16_t __s1 = __p1; \
30094		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
30095		__builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 32); \
	30115	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	30116	__builtin_neon_vst1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 32); \
30096	30117	})
30097	30118	#endif
30098	30119
30099	30120	#ifdef __LITTLE_ENDIAN__
30100	30121	#define vst1q_lane_f32(__p0, __p1, __p2) __extension__ ({ \
30101	30122	float32x4_t __s1 = __p1; \
30102		__builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__s1, __p2, 41); \
	30123	__builtin_neon_vst1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 41); \
30103	30124	})
30104	30125	#else
30105	30126	#define vst1q_lane_f32(__p0, __p1, __p2) __extension__ ({ \
30106	30127	float32x4_t __s1 = __p1; \
30107		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
30108		__builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 41); \
	30128	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	30129	__builtin_neon_vst1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 41); \
30109	30130	})
30110	30131	#endif
30111	30132
30112	30133	#ifdef __LITTLE_ENDIAN__
30113	30134	#define vst1q_lane_s32(__p0, __p1, __p2) __extension__ ({ \
30114	30135	int32x4_t __s1 = __p1; \
30115		__builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__s1, __p2, 34); \
	30136	__builtin_neon_vst1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 34); \
30116	30137	})
30117	30138	#else
30118	30139	#define vst1q_lane_s32(__p0, __p1, __p2) __extension__ ({ \
30119	30140	int32x4_t __s1 = __p1; \
30120		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
30121		__builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 34); \
	30141	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	30142	__builtin_neon_vst1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 34); \
30122	30143	})
30123	30144	#endif
30124	30145
30125	30146	#ifdef __LITTLE_ENDIAN__
30126	30147	#define vst1q_lane_s64(__p0, __p1, __p2) __extension__ ({ \
30127	30148	int64x2_t __s1 = __p1; \
30128		__builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__s1, __p2, 35); \
	30149	__builtin_neon_vst1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 35); \
30129	30150	})
30130	30151	#else
30131	30152	#define vst1q_lane_s64(__p0, __p1, __p2) __extension__ ({ \
30132	30153	int64x2_t __s1 = __p1; \
30133		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
30134		__builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 35); \
	30154	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	30155	__builtin_neon_vst1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 35); \
30135	30156	})
30136	30157	#endif
30137	30158
30138	30159	#ifdef __LITTLE_ENDIAN__
30139	30160	#define vst1q_lane_s16(__p0, __p1, __p2) __extension__ ({ \
30140	30161	int16x8_t __s1 = __p1; \
30141		__builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__s1, __p2, 33); \
	30162	__builtin_neon_vst1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 33); \
30142	30163	})
30143	30164	#else
30144	30165	#define vst1q_lane_s16(__p0, __p1, __p2) __extension__ ({ \
30145	30166	int16x8_t __s1 = __p1; \
30146		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
30147		__builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 33); \
	30167	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	30168	__builtin_neon_vst1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 33); \
30148	30169	})
30149	30170	#endif
30150	30171
30151	30172	#ifdef __LITTLE_ENDIAN__
30152	30173	#define vst1_lane_u8(__p0, __p1, __p2) __extension__ ({ \
30153	30174	uint8x8_t __s1 = __p1; \
30154		__builtin_neon_vst1_lane_v(__p0, (int8x8_t)__s1, __p2, 16); \
	30175	__builtin_neon_vst1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 16); \
30155	30176	})
30156	30177	#else
30157	30178	#define vst1_lane_u8(__p0, __p1, __p2) __extension__ ({ \
30158	30179	uint8x8_t __s1 = __p1; \
30159		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
30160		__builtin_neon_vst1_lane_v(__p0, (int8x8_t)__rev1, __p2, 16); \
	30180	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	30181	__builtin_neon_vst1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), __p2, 16); \
30161	30182	})
30162	30183	#endif
30163	30184
30164	30185	#ifdef __LITTLE_ENDIAN__
30165	30186	#define vst1_lane_u32(__p0, __p1, __p2) __extension__ ({ \
30166	30187	uint32x2_t __s1 = __p1; \
30167		__builtin_neon_vst1_lane_v(__p0, (int8x8_t)__s1, __p2, 18); \
	30188	__builtin_neon_vst1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 18); \
30168	30189	})
30169	30190	#else
30170	30191	#define vst1_lane_u32(__p0, __p1, __p2) __extension__ ({ \
30171	30192	uint32x2_t __s1 = __p1; \
30172		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
30173		__builtin_neon_vst1_lane_v(__p0, (int8x8_t)__rev1, __p2, 18); \
	30193	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_32); \
	30194	__builtin_neon_vst1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), __p2, 18); \
30174	30195	})
30175	30196	#endif
30176	30197
30177	30198	#define vst1_lane_u64(__p0, __p1, __p2) __extension__ ({ \
30178	30199	uint64x1_t __s1 = __p1; \
30179		__builtin_neon_vst1_lane_v(__p0, (int8x8_t)__s1, __p2, 19); \
	30200	__builtin_neon_vst1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 19); \
30180	30201	})
30181	30202	#ifdef __LITTLE_ENDIAN__
30182	30203	#define vst1_lane_u16(__p0, __p1, __p2) __extension__ ({ \
30183	30204	uint16x4_t __s1 = __p1; \
30184		__builtin_neon_vst1_lane_v(__p0, (int8x8_t)__s1, __p2, 17); \
	30205	__builtin_neon_vst1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 17); \
30185	30206	})
30186	30207	#else
30187	30208	#define vst1_lane_u16(__p0, __p1, __p2) __extension__ ({ \
30188	30209	uint16x4_t __s1 = __p1; \
30189		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
30190		__builtin_neon_vst1_lane_v(__p0, (int8x8_t)__rev1, __p2, 17); \
	30210	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	30211	__builtin_neon_vst1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), __p2, 17); \
30191	30212	})
30192	30213	#endif
30193	30214
30194	30215	#ifdef __LITTLE_ENDIAN__
30195	30216	#define vst1_lane_s8(__p0, __p1, __p2) __extension__ ({ \
30196	30217	int8x8_t __s1 = __p1; \
30197		__builtin_neon_vst1_lane_v(__p0, (int8x8_t)__s1, __p2, 0); \
	30218	__builtin_neon_vst1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 0); \
30198	30219	})
30199	30220	#else
30200	30221	#define vst1_lane_s8(__p0, __p1, __p2) __extension__ ({ \
30201	30222	int8x8_t __s1 = __p1; \
30202		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
30203		__builtin_neon_vst1_lane_v(__p0, (int8x8_t)__rev1, __p2, 0); \
	30223	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	30224	__builtin_neon_vst1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), __p2, 0); \
30204	30225	})
30205	30226	#endif
30206	30227
30207	30228	#ifdef __LITTLE_ENDIAN__
30208	30229	#define vst1_lane_f32(__p0, __p1, __p2) __extension__ ({ \
30209	30230	float32x2_t __s1 = __p1; \
30210		__builtin_neon_vst1_lane_v(__p0, (int8x8_t)__s1, __p2, 9); \
	30231	__builtin_neon_vst1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 9); \
30211	30232	})
30212	30233	#else
30213	30234	#define vst1_lane_f32(__p0, __p1, __p2) __extension__ ({ \
30214	30235	float32x2_t __s1 = __p1; \
30215		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
30216		__builtin_neon_vst1_lane_v(__p0, (int8x8_t)__rev1, __p2, 9); \
	30236	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_32); \
	30237	__builtin_neon_vst1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), __p2, 9); \
30217	30238	})
30218	30239	#endif
30219	30240
30220	30241	#ifdef __LITTLE_ENDIAN__
30221	30242	#define vst1_lane_s32(__p0, __p1, __p2) __extension__ ({ \
30222	30243	int32x2_t __s1 = __p1; \
30223		__builtin_neon_vst1_lane_v(__p0, (int8x8_t)__s1, __p2, 2); \
	30244	__builtin_neon_vst1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 2); \
30224	30245	})
30225	30246	#else
30226	30247	#define vst1_lane_s32(__p0, __p1, __p2) __extension__ ({ \
30227	30248	int32x2_t __s1 = __p1; \
30228		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
30229		__builtin_neon_vst1_lane_v(__p0, (int8x8_t)__rev1, __p2, 2); \
	30249	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_32); \
	30250	__builtin_neon_vst1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), __p2, 2); \
30230	30251	})
30231	30252	#endif
30232	30253
30233	30254	#define vst1_lane_s64(__p0, __p1, __p2) __extension__ ({ \
30234	30255	int64x1_t __s1 = __p1; \
30235		__builtin_neon_vst1_lane_v(__p0, (int8x8_t)__s1, __p2, 3); \
	30256	__builtin_neon_vst1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 3); \
30236	30257	})
30237	30258	#ifdef __LITTLE_ENDIAN__
30238	30259	#define vst1_lane_s16(__p0, __p1, __p2) __extension__ ({ \
30239	30260	int16x4_t __s1 = __p1; \
30240		__builtin_neon_vst1_lane_v(__p0, (int8x8_t)__s1, __p2, 1); \
	30261	__builtin_neon_vst1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 1); \
30241	30262	})
30242	30263	#else
30243	30264	#define vst1_lane_s16(__p0, __p1, __p2) __extension__ ({ \
30244	30265	int16x4_t __s1 = __p1; \
30245		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
30246		__builtin_neon_vst1_lane_v(__p0, (int8x8_t)__rev1, __p2, 1); \
	30266	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	30267	__builtin_neon_vst1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), __p2, 1); \
30247	30268	})
30248	30269	#endif
30249	30270
30250	30271	#ifdef __LITTLE_ENDIAN__
30251	30272	#define vst1_p8_x2(__p0, __p1) __extension__ ({ \
30252	30273	poly8x8x2_t __s1 = __p1; \
30253		__builtin_neon_vst1_x2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 4); \
	30274	__builtin_neon_vst1_x2_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 4); \
30254	30275	})
30255	30276	#else
30256	30277	#define vst1_p8_x2(__p0, __p1) __extension__ ({ \
30257	30278	poly8x8x2_t __s1 = __p1; \
30258	30279	poly8x8x2_t __rev1; \
30259		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
30260		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
30261		__builtin_neon_vst1_x2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 4); \
	30280	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	30281	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	30282	__builtin_neon_vst1_x2_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), 4); \
30262	30283	})
30263	30284	#endif
30264	30285
30265	30286	#ifdef __LITTLE_ENDIAN__
30266	30287	#define vst1_p16_x2(__p0, __p1) __extension__ ({ \
30267	30288	poly16x4x2_t __s1 = __p1; \
30268		__builtin_neon_vst1_x2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 5); \
	30289	__builtin_neon_vst1_x2_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 5); \
30269	30290	})
30270	30291	#else
30271	30292	#define vst1_p16_x2(__p0, __p1) __extension__ ({ \
30272	30293	poly16x4x2_t __s1 = __p1; \
30273	30294	poly16x4x2_t __rev1; \
30274		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
30275		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
30276		__builtin_neon_vst1_x2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 5); \
	30295	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	30296	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	30297	__builtin_neon_vst1_x2_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), 5); \
30277	30298	})
30278	30299	#endif
30279	30300
30280	30301	#ifdef __LITTLE_ENDIAN__
30281	30302	#define vst1q_p8_x2(__p0, __p1) __extension__ ({ \
30282	30303	poly8x16x2_t __s1 = __p1; \
30283		__builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 36); \
	30304	__builtin_neon_vst1q_x2_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 36); \
30284	30305	})
30285	30306	#else
30286	30307	#define vst1q_p8_x2(__p0, __p1) __extension__ ({ \
30287	30308	poly8x16x2_t __s1 = __p1; \
30288	30309	poly8x16x2_t __rev1; \
30289		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
30290		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
30291		__builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 36); \
	30310	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	30311	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	30312	__builtin_neon_vst1q_x2_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 36); \
30292	30313	})
30293	30314	#endif
30294	30315
30295	30316	#ifdef __LITTLE_ENDIAN__
30296	30317	#define vst1q_p16_x2(__p0, __p1) __extension__ ({ \
30297	30318	poly16x8x2_t __s1 = __p1; \
30298		__builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 37); \
	30319	__builtin_neon_vst1q_x2_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 37); \
30299	30320	})
30300	30321	#else
30301	30322	#define vst1q_p16_x2(__p0, __p1) __extension__ ({ \
30302	30323	poly16x8x2_t __s1 = __p1; \
30303	30324	poly16x8x2_t __rev1; \
30304		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
30305		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
30306		__builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 37); \
	30325	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	30326	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	30327	__builtin_neon_vst1q_x2_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 37); \
30307	30328	})
30308	30329	#endif
30309	30330
30310	30331	#ifdef __LITTLE_ENDIAN__
30311	30332	#define vst1q_u8_x2(__p0, __p1) __extension__ ({ \
30312	30333	uint8x16x2_t __s1 = __p1; \
30313		__builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 48); \
	30334	__builtin_neon_vst1q_x2_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 48); \
30314	30335	})
30315	30336	#else
30316	30337	#define vst1q_u8_x2(__p0, __p1) __extension__ ({ \
30317	30338	uint8x16x2_t __s1 = __p1; \
30318	30339	uint8x16x2_t __rev1; \
30319		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
30320		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
30321		__builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 48); \
	30340	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	30341	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	30342	__builtin_neon_vst1q_x2_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 48); \
30322	30343	})
30323	30344	#endif
30324	30345
30325	30346	#ifdef __LITTLE_ENDIAN__
30326	30347	#define vst1q_u32_x2(__p0, __p1) __extension__ ({ \
30327	30348	uint32x4x2_t __s1 = __p1; \
30328		__builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 50); \
	30349	__builtin_neon_vst1q_x2_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 50); \
30329	30350	})
30330	30351	#else
30331	30352	#define vst1q_u32_x2(__p0, __p1) __extension__ ({ \
30332	30353	uint32x4x2_t __s1 = __p1; \
30333	30354	uint32x4x2_t __rev1; \
30334		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
30335		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
30336		__builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 50); \
	30355	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	30356	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	30357	__builtin_neon_vst1q_x2_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 50); \
30337	30358	})
30338	30359	#endif
30339	30360
30340	30361	#ifdef __LITTLE_ENDIAN__
30341	30362	#define vst1q_u64_x2(__p0, __p1) __extension__ ({ \
30342	30363	uint64x2x2_t __s1 = __p1; \
30343		__builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 51); \
	30364	__builtin_neon_vst1q_x2_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 51); \
30344	30365	})
30345	30366	#else
30346	30367	#define vst1q_u64_x2(__p0, __p1) __extension__ ({ \
30347	30368	uint64x2x2_t __s1 = __p1; \
30348	30369	uint64x2x2_t __rev1; \
30349		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
30350		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
30351		__builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 51); \
	30370	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	30371	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	30372	__builtin_neon_vst1q_x2_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 51); \
30352	30373	})
30353	30374	#endif
30354	30375
30355	30376	#ifdef __LITTLE_ENDIAN__
30356	30377	#define vst1q_u16_x2(__p0, __p1) __extension__ ({ \
30357	30378	uint16x8x2_t __s1 = __p1; \
30358		__builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 49); \
	30379	__builtin_neon_vst1q_x2_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 49); \
30359	30380	})
30360	30381	#else
30361	30382	#define vst1q_u16_x2(__p0, __p1) __extension__ ({ \
30362	30383	uint16x8x2_t __s1 = __p1; \
30363	30384	uint16x8x2_t __rev1; \
30364		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
30365		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
30366		__builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 49); \
	30385	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	30386	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	30387	__builtin_neon_vst1q_x2_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 49); \
30367	30388	})
30368	30389	#endif
30369	30390
30370	30391	#ifdef __LITTLE_ENDIAN__
30371	30392	#define vst1q_s8_x2(__p0, __p1) __extension__ ({ \
30372	30393	int8x16x2_t __s1 = __p1; \
30373		__builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 32); \
	30394	__builtin_neon_vst1q_x2_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 32); \
30374	30395	})
30375	30396	#else
30376	30397	#define vst1q_s8_x2(__p0, __p1) __extension__ ({ \
30377	30398	int8x16x2_t __s1 = __p1; \
30378	30399	int8x16x2_t __rev1; \
30379		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
30380		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
30381		__builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 32); \
	30400	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	30401	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	30402	__builtin_neon_vst1q_x2_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 32); \
30382	30403	})
30383	30404	#endif
30384	30405
30385	30406	#ifdef __LITTLE_ENDIAN__
30386	30407	#define vst1q_f32_x2(__p0, __p1) __extension__ ({ \
30387	30408	float32x4x2_t __s1 = __p1; \
30388		__builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 41); \
	30409	__builtin_neon_vst1q_x2_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 41); \
30389	30410	})
30390	30411	#else
30391	30412	#define vst1q_f32_x2(__p0, __p1) __extension__ ({ \
30392	30413	float32x4x2_t __s1 = __p1; \
30393	30414	float32x4x2_t __rev1; \
30394		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
30395		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
30396		__builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 41); \
	30415	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	30416	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	30417	__builtin_neon_vst1q_x2_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 41); \
30397	30418	})
30398	30419	#endif
30399	30420
30400	30421	#ifdef __LITTLE_ENDIAN__
30401	30422	#define vst1q_s32_x2(__p0, __p1) __extension__ ({ \
30402	30423	int32x4x2_t __s1 = __p1; \
30403		__builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 34); \
	30424	__builtin_neon_vst1q_x2_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 34); \
30404	30425	})
30405	30426	#else
30406	30427	#define vst1q_s32_x2(__p0, __p1) __extension__ ({ \
30407	30428	int32x4x2_t __s1 = __p1; \
30408	30429	int32x4x2_t __rev1; \
30409		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
30410		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
30411		__builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 34); \
	30430	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	30431	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	30432	__builtin_neon_vst1q_x2_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 34); \
30412	30433	})
30413	30434	#endif
30414	30435
30415	30436	#ifdef __LITTLE_ENDIAN__
30416	30437	#define vst1q_s64_x2(__p0, __p1) __extension__ ({ \
30417	30438	int64x2x2_t __s1 = __p1; \
30418		__builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 35); \
	30439	__builtin_neon_vst1q_x2_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 35); \
30419	30440	})
30420	30441	#else
30421	30442	#define vst1q_s64_x2(__p0, __p1) __extension__ ({ \
30422	30443	int64x2x2_t __s1 = __p1; \
30423	30444	int64x2x2_t __rev1; \
30424		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
30425		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
30426		__builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 35); \
	30445	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	30446	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	30447	__builtin_neon_vst1q_x2_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 35); \
30427	30448	})
30428	30449	#endif
30429	30450
30430	30451	#ifdef __LITTLE_ENDIAN__
30431	30452	#define vst1q_s16_x2(__p0, __p1) __extension__ ({ \
30432	30453	int16x8x2_t __s1 = __p1; \
30433		__builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 33); \
	30454	__builtin_neon_vst1q_x2_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 33); \
30434	30455	})
30435	30456	#else
30436	30457	#define vst1q_s16_x2(__p0, __p1) __extension__ ({ \
30437	30458	int16x8x2_t __s1 = __p1; \
30438	30459	int16x8x2_t __rev1; \
30439		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
30440		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
30441		__builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 33); \
	30460	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	30461	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	30462	__builtin_neon_vst1q_x2_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 33); \
30442	30463	})
30443	30464	#endif
30444	30465
30445	30466	#ifdef __LITTLE_ENDIAN__
30446	30467	#define vst1_u8_x2(__p0, __p1) __extension__ ({ \
30447	30468	uint8x8x2_t __s1 = __p1; \
30448		__builtin_neon_vst1_x2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 16); \
	30469	__builtin_neon_vst1_x2_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 16); \
30449	30470	})
30450	30471	#else
30451	30472	#define vst1_u8_x2(__p0, __p1) __extension__ ({ \
30452	30473	uint8x8x2_t __s1 = __p1; \
30453	30474	uint8x8x2_t __rev1; \
30454		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
30455		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
30456		__builtin_neon_vst1_x2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 16); \
	30475	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	30476	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	30477	__builtin_neon_vst1_x2_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), 16); \
30457	30478	})
30458	30479	#endif
30459	30480
30460	30481	#ifdef __LITTLE_ENDIAN__
30461	30482	#define vst1_u32_x2(__p0, __p1) __extension__ ({ \
30462	30483	uint32x2x2_t __s1 = __p1; \
30463		__builtin_neon_vst1_x2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 18); \
	30484	__builtin_neon_vst1_x2_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 18); \
30464	30485	})
30465	30486	#else
30466	30487	#define vst1_u32_x2(__p0, __p1) __extension__ ({ \
30467	30488	uint32x2x2_t __s1 = __p1; \
30468	30489	uint32x2x2_t __rev1; \
30469		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
30470		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
30471		__builtin_neon_vst1_x2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 18); \
	30490	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	30491	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	30492	__builtin_neon_vst1_x2_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), 18); \
30472	30493	})
30473	30494	#endif
30474	30495
30475	30496	#define vst1_u64_x2(__p0, __p1) __extension__ ({ \
30476	30497	uint64x1x2_t __s1 = __p1; \
30477		__builtin_neon_vst1_x2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 19); \
	30498	__builtin_neon_vst1_x2_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 19); \
30478	30499	})
30479	30500	#ifdef __LITTLE_ENDIAN__
30480	30501	#define vst1_u16_x2(__p0, __p1) __extension__ ({ \
30481	30502	uint16x4x2_t __s1 = __p1; \
30482		__builtin_neon_vst1_x2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 17); \
	30503	__builtin_neon_vst1_x2_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 17); \
30483	30504	})
30484	30505	#else
30485	30506	#define vst1_u16_x2(__p0, __p1) __extension__ ({ \
30486	30507	uint16x4x2_t __s1 = __p1; \
30487	30508	uint16x4x2_t __rev1; \
30488		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
30489		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
30490		__builtin_neon_vst1_x2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 17); \
	30509	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	30510	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	30511	__builtin_neon_vst1_x2_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), 17); \
30491	30512	})
30492	30513	#endif
30493	30514
30494	30515	#ifdef __LITTLE_ENDIAN__
30495	30516	#define vst1_s8_x2(__p0, __p1) __extension__ ({ \
30496	30517	int8x8x2_t __s1 = __p1; \
30497		__builtin_neon_vst1_x2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 0); \
	30518	__builtin_neon_vst1_x2_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 0); \
30498	30519	})
30499	30520	#else
30500	30521	#define vst1_s8_x2(__p0, __p1) __extension__ ({ \
30501	30522	int8x8x2_t __s1 = __p1; \
30502	30523	int8x8x2_t __rev1; \
30503		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
30504		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
30505		__builtin_neon_vst1_x2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 0); \
	30524	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	30525	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	30526	__builtin_neon_vst1_x2_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), 0); \
30506	30527	})
30507	30528	#endif
30508	30529
30509	30530	#ifdef __LITTLE_ENDIAN__
30510	30531	#define vst1_f32_x2(__p0, __p1) __extension__ ({ \
30511	30532	float32x2x2_t __s1 = __p1; \
30512		__builtin_neon_vst1_x2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 9); \
	30533	__builtin_neon_vst1_x2_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 9); \
30513	30534	})
30514	30535	#else
30515	30536	#define vst1_f32_x2(__p0, __p1) __extension__ ({ \
30516	30537	float32x2x2_t __s1 = __p1; \
30517	30538	float32x2x2_t __rev1; \
30518		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
30519		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
30520		__builtin_neon_vst1_x2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 9); \
	30539	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	30540	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	30541	__builtin_neon_vst1_x2_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), 9); \
30521	30542	})
30522	30543	#endif
30523	30544
30524	30545	#ifdef __LITTLE_ENDIAN__
30525	30546	#define vst1_s32_x2(__p0, __p1) __extension__ ({ \
30526	30547	int32x2x2_t __s1 = __p1; \
30527		__builtin_neon_vst1_x2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 2); \
	30548	__builtin_neon_vst1_x2_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 2); \
30528	30549	})
30529	30550	#else
30530	30551	#define vst1_s32_x2(__p0, __p1) __extension__ ({ \
30531	30552	int32x2x2_t __s1 = __p1; \
30532	30553	int32x2x2_t __rev1; \
30533		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
30534		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
30535		__builtin_neon_vst1_x2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 2); \
	30554	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	30555	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	30556	__builtin_neon_vst1_x2_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), 2); \
30536	30557	})
30537	30558	#endif
30538	30559
30539	30560	#define vst1_s64_x2(__p0, __p1) __extension__ ({ \
30540	30561	int64x1x2_t __s1 = __p1; \
30541		__builtin_neon_vst1_x2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 3); \
	30562	__builtin_neon_vst1_x2_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 3); \
30542	30563	})
30543	30564	#ifdef __LITTLE_ENDIAN__
30544	30565	#define vst1_s16_x2(__p0, __p1) __extension__ ({ \
30545	30566	int16x4x2_t __s1 = __p1; \
30546		__builtin_neon_vst1_x2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 1); \
	30567	__builtin_neon_vst1_x2_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 1); \
30547	30568	})
30548	30569	#else
30549	30570	#define vst1_s16_x2(__p0, __p1) __extension__ ({ \
30550	30571	int16x4x2_t __s1 = __p1; \
30551	30572	int16x4x2_t __rev1; \
30552		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
30553		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
30554		__builtin_neon_vst1_x2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 1); \
	30573	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	30574	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	30575	__builtin_neon_vst1_x2_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), 1); \
30555	30576	})
30556	30577	#endif
30557	30578
30558	30579	#ifdef __LITTLE_ENDIAN__
30559	30580	#define vst1_p8_x3(__p0, __p1) __extension__ ({ \
30560	30581	poly8x8x3_t __s1 = __p1; \
30561		__builtin_neon_vst1_x3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 4); \
	30582	__builtin_neon_vst1_x3_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 4); \
30562	30583	})
30563	30584	#else
30564	30585	#define vst1_p8_x3(__p0, __p1) __extension__ ({ \
30565	30586	poly8x8x3_t __s1 = __p1; \
30566	30587	poly8x8x3_t __rev1; \
30567		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
30568		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
30569		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
30570		__builtin_neon_vst1_x3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 4); \
	30588	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	30589	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	30590	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	30591	__builtin_neon_vst1_x3_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), 4); \
30571	30592	})
30572	30593	#endif
30573	30594
30574	30595	#ifdef __LITTLE_ENDIAN__
30575	30596	#define vst1_p16_x3(__p0, __p1) __extension__ ({ \
30576	30597	poly16x4x3_t __s1 = __p1; \
30577		__builtin_neon_vst1_x3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 5); \
	30598	__builtin_neon_vst1_x3_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 5); \
30578	30599	})
30579	30600	#else
30580	30601	#define vst1_p16_x3(__p0, __p1) __extension__ ({ \
30581	30602	poly16x4x3_t __s1 = __p1; \
30582	30603	poly16x4x3_t __rev1; \
30583		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
30584		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
30585		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
30586		__builtin_neon_vst1_x3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 5); \
	30604	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	30605	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	30606	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	30607	__builtin_neon_vst1_x3_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), 5); \
30587	30608	})
30588	30609	#endif
30589	30610
30590	30611	#ifdef __LITTLE_ENDIAN__
30591	30612	#define vst1q_p8_x3(__p0, __p1) __extension__ ({ \
30592	30613	poly8x16x3_t __s1 = __p1; \
30593		__builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 36); \
	30614	__builtin_neon_vst1q_x3_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 36); \
30594	30615	})
30595	30616	#else
30596	30617	#define vst1q_p8_x3(__p0, __p1) __extension__ ({ \
30597	30618	poly8x16x3_t __s1 = __p1; \
30598	30619	poly8x16x3_t __rev1; \
30599		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
30600		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
30601		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
30602		__builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 36); \
	30620	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	30621	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	30622	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	30623	__builtin_neon_vst1q_x3_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 36); \
30603	30624	})
30604	30625	#endif
30605	30626
30606	30627	#ifdef __LITTLE_ENDIAN__
30607	30628	#define vst1q_p16_x3(__p0, __p1) __extension__ ({ \
30608	30629	poly16x8x3_t __s1 = __p1; \
30609		__builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 37); \
	30630	__builtin_neon_vst1q_x3_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 37); \
30610	30631	})
30611	30632	#else
30612	30633	#define vst1q_p16_x3(__p0, __p1) __extension__ ({ \
30613	30634	poly16x8x3_t __s1 = __p1; \
30614	30635	poly16x8x3_t __rev1; \
30615		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
30616		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
30617		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
30618		__builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 37); \
	30636	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	30637	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	30638	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	30639	__builtin_neon_vst1q_x3_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 37); \
30619	30640	})
30620	30641	#endif
30621	30642
30622	30643	#ifdef __LITTLE_ENDIAN__
30623	30644	#define vst1q_u8_x3(__p0, __p1) __extension__ ({ \
30624	30645	uint8x16x3_t __s1 = __p1; \
30625		__builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 48); \
	30646	__builtin_neon_vst1q_x3_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 48); \
30626	30647	})
30627	30648	#else
30628	30649	#define vst1q_u8_x3(__p0, __p1) __extension__ ({ \
30629	30650	uint8x16x3_t __s1 = __p1; \
30630	30651	uint8x16x3_t __rev1; \
30631		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
30632		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
30633		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
30634		__builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 48); \
	30652	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	30653	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	30654	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	30655	__builtin_neon_vst1q_x3_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 48); \
30635	30656	})
30636	30657	#endif
30637	30658
30638	30659	#ifdef __LITTLE_ENDIAN__
30639	30660	#define vst1q_u32_x3(__p0, __p1) __extension__ ({ \
30640	30661	uint32x4x3_t __s1 = __p1; \
30641		__builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 50); \
	30662	__builtin_neon_vst1q_x3_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 50); \
30642	30663	})
30643	30664	#else
30644	30665	#define vst1q_u32_x3(__p0, __p1) __extension__ ({ \
30645	30666	uint32x4x3_t __s1 = __p1; \
30646	30667	uint32x4x3_t __rev1; \
30647		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
30648		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
30649		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
30650		__builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 50); \
	30668	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	30669	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	30670	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_32); \
	30671	__builtin_neon_vst1q_x3_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 50); \
30651	30672	})
30652	30673	#endif
30653	30674
30654	30675	#ifdef __LITTLE_ENDIAN__
30655	30676	#define vst1q_u64_x3(__p0, __p1) __extension__ ({ \
30656	30677	uint64x2x3_t __s1 = __p1; \
30657		__builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 51); \
	30678	__builtin_neon_vst1q_x3_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 51); \
30658	30679	})
30659	30680	#else
30660	30681	#define vst1q_u64_x3(__p0, __p1) __extension__ ({ \
30661	30682	uint64x2x3_t __s1 = __p1; \
30662	30683	uint64x2x3_t __rev1; \
30663		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
30664		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
30665		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
30666		__builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 51); \
	30684	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	30685	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	30686	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	30687	__builtin_neon_vst1q_x3_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 51); \
30667	30688	})
30668	30689	#endif
30669	30690
30670	30691	#ifdef __LITTLE_ENDIAN__
30671	30692	#define vst1q_u16_x3(__p0, __p1) __extension__ ({ \
30672	30693	uint16x8x3_t __s1 = __p1; \
30673		__builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 49); \
	30694	__builtin_neon_vst1q_x3_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 49); \
30674	30695	})
30675	30696	#else
30676	30697	#define vst1q_u16_x3(__p0, __p1) __extension__ ({ \
30677	30698	uint16x8x3_t __s1 = __p1; \
30678	30699	uint16x8x3_t __rev1; \
30679		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
30680		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
30681		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
30682		__builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 49); \
	30700	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	30701	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	30702	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	30703	__builtin_neon_vst1q_x3_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 49); \
30683	30704	})
30684	30705	#endif
30685	30706
30686	30707	#ifdef __LITTLE_ENDIAN__
30687	30708	#define vst1q_s8_x3(__p0, __p1) __extension__ ({ \
30688	30709	int8x16x3_t __s1 = __p1; \
30689		__builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 32); \
	30710	__builtin_neon_vst1q_x3_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 32); \
30690	30711	})
30691	30712	#else
30692	30713	#define vst1q_s8_x3(__p0, __p1) __extension__ ({ \
30693	30714	int8x16x3_t __s1 = __p1; \
30694	30715	int8x16x3_t __rev1; \
30695		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
30696		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
30697		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
30698		__builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 32); \
	30716	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	30717	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	30718	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	30719	__builtin_neon_vst1q_x3_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 32); \
30699	30720	})
30700	30721	#endif
30701	30722
30702	30723	#ifdef __LITTLE_ENDIAN__
30703	30724	#define vst1q_f32_x3(__p0, __p1) __extension__ ({ \
30704	30725	float32x4x3_t __s1 = __p1; \
30705		__builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 41); \
	30726	__builtin_neon_vst1q_x3_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 41); \
30706	30727	})
30707	30728	#else
30708	30729	#define vst1q_f32_x3(__p0, __p1) __extension__ ({ \
30709	30730	float32x4x3_t __s1 = __p1; \
30710	30731	float32x4x3_t __rev1; \
30711		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
30712		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
30713		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
30714		__builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 41); \
	30732	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	30733	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	30734	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_32); \
	30735	__builtin_neon_vst1q_x3_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 41); \
30715	30736	})
30716	30737	#endif
30717	30738
30718	30739	#ifdef __LITTLE_ENDIAN__
30719	30740	#define vst1q_s32_x3(__p0, __p1) __extension__ ({ \
30720	30741	int32x4x3_t __s1 = __p1; \
30721		__builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 34); \
	30742	__builtin_neon_vst1q_x3_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 34); \
30722	30743	})
30723	30744	#else
30724	30745	#define vst1q_s32_x3(__p0, __p1) __extension__ ({ \
30725	30746	int32x4x3_t __s1 = __p1; \
30726	30747	int32x4x3_t __rev1; \
30727		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
30728		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
30729		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
30730		__builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 34); \
	30748	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	30749	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	30750	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_32); \
	30751	__builtin_neon_vst1q_x3_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 34); \
30731	30752	})
30732	30753	#endif
30733	30754
30734	30755	#ifdef __LITTLE_ENDIAN__
30735	30756	#define vst1q_s64_x3(__p0, __p1) __extension__ ({ \
30736	30757	int64x2x3_t __s1 = __p1; \
30737		__builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 35); \
	30758	__builtin_neon_vst1q_x3_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 35); \
30738	30759	})
30739	30760	#else
30740	30761	#define vst1q_s64_x3(__p0, __p1) __extension__ ({ \
30741	30762	int64x2x3_t __s1 = __p1; \
30742	30763	int64x2x3_t __rev1; \
30743		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
30744		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
30745		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
30746		__builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 35); \
	30764	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	30765	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	30766	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	30767	__builtin_neon_vst1q_x3_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 35); \
30747	30768	})
30748	30769	#endif
30749	30770
30750	30771	#ifdef __LITTLE_ENDIAN__
30751	30772	#define vst1q_s16_x3(__p0, __p1) __extension__ ({ \
30752	30773	int16x8x3_t __s1 = __p1; \
30753		__builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 33); \
	30774	__builtin_neon_vst1q_x3_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 33); \
30754	30775	})
30755	30776	#else
30756	30777	#define vst1q_s16_x3(__p0, __p1) __extension__ ({ \
30757	30778	int16x8x3_t __s1 = __p1; \
30758	30779	int16x8x3_t __rev1; \
30759		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
30760		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
30761		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
30762		__builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 33); \
	30780	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	30781	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	30782	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	30783	__builtin_neon_vst1q_x3_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 33); \
30763	30784	})
30764	30785	#endif
30765	30786
30766	30787	#ifdef __LITTLE_ENDIAN__
30767	30788	#define vst1_u8_x3(__p0, __p1) __extension__ ({ \
30768	30789	uint8x8x3_t __s1 = __p1; \
30769		__builtin_neon_vst1_x3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 16); \
	30790	__builtin_neon_vst1_x3_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 16); \
30770	30791	})
30771	30792	#else
30772	30793	#define vst1_u8_x3(__p0, __p1) __extension__ ({ \
30773	30794	uint8x8x3_t __s1 = __p1; \
30774	30795	uint8x8x3_t __rev1; \
30775		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
30776		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
30777		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
30778		__builtin_neon_vst1_x3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 16); \
	30796	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	30797	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	30798	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	30799	__builtin_neon_vst1_x3_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), 16); \
30779	30800	})
30780	30801	#endif
30781	30802
30782	30803	#ifdef __LITTLE_ENDIAN__
30783	30804	#define vst1_u32_x3(__p0, __p1) __extension__ ({ \
30784	30805	uint32x2x3_t __s1 = __p1; \
30785		__builtin_neon_vst1_x3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 18); \
	30806	__builtin_neon_vst1_x3_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 18); \
30786	30807	})
30787	30808	#else
30788	30809	#define vst1_u32_x3(__p0, __p1) __extension__ ({ \
30789	30810	uint32x2x3_t __s1 = __p1; \
30790	30811	uint32x2x3_t __rev1; \
30791		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
30792		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
30793		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
30794		__builtin_neon_vst1_x3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 18); \
	30812	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	30813	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	30814	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_32); \
	30815	__builtin_neon_vst1_x3_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), 18); \
30795	30816	})
30796	30817	#endif
30797	30818
30798	30819	#define vst1_u64_x3(__p0, __p1) __extension__ ({ \
30799	30820	uint64x1x3_t __s1 = __p1; \
30800		__builtin_neon_vst1_x3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 19); \
	30821	__builtin_neon_vst1_x3_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 19); \
30801	30822	})
30802	30823	#ifdef __LITTLE_ENDIAN__
30803	30824	#define vst1_u16_x3(__p0, __p1) __extension__ ({ \
30804	30825	uint16x4x3_t __s1 = __p1; \
30805		__builtin_neon_vst1_x3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 17); \
	30826	__builtin_neon_vst1_x3_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 17); \
30806	30827	})
30807	30828	#else
30808	30829	#define vst1_u16_x3(__p0, __p1) __extension__ ({ \
30809	30830	uint16x4x3_t __s1 = __p1; \
30810	30831	uint16x4x3_t __rev1; \
30811		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
30812		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
30813		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
30814		__builtin_neon_vst1_x3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 17); \
	30832	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	30833	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	30834	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	30835	__builtin_neon_vst1_x3_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), 17); \
30815	30836	})
30816	30837	#endif
30817	30838
30818	30839	#ifdef __LITTLE_ENDIAN__
30819	30840	#define vst1_s8_x3(__p0, __p1) __extension__ ({ \
30820	30841	int8x8x3_t __s1 = __p1; \
30821		__builtin_neon_vst1_x3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 0); \
	30842	__builtin_neon_vst1_x3_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 0); \
30822	30843	})
30823	30844	#else
30824	30845	#define vst1_s8_x3(__p0, __p1) __extension__ ({ \
30825	30846	int8x8x3_t __s1 = __p1; \
30826	30847	int8x8x3_t __rev1; \
30827		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
30828		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
30829		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
30830		__builtin_neon_vst1_x3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 0); \
	30848	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	30849	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	30850	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	30851	__builtin_neon_vst1_x3_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), 0); \
30831	30852	})
30832	30853	#endif
30833	30854
30834	30855	#ifdef __LITTLE_ENDIAN__
30835	30856	#define vst1_f32_x3(__p0, __p1) __extension__ ({ \
30836	30857	float32x2x3_t __s1 = __p1; \
30837		__builtin_neon_vst1_x3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 9); \
	30858	__builtin_neon_vst1_x3_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 9); \
30838	30859	})
30839	30860	#else
30840	30861	#define vst1_f32_x3(__p0, __p1) __extension__ ({ \
30841	30862	float32x2x3_t __s1 = __p1; \
30842	30863	float32x2x3_t __rev1; \
30843		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
30844		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
30845		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
30846		__builtin_neon_vst1_x3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 9); \
	30864	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	30865	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	30866	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_32); \
	30867	__builtin_neon_vst1_x3_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), 9); \
30847	30868	})
30848	30869	#endif
30849	30870
30850	30871	#ifdef __LITTLE_ENDIAN__
30851	30872	#define vst1_s32_x3(__p0, __p1) __extension__ ({ \
30852	30873	int32x2x3_t __s1 = __p1; \
30853		__builtin_neon_vst1_x3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 2); \
	30874	__builtin_neon_vst1_x3_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 2); \
30854	30875	})
30855	30876	#else
30856	30877	#define vst1_s32_x3(__p0, __p1) __extension__ ({ \
30857	30878	int32x2x3_t __s1 = __p1; \
30858	30879	int32x2x3_t __rev1; \
30859		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
30860		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
30861		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
30862		__builtin_neon_vst1_x3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 2); \
	30880	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	30881	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	30882	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_32); \
	30883	__builtin_neon_vst1_x3_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), 2); \
30863	30884	})
30864	30885	#endif
30865	30886
30866	30887	#define vst1_s64_x3(__p0, __p1) __extension__ ({ \
30867	30888	int64x1x3_t __s1 = __p1; \
30868		__builtin_neon_vst1_x3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 3); \
	30889	__builtin_neon_vst1_x3_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 3); \
30869	30890	})
30870	30891	#ifdef __LITTLE_ENDIAN__
30871	30892	#define vst1_s16_x3(__p0, __p1) __extension__ ({ \
30872	30893	int16x4x3_t __s1 = __p1; \
30873		__builtin_neon_vst1_x3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 1); \
	30894	__builtin_neon_vst1_x3_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 1); \
30874	30895	})
30875	30896	#else
30876	30897	#define vst1_s16_x3(__p0, __p1) __extension__ ({ \
30877	30898	int16x4x3_t __s1 = __p1; \
30878	30899	int16x4x3_t __rev1; \
30879		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
30880		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
30881		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
30882		__builtin_neon_vst1_x3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 1); \
	30900	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	30901	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	30902	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	30903	__builtin_neon_vst1_x3_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), 1); \
30883	30904	})
30884	30905	#endif
30885	30906
30886	30907	#ifdef __LITTLE_ENDIAN__
30887	30908	#define vst1_p8_x4(__p0, __p1) __extension__ ({ \
30888	30909	poly8x8x4_t __s1 = __p1; \
30889		__builtin_neon_vst1_x4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 4); \
	30910	__builtin_neon_vst1_x4_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 4); \
30890	30911	})
30891	30912	#else
30892	30913	#define vst1_p8_x4(__p0, __p1) __extension__ ({ \
30893	30914	poly8x8x4_t __s1 = __p1; \
30894	30915	poly8x8x4_t __rev1; \
30895		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
30896		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
30897		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
30898		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
30899		__builtin_neon_vst1_x4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 4); \
	30916	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	30917	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	30918	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	30919	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_8); \
	30920	__builtin_neon_vst1_x4_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), 4); \
30900	30921	})
30901	30922	#endif
30902	30923
30903	30924	#ifdef __LITTLE_ENDIAN__
30904	30925	#define vst1_p16_x4(__p0, __p1) __extension__ ({ \
30905	30926	poly16x4x4_t __s1 = __p1; \
30906		__builtin_neon_vst1_x4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 5); \
	30927	__builtin_neon_vst1_x4_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 5); \
30907	30928	})
30908	30929	#else
30909	30930	#define vst1_p16_x4(__p0, __p1) __extension__ ({ \
30910	30931	poly16x4x4_t __s1 = __p1; \
30911	30932	poly16x4x4_t __rev1; \
30912		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
30913		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
30914		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
30915		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
30916		__builtin_neon_vst1_x4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 5); \
	30933	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	30934	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	30935	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	30936	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_16); \
	30937	__builtin_neon_vst1_x4_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), 5); \
30917	30938	})
30918	30939	#endif
30919	30940
30920	30941	#ifdef __LITTLE_ENDIAN__
30921	30942	#define vst1q_p8_x4(__p0, __p1) __extension__ ({ \
30922	30943	poly8x16x4_t __s1 = __p1; \
30923		__builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 36); \
	30944	__builtin_neon_vst1q_x4_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 36); \
30924	30945	})
30925	30946	#else
30926	30947	#define vst1q_p8_x4(__p0, __p1) __extension__ ({ \
30927	30948	poly8x16x4_t __s1 = __p1; \
30928	30949	poly8x16x4_t __rev1; \
30929		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
30930		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
30931		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
30932		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
30933		__builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 36); \
	30950	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	30951	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	30952	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	30953	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_8); \
	30954	__builtin_neon_vst1q_x4_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 36); \
30934	30955	})
30935	30956	#endif
30936	30957
30937	30958	#ifdef __LITTLE_ENDIAN__
30938	30959	#define vst1q_p16_x4(__p0, __p1) __extension__ ({ \
30939	30960	poly16x8x4_t __s1 = __p1; \
30940		__builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 37); \
	30961	__builtin_neon_vst1q_x4_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 37); \
30941	30962	})
30942	30963	#else
30943	30964	#define vst1q_p16_x4(__p0, __p1) __extension__ ({ \
30944	30965	poly16x8x4_t __s1 = __p1; \
30945	30966	poly16x8x4_t __rev1; \
30946		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
30947		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
30948		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
30949		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
30950		__builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 37); \
	30967	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	30968	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	30969	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	30970	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_16); \
	30971	__builtin_neon_vst1q_x4_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 37); \
30951	30972	})
30952	30973	#endif
30953	30974
30954	30975	#ifdef __LITTLE_ENDIAN__
30955	30976	#define vst1q_u8_x4(__p0, __p1) __extension__ ({ \
30956	30977	uint8x16x4_t __s1 = __p1; \
30957		__builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 48); \
	30978	__builtin_neon_vst1q_x4_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 48); \
30958	30979	})
30959	30980	#else
30960	30981	#define vst1q_u8_x4(__p0, __p1) __extension__ ({ \
30961	30982	uint8x16x4_t __s1 = __p1; \
30962	30983	uint8x16x4_t __rev1; \
30963		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
30964		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
30965		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
30966		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
30967		__builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 48); \
	30984	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	30985	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	30986	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	30987	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_8); \
	30988	__builtin_neon_vst1q_x4_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 48); \
30968	30989	})
30969	30990	#endif
30970	30991
30971	30992	#ifdef __LITTLE_ENDIAN__
30972	30993	#define vst1q_u32_x4(__p0, __p1) __extension__ ({ \
30973	30994	uint32x4x4_t __s1 = __p1; \
30974		__builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 50); \
	30995	__builtin_neon_vst1q_x4_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 50); \
30975	30996	})
30976	30997	#else
30977	30998	#define vst1q_u32_x4(__p0, __p1) __extension__ ({ \
30978	30999	uint32x4x4_t __s1 = __p1; \
30979	31000	uint32x4x4_t __rev1; \
30980		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
30981		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
30982		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
30983		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
30984		__builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 50); \
	31001	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	31002	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	31003	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_32); \
	31004	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_32); \
	31005	__builtin_neon_vst1q_x4_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 50); \
30985	31006	})
30986	31007	#endif
30987	31008
30988	31009	#ifdef __LITTLE_ENDIAN__
30989	31010	#define vst1q_u64_x4(__p0, __p1) __extension__ ({ \
30990	31011	uint64x2x4_t __s1 = __p1; \
30991		__builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 51); \
	31012	__builtin_neon_vst1q_x4_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 51); \
30992	31013	})
30993	31014	#else
30994	31015	#define vst1q_u64_x4(__p0, __p1) __extension__ ({ \
30995	31016	uint64x2x4_t __s1 = __p1; \
30996	31017	uint64x2x4_t __rev1; \
30997		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
30998		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
30999		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
31000		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
31001		__builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 51); \
	31018	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	31019	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	31020	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	31021	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_64); \
	31022	__builtin_neon_vst1q_x4_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 51); \
31002	31023	})
31003	31024	#endif
31004	31025
31005	31026	#ifdef __LITTLE_ENDIAN__
31006	31027	#define vst1q_u16_x4(__p0, __p1) __extension__ ({ \
31007	31028	uint16x8x4_t __s1 = __p1; \
31008		__builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 49); \
	31029	__builtin_neon_vst1q_x4_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 49); \
31009	31030	})
31010	31031	#else
31011	31032	#define vst1q_u16_x4(__p0, __p1) __extension__ ({ \
31012	31033	uint16x8x4_t __s1 = __p1; \
31013	31034	uint16x8x4_t __rev1; \
31014		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
31015		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
31016		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
31017		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
31018		__builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 49); \
	31035	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	31036	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	31037	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	31038	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_16); \
	31039	__builtin_neon_vst1q_x4_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 49); \
31019	31040	})
31020	31041	#endif
31021	31042
31022	31043	#ifdef __LITTLE_ENDIAN__
31023	31044	#define vst1q_s8_x4(__p0, __p1) __extension__ ({ \
31024	31045	int8x16x4_t __s1 = __p1; \
31025		__builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 32); \
	31046	__builtin_neon_vst1q_x4_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 32); \
31026	31047	})
31027	31048	#else
31028	31049	#define vst1q_s8_x4(__p0, __p1) __extension__ ({ \
31029	31050	int8x16x4_t __s1 = __p1; \
31030	31051	int8x16x4_t __rev1; \
31031		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
31032		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
31033		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
31034		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
31035		__builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 32); \
	31052	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	31053	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	31054	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	31055	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_8); \
	31056	__builtin_neon_vst1q_x4_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 32); \
31036	31057	})
31037	31058	#endif
31038	31059
31039	31060	#ifdef __LITTLE_ENDIAN__
31040	31061	#define vst1q_f32_x4(__p0, __p1) __extension__ ({ \
31041	31062	float32x4x4_t __s1 = __p1; \
31042		__builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 41); \
	31063	__builtin_neon_vst1q_x4_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 41); \
31043	31064	})
31044	31065	#else
31045	31066	#define vst1q_f32_x4(__p0, __p1) __extension__ ({ \
31046	31067	float32x4x4_t __s1 = __p1; \
31047	31068	float32x4x4_t __rev1; \
31048		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
31049		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
31050		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
31051		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
31052		__builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 41); \
	31069	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	31070	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	31071	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_32); \
	31072	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_32); \
	31073	__builtin_neon_vst1q_x4_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 41); \
31053	31074	})
31054	31075	#endif
31055	31076
31056	31077	#ifdef __LITTLE_ENDIAN__
31057	31078	#define vst1q_s32_x4(__p0, __p1) __extension__ ({ \
31058	31079	int32x4x4_t __s1 = __p1; \
31059		__builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 34); \
	31080	__builtin_neon_vst1q_x4_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 34); \
31060	31081	})
31061	31082	#else
31062	31083	#define vst1q_s32_x4(__p0, __p1) __extension__ ({ \
31063	31084	int32x4x4_t __s1 = __p1; \
31064	31085	int32x4x4_t __rev1; \
31065		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
31066		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
31067		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
31068		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
31069		__builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 34); \
	31086	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	31087	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	31088	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_32); \
	31089	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_32); \
	31090	__builtin_neon_vst1q_x4_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 34); \
31070	31091	})
31071	31092	#endif
31072	31093
31073	31094	#ifdef __LITTLE_ENDIAN__
31074	31095	#define vst1q_s64_x4(__p0, __p1) __extension__ ({ \
31075	31096	int64x2x4_t __s1 = __p1; \
31076		__builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 35); \
	31097	__builtin_neon_vst1q_x4_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 35); \
31077	31098	})
31078	31099	#else
31079	31100	#define vst1q_s64_x4(__p0, __p1) __extension__ ({ \
31080	31101	int64x2x4_t __s1 = __p1; \
31081	31102	int64x2x4_t __rev1; \
31082		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
31083		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
31084		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
31085		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
31086		__builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 35); \
	31103	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	31104	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	31105	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	31106	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_64); \
	31107	__builtin_neon_vst1q_x4_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 35); \
31087	31108	})
31088	31109	#endif
31089	31110
31090	31111	#ifdef __LITTLE_ENDIAN__
31091	31112	#define vst1q_s16_x4(__p0, __p1) __extension__ ({ \
31092	31113	int16x8x4_t __s1 = __p1; \
31093		__builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 33); \
	31114	__builtin_neon_vst1q_x4_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 33); \
31094	31115	})
31095	31116	#else
31096	31117	#define vst1q_s16_x4(__p0, __p1) __extension__ ({ \
31097	31118	int16x8x4_t __s1 = __p1; \
31098	31119	int16x8x4_t __rev1; \
31099		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
31100		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
31101		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
31102		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
31103		__builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 33); \
	31120	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	31121	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	31122	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	31123	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_16); \
	31124	__builtin_neon_vst1q_x4_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 33); \
31104	31125	})
31105	31126	#endif
31106	31127
31107	31128	#ifdef __LITTLE_ENDIAN__
31108	31129	#define vst1_u8_x4(__p0, __p1) __extension__ ({ \
31109	31130	uint8x8x4_t __s1 = __p1; \
31110		__builtin_neon_vst1_x4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 16); \
	31131	__builtin_neon_vst1_x4_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 16); \
31111	31132	})
31112	31133	#else
31113	31134	#define vst1_u8_x4(__p0, __p1) __extension__ ({ \
31114	31135	uint8x8x4_t __s1 = __p1; \
31115	31136	uint8x8x4_t __rev1; \
31116		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
31117		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
31118		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
31119		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
31120		__builtin_neon_vst1_x4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 16); \
	31137	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	31138	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	31139	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	31140	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_8); \
	31141	__builtin_neon_vst1_x4_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), 16); \
31121	31142	})
31122	31143	#endif
31123	31144
31124	31145	#ifdef __LITTLE_ENDIAN__
31125	31146	#define vst1_u32_x4(__p0, __p1) __extension__ ({ \
31126	31147	uint32x2x4_t __s1 = __p1; \
31127		__builtin_neon_vst1_x4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 18); \
	31148	__builtin_neon_vst1_x4_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 18); \
31128	31149	})
31129	31150	#else
31130	31151	#define vst1_u32_x4(__p0, __p1) __extension__ ({ \
31131	31152	uint32x2x4_t __s1 = __p1; \
31132	31153	uint32x2x4_t __rev1; \
31133		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
31134		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
31135		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
31136		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
31137		__builtin_neon_vst1_x4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 18); \
	31154	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	31155	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	31156	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_32); \
	31157	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_32); \
	31158	__builtin_neon_vst1_x4_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), 18); \
31138	31159	})
31139	31160	#endif
31140	31161
31141	31162	#define vst1_u64_x4(__p0, __p1) __extension__ ({ \
31142	31163	uint64x1x4_t __s1 = __p1; \
31143		__builtin_neon_vst1_x4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 19); \
	31164	__builtin_neon_vst1_x4_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 19); \
31144	31165	})
31145	31166	#ifdef __LITTLE_ENDIAN__
31146	31167	#define vst1_u16_x4(__p0, __p1) __extension__ ({ \
31147	31168	uint16x4x4_t __s1 = __p1; \
31148		__builtin_neon_vst1_x4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 17); \
	31169	__builtin_neon_vst1_x4_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 17); \
31149	31170	})
31150	31171	#else
31151	31172	#define vst1_u16_x4(__p0, __p1) __extension__ ({ \
31152	31173	uint16x4x4_t __s1 = __p1; \
31153	31174	uint16x4x4_t __rev1; \
31154		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
31155		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
31156		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
31157		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
31158		__builtin_neon_vst1_x4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 17); \
	31175	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	31176	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	31177	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	31178	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_16); \
	31179	__builtin_neon_vst1_x4_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), 17); \
31159	31180	})
31160	31181	#endif
31161	31182
31162	31183	#ifdef __LITTLE_ENDIAN__
31163	31184	#define vst1_s8_x4(__p0, __p1) __extension__ ({ \
31164	31185	int8x8x4_t __s1 = __p1; \
31165		__builtin_neon_vst1_x4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 0); \
	31186	__builtin_neon_vst1_x4_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 0); \
31166	31187	})
31167	31188	#else
31168	31189	#define vst1_s8_x4(__p0, __p1) __extension__ ({ \
31169	31190	int8x8x4_t __s1 = __p1; \
31170	31191	int8x8x4_t __rev1; \
31171		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
31172		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
31173		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
31174		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
31175		__builtin_neon_vst1_x4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 0); \
	31192	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	31193	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	31194	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	31195	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_8); \
	31196	__builtin_neon_vst1_x4_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), 0); \
31176	31197	})
31177	31198	#endif
31178	31199
31179	31200	#ifdef __LITTLE_ENDIAN__
31180	31201	#define vst1_f32_x4(__p0, __p1) __extension__ ({ \
31181	31202	float32x2x4_t __s1 = __p1; \
31182		__builtin_neon_vst1_x4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 9); \
	31203	__builtin_neon_vst1_x4_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 9); \
31183	31204	})
31184	31205	#else
31185	31206	#define vst1_f32_x4(__p0, __p1) __extension__ ({ \
31186	31207	float32x2x4_t __s1 = __p1; \
31187	31208	float32x2x4_t __rev1; \
31188		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
31189		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
31190		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
31191		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
31192		__builtin_neon_vst1_x4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 9); \
	31209	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	31210	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	31211	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_32); \
	31212	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_32); \
	31213	__builtin_neon_vst1_x4_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), 9); \
31193	31214	})
31194	31215	#endif
31195	31216
31196	31217	#ifdef __LITTLE_ENDIAN__
31197	31218	#define vst1_s32_x4(__p0, __p1) __extension__ ({ \
31198	31219	int32x2x4_t __s1 = __p1; \
31199		__builtin_neon_vst1_x4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 2); \
	31220	__builtin_neon_vst1_x4_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 2); \
31200	31221	})
31201	31222	#else
31202	31223	#define vst1_s32_x4(__p0, __p1) __extension__ ({ \
31203	31224	int32x2x4_t __s1 = __p1; \
31204	31225	int32x2x4_t __rev1; \
31205		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
31206		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
31207		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
31208		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
31209		__builtin_neon_vst1_x4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 2); \
	31226	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	31227	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	31228	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_32); \
	31229	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_32); \
	31230	__builtin_neon_vst1_x4_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), 2); \
31210	31231	})
31211	31232	#endif
31212	31233
31213	31234	#define vst1_s64_x4(__p0, __p1) __extension__ ({ \
31214	31235	int64x1x4_t __s1 = __p1; \
31215		__builtin_neon_vst1_x4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 3); \
	31236	__builtin_neon_vst1_x4_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 3); \
31216	31237	})
31217	31238	#ifdef __LITTLE_ENDIAN__
31218	31239	#define vst1_s16_x4(__p0, __p1) __extension__ ({ \
31219	31240	int16x4x4_t __s1 = __p1; \
31220		__builtin_neon_vst1_x4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 1); \
	31241	__builtin_neon_vst1_x4_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 1); \
31221	31242	})
31222	31243	#else
31223	31244	#define vst1_s16_x4(__p0, __p1) __extension__ ({ \
31224	31245	int16x4x4_t __s1 = __p1; \
31225	31246	int16x4x4_t __rev1; \
31226		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
31227		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
31228		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
31229		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
31230		__builtin_neon_vst1_x4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 1); \
	31247	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	31248	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	31249	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	31250	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_16); \
	31251	__builtin_neon_vst1_x4_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), 1); \
31231	31252	})
31232	31253	#endif
31233	31254
31234	31255	#ifdef __LITTLE_ENDIAN__
31235	31256	#define vst2_p8(__p0, __p1) __extension__ ({ \
31236	31257	poly8x8x2_t __s1 = __p1; \
31237		__builtin_neon_vst2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 4); \
	31258	__builtin_neon_vst2_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 4); \
31238	31259	})
31239	31260	#else
31240	31261	#define vst2_p8(__p0, __p1) __extension__ ({ \
31241	31262	poly8x8x2_t __s1 = __p1; \
31242	31263	poly8x8x2_t __rev1; \
31243		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
31244		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
31245		__builtin_neon_vst2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 4); \
	31264	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	31265	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	31266	__builtin_neon_vst2_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), 4); \
31246	31267	})
31247	31268	#endif
31248	31269
31249	31270	#ifdef __LITTLE_ENDIAN__
31250	31271	#define vst2_p16(__p0, __p1) __extension__ ({ \
31251	31272	poly16x4x2_t __s1 = __p1; \
31252		__builtin_neon_vst2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 5); \
	31273	__builtin_neon_vst2_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 5); \
31253	31274	})
31254	31275	#else
31255	31276	#define vst2_p16(__p0, __p1) __extension__ ({ \
31256	31277	poly16x4x2_t __s1 = __p1; \
31257	31278	poly16x4x2_t __rev1; \
31258		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
31259		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
31260		__builtin_neon_vst2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 5); \
	31279	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	31280	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	31281	__builtin_neon_vst2_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), 5); \
31261	31282	})
31262	31283	#endif
31263	31284
31264	31285	#ifdef __LITTLE_ENDIAN__
31265	31286	#define vst2q_p8(__p0, __p1) __extension__ ({ \
31266	31287	poly8x16x2_t __s1 = __p1; \
31267		__builtin_neon_vst2q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 36); \
	31288	__builtin_neon_vst2q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 36); \
31268	31289	})
31269	31290	#else
31270	31291	#define vst2q_p8(__p0, __p1) __extension__ ({ \
31271	31292	poly8x16x2_t __s1 = __p1; \
31272	31293	poly8x16x2_t __rev1; \
31273		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
31274		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
31275		__builtin_neon_vst2q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 36); \
	31294	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	31295	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	31296	__builtin_neon_vst2q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 36); \
31276	31297	})
31277	31298	#endif
31278	31299
31279	31300	#ifdef __LITTLE_ENDIAN__
31280	31301	#define vst2q_p16(__p0, __p1) __extension__ ({ \
31281	31302	poly16x8x2_t __s1 = __p1; \
31282		__builtin_neon_vst2q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 37); \
	31303	__builtin_neon_vst2q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 37); \
31283	31304	})
31284	31305	#else
31285	31306	#define vst2q_p16(__p0, __p1) __extension__ ({ \
31286	31307	poly16x8x2_t __s1 = __p1; \
31287	31308	poly16x8x2_t __rev1; \
31288		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
31289		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
31290		__builtin_neon_vst2q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 37); \
	31309	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	31310	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	31311	__builtin_neon_vst2q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 37); \
31291	31312	})
31292	31313	#endif
31293	31314
31294	31315	#ifdef __LITTLE_ENDIAN__
31295	31316	#define vst2q_u8(__p0, __p1) __extension__ ({ \
31296	31317	uint8x16x2_t __s1 = __p1; \
31297		__builtin_neon_vst2q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 48); \
	31318	__builtin_neon_vst2q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 48); \
31298	31319	})
31299	31320	#else
31300	31321	#define vst2q_u8(__p0, __p1) __extension__ ({ \
31301	31322	uint8x16x2_t __s1 = __p1; \
31302	31323	uint8x16x2_t __rev1; \
31303		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
31304		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
31305		__builtin_neon_vst2q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 48); \
	31324	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	31325	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	31326	__builtin_neon_vst2q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 48); \
31306	31327	})
31307	31328	#endif
31308	31329
31309	31330	#ifdef __LITTLE_ENDIAN__
31310	31331	#define vst2q_u32(__p0, __p1) __extension__ ({ \
31311	31332	uint32x4x2_t __s1 = __p1; \
31312		__builtin_neon_vst2q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 50); \
	31333	__builtin_neon_vst2q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 50); \
31313	31334	})
31314	31335	#else
31315	31336	#define vst2q_u32(__p0, __p1) __extension__ ({ \
31316	31337	uint32x4x2_t __s1 = __p1; \
31317	31338	uint32x4x2_t __rev1; \
31318		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
31319		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
31320		__builtin_neon_vst2q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 50); \
	31339	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	31340	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	31341	__builtin_neon_vst2q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 50); \
31321	31342	})
31322	31343	#endif
31323	31344
31324	31345	#ifdef __LITTLE_ENDIAN__
31325	31346	#define vst2q_u16(__p0, __p1) __extension__ ({ \
31326	31347	uint16x8x2_t __s1 = __p1; \
31327		__builtin_neon_vst2q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 49); \
	31348	__builtin_neon_vst2q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 49); \
31328	31349	})
31329	31350	#else
31330	31351	#define vst2q_u16(__p0, __p1) __extension__ ({ \
31331	31352	uint16x8x2_t __s1 = __p1; \
31332	31353	uint16x8x2_t __rev1; \
31333		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
31334		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
31335		__builtin_neon_vst2q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 49); \
	31354	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	31355	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	31356	__builtin_neon_vst2q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 49); \
31336	31357	})
31337	31358	#endif
31338	31359
31339	31360	#ifdef __LITTLE_ENDIAN__
31340	31361	#define vst2q_s8(__p0, __p1) __extension__ ({ \
31341	31362	int8x16x2_t __s1 = __p1; \
31342		__builtin_neon_vst2q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 32); \
	31363	__builtin_neon_vst2q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 32); \
31343	31364	})
31344	31365	#else
31345	31366	#define vst2q_s8(__p0, __p1) __extension__ ({ \
31346	31367	int8x16x2_t __s1 = __p1; \
31347	31368	int8x16x2_t __rev1; \
31348		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
31349		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
31350		__builtin_neon_vst2q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 32); \
	31369	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	31370	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	31371	__builtin_neon_vst2q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 32); \
31351	31372	})
31352	31373	#endif
31353	31374
31354	31375	#ifdef __LITTLE_ENDIAN__
31355	31376	#define vst2q_f32(__p0, __p1) __extension__ ({ \
31356	31377	float32x4x2_t __s1 = __p1; \
31357		__builtin_neon_vst2q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 41); \
	31378	__builtin_neon_vst2q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 41); \
31358	31379	})
31359	31380	#else
31360	31381	#define vst2q_f32(__p0, __p1) __extension__ ({ \
31361	31382	float32x4x2_t __s1 = __p1; \
31362	31383	float32x4x2_t __rev1; \
31363		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
31364		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
31365		__builtin_neon_vst2q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 41); \
	31384	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	31385	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	31386	__builtin_neon_vst2q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 41); \
31366	31387	})
31367	31388	#endif
31368	31389
31369	31390	#ifdef __LITTLE_ENDIAN__
31370	31391	#define vst2q_s32(__p0, __p1) __extension__ ({ \
31371	31392	int32x4x2_t __s1 = __p1; \
31372		__builtin_neon_vst2q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 34); \
	31393	__builtin_neon_vst2q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 34); \
31373	31394	})
31374	31395	#else
31375	31396	#define vst2q_s32(__p0, __p1) __extension__ ({ \
31376	31397	int32x4x2_t __s1 = __p1; \
31377	31398	int32x4x2_t __rev1; \
31378		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
31379		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
31380		__builtin_neon_vst2q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 34); \
	31399	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	31400	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	31401	__builtin_neon_vst2q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 34); \
31381	31402	})
31382	31403	#endif
31383	31404
31384	31405	#ifdef __LITTLE_ENDIAN__
31385	31406	#define vst2q_s16(__p0, __p1) __extension__ ({ \
31386	31407	int16x8x2_t __s1 = __p1; \
31387		__builtin_neon_vst2q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 33); \
	31408	__builtin_neon_vst2q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 33); \
31388	31409	})
31389	31410	#else
31390	31411	#define vst2q_s16(__p0, __p1) __extension__ ({ \
31391	31412	int16x8x2_t __s1 = __p1; \
31392	31413	int16x8x2_t __rev1; \
31393		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
31394		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
31395		__builtin_neon_vst2q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 33); \
	31414	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	31415	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	31416	__builtin_neon_vst2q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 33); \
31396	31417	})
31397	31418	#endif
31398	31419
31399	31420	#ifdef __LITTLE_ENDIAN__
31400	31421	#define vst2_u8(__p0, __p1) __extension__ ({ \
31401	31422	uint8x8x2_t __s1 = __p1; \
31402		__builtin_neon_vst2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 16); \
	31423	__builtin_neon_vst2_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 16); \
31403	31424	})
31404	31425	#else
31405	31426	#define vst2_u8(__p0, __p1) __extension__ ({ \
31406	31427	uint8x8x2_t __s1 = __p1; \
31407	31428	uint8x8x2_t __rev1; \
31408		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
31409		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
31410		__builtin_neon_vst2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 16); \
	31429	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	31430	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	31431	__builtin_neon_vst2_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), 16); \
31411	31432	})
31412	31433	#endif
31413	31434
31414	31435	#ifdef __LITTLE_ENDIAN__
31415	31436	#define vst2_u32(__p0, __p1) __extension__ ({ \
31416	31437	uint32x2x2_t __s1 = __p1; \
31417		__builtin_neon_vst2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 18); \
	31438	__builtin_neon_vst2_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 18); \
31418	31439	})
31419	31440	#else
31420	31441	#define vst2_u32(__p0, __p1) __extension__ ({ \
31421	31442	uint32x2x2_t __s1 = __p1; \
31422	31443	uint32x2x2_t __rev1; \
31423		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
31424		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
31425		__builtin_neon_vst2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 18); \
	31444	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	31445	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	31446	__builtin_neon_vst2_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), 18); \
31426	31447	})
31427	31448	#endif
31428	31449
31429	31450	#define vst2_u64(__p0, __p1) __extension__ ({ \
31430	31451	uint64x1x2_t __s1 = __p1; \
31431		__builtin_neon_vst2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 19); \
	31452	__builtin_neon_vst2_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 19); \
31432	31453	})
31433	31454	#ifdef __LITTLE_ENDIAN__
31434	31455	#define vst2_u16(__p0, __p1) __extension__ ({ \
31435	31456	uint16x4x2_t __s1 = __p1; \
31436		__builtin_neon_vst2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 17); \
	31457	__builtin_neon_vst2_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 17); \
31437	31458	})
31438	31459	#else
31439	31460	#define vst2_u16(__p0, __p1) __extension__ ({ \
31440	31461	uint16x4x2_t __s1 = __p1; \
31441	31462	uint16x4x2_t __rev1; \
31442		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
31443		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
31444		__builtin_neon_vst2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 17); \
	31463	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	31464	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	31465	__builtin_neon_vst2_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), 17); \
31445	31466	})
31446	31467	#endif
31447	31468
31448	31469	#ifdef __LITTLE_ENDIAN__
31449	31470	#define vst2_s8(__p0, __p1) __extension__ ({ \
31450	31471	int8x8x2_t __s1 = __p1; \
31451		__builtin_neon_vst2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 0); \
	31472	__builtin_neon_vst2_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 0); \
31452	31473	})
31453	31474	#else
31454	31475	#define vst2_s8(__p0, __p1) __extension__ ({ \
31455	31476	int8x8x2_t __s1 = __p1; \
31456	31477	int8x8x2_t __rev1; \
31457		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
31458		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
31459		__builtin_neon_vst2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 0); \
	31478	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	31479	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	31480	__builtin_neon_vst2_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), 0); \
31460	31481	})
31461	31482	#endif
31462	31483
31463	31484	#ifdef __LITTLE_ENDIAN__
31464	31485	#define vst2_f32(__p0, __p1) __extension__ ({ \
31465	31486	float32x2x2_t __s1 = __p1; \
31466		__builtin_neon_vst2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 9); \
	31487	__builtin_neon_vst2_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 9); \
31467	31488	})
31468	31489	#else
31469	31490	#define vst2_f32(__p0, __p1) __extension__ ({ \
31470	31491	float32x2x2_t __s1 = __p1; \
31471	31492	float32x2x2_t __rev1; \
31472		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
31473		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
31474		__builtin_neon_vst2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 9); \
	31493	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	31494	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	31495	__builtin_neon_vst2_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), 9); \
31475	31496	})
31476	31497	#endif
31477	31498
31478	31499	#ifdef __LITTLE_ENDIAN__
31479	31500	#define vst2_s32(__p0, __p1) __extension__ ({ \
31480	31501	int32x2x2_t __s1 = __p1; \
31481		__builtin_neon_vst2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 2); \
	31502	__builtin_neon_vst2_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 2); \
31482	31503	})
31483	31504	#else
31484	31505	#define vst2_s32(__p0, __p1) __extension__ ({ \
31485	31506	int32x2x2_t __s1 = __p1; \
31486	31507	int32x2x2_t __rev1; \
31487		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
31488		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
31489		__builtin_neon_vst2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 2); \
	31508	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	31509	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	31510	__builtin_neon_vst2_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), 2); \
31490	31511	})
31491	31512	#endif
31492	31513
31493	31514	#define vst2_s64(__p0, __p1) __extension__ ({ \
31494	31515	int64x1x2_t __s1 = __p1; \
31495		__builtin_neon_vst2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 3); \
	31516	__builtin_neon_vst2_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 3); \
31496	31517	})
31497	31518	#ifdef __LITTLE_ENDIAN__
31498	31519	#define vst2_s16(__p0, __p1) __extension__ ({ \
31499	31520	int16x4x2_t __s1 = __p1; \
31500		__builtin_neon_vst2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 1); \
	31521	__builtin_neon_vst2_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 1); \
31501	31522	})
31502	31523	#else
31503	31524	#define vst2_s16(__p0, __p1) __extension__ ({ \
31504	31525	int16x4x2_t __s1 = __p1; \
31505	31526	int16x4x2_t __rev1; \
31506		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
31507		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
31508		__builtin_neon_vst2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 1); \
	31527	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	31528	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	31529	__builtin_neon_vst2_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), 1); \
31509	31530	})
31510	31531	#endif
31511	31532
31512	31533	#ifdef __LITTLE_ENDIAN__
31513	31534	#define vst2_lane_p8(__p0, __p1, __p2) __extension__ ({ \
31514	31535	poly8x8x2_t __s1 = __p1; \
31515		__builtin_neon_vst2_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 4); \
	31536	__builtin_neon_vst2_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 4); \
31516	31537	})
31517	31538	#else
31518	31539	#define vst2_lane_p8(__p0, __p1, __p2) __extension__ ({ \
31519	31540	poly8x8x2_t __s1 = __p1; \
31520	31541	poly8x8x2_t __rev1; \
31521		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
31522		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
31523		__builtin_neon_vst2_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 4); \
	31542	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	31543	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	31544	__builtin_neon_vst2_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __p2, 4); \
31524	31545	})
31525	31546	#endif
31526	31547
31527	31548	#ifdef __LITTLE_ENDIAN__
31528	31549	#define vst2_lane_p16(__p0, __p1, __p2) __extension__ ({ \
31529	31550	poly16x4x2_t __s1 = __p1; \
31530		__builtin_neon_vst2_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 5); \
	31551	__builtin_neon_vst2_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 5); \
31531	31552	})
31532	31553	#else
31533	31554	#define vst2_lane_p16(__p0, __p1, __p2) __extension__ ({ \
31534	31555	poly16x4x2_t __s1 = __p1; \
31535	31556	poly16x4x2_t __rev1; \
31536		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
31537		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
31538		__builtin_neon_vst2_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 5); \
	31557	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	31558	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	31559	__builtin_neon_vst2_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __p2, 5); \
31539	31560	})
31540	31561	#endif
31541	31562
31542	31563	#ifdef __LITTLE_ENDIAN__
31543	31564	#define vst2q_lane_p16(__p0, __p1, __p2) __extension__ ({ \
31544	31565	poly16x8x2_t __s1 = __p1; \
31545		__builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 37); \
	31566	__builtin_neon_vst2q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 37); \
31546	31567	})
31547	31568	#else
31548	31569	#define vst2q_lane_p16(__p0, __p1, __p2) __extension__ ({ \
31549	31570	poly16x8x2_t __s1 = __p1; \
31550	31571	poly16x8x2_t __rev1; \
31551		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
31552		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
31553		__builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 37); \
	31572	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	31573	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	31574	__builtin_neon_vst2q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 37); \
31554	31575	})
31555	31576	#endif
31556	31577
31557	31578	#ifdef __LITTLE_ENDIAN__
31558	31579	#define vst2q_lane_u32(__p0, __p1, __p2) __extension__ ({ \
31559	31580	uint32x4x2_t __s1 = __p1; \
31560		__builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 50); \
	31581	__builtin_neon_vst2q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 50); \
31561	31582	})
31562	31583	#else
31563	31584	#define vst2q_lane_u32(__p0, __p1, __p2) __extension__ ({ \
31564	31585	uint32x4x2_t __s1 = __p1; \
31565	31586	uint32x4x2_t __rev1; \
31566		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
31567		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
31568		__builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 50); \
	31587	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	31588	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	31589	__builtin_neon_vst2q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 50); \
31569	31590	})
31570	31591	#endif
31571	31592
31572	31593	#ifdef __LITTLE_ENDIAN__
31573	31594	#define vst2q_lane_u16(__p0, __p1, __p2) __extension__ ({ \
31574	31595	uint16x8x2_t __s1 = __p1; \
31575		__builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 49); \
	31596	__builtin_neon_vst2q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 49); \
31576	31597	})
31577	31598	#else
31578	31599	#define vst2q_lane_u16(__p0, __p1, __p2) __extension__ ({ \
31579	31600	uint16x8x2_t __s1 = __p1; \
31580	31601	uint16x8x2_t __rev1; \
31581		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
31582		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
31583		__builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 49); \
	31602	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	31603	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	31604	__builtin_neon_vst2q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 49); \
31584	31605	})
31585	31606	#endif
31586	31607
31587	31608	#ifdef __LITTLE_ENDIAN__
31588	31609	#define vst2q_lane_f32(__p0, __p1, __p2) __extension__ ({ \
31589	31610	float32x4x2_t __s1 = __p1; \
31590		__builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 41); \
	31611	__builtin_neon_vst2q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 41); \
31591	31612	})
31592	31613	#else
31593	31614	#define vst2q_lane_f32(__p0, __p1, __p2) __extension__ ({ \
31594	31615	float32x4x2_t __s1 = __p1; \
31595	31616	float32x4x2_t __rev1; \
31596		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
31597		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
31598		__builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 41); \
	31617	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	31618	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	31619	__builtin_neon_vst2q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 41); \
31599	31620	})
31600	31621	#endif
31601	31622
31602	31623	#ifdef __LITTLE_ENDIAN__
31603	31624	#define vst2q_lane_s32(__p0, __p1, __p2) __extension__ ({ \
31604	31625	int32x4x2_t __s1 = __p1; \
31605		__builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 34); \
	31626	__builtin_neon_vst2q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 34); \
31606	31627	})
31607	31628	#else
31608	31629	#define vst2q_lane_s32(__p0, __p1, __p2) __extension__ ({ \
31609	31630	int32x4x2_t __s1 = __p1; \
31610	31631	int32x4x2_t __rev1; \
31611		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
31612		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
31613		__builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 34); \
	31632	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	31633	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	31634	__builtin_neon_vst2q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 34); \
31614	31635	})
31615	31636	#endif
31616	31637
31617	31638	#ifdef __LITTLE_ENDIAN__
31618	31639	#define vst2q_lane_s16(__p0, __p1, __p2) __extension__ ({ \
31619	31640	int16x8x2_t __s1 = __p1; \
31620		__builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 33); \
	31641	__builtin_neon_vst2q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 33); \
31621	31642	})
31622	31643	#else
31623	31644	#define vst2q_lane_s16(__p0, __p1, __p2) __extension__ ({ \
31624	31645	int16x8x2_t __s1 = __p1; \
31625	31646	int16x8x2_t __rev1; \
31626		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
31627		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
31628		__builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 33); \
	31647	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	31648	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	31649	__builtin_neon_vst2q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 33); \
31629	31650	})
31630	31651	#endif
31631	31652
31632	31653	#ifdef __LITTLE_ENDIAN__
31633	31654	#define vst2_lane_u8(__p0, __p1, __p2) __extension__ ({ \
31634	31655	uint8x8x2_t __s1 = __p1; \
31635		__builtin_neon_vst2_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 16); \
	31656	__builtin_neon_vst2_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 16); \
31636	31657	})
31637	31658	#else
31638	31659	#define vst2_lane_u8(__p0, __p1, __p2) __extension__ ({ \
31639	31660	uint8x8x2_t __s1 = __p1; \
31640	31661	uint8x8x2_t __rev1; \
31641		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
31642		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
31643		__builtin_neon_vst2_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 16); \
	31662	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	31663	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	31664	__builtin_neon_vst2_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __p2, 16); \
31644	31665	})
31645	31666	#endif
31646	31667
31647	31668	#ifdef __LITTLE_ENDIAN__
31648	31669	#define vst2_lane_u32(__p0, __p1, __p2) __extension__ ({ \
31649	31670	uint32x2x2_t __s1 = __p1; \
31650		__builtin_neon_vst2_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 18); \
	31671	__builtin_neon_vst2_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 18); \
31651	31672	})
31652	31673	#else
31653	31674	#define vst2_lane_u32(__p0, __p1, __p2) __extension__ ({ \
31654	31675	uint32x2x2_t __s1 = __p1; \
31655	31676	uint32x2x2_t __rev1; \
31656		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
31657		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
31658		__builtin_neon_vst2_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 18); \
	31677	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	31678	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	31679	__builtin_neon_vst2_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __p2, 18); \
31659	31680	})
31660	31681	#endif
31661	31682
31662	31683	#ifdef __LITTLE_ENDIAN__
31663	31684	#define vst2_lane_u16(__p0, __p1, __p2) __extension__ ({ \
31664	31685	uint16x4x2_t __s1 = __p1; \
31665		__builtin_neon_vst2_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 17); \
	31686	__builtin_neon_vst2_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 17); \
31666	31687	})
31667	31688	#else
31668	31689	#define vst2_lane_u16(__p0, __p1, __p2) __extension__ ({ \
31669	31690	uint16x4x2_t __s1 = __p1; \
31670	31691	uint16x4x2_t __rev1; \
31671		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
31672		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
31673		__builtin_neon_vst2_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 17); \
	31692	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	31693	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	31694	__builtin_neon_vst2_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __p2, 17); \
31674	31695	})
31675	31696	#endif
31676	31697
31677	31698	#ifdef __LITTLE_ENDIAN__
31678	31699	#define vst2_lane_s8(__p0, __p1, __p2) __extension__ ({ \
31679	31700	int8x8x2_t __s1 = __p1; \
31680		__builtin_neon_vst2_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 0); \
	31701	__builtin_neon_vst2_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 0); \
31681	31702	})
31682	31703	#else
31683	31704	#define vst2_lane_s8(__p0, __p1, __p2) __extension__ ({ \
31684	31705	int8x8x2_t __s1 = __p1; \
31685	31706	int8x8x2_t __rev1; \
31686		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
31687		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
31688		__builtin_neon_vst2_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 0); \
	31707	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	31708	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	31709	__builtin_neon_vst2_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __p2, 0); \
31689	31710	})
31690	31711	#endif
31691	31712
31692	31713	#ifdef __LITTLE_ENDIAN__
31693	31714	#define vst2_lane_f32(__p0, __p1, __p2) __extension__ ({ \
31694	31715	float32x2x2_t __s1 = __p1; \
31695		__builtin_neon_vst2_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 9); \
	31716	__builtin_neon_vst2_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 9); \
31696	31717	})
31697	31718	#else
31698	31719	#define vst2_lane_f32(__p0, __p1, __p2) __extension__ ({ \
31699	31720	float32x2x2_t __s1 = __p1; \
31700	31721	float32x2x2_t __rev1; \
31701		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
31702		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
31703		__builtin_neon_vst2_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 9); \
	31722	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	31723	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	31724	__builtin_neon_vst2_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __p2, 9); \
31704	31725	})
31705	31726	#endif
31706	31727
31707	31728	#ifdef __LITTLE_ENDIAN__
31708	31729	#define vst2_lane_s32(__p0, __p1, __p2) __extension__ ({ \
31709	31730	int32x2x2_t __s1 = __p1; \
31710		__builtin_neon_vst2_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 2); \
	31731	__builtin_neon_vst2_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 2); \
31711	31732	})
31712	31733	#else
31713	31734	#define vst2_lane_s32(__p0, __p1, __p2) __extension__ ({ \
31714	31735	int32x2x2_t __s1 = __p1; \
31715	31736	int32x2x2_t __rev1; \
31716		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
31717		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
31718		__builtin_neon_vst2_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 2); \
	31737	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	31738	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	31739	__builtin_neon_vst2_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __p2, 2); \
31719	31740	})
31720	31741	#endif
31721	31742
31722	31743	#ifdef __LITTLE_ENDIAN__
31723	31744	#define vst2_lane_s16(__p0, __p1, __p2) __extension__ ({ \
31724	31745	int16x4x2_t __s1 = __p1; \
31725		__builtin_neon_vst2_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 1); \
	31746	__builtin_neon_vst2_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 1); \
31726	31747	})
31727	31748	#else
31728	31749	#define vst2_lane_s16(__p0, __p1, __p2) __extension__ ({ \
31729	31750	int16x4x2_t __s1 = __p1; \
31730	31751	int16x4x2_t __rev1; \
31731		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
31732		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
31733		__builtin_neon_vst2_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 1); \
	31752	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	31753	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	31754	__builtin_neon_vst2_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __p2, 1); \
31734	31755	})
31735	31756	#endif
31736	31757
31737	31758	#ifdef __LITTLE_ENDIAN__
31738	31759	#define vst3_p8(__p0, __p1) __extension__ ({ \
31739	31760	poly8x8x3_t __s1 = __p1; \
31740		__builtin_neon_vst3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 4); \
	31761	__builtin_neon_vst3_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 4); \
31741	31762	})
31742	31763	#else
31743	31764	#define vst3_p8(__p0, __p1) __extension__ ({ \
31744	31765	poly8x8x3_t __s1 = __p1; \
31745	31766	poly8x8x3_t __rev1; \
31746		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
31747		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
31748		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
31749		__builtin_neon_vst3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 4); \
	31767	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	31768	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	31769	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	31770	__builtin_neon_vst3_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), 4); \
31750	31771	})
31751	31772	#endif
31752	31773
31753	31774	#ifdef __LITTLE_ENDIAN__
31754	31775	#define vst3_p16(__p0, __p1) __extension__ ({ \
31755	31776	poly16x4x3_t __s1 = __p1; \
31756		__builtin_neon_vst3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 5); \
	31777	__builtin_neon_vst3_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 5); \
31757	31778	})
31758	31779	#else
31759	31780	#define vst3_p16(__p0, __p1) __extension__ ({ \
31760	31781	poly16x4x3_t __s1 = __p1; \
31761	31782	poly16x4x3_t __rev1; \
31762		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
31763		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
31764		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
31765		__builtin_neon_vst3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 5); \
	31783	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	31784	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	31785	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	31786	__builtin_neon_vst3_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), 5); \
31766	31787	})
31767	31788	#endif
31768	31789
31769	31790	#ifdef __LITTLE_ENDIAN__
31770	31791	#define vst3q_p8(__p0, __p1) __extension__ ({ \
31771	31792	poly8x16x3_t __s1 = __p1; \
31772		__builtin_neon_vst3q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 36); \
	31793	__builtin_neon_vst3q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 36); \
31773	31794	})
31774	31795	#else
31775	31796	#define vst3q_p8(__p0, __p1) __extension__ ({ \
31776	31797	poly8x16x3_t __s1 = __p1; \
31777	31798	poly8x16x3_t __rev1; \
31778		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
31779		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
31780		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
31781		__builtin_neon_vst3q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 36); \
	31799	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	31800	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	31801	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	31802	__builtin_neon_vst3q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 36); \
31782	31803	})
31783	31804	#endif
31784	31805
31785	31806	#ifdef __LITTLE_ENDIAN__
31786	31807	#define vst3q_p16(__p0, __p1) __extension__ ({ \
31787	31808	poly16x8x3_t __s1 = __p1; \
31788		__builtin_neon_vst3q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 37); \
	31809	__builtin_neon_vst3q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 37); \
31789	31810	})
31790	31811	#else
31791	31812	#define vst3q_p16(__p0, __p1) __extension__ ({ \
31792	31813	poly16x8x3_t __s1 = __p1; \
31793	31814	poly16x8x3_t __rev1; \
31794		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
31795		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
31796		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
31797		__builtin_neon_vst3q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 37); \
	31815	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	31816	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	31817	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	31818	__builtin_neon_vst3q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 37); \
31798	31819	})
31799	31820	#endif
31800	31821
31801	31822	#ifdef __LITTLE_ENDIAN__
31802	31823	#define vst3q_u8(__p0, __p1) __extension__ ({ \
31803	31824	uint8x16x3_t __s1 = __p1; \
31804		__builtin_neon_vst3q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 48); \
	31825	__builtin_neon_vst3q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 48); \
31805	31826	})
31806	31827	#else
31807	31828	#define vst3q_u8(__p0, __p1) __extension__ ({ \
31808	31829	uint8x16x3_t __s1 = __p1; \
31809	31830	uint8x16x3_t __rev1; \
31810		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
31811		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
31812		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
31813		__builtin_neon_vst3q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 48); \
	31831	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	31832	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	31833	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	31834	__builtin_neon_vst3q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 48); \
31814	31835	})
31815	31836	#endif
31816	31837
31817	31838	#ifdef __LITTLE_ENDIAN__
31818	31839	#define vst3q_u32(__p0, __p1) __extension__ ({ \
31819	31840	uint32x4x3_t __s1 = __p1; \
31820		__builtin_neon_vst3q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 50); \
	31841	__builtin_neon_vst3q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 50); \
31821	31842	})
31822	31843	#else
31823	31844	#define vst3q_u32(__p0, __p1) __extension__ ({ \
31824	31845	uint32x4x3_t __s1 = __p1; \
31825	31846	uint32x4x3_t __rev1; \
31826		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
31827		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
31828		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
31829		__builtin_neon_vst3q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 50); \
	31847	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	31848	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	31849	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_32); \
	31850	__builtin_neon_vst3q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 50); \
31830	31851	})
31831	31852	#endif
31832	31853
31833	31854	#ifdef __LITTLE_ENDIAN__
31834	31855	#define vst3q_u16(__p0, __p1) __extension__ ({ \
31835	31856	uint16x8x3_t __s1 = __p1; \
31836		__builtin_neon_vst3q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 49); \
	31857	__builtin_neon_vst3q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 49); \
31837	31858	})
31838	31859	#else
31839	31860	#define vst3q_u16(__p0, __p1) __extension__ ({ \
31840	31861	uint16x8x3_t __s1 = __p1; \
31841	31862	uint16x8x3_t __rev1; \
31842		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
31843		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
31844		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
31845		__builtin_neon_vst3q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 49); \
	31863	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	31864	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	31865	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	31866	__builtin_neon_vst3q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 49); \
31846	31867	})
31847	31868	#endif
31848	31869
31849	31870	#ifdef __LITTLE_ENDIAN__
31850	31871	#define vst3q_s8(__p0, __p1) __extension__ ({ \
31851	31872	int8x16x3_t __s1 = __p1; \
31852		__builtin_neon_vst3q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 32); \
	31873	__builtin_neon_vst3q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 32); \
31853	31874	})
31854	31875	#else
31855	31876	#define vst3q_s8(__p0, __p1) __extension__ ({ \
31856	31877	int8x16x3_t __s1 = __p1; \
31857	31878	int8x16x3_t __rev1; \
31858		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
31859		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
31860		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
31861		__builtin_neon_vst3q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 32); \
	31879	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	31880	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	31881	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	31882	__builtin_neon_vst3q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 32); \
31862	31883	})
31863	31884	#endif
31864	31885
31865	31886	#ifdef __LITTLE_ENDIAN__
31866	31887	#define vst3q_f32(__p0, __p1) __extension__ ({ \
31867	31888	float32x4x3_t __s1 = __p1; \
31868		__builtin_neon_vst3q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 41); \
	31889	__builtin_neon_vst3q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 41); \
31869	31890	})
31870	31891	#else
31871	31892	#define vst3q_f32(__p0, __p1) __extension__ ({ \
31872	31893	float32x4x3_t __s1 = __p1; \
31873	31894	float32x4x3_t __rev1; \
31874		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
31875		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
31876		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
31877		__builtin_neon_vst3q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 41); \
	31895	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	31896	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	31897	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_32); \
	31898	__builtin_neon_vst3q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 41); \
31878	31899	})
31879	31900	#endif
31880	31901
31881	31902	#ifdef __LITTLE_ENDIAN__
31882	31903	#define vst3q_s32(__p0, __p1) __extension__ ({ \
31883	31904	int32x4x3_t __s1 = __p1; \
31884		__builtin_neon_vst3q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 34); \
	31905	__builtin_neon_vst3q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 34); \
31885	31906	})
31886	31907	#else
31887	31908	#define vst3q_s32(__p0, __p1) __extension__ ({ \
31888	31909	int32x4x3_t __s1 = __p1; \
31889	31910	int32x4x3_t __rev1; \
31890		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
31891		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
31892		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
31893		__builtin_neon_vst3q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 34); \
	31911	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	31912	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	31913	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_32); \
	31914	__builtin_neon_vst3q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 34); \
31894	31915	})
31895	31916	#endif
31896	31917
31897	31918	#ifdef __LITTLE_ENDIAN__
31898	31919	#define vst3q_s16(__p0, __p1) __extension__ ({ \
31899	31920	int16x8x3_t __s1 = __p1; \
31900		__builtin_neon_vst3q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 33); \
	31921	__builtin_neon_vst3q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 33); \
31901	31922	})
31902	31923	#else
31903	31924	#define vst3q_s16(__p0, __p1) __extension__ ({ \
31904	31925	int16x8x3_t __s1 = __p1; \
31905	31926	int16x8x3_t __rev1; \
31906		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
31907		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
31908		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
31909		__builtin_neon_vst3q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 33); \
	31927	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	31928	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	31929	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	31930	__builtin_neon_vst3q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 33); \
31910	31931	})
31911	31932	#endif
31912	31933
31913	31934	#ifdef __LITTLE_ENDIAN__
31914	31935	#define vst3_u8(__p0, __p1) __extension__ ({ \
31915	31936	uint8x8x3_t __s1 = __p1; \
31916		__builtin_neon_vst3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 16); \
	31937	__builtin_neon_vst3_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 16); \
31917	31938	})
31918	31939	#else
31919	31940	#define vst3_u8(__p0, __p1) __extension__ ({ \
31920	31941	uint8x8x3_t __s1 = __p1; \
31921	31942	uint8x8x3_t __rev1; \
31922		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
31923		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
31924		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
31925		__builtin_neon_vst3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 16); \
	31943	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	31944	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	31945	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	31946	__builtin_neon_vst3_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), 16); \
31926	31947	})
31927	31948	#endif
31928	31949
31929	31950	#ifdef __LITTLE_ENDIAN__
31930	31951	#define vst3_u32(__p0, __p1) __extension__ ({ \
31931	31952	uint32x2x3_t __s1 = __p1; \
31932		__builtin_neon_vst3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 18); \
	31953	__builtin_neon_vst3_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 18); \
31933	31954	})
31934	31955	#else
31935	31956	#define vst3_u32(__p0, __p1) __extension__ ({ \
31936	31957	uint32x2x3_t __s1 = __p1; \
31937	31958	uint32x2x3_t __rev1; \
31938		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
31939		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
31940		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
31941		__builtin_neon_vst3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 18); \
	31959	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	31960	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	31961	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_32); \
	31962	__builtin_neon_vst3_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), 18); \
31942	31963	})
31943	31964	#endif
31944	31965
31945	31966	#define vst3_u64(__p0, __p1) __extension__ ({ \
31946	31967	uint64x1x3_t __s1 = __p1; \
31947		__builtin_neon_vst3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 19); \
	31968	__builtin_neon_vst3_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 19); \
31948	31969	})
31949	31970	#ifdef __LITTLE_ENDIAN__
31950	31971	#define vst3_u16(__p0, __p1) __extension__ ({ \
31951	31972	uint16x4x3_t __s1 = __p1; \
31952		__builtin_neon_vst3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 17); \
	31973	__builtin_neon_vst3_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 17); \
31953	31974	})
31954	31975	#else
31955	31976	#define vst3_u16(__p0, __p1) __extension__ ({ \
31956	31977	uint16x4x3_t __s1 = __p1; \
31957	31978	uint16x4x3_t __rev1; \
31958		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
31959		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
31960		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
31961		__builtin_neon_vst3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 17); \
	31979	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	31980	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	31981	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	31982	__builtin_neon_vst3_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), 17); \
31962	31983	})
31963	31984	#endif
31964	31985
31965	31986	#ifdef __LITTLE_ENDIAN__
31966	31987	#define vst3_s8(__p0, __p1) __extension__ ({ \
31967	31988	int8x8x3_t __s1 = __p1; \
31968		__builtin_neon_vst3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 0); \
	31989	__builtin_neon_vst3_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 0); \
31969	31990	})
31970	31991	#else
31971	31992	#define vst3_s8(__p0, __p1) __extension__ ({ \
31972	31993	int8x8x3_t __s1 = __p1; \
31973	31994	int8x8x3_t __rev1; \
31974		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
31975		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
31976		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
31977		__builtin_neon_vst3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 0); \
	31995	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	31996	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	31997	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	31998	__builtin_neon_vst3_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), 0); \
31978	31999	})
31979	32000	#endif
31980	32001
31981	32002	#ifdef __LITTLE_ENDIAN__
31982	32003	#define vst3_f32(__p0, __p1) __extension__ ({ \
31983	32004	float32x2x3_t __s1 = __p1; \
31984		__builtin_neon_vst3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 9); \
	32005	__builtin_neon_vst3_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 9); \
31985	32006	})
31986	32007	#else
31987	32008	#define vst3_f32(__p0, __p1) __extension__ ({ \
31988	32009	float32x2x3_t __s1 = __p1; \
31989	32010	float32x2x3_t __rev1; \
31990		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
31991		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
31992		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
31993		__builtin_neon_vst3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 9); \
	32011	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	32012	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	32013	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_32); \
	32014	__builtin_neon_vst3_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), 9); \
31994	32015	})
31995	32016	#endif
31996	32017
31997	32018	#ifdef __LITTLE_ENDIAN__
31998	32019	#define vst3_s32(__p0, __p1) __extension__ ({ \
31999	32020	int32x2x3_t __s1 = __p1; \
32000		__builtin_neon_vst3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 2); \
	32021	__builtin_neon_vst3_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 2); \
32001	32022	})
32002	32023	#else
32003	32024	#define vst3_s32(__p0, __p1) __extension__ ({ \
32004	32025	int32x2x3_t __s1 = __p1; \
32005	32026	int32x2x3_t __rev1; \
32006		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
32007		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
32008		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
32009		__builtin_neon_vst3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 2); \
	32027	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	32028	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	32029	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_32); \
	32030	__builtin_neon_vst3_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), 2); \
32010	32031	})
32011	32032	#endif
32012	32033
32013	32034	#define vst3_s64(__p0, __p1) __extension__ ({ \
32014	32035	int64x1x3_t __s1 = __p1; \
32015		__builtin_neon_vst3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 3); \
	32036	__builtin_neon_vst3_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 3); \
32016	32037	})
32017	32038	#ifdef __LITTLE_ENDIAN__
32018	32039	#define vst3_s16(__p0, __p1) __extension__ ({ \
32019	32040	int16x4x3_t __s1 = __p1; \
32020		__builtin_neon_vst3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 1); \
	32041	__builtin_neon_vst3_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 1); \
32021	32042	})
32022	32043	#else
32023	32044	#define vst3_s16(__p0, __p1) __extension__ ({ \
32024	32045	int16x4x3_t __s1 = __p1; \
32025	32046	int16x4x3_t __rev1; \
32026		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
32027		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
32028		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
32029		__builtin_neon_vst3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 1); \
	32047	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	32048	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	32049	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	32050	__builtin_neon_vst3_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), 1); \
32030	32051	})
32031	32052	#endif
32032	32053
32033	32054	#ifdef __LITTLE_ENDIAN__
32034	32055	#define vst3_lane_p8(__p0, __p1, __p2) __extension__ ({ \
32035	32056	poly8x8x3_t __s1 = __p1; \
32036		__builtin_neon_vst3_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 4); \
	32057	__builtin_neon_vst3_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 4); \
32037	32058	})
32038	32059	#else
32039	32060	#define vst3_lane_p8(__p0, __p1, __p2) __extension__ ({ \
32040	32061	poly8x8x3_t __s1 = __p1; \
32041	32062	poly8x8x3_t __rev1; \
32042		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
32043		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
32044		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
32045		__builtin_neon_vst3_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 4); \
	32063	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	32064	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	32065	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	32066	__builtin_neon_vst3_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __p2, 4); \
32046	32067	})
32047	32068	#endif
32048	32069
32049	32070	#ifdef __LITTLE_ENDIAN__
32050	32071	#define vst3_lane_p16(__p0, __p1, __p2) __extension__ ({ \
32051	32072	poly16x4x3_t __s1 = __p1; \
32052		__builtin_neon_vst3_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 5); \
	32073	__builtin_neon_vst3_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 5); \
32053	32074	})
32054	32075	#else
32055	32076	#define vst3_lane_p16(__p0, __p1, __p2) __extension__ ({ \
32056	32077	poly16x4x3_t __s1 = __p1; \
32057	32078	poly16x4x3_t __rev1; \
32058		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
32059		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
32060		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
32061		__builtin_neon_vst3_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 5); \
	32079	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	32080	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	32081	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	32082	__builtin_neon_vst3_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __p2, 5); \
32062	32083	})
32063	32084	#endif
32064	32085
32065	32086	#ifdef __LITTLE_ENDIAN__
32066	32087	#define vst3q_lane_p16(__p0, __p1, __p2) __extension__ ({ \
32067	32088	poly16x8x3_t __s1 = __p1; \
32068		__builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 37); \
	32089	__builtin_neon_vst3q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 37); \
32069	32090	})
32070	32091	#else
32071	32092	#define vst3q_lane_p16(__p0, __p1, __p2) __extension__ ({ \
32072	32093	poly16x8x3_t __s1 = __p1; \
32073	32094	poly16x8x3_t __rev1; \
32074		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
32075		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
32076		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
32077		__builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 37); \
	32095	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	32096	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	32097	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	32098	__builtin_neon_vst3q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 37); \
32078	32099	})
32079	32100	#endif
32080	32101
32081	32102	#ifdef __LITTLE_ENDIAN__
32082	32103	#define vst3q_lane_u32(__p0, __p1, __p2) __extension__ ({ \
32083	32104	uint32x4x3_t __s1 = __p1; \
32084		__builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 50); \
	32105	__builtin_neon_vst3q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 50); \
32085	32106	})
32086	32107	#else
32087	32108	#define vst3q_lane_u32(__p0, __p1, __p2) __extension__ ({ \
32088	32109	uint32x4x3_t __s1 = __p1; \
32089	32110	uint32x4x3_t __rev1; \
32090		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
32091		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
32092		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
32093		__builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 50); \
	32111	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	32112	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	32113	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_32); \
	32114	__builtin_neon_vst3q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 50); \
32094	32115	})
32095	32116	#endif
32096	32117
32097	32118	#ifdef __LITTLE_ENDIAN__
32098	32119	#define vst3q_lane_u16(__p0, __p1, __p2) __extension__ ({ \
32099	32120	uint16x8x3_t __s1 = __p1; \
32100		__builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 49); \
	32121	__builtin_neon_vst3q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 49); \
32101	32122	})
32102	32123	#else
32103	32124	#define vst3q_lane_u16(__p0, __p1, __p2) __extension__ ({ \
32104	32125	uint16x8x3_t __s1 = __p1; \
32105	32126	uint16x8x3_t __rev1; \
32106		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
32107		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
32108		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
32109		__builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 49); \
	32127	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	32128	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	32129	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	32130	__builtin_neon_vst3q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 49); \
32110	32131	})
32111	32132	#endif
32112	32133
32113	32134	#ifdef __LITTLE_ENDIAN__
32114	32135	#define vst3q_lane_f32(__p0, __p1, __p2) __extension__ ({ \
32115	32136	float32x4x3_t __s1 = __p1; \
32116		__builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 41); \
	32137	__builtin_neon_vst3q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 41); \
32117	32138	})
32118	32139	#else
32119	32140	#define vst3q_lane_f32(__p0, __p1, __p2) __extension__ ({ \
32120	32141	float32x4x3_t __s1 = __p1; \
32121	32142	float32x4x3_t __rev1; \
32122		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
32123		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
32124		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
32125		__builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 41); \
	32143	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	32144	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	32145	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_32); \
	32146	__builtin_neon_vst3q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 41); \
32126	32147	})
32127	32148	#endif
32128	32149
32129	32150	#ifdef __LITTLE_ENDIAN__
32130	32151	#define vst3q_lane_s32(__p0, __p1, __p2) __extension__ ({ \
32131	32152	int32x4x3_t __s1 = __p1; \
32132		__builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 34); \
	32153	__builtin_neon_vst3q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 34); \
32133	32154	})
32134	32155	#else
32135	32156	#define vst3q_lane_s32(__p0, __p1, __p2) __extension__ ({ \
32136	32157	int32x4x3_t __s1 = __p1; \
32137	32158	int32x4x3_t __rev1; \
32138		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
32139		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
32140		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
32141		__builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 34); \
	32159	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	32160	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	32161	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_32); \
	32162	__builtin_neon_vst3q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 34); \
32142	32163	})
32143	32164	#endif
32144	32165
32145	32166	#ifdef __LITTLE_ENDIAN__
32146	32167	#define vst3q_lane_s16(__p0, __p1, __p2) __extension__ ({ \
32147	32168	int16x8x3_t __s1 = __p1; \
32148		__builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 33); \
	32169	__builtin_neon_vst3q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 33); \
32149	32170	})
32150	32171	#else
32151	32172	#define vst3q_lane_s16(__p0, __p1, __p2) __extension__ ({ \
32152	32173	int16x8x3_t __s1 = __p1; \
32153	32174	int16x8x3_t __rev1; \
32154		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
32155		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
32156		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
32157		__builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 33); \
	32175	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	32176	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	32177	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	32178	__builtin_neon_vst3q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 33); \
32158	32179	})
32159	32180	#endif
32160	32181
32161	32182	#ifdef __LITTLE_ENDIAN__
32162	32183	#define vst3_lane_u8(__p0, __p1, __p2) __extension__ ({ \
32163	32184	uint8x8x3_t __s1 = __p1; \
32164		__builtin_neon_vst3_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 16); \
	32185	__builtin_neon_vst3_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 16); \
32165	32186	})
32166	32187	#else
32167	32188	#define vst3_lane_u8(__p0, __p1, __p2) __extension__ ({ \
32168	32189	uint8x8x3_t __s1 = __p1; \
32169	32190	uint8x8x3_t __rev1; \
32170		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
32171		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
32172		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
32173		__builtin_neon_vst3_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 16); \
	32191	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	32192	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	32193	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	32194	__builtin_neon_vst3_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __p2, 16); \
32174	32195	})
32175	32196	#endif
32176	32197
32177	32198	#ifdef __LITTLE_ENDIAN__
32178	32199	#define vst3_lane_u32(__p0, __p1, __p2) __extension__ ({ \
32179	32200	uint32x2x3_t __s1 = __p1; \
32180		__builtin_neon_vst3_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 18); \
	32201	__builtin_neon_vst3_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 18); \
32181	32202	})
32182	32203	#else
32183	32204	#define vst3_lane_u32(__p0, __p1, __p2) __extension__ ({ \
32184	32205	uint32x2x3_t __s1 = __p1; \
32185	32206	uint32x2x3_t __rev1; \
32186		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
32187		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
32188		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
32189		__builtin_neon_vst3_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 18); \
	32207	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	32208	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	32209	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_32); \
	32210	__builtin_neon_vst3_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __p2, 18); \
32190	32211	})
32191	32212	#endif
32192	32213
32193	32214	#ifdef __LITTLE_ENDIAN__
32194	32215	#define vst3_lane_u16(__p0, __p1, __p2) __extension__ ({ \
32195	32216	uint16x4x3_t __s1 = __p1; \
32196		__builtin_neon_vst3_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 17); \
	32217	__builtin_neon_vst3_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 17); \
32197	32218	})
32198	32219	#else
32199	32220	#define vst3_lane_u16(__p0, __p1, __p2) __extension__ ({ \
32200	32221	uint16x4x3_t __s1 = __p1; \
32201	32222	uint16x4x3_t __rev1; \
32202		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
32203		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
32204		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
32205		__builtin_neon_vst3_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 17); \
	32223	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	32224	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	32225	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	32226	__builtin_neon_vst3_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __p2, 17); \
32206	32227	})
32207	32228	#endif
32208	32229
32209	32230	#ifdef __LITTLE_ENDIAN__
32210	32231	#define vst3_lane_s8(__p0, __p1, __p2) __extension__ ({ \
32211	32232	int8x8x3_t __s1 = __p1; \
32212		__builtin_neon_vst3_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 0); \
	32233	__builtin_neon_vst3_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 0); \
32213	32234	})
32214	32235	#else
32215	32236	#define vst3_lane_s8(__p0, __p1, __p2) __extension__ ({ \
32216	32237	int8x8x3_t __s1 = __p1; \
32217	32238	int8x8x3_t __rev1; \
32218		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
32219		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
32220		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
32221		__builtin_neon_vst3_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 0); \
	32239	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	32240	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	32241	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	32242	__builtin_neon_vst3_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __p2, 0); \
32222	32243	})
32223	32244	#endif
32224	32245
32225	32246	#ifdef __LITTLE_ENDIAN__
32226	32247	#define vst3_lane_f32(__p0, __p1, __p2) __extension__ ({ \
32227	32248	float32x2x3_t __s1 = __p1; \
32228		__builtin_neon_vst3_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 9); \
	32249	__builtin_neon_vst3_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 9); \
32229	32250	})
32230	32251	#else
32231	32252	#define vst3_lane_f32(__p0, __p1, __p2) __extension__ ({ \
32232	32253	float32x2x3_t __s1 = __p1; \
32233	32254	float32x2x3_t __rev1; \
32234		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
32235		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
32236		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
32237		__builtin_neon_vst3_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 9); \
	32255	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	32256	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	32257	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_32); \
	32258	__builtin_neon_vst3_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __p2, 9); \
32238	32259	})
32239	32260	#endif
32240	32261
32241	32262	#ifdef __LITTLE_ENDIAN__
32242	32263	#define vst3_lane_s32(__p0, __p1, __p2) __extension__ ({ \
32243	32264	int32x2x3_t __s1 = __p1; \
32244		__builtin_neon_vst3_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 2); \
	32265	__builtin_neon_vst3_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 2); \
32245	32266	})
32246	32267	#else
32247	32268	#define vst3_lane_s32(__p0, __p1, __p2) __extension__ ({ \
32248	32269	int32x2x3_t __s1 = __p1; \
32249	32270	int32x2x3_t __rev1; \
32250		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
32251		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
32252		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
32253		__builtin_neon_vst3_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 2); \
	32271	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	32272	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	32273	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_32); \
	32274	__builtin_neon_vst3_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __p2, 2); \
32254	32275	})
32255	32276	#endif
32256	32277
32257	32278	#ifdef __LITTLE_ENDIAN__
32258	32279	#define vst3_lane_s16(__p0, __p1, __p2) __extension__ ({ \
32259	32280	int16x4x3_t __s1 = __p1; \
32260		__builtin_neon_vst3_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 1); \
	32281	__builtin_neon_vst3_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 1); \
32261	32282	})
32262	32283	#else
32263	32284	#define vst3_lane_s16(__p0, __p1, __p2) __extension__ ({ \
32264	32285	int16x4x3_t __s1 = __p1; \
32265	32286	int16x4x3_t __rev1; \
32266		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
32267		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
32268		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
32269		__builtin_neon_vst3_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 1); \
	32287	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	32288	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	32289	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	32290	__builtin_neon_vst3_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __p2, 1); \
32270	32291	})
32271	32292	#endif
32272	32293
32273	32294	#ifdef __LITTLE_ENDIAN__
32274	32295	#define vst4_p8(__p0, __p1) __extension__ ({ \
32275	32296	poly8x8x4_t __s1 = __p1; \
32276		__builtin_neon_vst4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 4); \
	32297	__builtin_neon_vst4_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 4); \
32277	32298	})
32278	32299	#else
32279	32300	#define vst4_p8(__p0, __p1) __extension__ ({ \
32280	32301	poly8x8x4_t __s1 = __p1; \
32281	32302	poly8x8x4_t __rev1; \
32282		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
32283		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
32284		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
32285		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
32286		__builtin_neon_vst4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 4); \
	32303	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	32304	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	32305	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	32306	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_8); \
	32307	__builtin_neon_vst4_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), 4); \
32287	32308	})
32288	32309	#endif
32289	32310
32290	32311	#ifdef __LITTLE_ENDIAN__
32291	32312	#define vst4_p16(__p0, __p1) __extension__ ({ \
32292	32313	poly16x4x4_t __s1 = __p1; \
32293		__builtin_neon_vst4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 5); \
	32314	__builtin_neon_vst4_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 5); \
32294	32315	})
32295	32316	#else
32296	32317	#define vst4_p16(__p0, __p1) __extension__ ({ \
32297	32318	poly16x4x4_t __s1 = __p1; \
32298	32319	poly16x4x4_t __rev1; \
32299		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
32300		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
32301		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
32302		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
32303		__builtin_neon_vst4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 5); \
	32320	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	32321	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	32322	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	32323	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_16); \
	32324	__builtin_neon_vst4_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), 5); \
32304	32325	})
32305	32326	#endif
32306	32327
32307	32328	#ifdef __LITTLE_ENDIAN__
32308	32329	#define vst4q_p8(__p0, __p1) __extension__ ({ \
32309	32330	poly8x16x4_t __s1 = __p1; \
32310		__builtin_neon_vst4q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 36); \
	32331	__builtin_neon_vst4q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 36); \
32311	32332	})
32312	32333	#else
32313	32334	#define vst4q_p8(__p0, __p1) __extension__ ({ \
32314	32335	poly8x16x4_t __s1 = __p1; \
32315	32336	poly8x16x4_t __rev1; \
32316		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
32317		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
32318		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
32319		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
32320		__builtin_neon_vst4q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 36); \
	32337	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	32338	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	32339	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	32340	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_8); \
	32341	__builtin_neon_vst4q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 36); \
32321	32342	})
32322	32343	#endif
32323	32344
32324	32345	#ifdef __LITTLE_ENDIAN__
32325	32346	#define vst4q_p16(__p0, __p1) __extension__ ({ \
32326	32347	poly16x8x4_t __s1 = __p1; \
32327		__builtin_neon_vst4q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 37); \
	32348	__builtin_neon_vst4q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 37); \
32328	32349	})
32329	32350	#else
32330	32351	#define vst4q_p16(__p0, __p1) __extension__ ({ \
32331	32352	poly16x8x4_t __s1 = __p1; \
32332	32353	poly16x8x4_t __rev1; \
32333		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
32334		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
32335		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
32336		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
32337		__builtin_neon_vst4q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 37); \
	32354	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	32355	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	32356	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	32357	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_16); \
	32358	__builtin_neon_vst4q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 37); \
32338	32359	})
32339	32360	#endif
32340	32361
32341	32362	#ifdef __LITTLE_ENDIAN__
32342	32363	#define vst4q_u8(__p0, __p1) __extension__ ({ \
32343	32364	uint8x16x4_t __s1 = __p1; \
32344		__builtin_neon_vst4q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 48); \
	32365	__builtin_neon_vst4q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 48); \
32345	32366	})
32346	32367	#else
32347	32368	#define vst4q_u8(__p0, __p1) __extension__ ({ \
32348	32369	uint8x16x4_t __s1 = __p1; \
32349	32370	uint8x16x4_t __rev1; \
32350		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
32351		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
32352		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
32353		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
32354		__builtin_neon_vst4q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 48); \
	32371	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	32372	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	32373	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	32374	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_8); \
	32375	__builtin_neon_vst4q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 48); \
32355	32376	})
32356	32377	#endif
32357	32378
32358	32379	#ifdef __LITTLE_ENDIAN__
32359	32380	#define vst4q_u32(__p0, __p1) __extension__ ({ \
32360	32381	uint32x4x4_t __s1 = __p1; \
32361		__builtin_neon_vst4q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 50); \
	32382	__builtin_neon_vst4q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 50); \
32362	32383	})
32363	32384	#else
32364	32385	#define vst4q_u32(__p0, __p1) __extension__ ({ \
32365	32386	uint32x4x4_t __s1 = __p1; \
32366	32387	uint32x4x4_t __rev1; \
32367		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
32368		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
32369		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
32370		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
32371		__builtin_neon_vst4q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 50); \
	32388	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	32389	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	32390	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_32); \
	32391	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_32); \
	32392	__builtin_neon_vst4q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 50); \
32372	32393	})
32373	32394	#endif
32374	32395
32375	32396	#ifdef __LITTLE_ENDIAN__
32376	32397	#define vst4q_u16(__p0, __p1) __extension__ ({ \
32377	32398	uint16x8x4_t __s1 = __p1; \
32378		__builtin_neon_vst4q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 49); \
	32399	__builtin_neon_vst4q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 49); \
32379	32400	})
32380	32401	#else
32381	32402	#define vst4q_u16(__p0, __p1) __extension__ ({ \
32382	32403	uint16x8x4_t __s1 = __p1; \
32383	32404	uint16x8x4_t __rev1; \
32384		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
32385		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
32386		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
32387		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
32388		__builtin_neon_vst4q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 49); \
	32405	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	32406	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	32407	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	32408	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_16); \
	32409	__builtin_neon_vst4q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 49); \
32389	32410	})
32390	32411	#endif
32391	32412
32392	32413	#ifdef __LITTLE_ENDIAN__
32393	32414	#define vst4q_s8(__p0, __p1) __extension__ ({ \
32394	32415	int8x16x4_t __s1 = __p1; \
32395		__builtin_neon_vst4q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 32); \
	32416	__builtin_neon_vst4q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 32); \
32396	32417	})
32397	32418	#else
32398	32419	#define vst4q_s8(__p0, __p1) __extension__ ({ \
32399	32420	int8x16x4_t __s1 = __p1; \
32400	32421	int8x16x4_t __rev1; \
32401		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
32402		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
32403		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
32404		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
32405		__builtin_neon_vst4q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 32); \
	32422	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	32423	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	32424	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	32425	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_8); \
	32426	__builtin_neon_vst4q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 32); \
32406	32427	})
32407	32428	#endif
32408	32429
32409	32430	#ifdef __LITTLE_ENDIAN__
32410	32431	#define vst4q_f32(__p0, __p1) __extension__ ({ \
32411	32432	float32x4x4_t __s1 = __p1; \
32412		__builtin_neon_vst4q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 41); \
	32433	__builtin_neon_vst4q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 41); \
32413	32434	})
32414	32435	#else
32415	32436	#define vst4q_f32(__p0, __p1) __extension__ ({ \
32416	32437	float32x4x4_t __s1 = __p1; \
32417	32438	float32x4x4_t __rev1; \
32418		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
32419		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
32420		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
32421		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
32422		__builtin_neon_vst4q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 41); \
	32439	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	32440	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	32441	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_32); \
	32442	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_32); \
	32443	__builtin_neon_vst4q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 41); \
32423	32444	})
32424	32445	#endif
32425	32446
32426	32447	#ifdef __LITTLE_ENDIAN__
32427	32448	#define vst4q_s32(__p0, __p1) __extension__ ({ \
32428	32449	int32x4x4_t __s1 = __p1; \
32429		__builtin_neon_vst4q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 34); \
	32450	__builtin_neon_vst4q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 34); \
32430	32451	})
32431	32452	#else
32432	32453	#define vst4q_s32(__p0, __p1) __extension__ ({ \
32433	32454	int32x4x4_t __s1 = __p1; \
32434	32455	int32x4x4_t __rev1; \
32435		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
32436		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
32437		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
32438		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
32439		__builtin_neon_vst4q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 34); \
	32456	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	32457	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	32458	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_32); \
	32459	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_32); \
	32460	__builtin_neon_vst4q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 34); \
32440	32461	})
32441	32462	#endif
32442	32463
32443	32464	#ifdef __LITTLE_ENDIAN__
32444	32465	#define vst4q_s16(__p0, __p1) __extension__ ({ \
32445	32466	int16x8x4_t __s1 = __p1; \
32446		__builtin_neon_vst4q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 33); \
	32467	__builtin_neon_vst4q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 33); \
32447	32468	})
32448	32469	#else
32449	32470	#define vst4q_s16(__p0, __p1) __extension__ ({ \
32450	32471	int16x8x4_t __s1 = __p1; \
32451	32472	int16x8x4_t __rev1; \
32452		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
32453		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
32454		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
32455		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
32456		__builtin_neon_vst4q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 33); \
	32473	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	32474	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	32475	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	32476	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_16); \
	32477	__builtin_neon_vst4q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 33); \
32457	32478	})
32458	32479	#endif
32459	32480
32460	32481	#ifdef __LITTLE_ENDIAN__
32461	32482	#define vst4_u8(__p0, __p1) __extension__ ({ \
32462	32483	uint8x8x4_t __s1 = __p1; \
32463		__builtin_neon_vst4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 16); \
	32484	__builtin_neon_vst4_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 16); \
32464	32485	})
32465	32486	#else
32466	32487	#define vst4_u8(__p0, __p1) __extension__ ({ \
32467	32488	uint8x8x4_t __s1 = __p1; \
32468	32489	uint8x8x4_t __rev1; \
32469		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
32470		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
32471		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
32472		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
32473		__builtin_neon_vst4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 16); \
	32490	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	32491	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	32492	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	32493	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_8); \
	32494	__builtin_neon_vst4_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), 16); \
32474	32495	})
32475	32496	#endif
32476	32497
32477	32498	#ifdef __LITTLE_ENDIAN__
32478	32499	#define vst4_u32(__p0, __p1) __extension__ ({ \
32479	32500	uint32x2x4_t __s1 = __p1; \
32480		__builtin_neon_vst4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 18); \
	32501	__builtin_neon_vst4_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 18); \
32481	32502	})
32482	32503	#else
32483	32504	#define vst4_u32(__p0, __p1) __extension__ ({ \
32484	32505	uint32x2x4_t __s1 = __p1; \
32485	32506	uint32x2x4_t __rev1; \
32486		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
32487		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
32488		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
32489		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
32490		__builtin_neon_vst4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 18); \
	32507	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	32508	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	32509	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_32); \
	32510	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_32); \
	32511	__builtin_neon_vst4_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), 18); \
32491	32512	})
32492	32513	#endif
32493	32514
32494	32515	#define vst4_u64(__p0, __p1) __extension__ ({ \
32495	32516	uint64x1x4_t __s1 = __p1; \
32496		__builtin_neon_vst4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 19); \
	32517	__builtin_neon_vst4_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 19); \
32497	32518	})
32498	32519	#ifdef __LITTLE_ENDIAN__
32499	32520	#define vst4_u16(__p0, __p1) __extension__ ({ \
32500	32521	uint16x4x4_t __s1 = __p1; \
32501		__builtin_neon_vst4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 17); \
	32522	__builtin_neon_vst4_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 17); \
32502	32523	})
32503	32524	#else
32504	32525	#define vst4_u16(__p0, __p1) __extension__ ({ \
32505	32526	uint16x4x4_t __s1 = __p1; \
32506	32527	uint16x4x4_t __rev1; \
32507		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
32508		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
32509		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
32510		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
32511		__builtin_neon_vst4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 17); \
	32528	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	32529	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	32530	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	32531	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_16); \
	32532	__builtin_neon_vst4_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), 17); \
32512	32533	})
32513	32534	#endif
32514	32535
32515	32536	#ifdef __LITTLE_ENDIAN__
32516	32537	#define vst4_s8(__p0, __p1) __extension__ ({ \
32517	32538	int8x8x4_t __s1 = __p1; \
32518		__builtin_neon_vst4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 0); \
	32539	__builtin_neon_vst4_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 0); \
32519	32540	})
32520	32541	#else
32521	32542	#define vst4_s8(__p0, __p1) __extension__ ({ \
32522	32543	int8x8x4_t __s1 = __p1; \
32523	32544	int8x8x4_t __rev1; \
32524		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
32525		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
32526		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
32527		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
32528		__builtin_neon_vst4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 0); \
	32545	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	32546	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	32547	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	32548	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_8); \
	32549	__builtin_neon_vst4_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), 0); \
32529	32550	})
32530	32551	#endif
32531	32552
32532	32553	#ifdef __LITTLE_ENDIAN__
32533	32554	#define vst4_f32(__p0, __p1) __extension__ ({ \
32534	32555	float32x2x4_t __s1 = __p1; \
32535		__builtin_neon_vst4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 9); \
	32556	__builtin_neon_vst4_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 9); \
32536	32557	})
32537	32558	#else
32538	32559	#define vst4_f32(__p0, __p1) __extension__ ({ \
32539	32560	float32x2x4_t __s1 = __p1; \
32540	32561	float32x2x4_t __rev1; \
32541		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
32542		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
32543		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
32544		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
32545		__builtin_neon_vst4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 9); \
	32562	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	32563	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	32564	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_32); \
	32565	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_32); \
	32566	__builtin_neon_vst4_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), 9); \
32546	32567	})
32547	32568	#endif
32548	32569
32549	32570	#ifdef __LITTLE_ENDIAN__
32550	32571	#define vst4_s32(__p0, __p1) __extension__ ({ \
32551	32572	int32x2x4_t __s1 = __p1; \
32552		__builtin_neon_vst4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 2); \
	32573	__builtin_neon_vst4_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 2); \
32553	32574	})
32554	32575	#else
32555	32576	#define vst4_s32(__p0, __p1) __extension__ ({ \
32556	32577	int32x2x4_t __s1 = __p1; \
32557	32578	int32x2x4_t __rev1; \
32558		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
32559		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
32560		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
32561		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
32562		__builtin_neon_vst4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 2); \
	32579	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	32580	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	32581	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_32); \
	32582	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_32); \
	32583	__builtin_neon_vst4_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), 2); \
32563	32584	})
32564	32585	#endif
32565	32586
32566	32587	#define vst4_s64(__p0, __p1) __extension__ ({ \
32567	32588	int64x1x4_t __s1 = __p1; \
32568		__builtin_neon_vst4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 3); \
	32589	__builtin_neon_vst4_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 3); \
32569	32590	})
32570	32591	#ifdef __LITTLE_ENDIAN__
32571	32592	#define vst4_s16(__p0, __p1) __extension__ ({ \
32572	32593	int16x4x4_t __s1 = __p1; \
32573		__builtin_neon_vst4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 1); \
	32594	__builtin_neon_vst4_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 1); \
32574	32595	})
32575	32596	#else
32576	32597	#define vst4_s16(__p0, __p1) __extension__ ({ \
32577	32598	int16x4x4_t __s1 = __p1; \
32578	32599	int16x4x4_t __rev1; \
32579		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
32580		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
32581		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
32582		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
32583		__builtin_neon_vst4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 1); \
	32600	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	32601	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	32602	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	32603	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_16); \
	32604	__builtin_neon_vst4_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), 1); \
32584	32605	})
32585	32606	#endif
32586	32607
32587	32608	#ifdef __LITTLE_ENDIAN__
32588	32609	#define vst4_lane_p8(__p0, __p1, __p2) __extension__ ({ \
32589	32610	poly8x8x4_t __s1 = __p1; \
32590		__builtin_neon_vst4_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 4); \
	32611	__builtin_neon_vst4_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 4); \
32591	32612	})
32592	32613	#else
32593	32614	#define vst4_lane_p8(__p0, __p1, __p2) __extension__ ({ \
32594	32615	poly8x8x4_t __s1 = __p1; \
32595	32616	poly8x8x4_t __rev1; \
32596		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
32597		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
32598		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
32599		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
32600		__builtin_neon_vst4_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 4); \
	32617	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	32618	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	32619	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	32620	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_8); \
	32621	__builtin_neon_vst4_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), __p2, 4); \
32601	32622	})
32602	32623	#endif
32603	32624
32604	32625	#ifdef __LITTLE_ENDIAN__
32605	32626	#define vst4_lane_p16(__p0, __p1, __p2) __extension__ ({ \
32606	32627	poly16x4x4_t __s1 = __p1; \
32607		__builtin_neon_vst4_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 5); \
	32628	__builtin_neon_vst4_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 5); \
32608	32629	})
32609	32630	#else
32610	32631	#define vst4_lane_p16(__p0, __p1, __p2) __extension__ ({ \
32611	32632	poly16x4x4_t __s1 = __p1; \
32612	32633	poly16x4x4_t __rev1; \
32613		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
32614		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
32615		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
32616		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
32617		__builtin_neon_vst4_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 5); \
	32634	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	32635	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	32636	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	32637	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_16); \
	32638	__builtin_neon_vst4_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), __p2, 5); \
32618	32639	})
32619	32640	#endif
32620	32641
32621	32642	#ifdef __LITTLE_ENDIAN__
32622	32643	#define vst4q_lane_p16(__p0, __p1, __p2) __extension__ ({ \
32623	32644	poly16x8x4_t __s1 = __p1; \
32624		__builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 37); \
	32645	__builtin_neon_vst4q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 37); \
32625	32646	})
32626	32647	#else
32627	32648	#define vst4q_lane_p16(__p0, __p1, __p2) __extension__ ({ \
32628	32649	poly16x8x4_t __s1 = __p1; \
32629	32650	poly16x8x4_t __rev1; \
32630		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
32631		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
32632		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
32633		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
32634		__builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 37); \
	32651	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	32652	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	32653	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	32654	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_16); \
	32655	__builtin_neon_vst4q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 37); \
32635	32656	})
32636	32657	#endif
32637	32658
32638	32659	#ifdef __LITTLE_ENDIAN__
32639	32660	#define vst4q_lane_u32(__p0, __p1, __p2) __extension__ ({ \
32640	32661	uint32x4x4_t __s1 = __p1; \
32641		__builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 50); \
	32662	__builtin_neon_vst4q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 50); \
32642	32663	})
32643	32664	#else
32644	32665	#define vst4q_lane_u32(__p0, __p1, __p2) __extension__ ({ \
32645	32666	uint32x4x4_t __s1 = __p1; \
32646	32667	uint32x4x4_t __rev1; \
32647		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
32648		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
32649		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
32650		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
32651		__builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 50); \
	32668	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	32669	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	32670	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_32); \
	32671	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_32); \
	32672	__builtin_neon_vst4q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 50); \
32652	32673	})
32653	32674	#endif
32654	32675
32655	32676	#ifdef __LITTLE_ENDIAN__
32656	32677	#define vst4q_lane_u16(__p0, __p1, __p2) __extension__ ({ \
32657	32678	uint16x8x4_t __s1 = __p1; \
32658		__builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 49); \
	32679	__builtin_neon_vst4q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 49); \
32659	32680	})
32660	32681	#else
32661	32682	#define vst4q_lane_u16(__p0, __p1, __p2) __extension__ ({ \
32662	32683	uint16x8x4_t __s1 = __p1; \
32663	32684	uint16x8x4_t __rev1; \
32664		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
32665		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
32666		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
32667		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
32668		__builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 49); \
	32685	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	32686	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	32687	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	32688	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_16); \
	32689	__builtin_neon_vst4q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 49); \
32669	32690	})
32670	32691	#endif
32671	32692
32672	32693	#ifdef __LITTLE_ENDIAN__
32673	32694	#define vst4q_lane_f32(__p0, __p1, __p2) __extension__ ({ \
32674	32695	float32x4x4_t __s1 = __p1; \
32675		__builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 41); \
	32696	__builtin_neon_vst4q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 41); \
32676	32697	})
32677	32698	#else
32678	32699	#define vst4q_lane_f32(__p0, __p1, __p2) __extension__ ({ \
32679	32700	float32x4x4_t __s1 = __p1; \
32680	32701	float32x4x4_t __rev1; \
32681		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
32682		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
32683		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
32684		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
32685		__builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 41); \
	32702	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	32703	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	32704	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_32); \
	32705	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_32); \
	32706	__builtin_neon_vst4q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 41); \
32686	32707	})
32687	32708	#endif
32688	32709
32689	32710	#ifdef __LITTLE_ENDIAN__
32690	32711	#define vst4q_lane_s32(__p0, __p1, __p2) __extension__ ({ \
32691	32712	int32x4x4_t __s1 = __p1; \
32692		__builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 34); \
	32713	__builtin_neon_vst4q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 34); \
32693	32714	})
32694	32715	#else
32695	32716	#define vst4q_lane_s32(__p0, __p1, __p2) __extension__ ({ \
32696	32717	int32x4x4_t __s1 = __p1; \
32697	32718	int32x4x4_t __rev1; \
32698		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
32699		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
32700		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
32701		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
32702		__builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 34); \
	32719	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_32); \
	32720	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_32); \
	32721	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_32); \
	32722	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_32); \
	32723	__builtin_neon_vst4q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 34); \
32703	32724	})
32704	32725	#endif
32705	32726
32706	32727	#ifdef __LITTLE_ENDIAN__
32707	32728	#define vst4q_lane_s16(__p0, __p1, __p2) __extension__ ({ \
32708	32729	int16x8x4_t __s1 = __p1; \
32709		__builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 33); \
	32730	__builtin_neon_vst4q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 33); \
32710	32731	})
32711	32732	#else
32712	32733	#define vst4q_lane_s16(__p0, __p1, __p2) __extension__ ({ \
32713	32734	int16x8x4_t __s1 = __p1; \
32714	32735	int16x8x4_t __rev1; \
32715		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
32716		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
32717		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
32718		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
32719		__builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 33); \
	32736	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	32737	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	32738	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	32739	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_16); \
	32740	__builtin_neon_vst4q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 33); \
32720	32741	})
32721	32742	#endif
32722	32743
32723	32744	#ifdef __LITTLE_ENDIAN__
32724	32745	#define vst4_lane_u8(__p0, __p1, __p2) __extension__ ({ \
32725	32746	uint8x8x4_t __s1 = __p1; \
32726		__builtin_neon_vst4_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 16); \
	32747	__builtin_neon_vst4_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 16); \
32727	32748	})
32728	32749	#else
32729	32750	#define vst4_lane_u8(__p0, __p1, __p2) __extension__ ({ \
32730	32751	uint8x8x4_t __s1 = __p1; \
32731	32752	uint8x8x4_t __rev1; \
32732		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
32733		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
32734		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
32735		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
32736		__builtin_neon_vst4_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 16); \
	32753	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	32754	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	32755	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	32756	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_8); \
	32757	__builtin_neon_vst4_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), __p2, 16); \
32737	32758	})
32738	32759	#endif
32739	32760
32740	32761	#ifdef __LITTLE_ENDIAN__
32741	32762	#define vst4_lane_u32(__p0, __p1, __p2) __extension__ ({ \
32742	32763	uint32x2x4_t __s1 = __p1; \
32743		__builtin_neon_vst4_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 18); \
	32764	__builtin_neon_vst4_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 18); \
32744	32765	})
32745	32766	#else
32746	32767	#define vst4_lane_u32(__p0, __p1, __p2) __extension__ ({ \
32747	32768	uint32x2x4_t __s1 = __p1; \
32748	32769	uint32x2x4_t __rev1; \
32749		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
32750		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
32751		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
32752		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
32753		__builtin_neon_vst4_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 18); \
	32770	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	32771	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	32772	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_32); \
	32773	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_32); \
	32774	__builtin_neon_vst4_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), __p2, 18); \
32754	32775	})
32755	32776	#endif
32756	32777
32757	32778	#ifdef __LITTLE_ENDIAN__
32758	32779	#define vst4_lane_u16(__p0, __p1, __p2) __extension__ ({ \
32759	32780	uint16x4x4_t __s1 = __p1; \
32760		__builtin_neon_vst4_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 17); \
	32781	__builtin_neon_vst4_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 17); \
32761	32782	})
32762	32783	#else
32763	32784	#define vst4_lane_u16(__p0, __p1, __p2) __extension__ ({ \
32764	32785	uint16x4x4_t __s1 = __p1; \
32765	32786	uint16x4x4_t __rev1; \
32766		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
32767		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
32768		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
32769		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
32770		__builtin_neon_vst4_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 17); \
	32787	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	32788	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	32789	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	32790	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_16); \
	32791	__builtin_neon_vst4_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), __p2, 17); \
32771	32792	})
32772	32793	#endif
32773	32794
32774	32795	#ifdef __LITTLE_ENDIAN__
32775	32796	#define vst4_lane_s8(__p0, __p1, __p2) __extension__ ({ \
32776	32797	int8x8x4_t __s1 = __p1; \
32777		__builtin_neon_vst4_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 0); \
	32798	__builtin_neon_vst4_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 0); \
32778	32799	})
32779	32800	#else
32780	32801	#define vst4_lane_s8(__p0, __p1, __p2) __extension__ ({ \
32781	32802	int8x8x4_t __s1 = __p1; \
32782	32803	int8x8x4_t __rev1; \
32783		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
32784		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
32785		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
32786		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
32787		__builtin_neon_vst4_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 0); \
	32804	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	32805	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	32806	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	32807	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_8); \
	32808	__builtin_neon_vst4_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), __p2, 0); \
32788	32809	})
32789	32810	#endif
32790	32811
32791	32812	#ifdef __LITTLE_ENDIAN__
32792	32813	#define vst4_lane_f32(__p0, __p1, __p2) __extension__ ({ \
32793	32814	float32x2x4_t __s1 = __p1; \
32794		__builtin_neon_vst4_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 9); \
	32815	__builtin_neon_vst4_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 9); \
32795	32816	})
32796	32817	#else
32797	32818	#define vst4_lane_f32(__p0, __p1, __p2) __extension__ ({ \
32798	32819	float32x2x4_t __s1 = __p1; \
32799	32820	float32x2x4_t __rev1; \
32800		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
32801		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
32802		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
32803		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
32804		__builtin_neon_vst4_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 9); \
	32821	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	32822	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	32823	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_32); \
	32824	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_32); \
	32825	__builtin_neon_vst4_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), __p2, 9); \
32805	32826	})
32806	32827	#endif
32807	32828
32808	32829	#ifdef __LITTLE_ENDIAN__
32809	32830	#define vst4_lane_s32(__p0, __p1, __p2) __extension__ ({ \
32810	32831	int32x2x4_t __s1 = __p1; \
32811		__builtin_neon_vst4_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 2); \
	32832	__builtin_neon_vst4_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 2); \
32812	32833	})
32813	32834	#else
32814	32835	#define vst4_lane_s32(__p0, __p1, __p2) __extension__ ({ \
32815	32836	int32x2x4_t __s1 = __p1; \
32816	32837	int32x2x4_t __rev1; \
32817		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
32818		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
32819		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
32820		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
32821		__builtin_neon_vst4_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 2); \
	32838	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_32); \
	32839	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_32); \
	32840	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_32); \
	32841	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_32); \
	32842	__builtin_neon_vst4_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), __p2, 2); \
32822	32843	})
32823	32844	#endif
32824	32845
32825	32846	#ifdef __LITTLE_ENDIAN__
32826	32847	#define vst4_lane_s16(__p0, __p1, __p2) __extension__ ({ \
32827	32848	int16x4x4_t __s1 = __p1; \
32828		__builtin_neon_vst4_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 1); \
	32849	__builtin_neon_vst4_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 1); \
32829	32850	})
32830	32851	#else
32831	32852	#define vst4_lane_s16(__p0, __p1, __p2) __extension__ ({ \
32832	32853	int16x4x4_t __s1 = __p1; \
32833	32854	int16x4x4_t __rev1; \
32834		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
32835		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
32836		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
32837		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
32838		__builtin_neon_vst4_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 1); \
	32855	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	32856	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	32857	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	32858	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_16); \
	32859	__builtin_neon_vst4_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), __p2, 1); \
32839	32860	})
32840	32861	#endif
32841	32862
...	...	@@ -32848,10 +32869,10 @@ __ai __attribute__((target("neon"))) uint8x16_t vsubq_u8(uint8x16_t __p0, uint8x
32848	32869	#else
32849	32870	__ai __attribute__((target("neon"))) uint8x16_t vsubq_u8(uint8x16_t __p0, uint8x16_t __p1) {
32850	32871	uint8x16_t __ret;
32851		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
32852		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	32872	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	32873	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
32853	32874	__ret = __rev0 - __rev1;
32854		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	32875	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
32855	32876	return __ret;
32856	32877	}
32857	32878	#endif
...	...	@@ -32865,10 +32886,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vsubq_u32(uint32x4_t __p0, uint3
32865	32886	#else
32866	32887	__ai __attribute__((target("neon"))) uint32x4_t vsubq_u32(uint32x4_t __p0, uint32x4_t __p1) {
32867	32888	uint32x4_t __ret;
32868		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
32869		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	32889	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	32890	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
32870	32891	__ret = __rev0 - __rev1;
32871		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	32892	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
32872	32893	return __ret;
32873	32894	}
32874	32895	#endif
...	...	@@ -32882,10 +32903,10 @@ __ai __attribute__((target("neon"))) uint64x2_t vsubq_u64(uint64x2_t __p0, uint6
32882	32903	#else
32883	32904	__ai __attribute__((target("neon"))) uint64x2_t vsubq_u64(uint64x2_t __p0, uint64x2_t __p1) {
32884	32905	uint64x2_t __ret;
32885		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
32886		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	32906	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	32907	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
32887	32908	__ret = __rev0 - __rev1;
32888		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	32909	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
32889	32910	return __ret;
32890	32911	}
32891	32912	#endif
...	...	@@ -32899,10 +32920,10 @@ __ai __attribute__((target("neon"))) uint16x8_t vsubq_u16(uint16x8_t __p0, uint1
32899	32920	#else
32900	32921	__ai __attribute__((target("neon"))) uint16x8_t vsubq_u16(uint16x8_t __p0, uint16x8_t __p1) {
32901	32922	uint16x8_t __ret;
32902		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
32903		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	32923	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	32924	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
32904	32925	__ret = __rev0 - __rev1;
32905		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	32926	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
32906	32927	return __ret;
32907	32928	}
32908	32929	#endif
...	...	@@ -32916,10 +32937,10 @@ __ai __attribute__((target("neon"))) int8x16_t vsubq_s8(int8x16_t __p0, int8x16_
32916	32937	#else
32917	32938	__ai __attribute__((target("neon"))) int8x16_t vsubq_s8(int8x16_t __p0, int8x16_t __p1) {
32918	32939	int8x16_t __ret;
32919		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
32920		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	32940	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	32941	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
32921	32942	__ret = __rev0 - __rev1;
32922		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	32943	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
32923	32944	return __ret;
32924	32945	}
32925	32946	#endif
...	...	@@ -32933,10 +32954,10 @@ __ai __attribute__((target("neon"))) float32x4_t vsubq_f32(float32x4_t __p0, flo
32933	32954	#else
32934	32955	__ai __attribute__((target("neon"))) float32x4_t vsubq_f32(float32x4_t __p0, float32x4_t __p1) {
32935	32956	float32x4_t __ret;
32936		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
32937		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	32957	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	32958	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
32938	32959	__ret = __rev0 - __rev1;
32939		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	32960	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
32940	32961	return __ret;
32941	32962	}
32942	32963	#endif
...	...	@@ -32950,10 +32971,10 @@ __ai __attribute__((target("neon"))) int32x4_t vsubq_s32(int32x4_t __p0, int32x4
32950	32971	#else
32951	32972	__ai __attribute__((target("neon"))) int32x4_t vsubq_s32(int32x4_t __p0, int32x4_t __p1) {
32952	32973	int32x4_t __ret;
32953		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
32954		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	32974	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	32975	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
32955	32976	__ret = __rev0 - __rev1;
32956		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	32977	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
32957	32978	return __ret;
32958	32979	}
32959	32980	#endif
...	...	@@ -32967,10 +32988,10 @@ __ai __attribute__((target("neon"))) int64x2_t vsubq_s64(int64x2_t __p0, int64x2
32967	32988	#else
32968	32989	__ai __attribute__((target("neon"))) int64x2_t vsubq_s64(int64x2_t __p0, int64x2_t __p1) {
32969	32990	int64x2_t __ret;
32970		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
32971		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	32991	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	32992	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
32972	32993	__ret = __rev0 - __rev1;
32973		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	32994	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
32974	32995	return __ret;
32975	32996	}
32976	32997	#endif
...	...	@@ -32984,10 +33005,10 @@ __ai __attribute__((target("neon"))) int16x8_t vsubq_s16(int16x8_t __p0, int16x8
32984	33005	#else
32985	33006	__ai __attribute__((target("neon"))) int16x8_t vsubq_s16(int16x8_t __p0, int16x8_t __p1) {
32986	33007	int16x8_t __ret;
32987		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
32988		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	33008	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	33009	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
32989	33010	__ret = __rev0 - __rev1;
32990		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33011	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
32991	33012	return __ret;
32992	33013	}
32993	33014	#endif
...	...	@@ -33001,10 +33022,10 @@ __ai __attribute__((target("neon"))) uint8x8_t vsub_u8(uint8x8_t __p0, uint8x8_t
33001	33022	#else
33002	33023	__ai __attribute__((target("neon"))) uint8x8_t vsub_u8(uint8x8_t __p0, uint8x8_t __p1) {
33003	33024	uint8x8_t __ret;
33004		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
33005		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	33025	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	33026	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
33006	33027	__ret = __rev0 - __rev1;
33007		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33028	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
33008	33029	return __ret;
33009	33030	}
33010	33031	#endif
...	...	@@ -33018,10 +33039,10 @@ __ai __attribute__((target("neon"))) uint32x2_t vsub_u32(uint32x2_t __p0, uint32
33018	33039	#else
33019	33040	__ai __attribute__((target("neon"))) uint32x2_t vsub_u32(uint32x2_t __p0, uint32x2_t __p1) {
33020	33041	uint32x2_t __ret;
33021		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
33022		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	33042	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	33043	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
33023	33044	__ret = __rev0 - __rev1;
33024		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	33045	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
33025	33046	return __ret;
33026	33047	}
33027	33048	#endif
...	...	@@ -33040,10 +33061,10 @@ __ai __attribute__((target("neon"))) uint16x4_t vsub_u16(uint16x4_t __p0, uint16
33040	33061	#else
33041	33062	__ai __attribute__((target("neon"))) uint16x4_t vsub_u16(uint16x4_t __p0, uint16x4_t __p1) {
33042	33063	uint16x4_t __ret;
33043		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
33044		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	33064	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	33065	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
33045	33066	__ret = __rev0 - __rev1;
33046		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	33067	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
33047	33068	return __ret;
33048	33069	}
33049	33070	#endif
...	...	@@ -33057,10 +33078,10 @@ __ai __attribute__((target("neon"))) int8x8_t vsub_s8(int8x8_t __p0, int8x8_t __
33057	33078	#else
33058	33079	__ai __attribute__((target("neon"))) int8x8_t vsub_s8(int8x8_t __p0, int8x8_t __p1) {
33059	33080	int8x8_t __ret;
33060		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
33061		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	33081	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	33082	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
33062	33083	__ret = __rev0 - __rev1;
33063		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33084	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
33064	33085	return __ret;
33065	33086	}
33066	33087	#endif
...	...	@@ -33074,10 +33095,10 @@ __ai __attribute__((target("neon"))) float32x2_t vsub_f32(float32x2_t __p0, floa
33074	33095	#else
33075	33096	__ai __attribute__((target("neon"))) float32x2_t vsub_f32(float32x2_t __p0, float32x2_t __p1) {
33076	33097	float32x2_t __ret;
33077		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
33078		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	33098	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	33099	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
33079	33100	__ret = __rev0 - __rev1;
33080		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	33101	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
33081	33102	return __ret;
33082	33103	}
33083	33104	#endif
...	...	@@ -33091,10 +33112,10 @@ __ai __attribute__((target("neon"))) int32x2_t vsub_s32(int32x2_t __p0, int32x2_
33091	33112	#else
33092	33113	__ai __attribute__((target("neon"))) int32x2_t vsub_s32(int32x2_t __p0, int32x2_t __p1) {
33093	33114	int32x2_t __ret;
33094		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
33095		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	33115	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	33116	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
33096	33117	__ret = __rev0 - __rev1;
33097		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	33118	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
33098	33119	return __ret;
33099	33120	}
33100	33121	#endif
...	...	@@ -33113,10 +33134,10 @@ __ai __attribute__((target("neon"))) int16x4_t vsub_s16(int16x4_t __p0, int16x4_
33113	33134	#else
33114	33135	__ai __attribute__((target("neon"))) int16x4_t vsub_s16(int16x4_t __p0, int16x4_t __p1) {
33115	33136	int16x4_t __ret;
33116		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
33117		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	33137	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	33138	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
33118	33139	__ret = __rev0 - __rev1;
33119		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	33140	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
33120	33141	return __ret;
33121	33142	}
33122	33143	#endif
...	...	@@ -33124,21 +33145,21 @@ __ai __attribute__((target("neon"))) int16x4_t vsub_s16(int16x4_t __p0, int16x4_
33124	33145	#ifdef __LITTLE_ENDIAN__
33125	33146	__ai __attribute__((target("neon"))) uint16x4_t vsubhn_u32(uint32x4_t __p0, uint32x4_t __p1) {
33126	33147	uint16x4_t __ret;
33127		__ret = (uint16x4_t) __builtin_neon_vsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 17);
	33148	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vsubhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 17));
33128	33149	return __ret;
33129	33150	}
33130	33151	#else
33131	33152	__ai __attribute__((target("neon"))) uint16x4_t vsubhn_u32(uint32x4_t __p0, uint32x4_t __p1) {
33132	33153	uint16x4_t __ret;
33133		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
33134		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
33135		__ret = (uint16x4_t) __builtin_neon_vsubhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 17);
33136		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	33154	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	33155	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	33156	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vsubhn_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 17));
	33157	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
33137	33158	return __ret;
33138	33159	}
33139	33160	__ai __attribute__((target("neon"))) uint16x4_t __noswap_vsubhn_u32(uint32x4_t __p0, uint32x4_t __p1) {
33140	33161	uint16x4_t __ret;
33141		__ret = (uint16x4_t) __builtin_neon_vsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 17);
	33162	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vsubhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 17));
33142	33163	return __ret;
33143	33164	}
33144	33165	#endif
...	...	@@ -33146,21 +33167,21 @@ __ai __attribute__((target("neon"))) uint16x4_t __noswap_vsubhn_u32(uint32x4_t _
33146	33167	#ifdef __LITTLE_ENDIAN__
33147	33168	__ai __attribute__((target("neon"))) uint32x2_t vsubhn_u64(uint64x2_t __p0, uint64x2_t __p1) {
33148	33169	uint32x2_t __ret;
33149		__ret = (uint32x2_t) __builtin_neon_vsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 18);
	33170	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vsubhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 18));
33150	33171	return __ret;
33151	33172	}
33152	33173	#else
33153	33174	__ai __attribute__((target("neon"))) uint32x2_t vsubhn_u64(uint64x2_t __p0, uint64x2_t __p1) {
33154	33175	uint32x2_t __ret;
33155		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
33156		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
33157		__ret = (uint32x2_t) __builtin_neon_vsubhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 18);
33158		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	33176	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	33177	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	33178	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vsubhn_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 18));
	33179	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
33159	33180	return __ret;
33160	33181	}
33161	33182	__ai __attribute__((target("neon"))) uint32x2_t __noswap_vsubhn_u64(uint64x2_t __p0, uint64x2_t __p1) {
33162	33183	uint32x2_t __ret;
33163		__ret = (uint32x2_t) __builtin_neon_vsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 18);
	33184	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vsubhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 18));
33164	33185	return __ret;
33165	33186	}
33166	33187	#endif
...	...	@@ -33168,21 +33189,21 @@ __ai __attribute__((target("neon"))) uint32x2_t __noswap_vsubhn_u64(uint64x2_t _
33168	33189	#ifdef __LITTLE_ENDIAN__
33169	33190	__ai __attribute__((target("neon"))) uint8x8_t vsubhn_u16(uint16x8_t __p0, uint16x8_t __p1) {
33170	33191	uint8x8_t __ret;
33171		__ret = (uint8x8_t) __builtin_neon_vsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 16);
	33192	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vsubhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 16));
33172	33193	return __ret;
33173	33194	}
33174	33195	#else
33175	33196	__ai __attribute__((target("neon"))) uint8x8_t vsubhn_u16(uint16x8_t __p0, uint16x8_t __p1) {
33176	33197	uint8x8_t __ret;
33177		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
33178		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
33179		__ret = (uint8x8_t) __builtin_neon_vsubhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 16);
33180		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33198	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	33199	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	33200	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vsubhn_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 16));
	33201	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
33181	33202	return __ret;
33182	33203	}
33183	33204	__ai __attribute__((target("neon"))) uint8x8_t __noswap_vsubhn_u16(uint16x8_t __p0, uint16x8_t __p1) {
33184	33205	uint8x8_t __ret;
33185		__ret = (uint8x8_t) __builtin_neon_vsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 16);
	33206	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vsubhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 16));
33186	33207	return __ret;
33187	33208	}
33188	33209	#endif
...	...	@@ -33190,21 +33211,21 @@ __ai __attribute__((target("neon"))) uint8x8_t __noswap_vsubhn_u16(uint16x8_t __
33190	33211	#ifdef __LITTLE_ENDIAN__
33191	33212	__ai __attribute__((target("neon"))) int16x4_t vsubhn_s32(int32x4_t __p0, int32x4_t __p1) {
33192	33213	int16x4_t __ret;
33193		__ret = (int16x4_t) __builtin_neon_vsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 1);
	33214	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vsubhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 1));
33194	33215	return __ret;
33195	33216	}
33196	33217	#else
33197	33218	__ai __attribute__((target("neon"))) int16x4_t vsubhn_s32(int32x4_t __p0, int32x4_t __p1) {
33198	33219	int16x4_t __ret;
33199		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
33200		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
33201		__ret = (int16x4_t) __builtin_neon_vsubhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 1);
33202		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	33220	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	33221	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	33222	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vsubhn_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 1));
	33223	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
33203	33224	return __ret;
33204	33225	}
33205	33226	__ai __attribute__((target("neon"))) int16x4_t __noswap_vsubhn_s32(int32x4_t __p0, int32x4_t __p1) {
33206	33227	int16x4_t __ret;
33207		__ret = (int16x4_t) __builtin_neon_vsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 1);
	33228	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vsubhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 1));
33208	33229	return __ret;
33209	33230	}
33210	33231	#endif
...	...	@@ -33212,21 +33233,21 @@ __ai __attribute__((target("neon"))) int16x4_t __noswap_vsubhn_s32(int32x4_t __p
33212	33233	#ifdef __LITTLE_ENDIAN__
33213	33234	__ai __attribute__((target("neon"))) int32x2_t vsubhn_s64(int64x2_t __p0, int64x2_t __p1) {
33214	33235	int32x2_t __ret;
33215		__ret = (int32x2_t) __builtin_neon_vsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 2);
	33236	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vsubhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 2));
33216	33237	return __ret;
33217	33238	}
33218	33239	#else
33219	33240	__ai __attribute__((target("neon"))) int32x2_t vsubhn_s64(int64x2_t __p0, int64x2_t __p1) {
33220	33241	int32x2_t __ret;
33221		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
33222		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
33223		__ret = (int32x2_t) __builtin_neon_vsubhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 2);
33224		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	33242	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	33243	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	33244	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vsubhn_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 2));
	33245	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
33225	33246	return __ret;
33226	33247	}
33227	33248	__ai __attribute__((target("neon"))) int32x2_t __noswap_vsubhn_s64(int64x2_t __p0, int64x2_t __p1) {
33228	33249	int32x2_t __ret;
33229		__ret = (int32x2_t) __builtin_neon_vsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 2);
	33250	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vsubhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 2));
33230	33251	return __ret;
33231	33252	}
33232	33253	#endif
...	...	@@ -33234,21 +33255,21 @@ __ai __attribute__((target("neon"))) int32x2_t __noswap_vsubhn_s64(int64x2_t __p
33234	33255	#ifdef __LITTLE_ENDIAN__
33235	33256	__ai __attribute__((target("neon"))) int8x8_t vsubhn_s16(int16x8_t __p0, int16x8_t __p1) {
33236	33257	int8x8_t __ret;
33237		__ret = (int8x8_t) __builtin_neon_vsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 0);
	33258	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vsubhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 0));
33238	33259	return __ret;
33239	33260	}
33240	33261	#else
33241	33262	__ai __attribute__((target("neon"))) int8x8_t vsubhn_s16(int16x8_t __p0, int16x8_t __p1) {
33242	33263	int8x8_t __ret;
33243		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
33244		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
33245		__ret = (int8x8_t) __builtin_neon_vsubhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 0);
33246		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33264	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	33265	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	33266	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vsubhn_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 0));
	33267	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
33247	33268	return __ret;
33248	33269	}
33249	33270	__ai __attribute__((target("neon"))) int8x8_t __noswap_vsubhn_s16(int16x8_t __p0, int16x8_t __p1) {
33250	33271	int8x8_t __ret;
33251		__ret = (int8x8_t) __builtin_neon_vsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 0);
	33272	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vsubhn_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 0));
33252	33273	return __ret;
33253	33274	}
33254	33275	#endif
...	...	@@ -33262,10 +33283,10 @@ __ai __attribute__((target("neon"))) uint16x8_t vsubl_u8(uint8x8_t __p0, uint8x8
33262	33283	#else
33263	33284	__ai __attribute__((target("neon"))) uint16x8_t vsubl_u8(uint8x8_t __p0, uint8x8_t __p1) {
33264	33285	uint16x8_t __ret;
33265		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
33266		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	33286	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	33287	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
33267	33288	__ret = __noswap_vmovl_u8(__rev0) - __noswap_vmovl_u8(__rev1);
33268		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33289	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
33269	33290	return __ret;
33270	33291	}
33271	33292	#endif
...	...	@@ -33279,10 +33300,10 @@ __ai __attribute__((target("neon"))) uint64x2_t vsubl_u32(uint32x2_t __p0, uint3
33279	33300	#else
33280	33301	__ai __attribute__((target("neon"))) uint64x2_t vsubl_u32(uint32x2_t __p0, uint32x2_t __p1) {
33281	33302	uint64x2_t __ret;
33282		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
33283		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	33303	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	33304	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
33284	33305	__ret = __noswap_vmovl_u32(__rev0) - __noswap_vmovl_u32(__rev1);
33285		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	33306	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
33286	33307	return __ret;
33287	33308	}
33288	33309	#endif
...	...	@@ -33296,10 +33317,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vsubl_u16(uint16x4_t __p0, uint1
33296	33317	#else
33297	33318	__ai __attribute__((target("neon"))) uint32x4_t vsubl_u16(uint16x4_t __p0, uint16x4_t __p1) {
33298	33319	uint32x4_t __ret;
33299		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
33300		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	33320	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	33321	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
33301	33322	__ret = __noswap_vmovl_u16(__rev0) - __noswap_vmovl_u16(__rev1);
33302		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	33323	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
33303	33324	return __ret;
33304	33325	}
33305	33326	#endif
...	...	@@ -33313,10 +33334,10 @@ __ai __attribute__((target("neon"))) int16x8_t vsubl_s8(int8x8_t __p0, int8x8_t
33313	33334	#else
33314	33335	__ai __attribute__((target("neon"))) int16x8_t vsubl_s8(int8x8_t __p0, int8x8_t __p1) {
33315	33336	int16x8_t __ret;
33316		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
33317		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	33337	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	33338	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
33318	33339	__ret = __noswap_vmovl_s8(__rev0) - __noswap_vmovl_s8(__rev1);
33319		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33340	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
33320	33341	return __ret;
33321	33342	}
33322	33343	#endif
...	...	@@ -33330,10 +33351,10 @@ __ai __attribute__((target("neon"))) int64x2_t vsubl_s32(int32x2_t __p0, int32x2
33330	33351	#else
33331	33352	__ai __attribute__((target("neon"))) int64x2_t vsubl_s32(int32x2_t __p0, int32x2_t __p1) {
33332	33353	int64x2_t __ret;
33333		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
33334		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	33354	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	33355	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
33335	33356	__ret = __noswap_vmovl_s32(__rev0) - __noswap_vmovl_s32(__rev1);
33336		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	33357	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
33337	33358	return __ret;
33338	33359	}
33339	33360	#endif
...	...	@@ -33347,10 +33368,10 @@ __ai __attribute__((target("neon"))) int32x4_t vsubl_s16(int16x4_t __p0, int16x4
33347	33368	#else
33348	33369	__ai __attribute__((target("neon"))) int32x4_t vsubl_s16(int16x4_t __p0, int16x4_t __p1) {
33349	33370	int32x4_t __ret;
33350		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
33351		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	33371	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	33372	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
33352	33373	__ret = __noswap_vmovl_s16(__rev0) - __noswap_vmovl_s16(__rev1);
33353		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	33374	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
33354	33375	return __ret;
33355	33376	}
33356	33377	#endif
...	...	@@ -33364,10 +33385,10 @@ __ai __attribute__((target("neon"))) uint16x8_t vsubw_u8(uint16x8_t __p0, uint8x
33364	33385	#else
33365	33386	__ai __attribute__((target("neon"))) uint16x8_t vsubw_u8(uint16x8_t __p0, uint8x8_t __p1) {
33366	33387	uint16x8_t __ret;
33367		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
33368		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	33388	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	33389	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
33369	33390	__ret = __rev0 - __noswap_vmovl_u8(__rev1);
33370		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33391	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
33371	33392	return __ret;
33372	33393	}
33373	33394	#endif
...	...	@@ -33381,10 +33402,10 @@ __ai __attribute__((target("neon"))) uint64x2_t vsubw_u32(uint64x2_t __p0, uint3
33381	33402	#else
33382	33403	__ai __attribute__((target("neon"))) uint64x2_t vsubw_u32(uint64x2_t __p0, uint32x2_t __p1) {
33383	33404	uint64x2_t __ret;
33384		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
33385		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	33405	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	33406	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
33386	33407	__ret = __rev0 - __noswap_vmovl_u32(__rev1);
33387		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	33408	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
33388	33409	return __ret;
33389	33410	}
33390	33411	#endif
...	...	@@ -33398,10 +33419,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vsubw_u16(uint32x4_t __p0, uint1
33398	33419	#else
33399	33420	__ai __attribute__((target("neon"))) uint32x4_t vsubw_u16(uint32x4_t __p0, uint16x4_t __p1) {
33400	33421	uint32x4_t __ret;
33401		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
33402		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	33422	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	33423	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
33403	33424	__ret = __rev0 - __noswap_vmovl_u16(__rev1);
33404		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	33425	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
33405	33426	return __ret;
33406	33427	}
33407	33428	#endif
...	...	@@ -33415,10 +33436,10 @@ __ai __attribute__((target("neon"))) int16x8_t vsubw_s8(int16x8_t __p0, int8x8_t
33415	33436	#else
33416	33437	__ai __attribute__((target("neon"))) int16x8_t vsubw_s8(int16x8_t __p0, int8x8_t __p1) {
33417	33438	int16x8_t __ret;
33418		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
33419		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	33439	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	33440	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
33420	33441	__ret = __rev0 - __noswap_vmovl_s8(__rev1);
33421		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33442	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
33422	33443	return __ret;
33423	33444	}
33424	33445	#endif
...	...	@@ -33432,10 +33453,10 @@ __ai __attribute__((target("neon"))) int64x2_t vsubw_s32(int64x2_t __p0, int32x2
33432	33453	#else
33433	33454	__ai __attribute__((target("neon"))) int64x2_t vsubw_s32(int64x2_t __p0, int32x2_t __p1) {
33434	33455	int64x2_t __ret;
33435		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
33436		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	33456	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	33457	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
33437	33458	__ret = __rev0 - __noswap_vmovl_s32(__rev1);
33438		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	33459	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
33439	33460	return __ret;
33440	33461	}
33441	33462	#endif
...	...	@@ -33449,10 +33470,10 @@ __ai __attribute__((target("neon"))) int32x4_t vsubw_s16(int32x4_t __p0, int16x4
33449	33470	#else
33450	33471	__ai __attribute__((target("neon"))) int32x4_t vsubw_s16(int32x4_t __p0, int16x4_t __p1) {
33451	33472	int32x4_t __ret;
33452		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
33453		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	33473	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	33474	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
33454	33475	__ret = __rev0 - __noswap_vmovl_s16(__rev1);
33455		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	33476	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
33456	33477	return __ret;
33457	33478	}
33458	33479	#endif
...	...	@@ -33460,16 +33481,16 @@ __ai __attribute__((target("neon"))) int32x4_t vsubw_s16(int32x4_t __p0, int16x4
33460	33481	#ifdef __LITTLE_ENDIAN__
33461	33482	__ai __attribute__((target("neon"))) poly8x8_t vtbl1_p8(poly8x8_t __p0, uint8x8_t __p1) {
33462	33483	poly8x8_t __ret;
33463		__ret = (poly8x8_t) __builtin_neon_vtbl1_v((int8x8_t)__p0, (int8x8_t)__p1, 4);
	33484	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vtbl1_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 4));
33464	33485	return __ret;
33465	33486	}
33466	33487	#else
33467	33488	__ai __attribute__((target("neon"))) poly8x8_t vtbl1_p8(poly8x8_t __p0, uint8x8_t __p1) {
33468	33489	poly8x8_t __ret;
33469		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
33470		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
33471		__ret = (poly8x8_t) __builtin_neon_vtbl1_v((int8x8_t)__rev0, (int8x8_t)__rev1, 4);
33472		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33490	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	33491	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	33492	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vtbl1_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 4));
	33493	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
33473	33494	return __ret;
33474	33495	}
33475	33496	#endif
...	...	@@ -33477,16 +33498,16 @@ __ai __attribute__((target("neon"))) poly8x8_t vtbl1_p8(poly8x8_t __p0, uint8x8_
33477	33498	#ifdef __LITTLE_ENDIAN__
33478	33499	__ai __attribute__((target("neon"))) uint8x8_t vtbl1_u8(uint8x8_t __p0, uint8x8_t __p1) {
33479	33500	uint8x8_t __ret;
33480		__ret = (uint8x8_t) __builtin_neon_vtbl1_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
	33501	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vtbl1_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 16));
33481	33502	return __ret;
33482	33503	}
33483	33504	#else
33484	33505	__ai __attribute__((target("neon"))) uint8x8_t vtbl1_u8(uint8x8_t __p0, uint8x8_t __p1) {
33485	33506	uint8x8_t __ret;
33486		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
33487		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
33488		__ret = (uint8x8_t) __builtin_neon_vtbl1_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
33489		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33507	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	33508	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	33509	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vtbl1_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 16));
	33510	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
33490	33511	return __ret;
33491	33512	}
33492	33513	#endif
...	...	@@ -33494,16 +33515,16 @@ __ai __attribute__((target("neon"))) uint8x8_t vtbl1_u8(uint8x8_t __p0, uint8x8_
33494	33515	#ifdef __LITTLE_ENDIAN__
33495	33516	__ai __attribute__((target("neon"))) int8x8_t vtbl1_s8(int8x8_t __p0, int8x8_t __p1) {
33496	33517	int8x8_t __ret;
33497		__ret = (int8x8_t) __builtin_neon_vtbl1_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
	33518	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vtbl1_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 0));
33498	33519	return __ret;
33499	33520	}
33500	33521	#else
33501	33522	__ai __attribute__((target("neon"))) int8x8_t vtbl1_s8(int8x8_t __p0, int8x8_t __p1) {
33502	33523	int8x8_t __ret;
33503		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
33504		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
33505		__ret = (int8x8_t) __builtin_neon_vtbl1_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
33506		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33524	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	33525	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	33526	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vtbl1_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 0));
	33527	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
33507	33528	return __ret;
33508	33529	}
33509	33530	#endif
...	...	@@ -33511,18 +33532,18 @@ __ai __attribute__((target("neon"))) int8x8_t vtbl1_s8(int8x8_t __p0, int8x8_t _
33511	33532	#ifdef __LITTLE_ENDIAN__
33512	33533	__ai __attribute__((target("neon"))) poly8x8_t vtbl2_p8(poly8x8x2_t __p0, uint8x8_t __p1) {
33513	33534	poly8x8_t __ret;
33514		__ret = (poly8x8_t) __builtin_neon_vtbl2_v((int8x8_t)__p0.val[0], (int8x8_t)__p0.val[1], (int8x8_t)__p1, 4);
	33535	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vtbl2_v(__builtin_bit_cast(int8x8_t, __p0.val[0]), __builtin_bit_cast(int8x8_t, __p0.val[1]), __builtin_bit_cast(int8x8_t, __p1), 4));
33515	33536	return __ret;
33516	33537	}
33517	33538	#else
33518	33539	__ai __attribute__((target("neon"))) poly8x8_t vtbl2_p8(poly8x8x2_t __p0, uint8x8_t __p1) {
33519	33540	poly8x8_t __ret;
33520	33541	poly8x8x2_t __rev0;
33521		__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
33522		__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
33523		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
33524		__ret = (poly8x8_t) __builtin_neon_vtbl2_v((int8x8_t)__rev0.val[0], (int8x8_t)__rev0.val[1], (int8x8_t)__rev1, 4);
33525		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33542	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_64_8);
	33543	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_64_8);
	33544	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	33545	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vtbl2_v(__builtin_bit_cast(int8x8_t, __rev0.val[0]), __builtin_bit_cast(int8x8_t, __rev0.val[1]), __builtin_bit_cast(int8x8_t, __rev1), 4));
	33546	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
33526	33547	return __ret;
33527	33548	}
33528	33549	#endif
...	...	@@ -33530,18 +33551,18 @@ __ai __attribute__((target("neon"))) poly8x8_t vtbl2_p8(poly8x8x2_t __p0, uint8x
33530	33551	#ifdef __LITTLE_ENDIAN__
33531	33552	__ai __attribute__((target("neon"))) uint8x8_t vtbl2_u8(uint8x8x2_t __p0, uint8x8_t __p1) {
33532	33553	uint8x8_t __ret;
33533		__ret = (uint8x8_t) __builtin_neon_vtbl2_v((int8x8_t)__p0.val[0], (int8x8_t)__p0.val[1], (int8x8_t)__p1, 16);
	33554	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vtbl2_v(__builtin_bit_cast(int8x8_t, __p0.val[0]), __builtin_bit_cast(int8x8_t, __p0.val[1]), __builtin_bit_cast(int8x8_t, __p1), 16));
33534	33555	return __ret;
33535	33556	}
33536	33557	#else
33537	33558	__ai __attribute__((target("neon"))) uint8x8_t vtbl2_u8(uint8x8x2_t __p0, uint8x8_t __p1) {
33538	33559	uint8x8_t __ret;
33539	33560	uint8x8x2_t __rev0;
33540		__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
33541		__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
33542		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
33543		__ret = (uint8x8_t) __builtin_neon_vtbl2_v((int8x8_t)__rev0.val[0], (int8x8_t)__rev0.val[1], (int8x8_t)__rev1, 16);
33544		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33561	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_64_8);
	33562	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_64_8);
	33563	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	33564	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vtbl2_v(__builtin_bit_cast(int8x8_t, __rev0.val[0]), __builtin_bit_cast(int8x8_t, __rev0.val[1]), __builtin_bit_cast(int8x8_t, __rev1), 16));
	33565	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
33545	33566	return __ret;
33546	33567	}
33547	33568	#endif
...	...	@@ -33549,18 +33570,18 @@ __ai __attribute__((target("neon"))) uint8x8_t vtbl2_u8(uint8x8x2_t __p0, uint8x
33549	33570	#ifdef __LITTLE_ENDIAN__
33550	33571	__ai __attribute__((target("neon"))) int8x8_t vtbl2_s8(int8x8x2_t __p0, int8x8_t __p1) {
33551	33572	int8x8_t __ret;
33552		__ret = (int8x8_t) __builtin_neon_vtbl2_v((int8x8_t)__p0.val[0], (int8x8_t)__p0.val[1], (int8x8_t)__p1, 0);
	33573	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vtbl2_v(__builtin_bit_cast(int8x8_t, __p0.val[0]), __builtin_bit_cast(int8x8_t, __p0.val[1]), __builtin_bit_cast(int8x8_t, __p1), 0));
33553	33574	return __ret;
33554	33575	}
33555	33576	#else
33556	33577	__ai __attribute__((target("neon"))) int8x8_t vtbl2_s8(int8x8x2_t __p0, int8x8_t __p1) {
33557	33578	int8x8_t __ret;
33558	33579	int8x8x2_t __rev0;
33559		__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
33560		__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
33561		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
33562		__ret = (int8x8_t) __builtin_neon_vtbl2_v((int8x8_t)__rev0.val[0], (int8x8_t)__rev0.val[1], (int8x8_t)__rev1, 0);
33563		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33580	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_64_8);
	33581	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_64_8);
	33582	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	33583	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vtbl2_v(__builtin_bit_cast(int8x8_t, __rev0.val[0]), __builtin_bit_cast(int8x8_t, __rev0.val[1]), __builtin_bit_cast(int8x8_t, __rev1), 0));
	33584	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
33564	33585	return __ret;
33565	33586	}
33566	33587	#endif
...	...	@@ -33568,19 +33589,19 @@ __ai __attribute__((target("neon"))) int8x8_t vtbl2_s8(int8x8x2_t __p0, int8x8_t
33568	33589	#ifdef __LITTLE_ENDIAN__
33569	33590	__ai __attribute__((target("neon"))) poly8x8_t vtbl3_p8(poly8x8x3_t __p0, uint8x8_t __p1) {
33570	33591	poly8x8_t __ret;
33571		__ret = (poly8x8_t) __builtin_neon_vtbl3_v((int8x8_t)__p0.val[0], (int8x8_t)__p0.val[1], (int8x8_t)__p0.val[2], (int8x8_t)__p1, 4);
	33592	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vtbl3_v(__builtin_bit_cast(int8x8_t, __p0.val[0]), __builtin_bit_cast(int8x8_t, __p0.val[1]), __builtin_bit_cast(int8x8_t, __p0.val[2]), __builtin_bit_cast(int8x8_t, __p1), 4));
33572	33593	return __ret;
33573	33594	}
33574	33595	#else
33575	33596	__ai __attribute__((target("neon"))) poly8x8_t vtbl3_p8(poly8x8x3_t __p0, uint8x8_t __p1) {
33576	33597	poly8x8_t __ret;
33577	33598	poly8x8x3_t __rev0;
33578		__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
33579		__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
33580		__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], 7, 6, 5, 4, 3, 2, 1, 0);
33581		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
33582		__ret = (poly8x8_t) __builtin_neon_vtbl3_v((int8x8_t)__rev0.val[0], (int8x8_t)__rev0.val[1], (int8x8_t)__rev0.val[2], (int8x8_t)__rev1, 4);
33583		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33599	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_64_8);
	33600	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_64_8);
	33601	__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], __lane_reverse_64_8);
	33602	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	33603	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vtbl3_v(__builtin_bit_cast(int8x8_t, __rev0.val[0]), __builtin_bit_cast(int8x8_t, __rev0.val[1]), __builtin_bit_cast(int8x8_t, __rev0.val[2]), __builtin_bit_cast(int8x8_t, __rev1), 4));
	33604	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
33584	33605	return __ret;
33585	33606	}
33586	33607	#endif
...	...	@@ -33588,19 +33609,19 @@ __ai __attribute__((target("neon"))) poly8x8_t vtbl3_p8(poly8x8x3_t __p0, uint8x
33588	33609	#ifdef __LITTLE_ENDIAN__
33589	33610	__ai __attribute__((target("neon"))) uint8x8_t vtbl3_u8(uint8x8x3_t __p0, uint8x8_t __p1) {
33590	33611	uint8x8_t __ret;
33591		__ret = (uint8x8_t) __builtin_neon_vtbl3_v((int8x8_t)__p0.val[0], (int8x8_t)__p0.val[1], (int8x8_t)__p0.val[2], (int8x8_t)__p1, 16);
	33612	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vtbl3_v(__builtin_bit_cast(int8x8_t, __p0.val[0]), __builtin_bit_cast(int8x8_t, __p0.val[1]), __builtin_bit_cast(int8x8_t, __p0.val[2]), __builtin_bit_cast(int8x8_t, __p1), 16));
33592	33613	return __ret;
33593	33614	}
33594	33615	#else
33595	33616	__ai __attribute__((target("neon"))) uint8x8_t vtbl3_u8(uint8x8x3_t __p0, uint8x8_t __p1) {
33596	33617	uint8x8_t __ret;
33597	33618	uint8x8x3_t __rev0;
33598		__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
33599		__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
33600		__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], 7, 6, 5, 4, 3, 2, 1, 0);
33601		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
33602		__ret = (uint8x8_t) __builtin_neon_vtbl3_v((int8x8_t)__rev0.val[0], (int8x8_t)__rev0.val[1], (int8x8_t)__rev0.val[2], (int8x8_t)__rev1, 16);
33603		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33619	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_64_8);
	33620	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_64_8);
	33621	__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], __lane_reverse_64_8);
	33622	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	33623	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vtbl3_v(__builtin_bit_cast(int8x8_t, __rev0.val[0]), __builtin_bit_cast(int8x8_t, __rev0.val[1]), __builtin_bit_cast(int8x8_t, __rev0.val[2]), __builtin_bit_cast(int8x8_t, __rev1), 16));
	33624	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
33604	33625	return __ret;
33605	33626	}
33606	33627	#endif
...	...	@@ -33608,19 +33629,19 @@ __ai __attribute__((target("neon"))) uint8x8_t vtbl3_u8(uint8x8x3_t __p0, uint8x
33608	33629	#ifdef __LITTLE_ENDIAN__
33609	33630	__ai __attribute__((target("neon"))) int8x8_t vtbl3_s8(int8x8x3_t __p0, int8x8_t __p1) {
33610	33631	int8x8_t __ret;
33611		__ret = (int8x8_t) __builtin_neon_vtbl3_v((int8x8_t)__p0.val[0], (int8x8_t)__p0.val[1], (int8x8_t)__p0.val[2], (int8x8_t)__p1, 0);
	33632	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vtbl3_v(__builtin_bit_cast(int8x8_t, __p0.val[0]), __builtin_bit_cast(int8x8_t, __p0.val[1]), __builtin_bit_cast(int8x8_t, __p0.val[2]), __builtin_bit_cast(int8x8_t, __p1), 0));
33612	33633	return __ret;
33613	33634	}
33614	33635	#else
33615	33636	__ai __attribute__((target("neon"))) int8x8_t vtbl3_s8(int8x8x3_t __p0, int8x8_t __p1) {
33616	33637	int8x8_t __ret;
33617	33638	int8x8x3_t __rev0;
33618		__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
33619		__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
33620		__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], 7, 6, 5, 4, 3, 2, 1, 0);
33621		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
33622		__ret = (int8x8_t) __builtin_neon_vtbl3_v((int8x8_t)__rev0.val[0], (int8x8_t)__rev0.val[1], (int8x8_t)__rev0.val[2], (int8x8_t)__rev1, 0);
33623		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33639	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_64_8);
	33640	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_64_8);
	33641	__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], __lane_reverse_64_8);
	33642	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	33643	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vtbl3_v(__builtin_bit_cast(int8x8_t, __rev0.val[0]), __builtin_bit_cast(int8x8_t, __rev0.val[1]), __builtin_bit_cast(int8x8_t, __rev0.val[2]), __builtin_bit_cast(int8x8_t, __rev1), 0));
	33644	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
33624	33645	return __ret;
33625	33646	}
33626	33647	#endif
...	...	@@ -33628,20 +33649,20 @@ __ai __attribute__((target("neon"))) int8x8_t vtbl3_s8(int8x8x3_t __p0, int8x8_t
33628	33649	#ifdef __LITTLE_ENDIAN__
33629	33650	__ai __attribute__((target("neon"))) poly8x8_t vtbl4_p8(poly8x8x4_t __p0, uint8x8_t __p1) {
33630	33651	poly8x8_t __ret;
33631		__ret = (poly8x8_t) __builtin_neon_vtbl4_v((int8x8_t)__p0.val[0], (int8x8_t)__p0.val[1], (int8x8_t)__p0.val[2], (int8x8_t)__p0.val[3], (int8x8_t)__p1, 4);
	33652	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vtbl4_v(__builtin_bit_cast(int8x8_t, __p0.val[0]), __builtin_bit_cast(int8x8_t, __p0.val[1]), __builtin_bit_cast(int8x8_t, __p0.val[2]), __builtin_bit_cast(int8x8_t, __p0.val[3]), __builtin_bit_cast(int8x8_t, __p1), 4));
33632	33653	return __ret;
33633	33654	}
33634	33655	#else
33635	33656	__ai __attribute__((target("neon"))) poly8x8_t vtbl4_p8(poly8x8x4_t __p0, uint8x8_t __p1) {
33636	33657	poly8x8_t __ret;
33637	33658	poly8x8x4_t __rev0;
33638		__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
33639		__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
33640		__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], 7, 6, 5, 4, 3, 2, 1, 0);
33641		__rev0.val[3] = __builtin_shufflevector(__p0.val[3], __p0.val[3], 7, 6, 5, 4, 3, 2, 1, 0);
33642		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
33643		__ret = (poly8x8_t) __builtin_neon_vtbl4_v((int8x8_t)__rev0.val[0], (int8x8_t)__rev0.val[1], (int8x8_t)__rev0.val[2], (int8x8_t)__rev0.val[3], (int8x8_t)__rev1, 4);
33644		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33659	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_64_8);
	33660	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_64_8);
	33661	__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], __lane_reverse_64_8);
	33662	__rev0.val[3] = __builtin_shufflevector(__p0.val[3], __p0.val[3], __lane_reverse_64_8);
	33663	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	33664	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vtbl4_v(__builtin_bit_cast(int8x8_t, __rev0.val[0]), __builtin_bit_cast(int8x8_t, __rev0.val[1]), __builtin_bit_cast(int8x8_t, __rev0.val[2]), __builtin_bit_cast(int8x8_t, __rev0.val[3]), __builtin_bit_cast(int8x8_t, __rev1), 4));
	33665	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
33645	33666	return __ret;
33646	33667	}
33647	33668	#endif
...	...	@@ -33649,20 +33670,20 @@ __ai __attribute__((target("neon"))) poly8x8_t vtbl4_p8(poly8x8x4_t __p0, uint8x
33649	33670	#ifdef __LITTLE_ENDIAN__
33650	33671	__ai __attribute__((target("neon"))) uint8x8_t vtbl4_u8(uint8x8x4_t __p0, uint8x8_t __p1) {
33651	33672	uint8x8_t __ret;
33652		__ret = (uint8x8_t) __builtin_neon_vtbl4_v((int8x8_t)__p0.val[0], (int8x8_t)__p0.val[1], (int8x8_t)__p0.val[2], (int8x8_t)__p0.val[3], (int8x8_t)__p1, 16);
	33673	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vtbl4_v(__builtin_bit_cast(int8x8_t, __p0.val[0]), __builtin_bit_cast(int8x8_t, __p0.val[1]), __builtin_bit_cast(int8x8_t, __p0.val[2]), __builtin_bit_cast(int8x8_t, __p0.val[3]), __builtin_bit_cast(int8x8_t, __p1), 16));
33653	33674	return __ret;
33654	33675	}
33655	33676	#else
33656	33677	__ai __attribute__((target("neon"))) uint8x8_t vtbl4_u8(uint8x8x4_t __p0, uint8x8_t __p1) {
33657	33678	uint8x8_t __ret;
33658	33679	uint8x8x4_t __rev0;
33659		__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
33660		__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
33661		__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], 7, 6, 5, 4, 3, 2, 1, 0);
33662		__rev0.val[3] = __builtin_shufflevector(__p0.val[3], __p0.val[3], 7, 6, 5, 4, 3, 2, 1, 0);
33663		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
33664		__ret = (uint8x8_t) __builtin_neon_vtbl4_v((int8x8_t)__rev0.val[0], (int8x8_t)__rev0.val[1], (int8x8_t)__rev0.val[2], (int8x8_t)__rev0.val[3], (int8x8_t)__rev1, 16);
33665		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33680	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_64_8);
	33681	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_64_8);
	33682	__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], __lane_reverse_64_8);
	33683	__rev0.val[3] = __builtin_shufflevector(__p0.val[3], __p0.val[3], __lane_reverse_64_8);
	33684	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	33685	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vtbl4_v(__builtin_bit_cast(int8x8_t, __rev0.val[0]), __builtin_bit_cast(int8x8_t, __rev0.val[1]), __builtin_bit_cast(int8x8_t, __rev0.val[2]), __builtin_bit_cast(int8x8_t, __rev0.val[3]), __builtin_bit_cast(int8x8_t, __rev1), 16));
	33686	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
33666	33687	return __ret;
33667	33688	}
33668	33689	#endif
...	...	@@ -33670,20 +33691,20 @@ __ai __attribute__((target("neon"))) uint8x8_t vtbl4_u8(uint8x8x4_t __p0, uint8x
33670	33691	#ifdef __LITTLE_ENDIAN__
33671	33692	__ai __attribute__((target("neon"))) int8x8_t vtbl4_s8(int8x8x4_t __p0, int8x8_t __p1) {
33672	33693	int8x8_t __ret;
33673		__ret = (int8x8_t) __builtin_neon_vtbl4_v((int8x8_t)__p0.val[0], (int8x8_t)__p0.val[1], (int8x8_t)__p0.val[2], (int8x8_t)__p0.val[3], (int8x8_t)__p1, 0);
	33694	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vtbl4_v(__builtin_bit_cast(int8x8_t, __p0.val[0]), __builtin_bit_cast(int8x8_t, __p0.val[1]), __builtin_bit_cast(int8x8_t, __p0.val[2]), __builtin_bit_cast(int8x8_t, __p0.val[3]), __builtin_bit_cast(int8x8_t, __p1), 0));
33674	33695	return __ret;
33675	33696	}
33676	33697	#else
33677	33698	__ai __attribute__((target("neon"))) int8x8_t vtbl4_s8(int8x8x4_t __p0, int8x8_t __p1) {
33678	33699	int8x8_t __ret;
33679	33700	int8x8x4_t __rev0;
33680		__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
33681		__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
33682		__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], 7, 6, 5, 4, 3, 2, 1, 0);
33683		__rev0.val[3] = __builtin_shufflevector(__p0.val[3], __p0.val[3], 7, 6, 5, 4, 3, 2, 1, 0);
33684		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
33685		__ret = (int8x8_t) __builtin_neon_vtbl4_v((int8x8_t)__rev0.val[0], (int8x8_t)__rev0.val[1], (int8x8_t)__rev0.val[2], (int8x8_t)__rev0.val[3], (int8x8_t)__rev1, 0);
33686		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33701	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_64_8);
	33702	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_64_8);
	33703	__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], __lane_reverse_64_8);
	33704	__rev0.val[3] = __builtin_shufflevector(__p0.val[3], __p0.val[3], __lane_reverse_64_8);
	33705	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	33706	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vtbl4_v(__builtin_bit_cast(int8x8_t, __rev0.val[0]), __builtin_bit_cast(int8x8_t, __rev0.val[1]), __builtin_bit_cast(int8x8_t, __rev0.val[2]), __builtin_bit_cast(int8x8_t, __rev0.val[3]), __builtin_bit_cast(int8x8_t, __rev1), 0));
	33707	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
33687	33708	return __ret;
33688	33709	}
33689	33710	#endif
...	...	@@ -33691,17 +33712,17 @@ __ai __attribute__((target("neon"))) int8x8_t vtbl4_s8(int8x8x4_t __p0, int8x8_t
33691	33712	#ifdef __LITTLE_ENDIAN__
33692	33713	__ai __attribute__((target("neon"))) poly8x8_t vtbx1_p8(poly8x8_t __p0, poly8x8_t __p1, uint8x8_t __p2) {
33693	33714	poly8x8_t __ret;
33694		__ret = (poly8x8_t) __builtin_neon_vtbx1_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 4);
	33715	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vtbx1_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 4));
33695	33716	return __ret;
33696	33717	}
33697	33718	#else
33698	33719	__ai __attribute__((target("neon"))) poly8x8_t vtbx1_p8(poly8x8_t __p0, poly8x8_t __p1, uint8x8_t __p2) {
33699	33720	poly8x8_t __ret;
33700		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
33701		poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
33702		uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
33703		__ret = (poly8x8_t) __builtin_neon_vtbx1_v((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 4);
33704		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33721	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	33722	poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	33723	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	33724	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vtbx1_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 4));
	33725	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
33705	33726	return __ret;
33706	33727	}
33707	33728	#endif
...	...	@@ -33709,17 +33730,17 @@ __ai __attribute__((target("neon"))) poly8x8_t vtbx1_p8(poly8x8_t __p0, poly8x8_
33709	33730	#ifdef __LITTLE_ENDIAN__
33710	33731	__ai __attribute__((target("neon"))) uint8x8_t vtbx1_u8(uint8x8_t __p0, uint8x8_t __p1, uint8x8_t __p2) {
33711	33732	uint8x8_t __ret;
33712		__ret = (uint8x8_t) __builtin_neon_vtbx1_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 16);
	33733	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vtbx1_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 16));
33713	33734	return __ret;
33714	33735	}
33715	33736	#else
33716	33737	__ai __attribute__((target("neon"))) uint8x8_t vtbx1_u8(uint8x8_t __p0, uint8x8_t __p1, uint8x8_t __p2) {
33717	33738	uint8x8_t __ret;
33718		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
33719		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
33720		uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
33721		__ret = (uint8x8_t) __builtin_neon_vtbx1_v((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 16);
33722		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33739	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	33740	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	33741	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	33742	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vtbx1_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 16));
	33743	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
33723	33744	return __ret;
33724	33745	}
33725	33746	#endif
...	...	@@ -33727,17 +33748,17 @@ __ai __attribute__((target("neon"))) uint8x8_t vtbx1_u8(uint8x8_t __p0, uint8x8_
33727	33748	#ifdef __LITTLE_ENDIAN__
33728	33749	__ai __attribute__((target("neon"))) int8x8_t vtbx1_s8(int8x8_t __p0, int8x8_t __p1, int8x8_t __p2) {
33729	33750	int8x8_t __ret;
33730		__ret = (int8x8_t) __builtin_neon_vtbx1_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 0);
	33751	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vtbx1_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 0));
33731	33752	return __ret;
33732	33753	}
33733	33754	#else
33734	33755	__ai __attribute__((target("neon"))) int8x8_t vtbx1_s8(int8x8_t __p0, int8x8_t __p1, int8x8_t __p2) {
33735	33756	int8x8_t __ret;
33736		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
33737		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
33738		int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
33739		__ret = (int8x8_t) __builtin_neon_vtbx1_v((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 0);
33740		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33757	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	33758	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	33759	int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	33760	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vtbx1_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 0));
	33761	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
33741	33762	return __ret;
33742	33763	}
33743	33764	#endif
...	...	@@ -33745,19 +33766,19 @@ __ai __attribute__((target("neon"))) int8x8_t vtbx1_s8(int8x8_t __p0, int8x8_t _
33745	33766	#ifdef __LITTLE_ENDIAN__
33746	33767	__ai __attribute__((target("neon"))) poly8x8_t vtbx2_p8(poly8x8_t __p0, poly8x8x2_t __p1, uint8x8_t __p2) {
33747	33768	poly8x8_t __ret;
33748		__ret = (poly8x8_t) __builtin_neon_vtbx2_v((int8x8_t)__p0, (int8x8_t)__p1.val[0], (int8x8_t)__p1.val[1], (int8x8_t)__p2, 4);
	33769	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vtbx2_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1.val[0]), __builtin_bit_cast(int8x8_t, __p1.val[1]), __builtin_bit_cast(int8x8_t, __p2), 4));
33749	33770	return __ret;
33750	33771	}
33751	33772	#else
33752	33773	__ai __attribute__((target("neon"))) poly8x8_t vtbx2_p8(poly8x8_t __p0, poly8x8x2_t __p1, uint8x8_t __p2) {
33753	33774	poly8x8_t __ret;
33754		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	33775	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
33755	33776	poly8x8x2_t __rev1;
33756		__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
33757		__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
33758		uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
33759		__ret = (poly8x8_t) __builtin_neon_vtbx2_v((int8x8_t)__rev0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev2, 4);
33760		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33777	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_64_8);
	33778	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_64_8);
	33779	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	33780	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vtbx2_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev2), 4));
	33781	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
33761	33782	return __ret;
33762	33783	}
33763	33784	#endif
...	...	@@ -33765,19 +33786,19 @@ __ai __attribute__((target("neon"))) poly8x8_t vtbx2_p8(poly8x8_t __p0, poly8x8x
33765	33786	#ifdef __LITTLE_ENDIAN__
33766	33787	__ai __attribute__((target("neon"))) uint8x8_t vtbx2_u8(uint8x8_t __p0, uint8x8x2_t __p1, uint8x8_t __p2) {
33767	33788	uint8x8_t __ret;
33768		__ret = (uint8x8_t) __builtin_neon_vtbx2_v((int8x8_t)__p0, (int8x8_t)__p1.val[0], (int8x8_t)__p1.val[1], (int8x8_t)__p2, 16);
	33789	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vtbx2_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1.val[0]), __builtin_bit_cast(int8x8_t, __p1.val[1]), __builtin_bit_cast(int8x8_t, __p2), 16));
33769	33790	return __ret;
33770	33791	}
33771	33792	#else
33772	33793	__ai __attribute__((target("neon"))) uint8x8_t vtbx2_u8(uint8x8_t __p0, uint8x8x2_t __p1, uint8x8_t __p2) {
33773	33794	uint8x8_t __ret;
33774		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	33795	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
33775	33796	uint8x8x2_t __rev1;
33776		__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
33777		__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
33778		uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
33779		__ret = (uint8x8_t) __builtin_neon_vtbx2_v((int8x8_t)__rev0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev2, 16);
33780		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33797	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_64_8);
	33798	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_64_8);
	33799	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	33800	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vtbx2_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev2), 16));
	33801	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
33781	33802	return __ret;
33782	33803	}
33783	33804	#endif
...	...	@@ -33785,19 +33806,19 @@ __ai __attribute__((target("neon"))) uint8x8_t vtbx2_u8(uint8x8_t __p0, uint8x8x
33785	33806	#ifdef __LITTLE_ENDIAN__
33786	33807	__ai __attribute__((target("neon"))) int8x8_t vtbx2_s8(int8x8_t __p0, int8x8x2_t __p1, int8x8_t __p2) {
33787	33808	int8x8_t __ret;
33788		__ret = (int8x8_t) __builtin_neon_vtbx2_v((int8x8_t)__p0, (int8x8_t)__p1.val[0], (int8x8_t)__p1.val[1], (int8x8_t)__p2, 0);
	33809	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vtbx2_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1.val[0]), __builtin_bit_cast(int8x8_t, __p1.val[1]), __builtin_bit_cast(int8x8_t, __p2), 0));
33789	33810	return __ret;
33790	33811	}
33791	33812	#else
33792	33813	__ai __attribute__((target("neon"))) int8x8_t vtbx2_s8(int8x8_t __p0, int8x8x2_t __p1, int8x8_t __p2) {
33793	33814	int8x8_t __ret;
33794		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	33815	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
33795	33816	int8x8x2_t __rev1;
33796		__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
33797		__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
33798		int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
33799		__ret = (int8x8_t) __builtin_neon_vtbx2_v((int8x8_t)__rev0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev2, 0);
33800		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33817	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_64_8);
	33818	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_64_8);
	33819	int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	33820	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vtbx2_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev2), 0));
	33821	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
33801	33822	return __ret;
33802	33823	}
33803	33824	#endif
...	...	@@ -33805,20 +33826,20 @@ __ai __attribute__((target("neon"))) int8x8_t vtbx2_s8(int8x8_t __p0, int8x8x2_t
33805	33826	#ifdef __LITTLE_ENDIAN__
33806	33827	__ai __attribute__((target("neon"))) poly8x8_t vtbx3_p8(poly8x8_t __p0, poly8x8x3_t __p1, uint8x8_t __p2) {
33807	33828	poly8x8_t __ret;
33808		__ret = (poly8x8_t) __builtin_neon_vtbx3_v((int8x8_t)__p0, (int8x8_t)__p1.val[0], (int8x8_t)__p1.val[1], (int8x8_t)__p1.val[2], (int8x8_t)__p2, 4);
	33829	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vtbx3_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1.val[0]), __builtin_bit_cast(int8x8_t, __p1.val[1]), __builtin_bit_cast(int8x8_t, __p1.val[2]), __builtin_bit_cast(int8x8_t, __p2), 4));
33809	33830	return __ret;
33810	33831	}
33811	33832	#else
33812	33833	__ai __attribute__((target("neon"))) poly8x8_t vtbx3_p8(poly8x8_t __p0, poly8x8x3_t __p1, uint8x8_t __p2) {
33813	33834	poly8x8_t __ret;
33814		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	33835	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
33815	33836	poly8x8x3_t __rev1;
33816		__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
33817		__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
33818		__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], 7, 6, 5, 4, 3, 2, 1, 0);
33819		uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
33820		__ret = (poly8x8_t) __builtin_neon_vtbx3_v((int8x8_t)__rev0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev2, 4);
33821		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33837	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_64_8);
	33838	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_64_8);
	33839	__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], __lane_reverse_64_8);
	33840	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	33841	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vtbx3_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev2), 4));
	33842	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
33822	33843	return __ret;
33823	33844	}
33824	33845	#endif
...	...	@@ -33826,20 +33847,20 @@ __ai __attribute__((target("neon"))) poly8x8_t vtbx3_p8(poly8x8_t __p0, poly8x8x
33826	33847	#ifdef __LITTLE_ENDIAN__
33827	33848	__ai __attribute__((target("neon"))) uint8x8_t vtbx3_u8(uint8x8_t __p0, uint8x8x3_t __p1, uint8x8_t __p2) {
33828	33849	uint8x8_t __ret;
33829		__ret = (uint8x8_t) __builtin_neon_vtbx3_v((int8x8_t)__p0, (int8x8_t)__p1.val[0], (int8x8_t)__p1.val[1], (int8x8_t)__p1.val[2], (int8x8_t)__p2, 16);
	33850	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vtbx3_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1.val[0]), __builtin_bit_cast(int8x8_t, __p1.val[1]), __builtin_bit_cast(int8x8_t, __p1.val[2]), __builtin_bit_cast(int8x8_t, __p2), 16));
33830	33851	return __ret;
33831	33852	}
33832	33853	#else
33833	33854	__ai __attribute__((target("neon"))) uint8x8_t vtbx3_u8(uint8x8_t __p0, uint8x8x3_t __p1, uint8x8_t __p2) {
33834	33855	uint8x8_t __ret;
33835		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	33856	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
33836	33857	uint8x8x3_t __rev1;
33837		__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
33838		__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
33839		__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], 7, 6, 5, 4, 3, 2, 1, 0);
33840		uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
33841		__ret = (uint8x8_t) __builtin_neon_vtbx3_v((int8x8_t)__rev0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev2, 16);
33842		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33858	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_64_8);
	33859	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_64_8);
	33860	__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], __lane_reverse_64_8);
	33861	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	33862	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vtbx3_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev2), 16));
	33863	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
33843	33864	return __ret;
33844	33865	}
33845	33866	#endif
...	...	@@ -33847,20 +33868,20 @@ __ai __attribute__((target("neon"))) uint8x8_t vtbx3_u8(uint8x8_t __p0, uint8x8x
33847	33868	#ifdef __LITTLE_ENDIAN__
33848	33869	__ai __attribute__((target("neon"))) int8x8_t vtbx3_s8(int8x8_t __p0, int8x8x3_t __p1, int8x8_t __p2) {
33849	33870	int8x8_t __ret;
33850		__ret = (int8x8_t) __builtin_neon_vtbx3_v((int8x8_t)__p0, (int8x8_t)__p1.val[0], (int8x8_t)__p1.val[1], (int8x8_t)__p1.val[2], (int8x8_t)__p2, 0);
	33871	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vtbx3_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1.val[0]), __builtin_bit_cast(int8x8_t, __p1.val[1]), __builtin_bit_cast(int8x8_t, __p1.val[2]), __builtin_bit_cast(int8x8_t, __p2), 0));
33851	33872	return __ret;
33852	33873	}
33853	33874	#else
33854	33875	__ai __attribute__((target("neon"))) int8x8_t vtbx3_s8(int8x8_t __p0, int8x8x3_t __p1, int8x8_t __p2) {
33855	33876	int8x8_t __ret;
33856		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	33877	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
33857	33878	int8x8x3_t __rev1;
33858		__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
33859		__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
33860		__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], 7, 6, 5, 4, 3, 2, 1, 0);
33861		int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
33862		__ret = (int8x8_t) __builtin_neon_vtbx3_v((int8x8_t)__rev0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev2, 0);
33863		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33879	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_64_8);
	33880	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_64_8);
	33881	__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], __lane_reverse_64_8);
	33882	int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	33883	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vtbx3_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev2), 0));
	33884	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
33864	33885	return __ret;
33865	33886	}
33866	33887	#endif
...	...	@@ -33868,21 +33889,21 @@ __ai __attribute__((target("neon"))) int8x8_t vtbx3_s8(int8x8_t __p0, int8x8x3_t
33868	33889	#ifdef __LITTLE_ENDIAN__
33869	33890	__ai __attribute__((target("neon"))) poly8x8_t vtbx4_p8(poly8x8_t __p0, poly8x8x4_t __p1, uint8x8_t __p2) {
33870	33891	poly8x8_t __ret;
33871		__ret = (poly8x8_t) __builtin_neon_vtbx4_v((int8x8_t)__p0, (int8x8_t)__p1.val[0], (int8x8_t)__p1.val[1], (int8x8_t)__p1.val[2], (int8x8_t)__p1.val[3], (int8x8_t)__p2, 4);
	33892	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vtbx4_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1.val[0]), __builtin_bit_cast(int8x8_t, __p1.val[1]), __builtin_bit_cast(int8x8_t, __p1.val[2]), __builtin_bit_cast(int8x8_t, __p1.val[3]), __builtin_bit_cast(int8x8_t, __p2), 4));
33872	33893	return __ret;
33873	33894	}
33874	33895	#else
33875	33896	__ai __attribute__((target("neon"))) poly8x8_t vtbx4_p8(poly8x8_t __p0, poly8x8x4_t __p1, uint8x8_t __p2) {
33876	33897	poly8x8_t __ret;
33877		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	33898	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
33878	33899	poly8x8x4_t __rev1;
33879		__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
33880		__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
33881		__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], 7, 6, 5, 4, 3, 2, 1, 0);
33882		__rev1.val[3] = __builtin_shufflevector(__p1.val[3], __p1.val[3], 7, 6, 5, 4, 3, 2, 1, 0);
33883		uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
33884		__ret = (poly8x8_t) __builtin_neon_vtbx4_v((int8x8_t)__rev0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], (int8x8_t)__rev2, 4);
33885		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33900	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_64_8);
	33901	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_64_8);
	33902	__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], __lane_reverse_64_8);
	33903	__rev1.val[3] = __builtin_shufflevector(__p1.val[3], __p1.val[3], __lane_reverse_64_8);
	33904	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	33905	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vtbx4_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), __builtin_bit_cast(int8x8_t, __rev2), 4));
	33906	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
33886	33907	return __ret;
33887	33908	}
33888	33909	#endif
...	...	@@ -33890,21 +33911,21 @@ __ai __attribute__((target("neon"))) poly8x8_t vtbx4_p8(poly8x8_t __p0, poly8x8x
33890	33911	#ifdef __LITTLE_ENDIAN__
33891	33912	__ai __attribute__((target("neon"))) uint8x8_t vtbx4_u8(uint8x8_t __p0, uint8x8x4_t __p1, uint8x8_t __p2) {
33892	33913	uint8x8_t __ret;
33893		__ret = (uint8x8_t) __builtin_neon_vtbx4_v((int8x8_t)__p0, (int8x8_t)__p1.val[0], (int8x8_t)__p1.val[1], (int8x8_t)__p1.val[2], (int8x8_t)__p1.val[3], (int8x8_t)__p2, 16);
	33914	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vtbx4_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1.val[0]), __builtin_bit_cast(int8x8_t, __p1.val[1]), __builtin_bit_cast(int8x8_t, __p1.val[2]), __builtin_bit_cast(int8x8_t, __p1.val[3]), __builtin_bit_cast(int8x8_t, __p2), 16));
33894	33915	return __ret;
33895	33916	}
33896	33917	#else
33897	33918	__ai __attribute__((target("neon"))) uint8x8_t vtbx4_u8(uint8x8_t __p0, uint8x8x4_t __p1, uint8x8_t __p2) {
33898	33919	uint8x8_t __ret;
33899		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	33920	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
33900	33921	uint8x8x4_t __rev1;
33901		__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
33902		__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
33903		__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], 7, 6, 5, 4, 3, 2, 1, 0);
33904		__rev1.val[3] = __builtin_shufflevector(__p1.val[3], __p1.val[3], 7, 6, 5, 4, 3, 2, 1, 0);
33905		uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
33906		__ret = (uint8x8_t) __builtin_neon_vtbx4_v((int8x8_t)__rev0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], (int8x8_t)__rev2, 16);
33907		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33922	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_64_8);
	33923	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_64_8);
	33924	__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], __lane_reverse_64_8);
	33925	__rev1.val[3] = __builtin_shufflevector(__p1.val[3], __p1.val[3], __lane_reverse_64_8);
	33926	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	33927	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vtbx4_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), __builtin_bit_cast(int8x8_t, __rev2), 16));
	33928	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
33908	33929	return __ret;
33909	33930	}
33910	33931	#endif
...	...	@@ -33912,21 +33933,21 @@ __ai __attribute__((target("neon"))) uint8x8_t vtbx4_u8(uint8x8_t __p0, uint8x8x
33912	33933	#ifdef __LITTLE_ENDIAN__
33913	33934	__ai __attribute__((target("neon"))) int8x8_t vtbx4_s8(int8x8_t __p0, int8x8x4_t __p1, int8x8_t __p2) {
33914	33935	int8x8_t __ret;
33915		__ret = (int8x8_t) __builtin_neon_vtbx4_v((int8x8_t)__p0, (int8x8_t)__p1.val[0], (int8x8_t)__p1.val[1], (int8x8_t)__p1.val[2], (int8x8_t)__p1.val[3], (int8x8_t)__p2, 0);
	33936	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vtbx4_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1.val[0]), __builtin_bit_cast(int8x8_t, __p1.val[1]), __builtin_bit_cast(int8x8_t, __p1.val[2]), __builtin_bit_cast(int8x8_t, __p1.val[3]), __builtin_bit_cast(int8x8_t, __p2), 0));
33916	33937	return __ret;
33917	33938	}
33918	33939	#else
33919	33940	__ai __attribute__((target("neon"))) int8x8_t vtbx4_s8(int8x8_t __p0, int8x8x4_t __p1, int8x8_t __p2) {
33920	33941	int8x8_t __ret;
33921		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	33942	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
33922	33943	int8x8x4_t __rev1;
33923		__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
33924		__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
33925		__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], 7, 6, 5, 4, 3, 2, 1, 0);
33926		__rev1.val[3] = __builtin_shufflevector(__p1.val[3], __p1.val[3], 7, 6, 5, 4, 3, 2, 1, 0);
33927		int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
33928		__ret = (int8x8_t) __builtin_neon_vtbx4_v((int8x8_t)__rev0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], (int8x8_t)__rev2, 0);
33929		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	33944	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_64_8);
	33945	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_64_8);
	33946	__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], __lane_reverse_64_8);
	33947	__rev1.val[3] = __builtin_shufflevector(__p1.val[3], __p1.val[3], __lane_reverse_64_8);
	33948	int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	33949	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vtbx4_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), __builtin_bit_cast(int8x8_t, __rev2), 0));
	33950	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
33930	33951	return __ret;
33931	33952	}
33932	33953	#endif
...	...	@@ -33934,18 +33955,18 @@ __ai __attribute__((target("neon"))) int8x8_t vtbx4_s8(int8x8_t __p0, int8x8x4_t
33934	33955	#ifdef __LITTLE_ENDIAN__
33935	33956	__ai __attribute__((target("neon"))) poly8x8x2_t vtrn_p8(poly8x8_t __p0, poly8x8_t __p1) {
33936	33957	poly8x8x2_t __ret;
33937		__builtin_neon_vtrn_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 4);
	33958	__builtin_neon_vtrn_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 4);
33938	33959	return __ret;
33939	33960	}
33940	33961	#else
33941	33962	__ai __attribute__((target("neon"))) poly8x8x2_t vtrn_p8(poly8x8_t __p0, poly8x8_t __p1) {
33942	33963	poly8x8x2_t __ret;
33943		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
33944		poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
33945		__builtin_neon_vtrn_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 4);
	33964	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	33965	poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	33966	__builtin_neon_vtrn_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 4);
33946	33967
33947		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
33948		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
	33968	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8);
	33969	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8);
33949	33970	return __ret;
33950	33971	}
33951	33972	#endif
...	...	@@ -33953,18 +33974,18 @@ __ai __attribute__((target("neon"))) poly8x8x2_t vtrn_p8(poly8x8_t __p0, poly8x8
33953	33974	#ifdef __LITTLE_ENDIAN__
33954	33975	__ai __attribute__((target("neon"))) poly16x4x2_t vtrn_p16(poly16x4_t __p0, poly16x4_t __p1) {
33955	33976	poly16x4x2_t __ret;
33956		__builtin_neon_vtrn_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 5);
	33977	__builtin_neon_vtrn_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 5);
33957	33978	return __ret;
33958	33979	}
33959	33980	#else
33960	33981	__ai __attribute__((target("neon"))) poly16x4x2_t vtrn_p16(poly16x4_t __p0, poly16x4_t __p1) {
33961	33982	poly16x4x2_t __ret;
33962		poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
33963		poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
33964		__builtin_neon_vtrn_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 5);
	33983	poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	33984	poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	33985	__builtin_neon_vtrn_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 5);
33965	33986
33966		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
33967		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
	33987	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16);
	33988	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16);
33968	33989	return __ret;
33969	33990	}
33970	33991	#endif
...	...	@@ -33972,18 +33993,18 @@ __ai __attribute__((target("neon"))) poly16x4x2_t vtrn_p16(poly16x4_t __p0, poly
33972	33993	#ifdef __LITTLE_ENDIAN__
33973	33994	__ai __attribute__((target("neon"))) poly8x16x2_t vtrnq_p8(poly8x16_t __p0, poly8x16_t __p1) {
33974	33995	poly8x16x2_t __ret;
33975		__builtin_neon_vtrnq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 36);
	33996	__builtin_neon_vtrnq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 36);
33976	33997	return __ret;
33977	33998	}
33978	33999	#else
33979	34000	__ai __attribute__((target("neon"))) poly8x16x2_t vtrnq_p8(poly8x16_t __p0, poly8x16_t __p1) {
33980	34001	poly8x16x2_t __ret;
33981		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
33982		poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
33983		__builtin_neon_vtrnq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 36);
	34002	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	34003	poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	34004	__builtin_neon_vtrnq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 36);
33984	34005
33985		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
33986		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	34006	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8);
	34007	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8);
33987	34008	return __ret;
33988	34009	}
33989	34010	#endif
...	...	@@ -33991,18 +34012,18 @@ __ai __attribute__((target("neon"))) poly8x16x2_t vtrnq_p8(poly8x16_t __p0, poly
33991	34012	#ifdef __LITTLE_ENDIAN__
33992	34013	__ai __attribute__((target("neon"))) poly16x8x2_t vtrnq_p16(poly16x8_t __p0, poly16x8_t __p1) {
33993	34014	poly16x8x2_t __ret;
33994		__builtin_neon_vtrnq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 37);
	34015	__builtin_neon_vtrnq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 37);
33995	34016	return __ret;
33996	34017	}
33997	34018	#else
33998	34019	__ai __attribute__((target("neon"))) poly16x8x2_t vtrnq_p16(poly16x8_t __p0, poly16x8_t __p1) {
33999	34020	poly16x8x2_t __ret;
34000		poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34001		poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
34002		__builtin_neon_vtrnq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 37);
	34021	poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	34022	poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	34023	__builtin_neon_vtrnq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 37);
34003	34024
34004		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
34005		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
	34025	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16);
	34026	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16);
34006	34027	return __ret;
34007	34028	}
34008	34029	#endif
...	...	@@ -34010,18 +34031,18 @@ __ai __attribute__((target("neon"))) poly16x8x2_t vtrnq_p16(poly16x8_t __p0, pol
34010	34031	#ifdef __LITTLE_ENDIAN__
34011	34032	__ai __attribute__((target("neon"))) uint8x16x2_t vtrnq_u8(uint8x16_t __p0, uint8x16_t __p1) {
34012	34033	uint8x16x2_t __ret;
34013		__builtin_neon_vtrnq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 48);
	34034	__builtin_neon_vtrnq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 48);
34014	34035	return __ret;
34015	34036	}
34016	34037	#else
34017	34038	__ai __attribute__((target("neon"))) uint8x16x2_t vtrnq_u8(uint8x16_t __p0, uint8x16_t __p1) {
34018	34039	uint8x16x2_t __ret;
34019		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
34020		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
34021		__builtin_neon_vtrnq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 48);
	34040	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	34041	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	34042	__builtin_neon_vtrnq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 48);
34022	34043
34023		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
34024		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	34044	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8);
	34045	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8);
34025	34046	return __ret;
34026	34047	}
34027	34048	#endif
...	...	@@ -34029,18 +34050,18 @@ __ai __attribute__((target("neon"))) uint8x16x2_t vtrnq_u8(uint8x16_t __p0, uint
34029	34050	#ifdef __LITTLE_ENDIAN__
34030	34051	__ai __attribute__((target("neon"))) uint32x4x2_t vtrnq_u32(uint32x4_t __p0, uint32x4_t __p1) {
34031	34052	uint32x4x2_t __ret;
34032		__builtin_neon_vtrnq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 50);
	34053	__builtin_neon_vtrnq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50);
34033	34054	return __ret;
34034	34055	}
34035	34056	#else
34036	34057	__ai __attribute__((target("neon"))) uint32x4x2_t vtrnq_u32(uint32x4_t __p0, uint32x4_t __p1) {
34037	34058	uint32x4x2_t __ret;
34038		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34039		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
34040		__builtin_neon_vtrnq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 50);
	34059	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	34060	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	34061	__builtin_neon_vtrnq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 50);
34041	34062
34042		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
34043		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
	34063	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32);
	34064	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32);
34044	34065	return __ret;
34045	34066	}
34046	34067	#endif
...	...	@@ -34048,18 +34069,18 @@ __ai __attribute__((target("neon"))) uint32x4x2_t vtrnq_u32(uint32x4_t __p0, uin
34048	34069	#ifdef __LITTLE_ENDIAN__
34049	34070	__ai __attribute__((target("neon"))) uint16x8x2_t vtrnq_u16(uint16x8_t __p0, uint16x8_t __p1) {
34050	34071	uint16x8x2_t __ret;
34051		__builtin_neon_vtrnq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 49);
	34072	__builtin_neon_vtrnq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 49);
34052	34073	return __ret;
34053	34074	}
34054	34075	#else
34055	34076	__ai __attribute__((target("neon"))) uint16x8x2_t vtrnq_u16(uint16x8_t __p0, uint16x8_t __p1) {
34056	34077	uint16x8x2_t __ret;
34057		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34058		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
34059		__builtin_neon_vtrnq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 49);
	34078	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	34079	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	34080	__builtin_neon_vtrnq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 49);
34060	34081
34061		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
34062		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
	34082	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16);
	34083	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16);
34063	34084	return __ret;
34064	34085	}
34065	34086	#endif
...	...	@@ -34067,18 +34088,18 @@ __ai __attribute__((target("neon"))) uint16x8x2_t vtrnq_u16(uint16x8_t __p0, uin
34067	34088	#ifdef __LITTLE_ENDIAN__
34068	34089	__ai __attribute__((target("neon"))) int8x16x2_t vtrnq_s8(int8x16_t __p0, int8x16_t __p1) {
34069	34090	int8x16x2_t __ret;
34070		__builtin_neon_vtrnq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 32);
	34091	__builtin_neon_vtrnq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 32);
34071	34092	return __ret;
34072	34093	}
34073	34094	#else
34074	34095	__ai __attribute__((target("neon"))) int8x16x2_t vtrnq_s8(int8x16_t __p0, int8x16_t __p1) {
34075	34096	int8x16x2_t __ret;
34076		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
34077		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
34078		__builtin_neon_vtrnq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 32);
	34097	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	34098	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	34099	__builtin_neon_vtrnq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 32);
34079	34100
34080		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
34081		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	34101	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8);
	34102	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8);
34082	34103	return __ret;
34083	34104	}
34084	34105	#endif
...	...	@@ -34086,18 +34107,18 @@ __ai __attribute__((target("neon"))) int8x16x2_t vtrnq_s8(int8x16_t __p0, int8x1
34086	34107	#ifdef __LITTLE_ENDIAN__
34087	34108	__ai __attribute__((target("neon"))) float32x4x2_t vtrnq_f32(float32x4_t __p0, float32x4_t __p1) {
34088	34109	float32x4x2_t __ret;
34089		__builtin_neon_vtrnq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 41);
	34110	__builtin_neon_vtrnq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 41);
34090	34111	return __ret;
34091	34112	}
34092	34113	#else
34093	34114	__ai __attribute__((target("neon"))) float32x4x2_t vtrnq_f32(float32x4_t __p0, float32x4_t __p1) {
34094	34115	float32x4x2_t __ret;
34095		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34096		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
34097		__builtin_neon_vtrnq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 41);
	34116	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	34117	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	34118	__builtin_neon_vtrnq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 41);
34098	34119
34099		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
34100		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
	34120	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32);
	34121	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32);
34101	34122	return __ret;
34102	34123	}
34103	34124	#endif
...	...	@@ -34105,18 +34126,18 @@ __ai __attribute__((target("neon"))) float32x4x2_t vtrnq_f32(float32x4_t __p0, f
34105	34126	#ifdef __LITTLE_ENDIAN__
34106	34127	__ai __attribute__((target("neon"))) int32x4x2_t vtrnq_s32(int32x4_t __p0, int32x4_t __p1) {
34107	34128	int32x4x2_t __ret;
34108		__builtin_neon_vtrnq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 34);
	34129	__builtin_neon_vtrnq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 34);
34109	34130	return __ret;
34110	34131	}
34111	34132	#else
34112	34133	__ai __attribute__((target("neon"))) int32x4x2_t vtrnq_s32(int32x4_t __p0, int32x4_t __p1) {
34113	34134	int32x4x2_t __ret;
34114		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34115		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
34116		__builtin_neon_vtrnq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 34);
	34135	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	34136	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	34137	__builtin_neon_vtrnq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 34);
34117	34138
34118		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
34119		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
	34139	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32);
	34140	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32);
34120	34141	return __ret;
34121	34142	}
34122	34143	#endif
...	...	@@ -34124,18 +34145,18 @@ __ai __attribute__((target("neon"))) int32x4x2_t vtrnq_s32(int32x4_t __p0, int32
34124	34145	#ifdef __LITTLE_ENDIAN__
34125	34146	__ai __attribute__((target("neon"))) int16x8x2_t vtrnq_s16(int16x8_t __p0, int16x8_t __p1) {
34126	34147	int16x8x2_t __ret;
34127		__builtin_neon_vtrnq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 33);
	34148	__builtin_neon_vtrnq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 33);
34128	34149	return __ret;
34129	34150	}
34130	34151	#else
34131	34152	__ai __attribute__((target("neon"))) int16x8x2_t vtrnq_s16(int16x8_t __p0, int16x8_t __p1) {
34132	34153	int16x8x2_t __ret;
34133		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34134		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
34135		__builtin_neon_vtrnq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 33);
	34154	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	34155	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	34156	__builtin_neon_vtrnq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 33);
34136	34157
34137		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
34138		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
	34158	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16);
	34159	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16);
34139	34160	return __ret;
34140	34161	}
34141	34162	#endif
...	...	@@ -34143,18 +34164,18 @@ __ai __attribute__((target("neon"))) int16x8x2_t vtrnq_s16(int16x8_t __p0, int16
34143	34164	#ifdef __LITTLE_ENDIAN__
34144	34165	__ai __attribute__((target("neon"))) uint8x8x2_t vtrn_u8(uint8x8_t __p0, uint8x8_t __p1) {
34145	34166	uint8x8x2_t __ret;
34146		__builtin_neon_vtrn_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 16);
	34167	__builtin_neon_vtrn_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 16);
34147	34168	return __ret;
34148	34169	}
34149	34170	#else
34150	34171	__ai __attribute__((target("neon"))) uint8x8x2_t vtrn_u8(uint8x8_t __p0, uint8x8_t __p1) {
34151	34172	uint8x8x2_t __ret;
34152		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34153		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
34154		__builtin_neon_vtrn_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 16);
	34173	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	34174	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	34175	__builtin_neon_vtrn_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 16);
34155	34176
34156		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
34157		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
	34177	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8);
	34178	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8);
34158	34179	return __ret;
34159	34180	}
34160	34181	#endif
...	...	@@ -34162,18 +34183,18 @@ __ai __attribute__((target("neon"))) uint8x8x2_t vtrn_u8(uint8x8_t __p0, uint8x8
34162	34183	#ifdef __LITTLE_ENDIAN__
34163	34184	__ai __attribute__((target("neon"))) uint32x2x2_t vtrn_u32(uint32x2_t __p0, uint32x2_t __p1) {
34164	34185	uint32x2x2_t __ret;
34165		__builtin_neon_vtrn_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 18);
	34186	__builtin_neon_vtrn_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 18);
34166	34187	return __ret;
34167	34188	}
34168	34189	#else
34169	34190	__ai __attribute__((target("neon"))) uint32x2x2_t vtrn_u32(uint32x2_t __p0, uint32x2_t __p1) {
34170	34191	uint32x2x2_t __ret;
34171		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
34172		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
34173		__builtin_neon_vtrn_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 18);
	34192	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	34193	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	34194	__builtin_neon_vtrn_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 18);
34174	34195
34175		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0);
34176		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0);
	34196	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32);
	34197	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32);
34177	34198	return __ret;
34178	34199	}
34179	34200	#endif
...	...	@@ -34181,18 +34202,18 @@ __ai __attribute__((target("neon"))) uint32x2x2_t vtrn_u32(uint32x2_t __p0, uint
34181	34202	#ifdef __LITTLE_ENDIAN__
34182	34203	__ai __attribute__((target("neon"))) uint16x4x2_t vtrn_u16(uint16x4_t __p0, uint16x4_t __p1) {
34183	34204	uint16x4x2_t __ret;
34184		__builtin_neon_vtrn_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 17);
	34205	__builtin_neon_vtrn_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 17);
34185	34206	return __ret;
34186	34207	}
34187	34208	#else
34188	34209	__ai __attribute__((target("neon"))) uint16x4x2_t vtrn_u16(uint16x4_t __p0, uint16x4_t __p1) {
34189	34210	uint16x4x2_t __ret;
34190		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34191		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
34192		__builtin_neon_vtrn_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 17);
	34211	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	34212	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	34213	__builtin_neon_vtrn_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 17);
34193	34214
34194		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
34195		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
	34215	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16);
	34216	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16);
34196	34217	return __ret;
34197	34218	}
34198	34219	#endif
...	...	@@ -34200,18 +34221,18 @@ __ai __attribute__((target("neon"))) uint16x4x2_t vtrn_u16(uint16x4_t __p0, uint
34200	34221	#ifdef __LITTLE_ENDIAN__
34201	34222	__ai __attribute__((target("neon"))) int8x8x2_t vtrn_s8(int8x8_t __p0, int8x8_t __p1) {
34202	34223	int8x8x2_t __ret;
34203		__builtin_neon_vtrn_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 0);
	34224	__builtin_neon_vtrn_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 0);
34204	34225	return __ret;
34205	34226	}
34206	34227	#else
34207	34228	__ai __attribute__((target("neon"))) int8x8x2_t vtrn_s8(int8x8_t __p0, int8x8_t __p1) {
34208	34229	int8x8x2_t __ret;
34209		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34210		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
34211		__builtin_neon_vtrn_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 0);
	34230	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	34231	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	34232	__builtin_neon_vtrn_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 0);
34212	34233
34213		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
34214		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
	34234	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8);
	34235	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8);
34215	34236	return __ret;
34216	34237	}
34217	34238	#endif
...	...	@@ -34219,18 +34240,18 @@ __ai __attribute__((target("neon"))) int8x8x2_t vtrn_s8(int8x8_t __p0, int8x8_t
34219	34240	#ifdef __LITTLE_ENDIAN__
34220	34241	__ai __attribute__((target("neon"))) float32x2x2_t vtrn_f32(float32x2_t __p0, float32x2_t __p1) {
34221	34242	float32x2x2_t __ret;
34222		__builtin_neon_vtrn_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 9);
	34243	__builtin_neon_vtrn_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 9);
34223	34244	return __ret;
34224	34245	}
34225	34246	#else
34226	34247	__ai __attribute__((target("neon"))) float32x2x2_t vtrn_f32(float32x2_t __p0, float32x2_t __p1) {
34227	34248	float32x2x2_t __ret;
34228		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
34229		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
34230		__builtin_neon_vtrn_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 9);
	34249	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	34250	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	34251	__builtin_neon_vtrn_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 9);
34231	34252
34232		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0);
34233		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0);
	34253	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32);
	34254	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32);
34234	34255	return __ret;
34235	34256	}
34236	34257	#endif
...	...	@@ -34238,18 +34259,18 @@ __ai __attribute__((target("neon"))) float32x2x2_t vtrn_f32(float32x2_t __p0, fl
34238	34259	#ifdef __LITTLE_ENDIAN__
34239	34260	__ai __attribute__((target("neon"))) int32x2x2_t vtrn_s32(int32x2_t __p0, int32x2_t __p1) {
34240	34261	int32x2x2_t __ret;
34241		__builtin_neon_vtrn_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 2);
	34262	__builtin_neon_vtrn_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 2);
34242	34263	return __ret;
34243	34264	}
34244	34265	#else
34245	34266	__ai __attribute__((target("neon"))) int32x2x2_t vtrn_s32(int32x2_t __p0, int32x2_t __p1) {
34246	34267	int32x2x2_t __ret;
34247		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
34248		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
34249		__builtin_neon_vtrn_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 2);
	34268	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	34269	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	34270	__builtin_neon_vtrn_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 2);
34250	34271
34251		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0);
34252		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0);
	34272	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32);
	34273	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32);
34253	34274	return __ret;
34254	34275	}
34255	34276	#endif
...	...	@@ -34257,18 +34278,18 @@ __ai __attribute__((target("neon"))) int32x2x2_t vtrn_s32(int32x2_t __p0, int32x
34257	34278	#ifdef __LITTLE_ENDIAN__
34258	34279	__ai __attribute__((target("neon"))) int16x4x2_t vtrn_s16(int16x4_t __p0, int16x4_t __p1) {
34259	34280	int16x4x2_t __ret;
34260		__builtin_neon_vtrn_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 1);
	34281	__builtin_neon_vtrn_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 1);
34261	34282	return __ret;
34262	34283	}
34263	34284	#else
34264	34285	__ai __attribute__((target("neon"))) int16x4x2_t vtrn_s16(int16x4_t __p0, int16x4_t __p1) {
34265	34286	int16x4x2_t __ret;
34266		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34267		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
34268		__builtin_neon_vtrn_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 1);
	34287	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	34288	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	34289	__builtin_neon_vtrn_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 1);
34269	34290
34270		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
34271		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
	34291	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16);
	34292	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16);
34272	34293	return __ret;
34273	34294	}
34274	34295	#endif
...	...	@@ -34276,18 +34297,18 @@ __ai __attribute__((target("neon"))) int16x4x2_t vtrn_s16(int16x4_t __p0, int16x
34276	34297	#ifdef __LITTLE_ENDIAN__
34277	34298	__ai __attribute__((target("neon"))) float16x8x2_t vtrnq_f16(float16x8_t __p0, float16x8_t __p1) {
34278	34299	float16x8x2_t __ret;
34279		__builtin_neon_vtrnq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 40);
	34300	__builtin_neon_vtrnq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 40);
34280	34301	return __ret;
34281	34302	}
34282	34303	#else
34283	34304	__ai __attribute__((target("neon"))) float16x8x2_t vtrnq_f16(float16x8_t __p0, float16x8_t __p1) {
34284	34305	float16x8x2_t __ret;
34285		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34286		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
34287		__builtin_neon_vtrnq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 40);
	34306	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	34307	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	34308	__builtin_neon_vtrnq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 40);
34288	34309
34289		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
34290		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
	34310	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16);
	34311	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16);
34291	34312	return __ret;
34292	34313	}
34293	34314	#endif
...	...	@@ -34295,18 +34316,18 @@ __ai __attribute__((target("neon"))) float16x8x2_t vtrnq_f16(float16x8_t __p0, f
34295	34316	#ifdef __LITTLE_ENDIAN__
34296	34317	__ai __attribute__((target("neon"))) float16x4x2_t vtrn_f16(float16x4_t __p0, float16x4_t __p1) {
34297	34318	float16x4x2_t __ret;
34298		__builtin_neon_vtrn_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 8);
	34319	__builtin_neon_vtrn_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 8);
34299	34320	return __ret;
34300	34321	}
34301	34322	#else
34302	34323	__ai __attribute__((target("neon"))) float16x4x2_t vtrn_f16(float16x4_t __p0, float16x4_t __p1) {
34303	34324	float16x4x2_t __ret;
34304		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34305		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
34306		__builtin_neon_vtrn_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 8);
	34325	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	34326	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	34327	__builtin_neon_vtrn_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 8);
34307	34328
34308		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
34309		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
	34329	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16);
	34330	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16);
34310	34331	return __ret;
34311	34332	}
34312	34333	#endif
...	...	@@ -34314,16 +34335,16 @@ __ai __attribute__((target("neon"))) float16x4x2_t vtrn_f16(float16x4_t __p0, fl
34314	34335	#ifdef __LITTLE_ENDIAN__
34315	34336	__ai __attribute__((target("neon"))) uint8x8_t vtst_p8(poly8x8_t __p0, poly8x8_t __p1) {
34316	34337	uint8x8_t __ret;
34317		__ret = (uint8x8_t) __builtin_neon_vtst_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
	34338	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vtst_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 16));
34318	34339	return __ret;
34319	34340	}
34320	34341	#else
34321	34342	__ai __attribute__((target("neon"))) uint8x8_t vtst_p8(poly8x8_t __p0, poly8x8_t __p1) {
34322	34343	uint8x8_t __ret;
34323		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34324		poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
34325		__ret = (uint8x8_t) __builtin_neon_vtst_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
34326		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	34344	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	34345	poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	34346	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vtst_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 16));
	34347	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
34327	34348	return __ret;
34328	34349	}
34329	34350	#endif
...	...	@@ -34331,16 +34352,16 @@ __ai __attribute__((target("neon"))) uint8x8_t vtst_p8(poly8x8_t __p0, poly8x8_t
34331	34352	#ifdef __LITTLE_ENDIAN__
34332	34353	__ai __attribute__((target("neon"))) uint16x4_t vtst_p16(poly16x4_t __p0, poly16x4_t __p1) {
34333	34354	uint16x4_t __ret;
34334		__ret = (uint16x4_t) __builtin_neon_vtst_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
	34355	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vtst_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 17));
34335	34356	return __ret;
34336	34357	}
34337	34358	#else
34338	34359	__ai __attribute__((target("neon"))) uint16x4_t vtst_p16(poly16x4_t __p0, poly16x4_t __p1) {
34339	34360	uint16x4_t __ret;
34340		poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34341		poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
34342		__ret = (uint16x4_t) __builtin_neon_vtst_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
34343		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	34361	poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	34362	poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	34363	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vtst_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 17));
	34364	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
34344	34365	return __ret;
34345	34366	}
34346	34367	#endif
...	...	@@ -34348,16 +34369,16 @@ __ai __attribute__((target("neon"))) uint16x4_t vtst_p16(poly16x4_t __p0, poly16
34348	34369	#ifdef __LITTLE_ENDIAN__
34349	34370	__ai __attribute__((target("neon"))) uint8x16_t vtstq_p8(poly8x16_t __p0, poly8x16_t __p1) {
34350	34371	uint8x16_t __ret;
34351		__ret = (uint8x16_t) __builtin_neon_vtstq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
	34372	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vtstq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 48));
34352	34373	return __ret;
34353	34374	}
34354	34375	#else
34355	34376	__ai __attribute__((target("neon"))) uint8x16_t vtstq_p8(poly8x16_t __p0, poly8x16_t __p1) {
34356	34377	uint8x16_t __ret;
34357		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
34358		poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
34359		__ret = (uint8x16_t) __builtin_neon_vtstq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
34360		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	34378	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	34379	poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	34380	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vtstq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 48));
	34381	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
34361	34382	return __ret;
34362	34383	}
34363	34384	#endif
...	...	@@ -34365,16 +34386,16 @@ __ai __attribute__((target("neon"))) uint8x16_t vtstq_p8(poly8x16_t __p0, poly8x
34365	34386	#ifdef __LITTLE_ENDIAN__
34366	34387	__ai __attribute__((target("neon"))) uint16x8_t vtstq_p16(poly16x8_t __p0, poly16x8_t __p1) {
34367	34388	uint16x8_t __ret;
34368		__ret = (uint16x8_t) __builtin_neon_vtstq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
	34389	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vtstq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 49));
34369	34390	return __ret;
34370	34391	}
34371	34392	#else
34372	34393	__ai __attribute__((target("neon"))) uint16x8_t vtstq_p16(poly16x8_t __p0, poly16x8_t __p1) {
34373	34394	uint16x8_t __ret;
34374		poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34375		poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
34376		__ret = (uint16x8_t) __builtin_neon_vtstq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
34377		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	34395	poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	34396	poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	34397	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vtstq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 49));
	34398	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
34378	34399	return __ret;
34379	34400	}
34380	34401	#endif
...	...	@@ -34382,16 +34403,16 @@ __ai __attribute__((target("neon"))) uint16x8_t vtstq_p16(poly16x8_t __p0, poly1
34382	34403	#ifdef __LITTLE_ENDIAN__
34383	34404	__ai __attribute__((target("neon"))) uint8x16_t vtstq_u8(uint8x16_t __p0, uint8x16_t __p1) {
34384	34405	uint8x16_t __ret;
34385		__ret = (uint8x16_t) __builtin_neon_vtstq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
	34406	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vtstq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 48));
34386	34407	return __ret;
34387	34408	}
34388	34409	#else
34389	34410	__ai __attribute__((target("neon"))) uint8x16_t vtstq_u8(uint8x16_t __p0, uint8x16_t __p1) {
34390	34411	uint8x16_t __ret;
34391		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
34392		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
34393		__ret = (uint8x16_t) __builtin_neon_vtstq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
34394		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	34412	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	34413	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	34414	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vtstq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 48));
	34415	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
34395	34416	return __ret;
34396	34417	}
34397	34418	#endif
...	...	@@ -34399,16 +34420,16 @@ __ai __attribute__((target("neon"))) uint8x16_t vtstq_u8(uint8x16_t __p0, uint8x
34399	34420	#ifdef __LITTLE_ENDIAN__
34400	34421	__ai __attribute__((target("neon"))) uint32x4_t vtstq_u32(uint32x4_t __p0, uint32x4_t __p1) {
34401	34422	uint32x4_t __ret;
34402		__ret = (uint32x4_t) __builtin_neon_vtstq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
	34423	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vtstq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50));
34403	34424	return __ret;
34404	34425	}
34405	34426	#else
34406	34427	__ai __attribute__((target("neon"))) uint32x4_t vtstq_u32(uint32x4_t __p0, uint32x4_t __p1) {
34407	34428	uint32x4_t __ret;
34408		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34409		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
34410		__ret = (uint32x4_t) __builtin_neon_vtstq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
34411		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	34429	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	34430	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	34431	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vtstq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 50));
	34432	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
34412	34433	return __ret;
34413	34434	}
34414	34435	#endif
...	...	@@ -34416,16 +34437,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vtstq_u32(uint32x4_t __p0, uint3
34416	34437	#ifdef __LITTLE_ENDIAN__
34417	34438	__ai __attribute__((target("neon"))) uint16x8_t vtstq_u16(uint16x8_t __p0, uint16x8_t __p1) {
34418	34439	uint16x8_t __ret;
34419		__ret = (uint16x8_t) __builtin_neon_vtstq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
	34440	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vtstq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 49));
34420	34441	return __ret;
34421	34442	}
34422	34443	#else
34423	34444	__ai __attribute__((target("neon"))) uint16x8_t vtstq_u16(uint16x8_t __p0, uint16x8_t __p1) {
34424	34445	uint16x8_t __ret;
34425		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34426		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
34427		__ret = (uint16x8_t) __builtin_neon_vtstq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
34428		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	34446	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	34447	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	34448	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vtstq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 49));
	34449	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
34429	34450	return __ret;
34430	34451	}
34431	34452	#endif
...	...	@@ -34433,16 +34454,16 @@ __ai __attribute__((target("neon"))) uint16x8_t vtstq_u16(uint16x8_t __p0, uint1
34433	34454	#ifdef __LITTLE_ENDIAN__
34434	34455	__ai __attribute__((target("neon"))) uint8x16_t vtstq_s8(int8x16_t __p0, int8x16_t __p1) {
34435	34456	uint8x16_t __ret;
34436		__ret = (uint8x16_t) __builtin_neon_vtstq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
	34457	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vtstq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 48));
34437	34458	return __ret;
34438	34459	}
34439	34460	#else
34440	34461	__ai __attribute__((target("neon"))) uint8x16_t vtstq_s8(int8x16_t __p0, int8x16_t __p1) {
34441	34462	uint8x16_t __ret;
34442		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
34443		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
34444		__ret = (uint8x16_t) __builtin_neon_vtstq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
34445		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	34463	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	34464	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	34465	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vtstq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 48));
	34466	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
34446	34467	return __ret;
34447	34468	}
34448	34469	#endif
...	...	@@ -34450,16 +34471,16 @@ __ai __attribute__((target("neon"))) uint8x16_t vtstq_s8(int8x16_t __p0, int8x16
34450	34471	#ifdef __LITTLE_ENDIAN__
34451	34472	__ai __attribute__((target("neon"))) uint32x4_t vtstq_s32(int32x4_t __p0, int32x4_t __p1) {
34452	34473	uint32x4_t __ret;
34453		__ret = (uint32x4_t) __builtin_neon_vtstq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
	34474	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vtstq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50));
34454	34475	return __ret;
34455	34476	}
34456	34477	#else
34457	34478	__ai __attribute__((target("neon"))) uint32x4_t vtstq_s32(int32x4_t __p0, int32x4_t __p1) {
34458	34479	uint32x4_t __ret;
34459		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34460		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
34461		__ret = (uint32x4_t) __builtin_neon_vtstq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
34462		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	34480	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	34481	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	34482	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vtstq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 50));
	34483	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
34463	34484	return __ret;
34464	34485	}
34465	34486	#endif
...	...	@@ -34467,16 +34488,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vtstq_s32(int32x4_t __p0, int32x
34467	34488	#ifdef __LITTLE_ENDIAN__
34468	34489	__ai __attribute__((target("neon"))) uint16x8_t vtstq_s16(int16x8_t __p0, int16x8_t __p1) {
34469	34490	uint16x8_t __ret;
34470		__ret = (uint16x8_t) __builtin_neon_vtstq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
	34491	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vtstq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 49));
34471	34492	return __ret;
34472	34493	}
34473	34494	#else
34474	34495	__ai __attribute__((target("neon"))) uint16x8_t vtstq_s16(int16x8_t __p0, int16x8_t __p1) {
34475	34496	uint16x8_t __ret;
34476		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34477		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
34478		__ret = (uint16x8_t) __builtin_neon_vtstq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
34479		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	34497	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	34498	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	34499	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vtstq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 49));
	34500	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
34480	34501	return __ret;
34481	34502	}
34482	34503	#endif
...	...	@@ -34484,16 +34505,16 @@ __ai __attribute__((target("neon"))) uint16x8_t vtstq_s16(int16x8_t __p0, int16x
34484	34505	#ifdef __LITTLE_ENDIAN__
34485	34506	__ai __attribute__((target("neon"))) uint8x8_t vtst_u8(uint8x8_t __p0, uint8x8_t __p1) {
34486	34507	uint8x8_t __ret;
34487		__ret = (uint8x8_t) __builtin_neon_vtst_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
	34508	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vtst_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 16));
34488	34509	return __ret;
34489	34510	}
34490	34511	#else
34491	34512	__ai __attribute__((target("neon"))) uint8x8_t vtst_u8(uint8x8_t __p0, uint8x8_t __p1) {
34492	34513	uint8x8_t __ret;
34493		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34494		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
34495		__ret = (uint8x8_t) __builtin_neon_vtst_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
34496		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	34514	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	34515	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	34516	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vtst_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 16));
	34517	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
34497	34518	return __ret;
34498	34519	}
34499	34520	#endif
...	...	@@ -34501,16 +34522,16 @@ __ai __attribute__((target("neon"))) uint8x8_t vtst_u8(uint8x8_t __p0, uint8x8_t
34501	34522	#ifdef __LITTLE_ENDIAN__
34502	34523	__ai __attribute__((target("neon"))) uint32x2_t vtst_u32(uint32x2_t __p0, uint32x2_t __p1) {
34503	34524	uint32x2_t __ret;
34504		__ret = (uint32x2_t) __builtin_neon_vtst_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
	34525	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vtst_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 18));
34505	34526	return __ret;
34506	34527	}
34507	34528	#else
34508	34529	__ai __attribute__((target("neon"))) uint32x2_t vtst_u32(uint32x2_t __p0, uint32x2_t __p1) {
34509	34530	uint32x2_t __ret;
34510		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
34511		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
34512		__ret = (uint32x2_t) __builtin_neon_vtst_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
34513		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	34531	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	34532	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	34533	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vtst_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 18));
	34534	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
34514	34535	return __ret;
34515	34536	}
34516	34537	#endif
...	...	@@ -34518,16 +34539,16 @@ __ai __attribute__((target("neon"))) uint32x2_t vtst_u32(uint32x2_t __p0, uint32
34518	34539	#ifdef __LITTLE_ENDIAN__
34519	34540	__ai __attribute__((target("neon"))) uint16x4_t vtst_u16(uint16x4_t __p0, uint16x4_t __p1) {
34520	34541	uint16x4_t __ret;
34521		__ret = (uint16x4_t) __builtin_neon_vtst_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
	34542	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vtst_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 17));
34522	34543	return __ret;
34523	34544	}
34524	34545	#else
34525	34546	__ai __attribute__((target("neon"))) uint16x4_t vtst_u16(uint16x4_t __p0, uint16x4_t __p1) {
34526	34547	uint16x4_t __ret;
34527		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34528		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
34529		__ret = (uint16x4_t) __builtin_neon_vtst_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
34530		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	34548	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	34549	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	34550	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vtst_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 17));
	34551	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
34531	34552	return __ret;
34532	34553	}
34533	34554	#endif
...	...	@@ -34535,16 +34556,16 @@ __ai __attribute__((target("neon"))) uint16x4_t vtst_u16(uint16x4_t __p0, uint16
34535	34556	#ifdef __LITTLE_ENDIAN__
34536	34557	__ai __attribute__((target("neon"))) uint8x8_t vtst_s8(int8x8_t __p0, int8x8_t __p1) {
34537	34558	uint8x8_t __ret;
34538		__ret = (uint8x8_t) __builtin_neon_vtst_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
	34559	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vtst_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 16));
34539	34560	return __ret;
34540	34561	}
34541	34562	#else
34542	34563	__ai __attribute__((target("neon"))) uint8x8_t vtst_s8(int8x8_t __p0, int8x8_t __p1) {
34543	34564	uint8x8_t __ret;
34544		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34545		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
34546		__ret = (uint8x8_t) __builtin_neon_vtst_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
34547		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	34565	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	34566	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	34567	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vtst_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 16));
	34568	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
34548	34569	return __ret;
34549	34570	}
34550	34571	#endif
...	...	@@ -34552,16 +34573,16 @@ __ai __attribute__((target("neon"))) uint8x8_t vtst_s8(int8x8_t __p0, int8x8_t _
34552	34573	#ifdef __LITTLE_ENDIAN__
34553	34574	__ai __attribute__((target("neon"))) uint32x2_t vtst_s32(int32x2_t __p0, int32x2_t __p1) {
34554	34575	uint32x2_t __ret;
34555		__ret = (uint32x2_t) __builtin_neon_vtst_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
	34576	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vtst_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 18));
34556	34577	return __ret;
34557	34578	}
34558	34579	#else
34559	34580	__ai __attribute__((target("neon"))) uint32x2_t vtst_s32(int32x2_t __p0, int32x2_t __p1) {
34560	34581	uint32x2_t __ret;
34561		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
34562		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
34563		__ret = (uint32x2_t) __builtin_neon_vtst_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
34564		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	34582	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	34583	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	34584	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vtst_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 18));
	34585	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
34565	34586	return __ret;
34566	34587	}
34567	34588	#endif
...	...	@@ -34569,16 +34590,16 @@ __ai __attribute__((target("neon"))) uint32x2_t vtst_s32(int32x2_t __p0, int32x2
34569	34590	#ifdef __LITTLE_ENDIAN__
34570	34591	__ai __attribute__((target("neon"))) uint16x4_t vtst_s16(int16x4_t __p0, int16x4_t __p1) {
34571	34592	uint16x4_t __ret;
34572		__ret = (uint16x4_t) __builtin_neon_vtst_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
	34593	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vtst_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 17));
34573	34594	return __ret;
34574	34595	}
34575	34596	#else
34576	34597	__ai __attribute__((target("neon"))) uint16x4_t vtst_s16(int16x4_t __p0, int16x4_t __p1) {
34577	34598	uint16x4_t __ret;
34578		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34579		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
34580		__ret = (uint16x4_t) __builtin_neon_vtst_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
34581		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	34599	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	34600	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	34601	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vtst_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 17));
	34602	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
34582	34603	return __ret;
34583	34604	}
34584	34605	#endif
...	...	@@ -34586,18 +34607,18 @@ __ai __attribute__((target("neon"))) uint16x4_t vtst_s16(int16x4_t __p0, int16x4
34586	34607	#ifdef __LITTLE_ENDIAN__
34587	34608	__ai __attribute__((target("neon"))) poly8x8x2_t vuzp_p8(poly8x8_t __p0, poly8x8_t __p1) {
34588	34609	poly8x8x2_t __ret;
34589		__builtin_neon_vuzp_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 4);
	34610	__builtin_neon_vuzp_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 4);
34590	34611	return __ret;
34591	34612	}
34592	34613	#else
34593	34614	__ai __attribute__((target("neon"))) poly8x8x2_t vuzp_p8(poly8x8_t __p0, poly8x8_t __p1) {
34594	34615	poly8x8x2_t __ret;
34595		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34596		poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
34597		__builtin_neon_vuzp_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 4);
	34616	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	34617	poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	34618	__builtin_neon_vuzp_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 4);
34598	34619
34599		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
34600		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
	34620	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8);
	34621	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8);
34601	34622	return __ret;
34602	34623	}
34603	34624	#endif
...	...	@@ -34605,18 +34626,18 @@ __ai __attribute__((target("neon"))) poly8x8x2_t vuzp_p8(poly8x8_t __p0, poly8x8
34605	34626	#ifdef __LITTLE_ENDIAN__
34606	34627	__ai __attribute__((target("neon"))) poly16x4x2_t vuzp_p16(poly16x4_t __p0, poly16x4_t __p1) {
34607	34628	poly16x4x2_t __ret;
34608		__builtin_neon_vuzp_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 5);
	34629	__builtin_neon_vuzp_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 5);
34609	34630	return __ret;
34610	34631	}
34611	34632	#else
34612	34633	__ai __attribute__((target("neon"))) poly16x4x2_t vuzp_p16(poly16x4_t __p0, poly16x4_t __p1) {
34613	34634	poly16x4x2_t __ret;
34614		poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34615		poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
34616		__builtin_neon_vuzp_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 5);
	34635	poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	34636	poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	34637	__builtin_neon_vuzp_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 5);
34617	34638
34618		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
34619		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
	34639	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16);
	34640	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16);
34620	34641	return __ret;
34621	34642	}
34622	34643	#endif
...	...	@@ -34624,18 +34645,18 @@ __ai __attribute__((target("neon"))) poly16x4x2_t vuzp_p16(poly16x4_t __p0, poly
34624	34645	#ifdef __LITTLE_ENDIAN__
34625	34646	__ai __attribute__((target("neon"))) poly8x16x2_t vuzpq_p8(poly8x16_t __p0, poly8x16_t __p1) {
34626	34647	poly8x16x2_t __ret;
34627		__builtin_neon_vuzpq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 36);
	34648	__builtin_neon_vuzpq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 36);
34628	34649	return __ret;
34629	34650	}
34630	34651	#else
34631	34652	__ai __attribute__((target("neon"))) poly8x16x2_t vuzpq_p8(poly8x16_t __p0, poly8x16_t __p1) {
34632	34653	poly8x16x2_t __ret;
34633		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
34634		poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
34635		__builtin_neon_vuzpq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 36);
	34654	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	34655	poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	34656	__builtin_neon_vuzpq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 36);
34636	34657
34637		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
34638		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	34658	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8);
	34659	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8);
34639	34660	return __ret;
34640	34661	}
34641	34662	#endif
...	...	@@ -34643,18 +34664,18 @@ __ai __attribute__((target("neon"))) poly8x16x2_t vuzpq_p8(poly8x16_t __p0, poly
34643	34664	#ifdef __LITTLE_ENDIAN__
34644	34665	__ai __attribute__((target("neon"))) poly16x8x2_t vuzpq_p16(poly16x8_t __p0, poly16x8_t __p1) {
34645	34666	poly16x8x2_t __ret;
34646		__builtin_neon_vuzpq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 37);
	34667	__builtin_neon_vuzpq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 37);
34647	34668	return __ret;
34648	34669	}
34649	34670	#else
34650	34671	__ai __attribute__((target("neon"))) poly16x8x2_t vuzpq_p16(poly16x8_t __p0, poly16x8_t __p1) {
34651	34672	poly16x8x2_t __ret;
34652		poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34653		poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
34654		__builtin_neon_vuzpq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 37);
	34673	poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	34674	poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	34675	__builtin_neon_vuzpq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 37);
34655	34676
34656		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
34657		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
	34677	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16);
	34678	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16);
34658	34679	return __ret;
34659	34680	}
34660	34681	#endif
...	...	@@ -34662,18 +34683,18 @@ __ai __attribute__((target("neon"))) poly16x8x2_t vuzpq_p16(poly16x8_t __p0, pol
34662	34683	#ifdef __LITTLE_ENDIAN__
34663	34684	__ai __attribute__((target("neon"))) uint8x16x2_t vuzpq_u8(uint8x16_t __p0, uint8x16_t __p1) {
34664	34685	uint8x16x2_t __ret;
34665		__builtin_neon_vuzpq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 48);
	34686	__builtin_neon_vuzpq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 48);
34666	34687	return __ret;
34667	34688	}
34668	34689	#else
34669	34690	__ai __attribute__((target("neon"))) uint8x16x2_t vuzpq_u8(uint8x16_t __p0, uint8x16_t __p1) {
34670	34691	uint8x16x2_t __ret;
34671		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
34672		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
34673		__builtin_neon_vuzpq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 48);
	34692	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	34693	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	34694	__builtin_neon_vuzpq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 48);
34674	34695
34675		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
34676		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	34696	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8);
	34697	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8);
34677	34698	return __ret;
34678	34699	}
34679	34700	#endif
...	...	@@ -34681,18 +34702,18 @@ __ai __attribute__((target("neon"))) uint8x16x2_t vuzpq_u8(uint8x16_t __p0, uint
34681	34702	#ifdef __LITTLE_ENDIAN__
34682	34703	__ai __attribute__((target("neon"))) uint32x4x2_t vuzpq_u32(uint32x4_t __p0, uint32x4_t __p1) {
34683	34704	uint32x4x2_t __ret;
34684		__builtin_neon_vuzpq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 50);
	34705	__builtin_neon_vuzpq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50);
34685	34706	return __ret;
34686	34707	}
34687	34708	#else
34688	34709	__ai __attribute__((target("neon"))) uint32x4x2_t vuzpq_u32(uint32x4_t __p0, uint32x4_t __p1) {
34689	34710	uint32x4x2_t __ret;
34690		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34691		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
34692		__builtin_neon_vuzpq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 50);
	34711	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	34712	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	34713	__builtin_neon_vuzpq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 50);
34693	34714
34694		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
34695		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
	34715	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32);
	34716	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32);
34696	34717	return __ret;
34697	34718	}
34698	34719	#endif
...	...	@@ -34700,18 +34721,18 @@ __ai __attribute__((target("neon"))) uint32x4x2_t vuzpq_u32(uint32x4_t __p0, uin
34700	34721	#ifdef __LITTLE_ENDIAN__
34701	34722	__ai __attribute__((target("neon"))) uint16x8x2_t vuzpq_u16(uint16x8_t __p0, uint16x8_t __p1) {
34702	34723	uint16x8x2_t __ret;
34703		__builtin_neon_vuzpq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 49);
	34724	__builtin_neon_vuzpq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 49);
34704	34725	return __ret;
34705	34726	}
34706	34727	#else
34707	34728	__ai __attribute__((target("neon"))) uint16x8x2_t vuzpq_u16(uint16x8_t __p0, uint16x8_t __p1) {
34708	34729	uint16x8x2_t __ret;
34709		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34710		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
34711		__builtin_neon_vuzpq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 49);
	34730	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	34731	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	34732	__builtin_neon_vuzpq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 49);
34712	34733
34713		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
34714		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
	34734	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16);
	34735	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16);
34715	34736	return __ret;
34716	34737	}
34717	34738	#endif
...	...	@@ -34719,18 +34740,18 @@ __ai __attribute__((target("neon"))) uint16x8x2_t vuzpq_u16(uint16x8_t __p0, uin
34719	34740	#ifdef __LITTLE_ENDIAN__
34720	34741	__ai __attribute__((target("neon"))) int8x16x2_t vuzpq_s8(int8x16_t __p0, int8x16_t __p1) {
34721	34742	int8x16x2_t __ret;
34722		__builtin_neon_vuzpq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 32);
	34743	__builtin_neon_vuzpq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 32);
34723	34744	return __ret;
34724	34745	}
34725	34746	#else
34726	34747	__ai __attribute__((target("neon"))) int8x16x2_t vuzpq_s8(int8x16_t __p0, int8x16_t __p1) {
34727	34748	int8x16x2_t __ret;
34728		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
34729		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
34730		__builtin_neon_vuzpq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 32);
	34749	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	34750	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	34751	__builtin_neon_vuzpq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 32);
34731	34752
34732		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
34733		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	34753	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8);
	34754	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8);
34734	34755	return __ret;
34735	34756	}
34736	34757	#endif
...	...	@@ -34738,18 +34759,18 @@ __ai __attribute__((target("neon"))) int8x16x2_t vuzpq_s8(int8x16_t __p0, int8x1
34738	34759	#ifdef __LITTLE_ENDIAN__
34739	34760	__ai __attribute__((target("neon"))) float32x4x2_t vuzpq_f32(float32x4_t __p0, float32x4_t __p1) {
34740	34761	float32x4x2_t __ret;
34741		__builtin_neon_vuzpq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 41);
	34762	__builtin_neon_vuzpq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 41);
34742	34763	return __ret;
34743	34764	}
34744	34765	#else
34745	34766	__ai __attribute__((target("neon"))) float32x4x2_t vuzpq_f32(float32x4_t __p0, float32x4_t __p1) {
34746	34767	float32x4x2_t __ret;
34747		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34748		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
34749		__builtin_neon_vuzpq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 41);
	34768	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	34769	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	34770	__builtin_neon_vuzpq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 41);
34750	34771
34751		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
34752		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
	34772	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32);
	34773	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32);
34753	34774	return __ret;
34754	34775	}
34755	34776	#endif
...	...	@@ -34757,18 +34778,18 @@ __ai __attribute__((target("neon"))) float32x4x2_t vuzpq_f32(float32x4_t __p0, f
34757	34778	#ifdef __LITTLE_ENDIAN__
34758	34779	__ai __attribute__((target("neon"))) int32x4x2_t vuzpq_s32(int32x4_t __p0, int32x4_t __p1) {
34759	34780	int32x4x2_t __ret;
34760		__builtin_neon_vuzpq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 34);
	34781	__builtin_neon_vuzpq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 34);
34761	34782	return __ret;
34762	34783	}
34763	34784	#else
34764	34785	__ai __attribute__((target("neon"))) int32x4x2_t vuzpq_s32(int32x4_t __p0, int32x4_t __p1) {
34765	34786	int32x4x2_t __ret;
34766		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34767		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
34768		__builtin_neon_vuzpq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 34);
	34787	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	34788	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	34789	__builtin_neon_vuzpq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 34);
34769	34790
34770		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
34771		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
	34791	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32);
	34792	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32);
34772	34793	return __ret;
34773	34794	}
34774	34795	#endif
...	...	@@ -34776,18 +34797,18 @@ __ai __attribute__((target("neon"))) int32x4x2_t vuzpq_s32(int32x4_t __p0, int32
34776	34797	#ifdef __LITTLE_ENDIAN__
34777	34798	__ai __attribute__((target("neon"))) int16x8x2_t vuzpq_s16(int16x8_t __p0, int16x8_t __p1) {
34778	34799	int16x8x2_t __ret;
34779		__builtin_neon_vuzpq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 33);
	34800	__builtin_neon_vuzpq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 33);
34780	34801	return __ret;
34781	34802	}
34782	34803	#else
34783	34804	__ai __attribute__((target("neon"))) int16x8x2_t vuzpq_s16(int16x8_t __p0, int16x8_t __p1) {
34784	34805	int16x8x2_t __ret;
34785		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34786		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
34787		__builtin_neon_vuzpq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 33);
	34806	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	34807	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	34808	__builtin_neon_vuzpq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 33);
34788	34809
34789		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
34790		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
	34810	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16);
	34811	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16);
34791	34812	return __ret;
34792	34813	}
34793	34814	#endif
...	...	@@ -34795,18 +34816,18 @@ __ai __attribute__((target("neon"))) int16x8x2_t vuzpq_s16(int16x8_t __p0, int16
34795	34816	#ifdef __LITTLE_ENDIAN__
34796	34817	__ai __attribute__((target("neon"))) uint8x8x2_t vuzp_u8(uint8x8_t __p0, uint8x8_t __p1) {
34797	34818	uint8x8x2_t __ret;
34798		__builtin_neon_vuzp_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 16);
	34819	__builtin_neon_vuzp_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 16);
34799	34820	return __ret;
34800	34821	}
34801	34822	#else
34802	34823	__ai __attribute__((target("neon"))) uint8x8x2_t vuzp_u8(uint8x8_t __p0, uint8x8_t __p1) {
34803	34824	uint8x8x2_t __ret;
34804		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34805		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
34806		__builtin_neon_vuzp_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 16);
	34825	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	34826	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	34827	__builtin_neon_vuzp_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 16);
34807	34828
34808		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
34809		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
	34829	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8);
	34830	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8);
34810	34831	return __ret;
34811	34832	}
34812	34833	#endif
...	...	@@ -34814,18 +34835,18 @@ __ai __attribute__((target("neon"))) uint8x8x2_t vuzp_u8(uint8x8_t __p0, uint8x8
34814	34835	#ifdef __LITTLE_ENDIAN__
34815	34836	__ai __attribute__((target("neon"))) uint32x2x2_t vuzp_u32(uint32x2_t __p0, uint32x2_t __p1) {
34816	34837	uint32x2x2_t __ret;
34817		__builtin_neon_vuzp_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 18);
	34838	__builtin_neon_vuzp_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 18);
34818	34839	return __ret;
34819	34840	}
34820	34841	#else
34821	34842	__ai __attribute__((target("neon"))) uint32x2x2_t vuzp_u32(uint32x2_t __p0, uint32x2_t __p1) {
34822	34843	uint32x2x2_t __ret;
34823		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
34824		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
34825		__builtin_neon_vuzp_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 18);
	34844	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	34845	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	34846	__builtin_neon_vuzp_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 18);
34826	34847
34827		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0);
34828		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0);
	34848	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32);
	34849	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32);
34829	34850	return __ret;
34830	34851	}
34831	34852	#endif
...	...	@@ -34833,18 +34854,18 @@ __ai __attribute__((target("neon"))) uint32x2x2_t vuzp_u32(uint32x2_t __p0, uint
34833	34854	#ifdef __LITTLE_ENDIAN__
34834	34855	__ai __attribute__((target("neon"))) uint16x4x2_t vuzp_u16(uint16x4_t __p0, uint16x4_t __p1) {
34835	34856	uint16x4x2_t __ret;
34836		__builtin_neon_vuzp_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 17);
	34857	__builtin_neon_vuzp_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 17);
34837	34858	return __ret;
34838	34859	}
34839	34860	#else
34840	34861	__ai __attribute__((target("neon"))) uint16x4x2_t vuzp_u16(uint16x4_t __p0, uint16x4_t __p1) {
34841	34862	uint16x4x2_t __ret;
34842		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34843		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
34844		__builtin_neon_vuzp_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 17);
	34863	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	34864	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	34865	__builtin_neon_vuzp_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 17);
34845	34866
34846		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
34847		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
	34867	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16);
	34868	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16);
34848	34869	return __ret;
34849	34870	}
34850	34871	#endif
...	...	@@ -34852,18 +34873,18 @@ __ai __attribute__((target("neon"))) uint16x4x2_t vuzp_u16(uint16x4_t __p0, uint
34852	34873	#ifdef __LITTLE_ENDIAN__
34853	34874	__ai __attribute__((target("neon"))) int8x8x2_t vuzp_s8(int8x8_t __p0, int8x8_t __p1) {
34854	34875	int8x8x2_t __ret;
34855		__builtin_neon_vuzp_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 0);
	34876	__builtin_neon_vuzp_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 0);
34856	34877	return __ret;
34857	34878	}
34858	34879	#else
34859	34880	__ai __attribute__((target("neon"))) int8x8x2_t vuzp_s8(int8x8_t __p0, int8x8_t __p1) {
34860	34881	int8x8x2_t __ret;
34861		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34862		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
34863		__builtin_neon_vuzp_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 0);
	34882	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	34883	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	34884	__builtin_neon_vuzp_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 0);
34864	34885
34865		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
34866		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
	34886	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8);
	34887	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8);
34867	34888	return __ret;
34868	34889	}
34869	34890	#endif
...	...	@@ -34871,18 +34892,18 @@ __ai __attribute__((target("neon"))) int8x8x2_t vuzp_s8(int8x8_t __p0, int8x8_t
34871	34892	#ifdef __LITTLE_ENDIAN__
34872	34893	__ai __attribute__((target("neon"))) float32x2x2_t vuzp_f32(float32x2_t __p0, float32x2_t __p1) {
34873	34894	float32x2x2_t __ret;
34874		__builtin_neon_vuzp_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 9);
	34895	__builtin_neon_vuzp_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 9);
34875	34896	return __ret;
34876	34897	}
34877	34898	#else
34878	34899	__ai __attribute__((target("neon"))) float32x2x2_t vuzp_f32(float32x2_t __p0, float32x2_t __p1) {
34879	34900	float32x2x2_t __ret;
34880		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
34881		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
34882		__builtin_neon_vuzp_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 9);
	34901	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	34902	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	34903	__builtin_neon_vuzp_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 9);
34883	34904
34884		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0);
34885		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0);
	34905	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32);
	34906	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32);
34886	34907	return __ret;
34887	34908	}
34888	34909	#endif
...	...	@@ -34890,18 +34911,18 @@ __ai __attribute__((target("neon"))) float32x2x2_t vuzp_f32(float32x2_t __p0, fl
34890	34911	#ifdef __LITTLE_ENDIAN__
34891	34912	__ai __attribute__((target("neon"))) int32x2x2_t vuzp_s32(int32x2_t __p0, int32x2_t __p1) {
34892	34913	int32x2x2_t __ret;
34893		__builtin_neon_vuzp_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 2);
	34914	__builtin_neon_vuzp_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 2);
34894	34915	return __ret;
34895	34916	}
34896	34917	#else
34897	34918	__ai __attribute__((target("neon"))) int32x2x2_t vuzp_s32(int32x2_t __p0, int32x2_t __p1) {
34898	34919	int32x2x2_t __ret;
34899		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
34900		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
34901		__builtin_neon_vuzp_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 2);
	34920	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	34921	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	34922	__builtin_neon_vuzp_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 2);
34902	34923
34903		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0);
34904		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0);
	34924	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32);
	34925	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32);
34905	34926	return __ret;
34906	34927	}
34907	34928	#endif
...	...	@@ -34909,18 +34930,18 @@ __ai __attribute__((target("neon"))) int32x2x2_t vuzp_s32(int32x2_t __p0, int32x
34909	34930	#ifdef __LITTLE_ENDIAN__
34910	34931	__ai __attribute__((target("neon"))) int16x4x2_t vuzp_s16(int16x4_t __p0, int16x4_t __p1) {
34911	34932	int16x4x2_t __ret;
34912		__builtin_neon_vuzp_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 1);
	34933	__builtin_neon_vuzp_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 1);
34913	34934	return __ret;
34914	34935	}
34915	34936	#else
34916	34937	__ai __attribute__((target("neon"))) int16x4x2_t vuzp_s16(int16x4_t __p0, int16x4_t __p1) {
34917	34938	int16x4x2_t __ret;
34918		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34919		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
34920		__builtin_neon_vuzp_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 1);
	34939	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	34940	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	34941	__builtin_neon_vuzp_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 1);
34921	34942
34922		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
34923		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
	34943	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16);
	34944	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16);
34924	34945	return __ret;
34925	34946	}
34926	34947	#endif
...	...	@@ -34928,18 +34949,18 @@ __ai __attribute__((target("neon"))) int16x4x2_t vuzp_s16(int16x4_t __p0, int16x
34928	34949	#ifdef __LITTLE_ENDIAN__
34929	34950	__ai __attribute__((target("neon"))) float16x8x2_t vuzpq_f16(float16x8_t __p0, float16x8_t __p1) {
34930	34951	float16x8x2_t __ret;
34931		__builtin_neon_vuzpq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 40);
	34952	__builtin_neon_vuzpq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 40);
34932	34953	return __ret;
34933	34954	}
34934	34955	#else
34935	34956	__ai __attribute__((target("neon"))) float16x8x2_t vuzpq_f16(float16x8_t __p0, float16x8_t __p1) {
34936	34957	float16x8x2_t __ret;
34937		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34938		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
34939		__builtin_neon_vuzpq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 40);
	34958	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	34959	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	34960	__builtin_neon_vuzpq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 40);
34940	34961
34941		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
34942		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
	34962	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16);
	34963	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16);
34943	34964	return __ret;
34944	34965	}
34945	34966	#endif
...	...	@@ -34947,18 +34968,18 @@ __ai __attribute__((target("neon"))) float16x8x2_t vuzpq_f16(float16x8_t __p0, f
34947	34968	#ifdef __LITTLE_ENDIAN__
34948	34969	__ai __attribute__((target("neon"))) float16x4x2_t vuzp_f16(float16x4_t __p0, float16x4_t __p1) {
34949	34970	float16x4x2_t __ret;
34950		__builtin_neon_vuzp_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 8);
	34971	__builtin_neon_vuzp_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 8);
34951	34972	return __ret;
34952	34973	}
34953	34974	#else
34954	34975	__ai __attribute__((target("neon"))) float16x4x2_t vuzp_f16(float16x4_t __p0, float16x4_t __p1) {
34955	34976	float16x4x2_t __ret;
34956		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34957		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
34958		__builtin_neon_vuzp_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 8);
	34977	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	34978	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	34979	__builtin_neon_vuzp_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 8);
34959	34980
34960		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
34961		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
	34981	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16);
	34982	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16);
34962	34983	return __ret;
34963	34984	}
34964	34985	#endif
...	...	@@ -34966,18 +34987,18 @@ __ai __attribute__((target("neon"))) float16x4x2_t vuzp_f16(float16x4_t __p0, fl
34966	34987	#ifdef __LITTLE_ENDIAN__
34967	34988	__ai __attribute__((target("neon"))) poly8x8x2_t vzip_p8(poly8x8_t __p0, poly8x8_t __p1) {
34968	34989	poly8x8x2_t __ret;
34969		__builtin_neon_vzip_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 4);
	34990	__builtin_neon_vzip_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 4);
34970	34991	return __ret;
34971	34992	}
34972	34993	#else
34973	34994	__ai __attribute__((target("neon"))) poly8x8x2_t vzip_p8(poly8x8_t __p0, poly8x8_t __p1) {
34974	34995	poly8x8x2_t __ret;
34975		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34976		poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
34977		__builtin_neon_vzip_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 4);
	34996	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	34997	poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	34998	__builtin_neon_vzip_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 4);
34978	34999
34979		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
34980		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
	35000	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8);
	35001	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8);
34981	35002	return __ret;
34982	35003	}
34983	35004	#endif
...	...	@@ -34985,18 +35006,18 @@ __ai __attribute__((target("neon"))) poly8x8x2_t vzip_p8(poly8x8_t __p0, poly8x8
34985	35006	#ifdef __LITTLE_ENDIAN__
34986	35007	__ai __attribute__((target("neon"))) poly16x4x2_t vzip_p16(poly16x4_t __p0, poly16x4_t __p1) {
34987	35008	poly16x4x2_t __ret;
34988		__builtin_neon_vzip_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 5);
	35009	__builtin_neon_vzip_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 5);
34989	35010	return __ret;
34990	35011	}
34991	35012	#else
34992	35013	__ai __attribute__((target("neon"))) poly16x4x2_t vzip_p16(poly16x4_t __p0, poly16x4_t __p1) {
34993	35014	poly16x4x2_t __ret;
34994		poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34995		poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
34996		__builtin_neon_vzip_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 5);
	35015	poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	35016	poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	35017	__builtin_neon_vzip_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 5);
34997	35018
34998		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
34999		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
	35019	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16);
	35020	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16);
35000	35021	return __ret;
35001	35022	}
35002	35023	#endif
...	...	@@ -35004,18 +35025,18 @@ __ai __attribute__((target("neon"))) poly16x4x2_t vzip_p16(poly16x4_t __p0, poly
35004	35025	#ifdef __LITTLE_ENDIAN__
35005	35026	__ai __attribute__((target("neon"))) poly8x16x2_t vzipq_p8(poly8x16_t __p0, poly8x16_t __p1) {
35006	35027	poly8x16x2_t __ret;
35007		__builtin_neon_vzipq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 36);
	35028	__builtin_neon_vzipq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 36);
35008	35029	return __ret;
35009	35030	}
35010	35031	#else
35011	35032	__ai __attribute__((target("neon"))) poly8x16x2_t vzipq_p8(poly8x16_t __p0, poly8x16_t __p1) {
35012	35033	poly8x16x2_t __ret;
35013		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
35014		poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
35015		__builtin_neon_vzipq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 36);
	35034	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	35035	poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	35036	__builtin_neon_vzipq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 36);
35016	35037
35017		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
35018		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	35038	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8);
	35039	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8);
35019	35040	return __ret;
35020	35041	}
35021	35042	#endif
...	...	@@ -35023,18 +35044,18 @@ __ai __attribute__((target("neon"))) poly8x16x2_t vzipq_p8(poly8x16_t __p0, poly
35023	35044	#ifdef __LITTLE_ENDIAN__
35024	35045	__ai __attribute__((target("neon"))) poly16x8x2_t vzipq_p16(poly16x8_t __p0, poly16x8_t __p1) {
35025	35046	poly16x8x2_t __ret;
35026		__builtin_neon_vzipq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 37);
	35047	__builtin_neon_vzipq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 37);
35027	35048	return __ret;
35028	35049	}
35029	35050	#else
35030	35051	__ai __attribute__((target("neon"))) poly16x8x2_t vzipq_p16(poly16x8_t __p0, poly16x8_t __p1) {
35031	35052	poly16x8x2_t __ret;
35032		poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
35033		poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
35034		__builtin_neon_vzipq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 37);
	35053	poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	35054	poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	35055	__builtin_neon_vzipq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 37);
35035	35056
35036		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
35037		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
	35057	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16);
	35058	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16);
35038	35059	return __ret;
35039	35060	}
35040	35061	#endif
...	...	@@ -35042,18 +35063,18 @@ __ai __attribute__((target("neon"))) poly16x8x2_t vzipq_p16(poly16x8_t __p0, pol
35042	35063	#ifdef __LITTLE_ENDIAN__
35043	35064	__ai __attribute__((target("neon"))) uint8x16x2_t vzipq_u8(uint8x16_t __p0, uint8x16_t __p1) {
35044	35065	uint8x16x2_t __ret;
35045		__builtin_neon_vzipq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 48);
	35066	__builtin_neon_vzipq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 48);
35046	35067	return __ret;
35047	35068	}
35048	35069	#else
35049	35070	__ai __attribute__((target("neon"))) uint8x16x2_t vzipq_u8(uint8x16_t __p0, uint8x16_t __p1) {
35050	35071	uint8x16x2_t __ret;
35051		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
35052		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
35053		__builtin_neon_vzipq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 48);
	35072	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	35073	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	35074	__builtin_neon_vzipq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 48);
35054	35075
35055		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
35056		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	35076	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8);
	35077	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8);
35057	35078	return __ret;
35058	35079	}
35059	35080	#endif
...	...	@@ -35061,18 +35082,18 @@ __ai __attribute__((target("neon"))) uint8x16x2_t vzipq_u8(uint8x16_t __p0, uint
35061	35082	#ifdef __LITTLE_ENDIAN__
35062	35083	__ai __attribute__((target("neon"))) uint32x4x2_t vzipq_u32(uint32x4_t __p0, uint32x4_t __p1) {
35063	35084	uint32x4x2_t __ret;
35064		__builtin_neon_vzipq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 50);
	35085	__builtin_neon_vzipq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50);
35065	35086	return __ret;
35066	35087	}
35067	35088	#else
35068	35089	__ai __attribute__((target("neon"))) uint32x4x2_t vzipq_u32(uint32x4_t __p0, uint32x4_t __p1) {
35069	35090	uint32x4x2_t __ret;
35070		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35071		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
35072		__builtin_neon_vzipq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 50);
	35091	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	35092	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	35093	__builtin_neon_vzipq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 50);
35073	35094
35074		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
35075		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
	35095	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32);
	35096	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32);
35076	35097	return __ret;
35077	35098	}
35078	35099	#endif
...	...	@@ -35080,18 +35101,18 @@ __ai __attribute__((target("neon"))) uint32x4x2_t vzipq_u32(uint32x4_t __p0, uin
35080	35101	#ifdef __LITTLE_ENDIAN__
35081	35102	__ai __attribute__((target("neon"))) uint16x8x2_t vzipq_u16(uint16x8_t __p0, uint16x8_t __p1) {
35082	35103	uint16x8x2_t __ret;
35083		__builtin_neon_vzipq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 49);
	35104	__builtin_neon_vzipq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 49);
35084	35105	return __ret;
35085	35106	}
35086	35107	#else
35087	35108	__ai __attribute__((target("neon"))) uint16x8x2_t vzipq_u16(uint16x8_t __p0, uint16x8_t __p1) {
35088	35109	uint16x8x2_t __ret;
35089		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
35090		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
35091		__builtin_neon_vzipq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 49);
	35110	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	35111	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	35112	__builtin_neon_vzipq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 49);
35092	35113
35093		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
35094		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
	35114	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16);
	35115	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16);
35095	35116	return __ret;
35096	35117	}
35097	35118	#endif
...	...	@@ -35099,18 +35120,18 @@ __ai __attribute__((target("neon"))) uint16x8x2_t vzipq_u16(uint16x8_t __p0, uin
35099	35120	#ifdef __LITTLE_ENDIAN__
35100	35121	__ai __attribute__((target("neon"))) int8x16x2_t vzipq_s8(int8x16_t __p0, int8x16_t __p1) {
35101	35122	int8x16x2_t __ret;
35102		__builtin_neon_vzipq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 32);
	35123	__builtin_neon_vzipq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 32);
35103	35124	return __ret;
35104	35125	}
35105	35126	#else
35106	35127	__ai __attribute__((target("neon"))) int8x16x2_t vzipq_s8(int8x16_t __p0, int8x16_t __p1) {
35107	35128	int8x16x2_t __ret;
35108		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
35109		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
35110		__builtin_neon_vzipq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 32);
	35129	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	35130	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	35131	__builtin_neon_vzipq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 32);
35111	35132
35112		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
35113		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	35133	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8);
	35134	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8);
35114	35135	return __ret;
35115	35136	}
35116	35137	#endif
...	...	@@ -35118,18 +35139,18 @@ __ai __attribute__((target("neon"))) int8x16x2_t vzipq_s8(int8x16_t __p0, int8x1
35118	35139	#ifdef __LITTLE_ENDIAN__
35119	35140	__ai __attribute__((target("neon"))) float32x4x2_t vzipq_f32(float32x4_t __p0, float32x4_t __p1) {
35120	35141	float32x4x2_t __ret;
35121		__builtin_neon_vzipq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 41);
	35142	__builtin_neon_vzipq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 41);
35122	35143	return __ret;
35123	35144	}
35124	35145	#else
35125	35146	__ai __attribute__((target("neon"))) float32x4x2_t vzipq_f32(float32x4_t __p0, float32x4_t __p1) {
35126	35147	float32x4x2_t __ret;
35127		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35128		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
35129		__builtin_neon_vzipq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 41);
	35148	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	35149	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	35150	__builtin_neon_vzipq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 41);
35130	35151
35131		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
35132		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
	35152	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32);
	35153	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32);
35133	35154	return __ret;
35134	35155	}
35135	35156	#endif
...	...	@@ -35137,18 +35158,18 @@ __ai __attribute__((target("neon"))) float32x4x2_t vzipq_f32(float32x4_t __p0, f
35137	35158	#ifdef __LITTLE_ENDIAN__
35138	35159	__ai __attribute__((target("neon"))) int32x4x2_t vzipq_s32(int32x4_t __p0, int32x4_t __p1) {
35139	35160	int32x4x2_t __ret;
35140		__builtin_neon_vzipq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 34);
	35161	__builtin_neon_vzipq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 34);
35141	35162	return __ret;
35142	35163	}
35143	35164	#else
35144	35165	__ai __attribute__((target("neon"))) int32x4x2_t vzipq_s32(int32x4_t __p0, int32x4_t __p1) {
35145	35166	int32x4x2_t __ret;
35146		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35147		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
35148		__builtin_neon_vzipq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 34);
	35167	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	35168	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	35169	__builtin_neon_vzipq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 34);
35149	35170
35150		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
35151		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
	35171	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_32);
	35172	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_32);
35152	35173	return __ret;
35153	35174	}
35154	35175	#endif
...	...	@@ -35156,18 +35177,18 @@ __ai __attribute__((target("neon"))) int32x4x2_t vzipq_s32(int32x4_t __p0, int32
35156	35177	#ifdef __LITTLE_ENDIAN__
35157	35178	__ai __attribute__((target("neon"))) int16x8x2_t vzipq_s16(int16x8_t __p0, int16x8_t __p1) {
35158	35179	int16x8x2_t __ret;
35159		__builtin_neon_vzipq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 33);
	35180	__builtin_neon_vzipq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 33);
35160	35181	return __ret;
35161	35182	}
35162	35183	#else
35163	35184	__ai __attribute__((target("neon"))) int16x8x2_t vzipq_s16(int16x8_t __p0, int16x8_t __p1) {
35164	35185	int16x8x2_t __ret;
35165		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
35166		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
35167		__builtin_neon_vzipq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 33);
	35186	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	35187	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	35188	__builtin_neon_vzipq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 33);
35168	35189
35169		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
35170		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
	35190	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16);
	35191	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16);
35171	35192	return __ret;
35172	35193	}
35173	35194	#endif
...	...	@@ -35175,18 +35196,18 @@ __ai __attribute__((target("neon"))) int16x8x2_t vzipq_s16(int16x8_t __p0, int16
35175	35196	#ifdef __LITTLE_ENDIAN__
35176	35197	__ai __attribute__((target("neon"))) uint8x8x2_t vzip_u8(uint8x8_t __p0, uint8x8_t __p1) {
35177	35198	uint8x8x2_t __ret;
35178		__builtin_neon_vzip_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 16);
	35199	__builtin_neon_vzip_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 16);
35179	35200	return __ret;
35180	35201	}
35181	35202	#else
35182	35203	__ai __attribute__((target("neon"))) uint8x8x2_t vzip_u8(uint8x8_t __p0, uint8x8_t __p1) {
35183	35204	uint8x8x2_t __ret;
35184		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
35185		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
35186		__builtin_neon_vzip_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 16);
	35205	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	35206	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	35207	__builtin_neon_vzip_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 16);
35187	35208
35188		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
35189		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
	35209	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8);
	35210	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8);
35190	35211	return __ret;
35191	35212	}
35192	35213	#endif
...	...	@@ -35194,18 +35215,18 @@ __ai __attribute__((target("neon"))) uint8x8x2_t vzip_u8(uint8x8_t __p0, uint8x8
35194	35215	#ifdef __LITTLE_ENDIAN__
35195	35216	__ai __attribute__((target("neon"))) uint32x2x2_t vzip_u32(uint32x2_t __p0, uint32x2_t __p1) {
35196	35217	uint32x2x2_t __ret;
35197		__builtin_neon_vzip_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 18);
	35218	__builtin_neon_vzip_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 18);
35198	35219	return __ret;
35199	35220	}
35200	35221	#else
35201	35222	__ai __attribute__((target("neon"))) uint32x2x2_t vzip_u32(uint32x2_t __p0, uint32x2_t __p1) {
35202	35223	uint32x2x2_t __ret;
35203		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
35204		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
35205		__builtin_neon_vzip_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 18);
	35224	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	35225	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	35226	__builtin_neon_vzip_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 18);
35206	35227
35207		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0);
35208		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0);
	35228	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32);
	35229	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32);
35209	35230	return __ret;
35210	35231	}
35211	35232	#endif
...	...	@@ -35213,18 +35234,18 @@ __ai __attribute__((target("neon"))) uint32x2x2_t vzip_u32(uint32x2_t __p0, uint
35213	35234	#ifdef __LITTLE_ENDIAN__
35214	35235	__ai __attribute__((target("neon"))) uint16x4x2_t vzip_u16(uint16x4_t __p0, uint16x4_t __p1) {
35215	35236	uint16x4x2_t __ret;
35216		__builtin_neon_vzip_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 17);
	35237	__builtin_neon_vzip_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 17);
35217	35238	return __ret;
35218	35239	}
35219	35240	#else
35220	35241	__ai __attribute__((target("neon"))) uint16x4x2_t vzip_u16(uint16x4_t __p0, uint16x4_t __p1) {
35221	35242	uint16x4x2_t __ret;
35222		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35223		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
35224		__builtin_neon_vzip_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 17);
	35243	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	35244	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	35245	__builtin_neon_vzip_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 17);
35225	35246
35226		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
35227		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
	35247	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16);
	35248	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16);
35228	35249	return __ret;
35229	35250	}
35230	35251	#endif
...	...	@@ -35232,18 +35253,18 @@ __ai __attribute__((target("neon"))) uint16x4x2_t vzip_u16(uint16x4_t __p0, uint
35232	35253	#ifdef __LITTLE_ENDIAN__
35233	35254	__ai __attribute__((target("neon"))) int8x8x2_t vzip_s8(int8x8_t __p0, int8x8_t __p1) {
35234	35255	int8x8x2_t __ret;
35235		__builtin_neon_vzip_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 0);
	35256	__builtin_neon_vzip_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 0);
35236	35257	return __ret;
35237	35258	}
35238	35259	#else
35239	35260	__ai __attribute__((target("neon"))) int8x8x2_t vzip_s8(int8x8_t __p0, int8x8_t __p1) {
35240	35261	int8x8x2_t __ret;
35241		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
35242		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
35243		__builtin_neon_vzip_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 0);
	35262	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	35263	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	35264	__builtin_neon_vzip_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 0);
35244	35265
35245		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
35246		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
	35266	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8);
	35267	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8);
35247	35268	return __ret;
35248	35269	}
35249	35270	#endif
...	...	@@ -35251,18 +35272,18 @@ __ai __attribute__((target("neon"))) int8x8x2_t vzip_s8(int8x8_t __p0, int8x8_t
35251	35272	#ifdef __LITTLE_ENDIAN__
35252	35273	__ai __attribute__((target("neon"))) float32x2x2_t vzip_f32(float32x2_t __p0, float32x2_t __p1) {
35253	35274	float32x2x2_t __ret;
35254		__builtin_neon_vzip_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 9);
	35275	__builtin_neon_vzip_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 9);
35255	35276	return __ret;
35256	35277	}
35257	35278	#else
35258	35279	__ai __attribute__((target("neon"))) float32x2x2_t vzip_f32(float32x2_t __p0, float32x2_t __p1) {
35259	35280	float32x2x2_t __ret;
35260		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
35261		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
35262		__builtin_neon_vzip_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 9);
	35281	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	35282	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	35283	__builtin_neon_vzip_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 9);
35263	35284
35264		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0);
35265		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0);
	35285	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32);
	35286	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32);
35266	35287	return __ret;
35267	35288	}
35268	35289	#endif
...	...	@@ -35270,18 +35291,18 @@ __ai __attribute__((target("neon"))) float32x2x2_t vzip_f32(float32x2_t __p0, fl
35270	35291	#ifdef __LITTLE_ENDIAN__
35271	35292	__ai __attribute__((target("neon"))) int32x2x2_t vzip_s32(int32x2_t __p0, int32x2_t __p1) {
35272	35293	int32x2x2_t __ret;
35273		__builtin_neon_vzip_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 2);
	35294	__builtin_neon_vzip_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 2);
35274	35295	return __ret;
35275	35296	}
35276	35297	#else
35277	35298	__ai __attribute__((target("neon"))) int32x2x2_t vzip_s32(int32x2_t __p0, int32x2_t __p1) {
35278	35299	int32x2x2_t __ret;
35279		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
35280		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
35281		__builtin_neon_vzip_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 2);
	35300	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	35301	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	35302	__builtin_neon_vzip_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 2);
35282	35303
35283		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0);
35284		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0);
	35304	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_32);
	35305	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_32);
35285	35306	return __ret;
35286	35307	}
35287	35308	#endif
...	...	@@ -35289,18 +35310,18 @@ __ai __attribute__((target("neon"))) int32x2x2_t vzip_s32(int32x2_t __p0, int32x
35289	35310	#ifdef __LITTLE_ENDIAN__
35290	35311	__ai __attribute__((target("neon"))) int16x4x2_t vzip_s16(int16x4_t __p0, int16x4_t __p1) {
35291	35312	int16x4x2_t __ret;
35292		__builtin_neon_vzip_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 1);
	35313	__builtin_neon_vzip_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 1);
35293	35314	return __ret;
35294	35315	}
35295	35316	#else
35296	35317	__ai __attribute__((target("neon"))) int16x4x2_t vzip_s16(int16x4_t __p0, int16x4_t __p1) {
35297	35318	int16x4x2_t __ret;
35298		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35299		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
35300		__builtin_neon_vzip_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 1);
	35319	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	35320	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	35321	__builtin_neon_vzip_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 1);
35301	35322
35302		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
35303		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
	35323	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16);
	35324	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16);
35304	35325	return __ret;
35305	35326	}
35306	35327	#endif
...	...	@@ -35308,18 +35329,18 @@ __ai __attribute__((target("neon"))) int16x4x2_t vzip_s16(int16x4_t __p0, int16x
35308	35329	#ifdef __LITTLE_ENDIAN__
35309	35330	__ai __attribute__((target("neon"))) float16x8x2_t vzipq_f16(float16x8_t __p0, float16x8_t __p1) {
35310	35331	float16x8x2_t __ret;
35311		__builtin_neon_vzipq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 40);
	35332	__builtin_neon_vzipq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 40);
35312	35333	return __ret;
35313	35334	}
35314	35335	#else
35315	35336	__ai __attribute__((target("neon"))) float16x8x2_t vzipq_f16(float16x8_t __p0, float16x8_t __p1) {
35316	35337	float16x8x2_t __ret;
35317		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
35318		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
35319		__builtin_neon_vzipq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 40);
	35338	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	35339	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	35340	__builtin_neon_vzipq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 40);
35320	35341
35321		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
35322		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
	35342	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16);
	35343	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16);
35323	35344	return __ret;
35324	35345	}
35325	35346	#endif
...	...	@@ -35327,18 +35348,18 @@ __ai __attribute__((target("neon"))) float16x8x2_t vzipq_f16(float16x8_t __p0, f
35327	35348	#ifdef __LITTLE_ENDIAN__
35328	35349	__ai __attribute__((target("neon"))) float16x4x2_t vzip_f16(float16x4_t __p0, float16x4_t __p1) {
35329	35350	float16x4x2_t __ret;
35330		__builtin_neon_vzip_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 8);
	35351	__builtin_neon_vzip_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 8);
35331	35352	return __ret;
35332	35353	}
35333	35354	#else
35334	35355	__ai __attribute__((target("neon"))) float16x4x2_t vzip_f16(float16x4_t __p0, float16x4_t __p1) {
35335	35356	float16x4x2_t __ret;
35336		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35337		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
35338		__builtin_neon_vzip_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 8);
	35357	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	35358	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	35359	__builtin_neon_vzip_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 8);
35339	35360
35340		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
35341		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
	35361	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16);
	35362	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16);
35342	35363	return __ret;
35343	35364	}
35344	35365	#endif
...	...	@@ -35346,22 +35367,22 @@ __ai __attribute__((target("neon"))) float16x4x2_t vzip_f16(float16x4_t __p0, fl
35346	35367	#ifdef __LITTLE_ENDIAN__
35347	35368	__ai __attribute__((target("v8.1a,neon"))) int32x4_t vqrdmlahq_s32(int32x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
35348	35369	int32x4_t __ret;
35349		__ret = (int32x4_t) __builtin_neon_vqrdmlahq_s32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 34);
	35370	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqrdmlahq_s32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 34));
35350	35371	return __ret;
35351	35372	}
35352	35373	#else
35353	35374	__ai __attribute__((target("v8.1a,neon"))) int32x4_t vqrdmlahq_s32(int32x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
35354	35375	int32x4_t __ret;
35355		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35356		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
35357		int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
35358		__ret = (int32x4_t) __builtin_neon_vqrdmlahq_s32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 34);
35359		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	35376	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	35377	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	35378	int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
	35379	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqrdmlahq_s32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 34));
	35380	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
35360	35381	return __ret;
35361	35382	}
35362	35383	__ai __attribute__((target("v8.1a,neon"))) int32x4_t __noswap_vqrdmlahq_s32(int32x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
35363	35384	int32x4_t __ret;
35364		__ret = (int32x4_t) __builtin_neon_vqrdmlahq_s32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 34);
	35385	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqrdmlahq_s32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 34));
35365	35386	return __ret;
35366	35387	}
35367	35388	#endif
...	...	@@ -35369,22 +35390,22 @@ __ai __attribute__((target("v8.1a,neon"))) int32x4_t __noswap_vqrdmlahq_s32(int3
35369	35390	#ifdef __LITTLE_ENDIAN__
35370	35391	__ai __attribute__((target("v8.1a,neon"))) int16x8_t vqrdmlahq_s16(int16x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
35371	35392	int16x8_t __ret;
35372		__ret = (int16x8_t) __builtin_neon_vqrdmlahq_s16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 33);
	35393	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqrdmlahq_s16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 33));
35373	35394	return __ret;
35374	35395	}
35375	35396	#else
35376	35397	__ai __attribute__((target("v8.1a,neon"))) int16x8_t vqrdmlahq_s16(int16x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
35377	35398	int16x8_t __ret;
35378		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
35379		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
35380		int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
35381		__ret = (int16x8_t) __builtin_neon_vqrdmlahq_s16((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 33);
35382		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	35399	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	35400	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	35401	int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
	35402	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqrdmlahq_s16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 33));
	35403	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
35383	35404	return __ret;
35384	35405	}
35385	35406	__ai __attribute__((target("v8.1a,neon"))) int16x8_t __noswap_vqrdmlahq_s16(int16x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
35386	35407	int16x8_t __ret;
35387		__ret = (int16x8_t) __builtin_neon_vqrdmlahq_s16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 33);
	35408	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqrdmlahq_s16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 33));
35388	35409	return __ret;
35389	35410	}
35390	35411	#endif
...	...	@@ -35392,22 +35413,22 @@ __ai __attribute__((target("v8.1a,neon"))) int16x8_t __noswap_vqrdmlahq_s16(int1
35392	35413	#ifdef __LITTLE_ENDIAN__
35393	35414	__ai __attribute__((target("v8.1a,neon"))) int32x2_t vqrdmlah_s32(int32x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
35394	35415	int32x2_t __ret;
35395		__ret = (int32x2_t) __builtin_neon_vqrdmlah_s32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 2);
	35416	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqrdmlah_s32(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 2));
35396	35417	return __ret;
35397	35418	}
35398	35419	#else
35399	35420	__ai __attribute__((target("v8.1a,neon"))) int32x2_t vqrdmlah_s32(int32x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
35400	35421	int32x2_t __ret;
35401		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
35402		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
35403		int32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
35404		__ret = (int32x2_t) __builtin_neon_vqrdmlah_s32((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 2);
35405		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	35422	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	35423	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	35424	int32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_32);
	35425	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqrdmlah_s32(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 2));
	35426	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
35406	35427	return __ret;
35407	35428	}
35408	35429	__ai __attribute__((target("v8.1a,neon"))) int32x2_t __noswap_vqrdmlah_s32(int32x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
35409	35430	int32x2_t __ret;
35410		__ret = (int32x2_t) __builtin_neon_vqrdmlah_s32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 2);
	35431	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqrdmlah_s32(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 2));
35411	35432	return __ret;
35412	35433	}
35413	35434	#endif
...	...	@@ -35415,22 +35436,22 @@ __ai __attribute__((target("v8.1a,neon"))) int32x2_t __noswap_vqrdmlah_s32(int32
35415	35436	#ifdef __LITTLE_ENDIAN__
35416	35437	__ai __attribute__((target("v8.1a,neon"))) int16x4_t vqrdmlah_s16(int16x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
35417	35438	int16x4_t __ret;
35418		__ret = (int16x4_t) __builtin_neon_vqrdmlah_s16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 1);
	35439	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqrdmlah_s16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 1));
35419	35440	return __ret;
35420	35441	}
35421	35442	#else
35422	35443	__ai __attribute__((target("v8.1a,neon"))) int16x4_t vqrdmlah_s16(int16x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
35423	35444	int16x4_t __ret;
35424		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35425		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
35426		int16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
35427		__ret = (int16x4_t) __builtin_neon_vqrdmlah_s16((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 1);
35428		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	35445	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	35446	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	35447	int16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
	35448	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqrdmlah_s16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 1));
	35449	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
35429	35450	return __ret;
35430	35451	}
35431	35452	__ai __attribute__((target("v8.1a,neon"))) int16x4_t __noswap_vqrdmlah_s16(int16x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
35432	35453	int16x4_t __ret;
35433		__ret = (int16x4_t) __builtin_neon_vqrdmlah_s16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 1);
	35454	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqrdmlah_s16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 1));
35434	35455	return __ret;
35435	35456	}
35436	35457	#endif
...	...	@@ -35450,11 +35471,11 @@ __ai __attribute__((target("v8.1a,neon"))) int16x4_t __noswap_vqrdmlah_s16(int16
35450	35471	int32x4_t __s0_135 = __p0_135; \
35451	35472	int32x4_t __s1_135 = __p1_135; \
35452	35473	int32x2_t __s2_135 = __p2_135; \
35453		int32x4_t __rev0_135; __rev0_135 = __builtin_shufflevector(__s0_135, __s0_135, 3, 2, 1, 0); \
35454		int32x4_t __rev1_135; __rev1_135 = __builtin_shufflevector(__s1_135, __s1_135, 3, 2, 1, 0); \
35455		int32x2_t __rev2_135; __rev2_135 = __builtin_shufflevector(__s2_135, __s2_135, 1, 0); \
	35474	int32x4_t __rev0_135; __rev0_135 = __builtin_shufflevector(__s0_135, __s0_135, __lane_reverse_128_32); \
	35475	int32x4_t __rev1_135; __rev1_135 = __builtin_shufflevector(__s1_135, __s1_135, __lane_reverse_128_32); \
	35476	int32x2_t __rev2_135; __rev2_135 = __builtin_shufflevector(__s2_135, __s2_135, __lane_reverse_64_32); \
35456	35477	__ret_135 = __noswap_vqrdmlahq_s32(__rev0_135, __rev1_135, __noswap_splatq_lane_s32(__rev2_135, __p3_135)); \
35457		__ret_135 = __builtin_shufflevector(__ret_135, __ret_135, 3, 2, 1, 0); \
	35478	__ret_135 = __builtin_shufflevector(__ret_135, __ret_135, __lane_reverse_128_32); \
35458	35479	__ret_135; \
35459	35480	})
35460	35481	#endif
...	...	@@ -35474,11 +35495,11 @@ __ai __attribute__((target("v8.1a,neon"))) int16x4_t __noswap_vqrdmlah_s16(int16
35474	35495	int16x8_t __s0_137 = __p0_137; \
35475	35496	int16x8_t __s1_137 = __p1_137; \
35476	35497	int16x4_t __s2_137 = __p2_137; \
35477		int16x8_t __rev0_137; __rev0_137 = __builtin_shufflevector(__s0_137, __s0_137, 7, 6, 5, 4, 3, 2, 1, 0); \
35478		int16x8_t __rev1_137; __rev1_137 = __builtin_shufflevector(__s1_137, __s1_137, 7, 6, 5, 4, 3, 2, 1, 0); \
35479		int16x4_t __rev2_137; __rev2_137 = __builtin_shufflevector(__s2_137, __s2_137, 3, 2, 1, 0); \
	35498	int16x8_t __rev0_137; __rev0_137 = __builtin_shufflevector(__s0_137, __s0_137, __lane_reverse_128_16); \
	35499	int16x8_t __rev1_137; __rev1_137 = __builtin_shufflevector(__s1_137, __s1_137, __lane_reverse_128_16); \
	35500	int16x4_t __rev2_137; __rev2_137 = __builtin_shufflevector(__s2_137, __s2_137, __lane_reverse_64_16); \
35480	35501	__ret_137 = __noswap_vqrdmlahq_s16(__rev0_137, __rev1_137, __noswap_splatq_lane_s16(__rev2_137, __p3_137)); \
35481		__ret_137 = __builtin_shufflevector(__ret_137, __ret_137, 7, 6, 5, 4, 3, 2, 1, 0); \
	35502	__ret_137 = __builtin_shufflevector(__ret_137, __ret_137, __lane_reverse_128_16); \
35482	35503	__ret_137; \
35483	35504	})
35484	35505	#endif
...	...	@@ -35498,11 +35519,11 @@ __ai __attribute__((target("v8.1a,neon"))) int16x4_t __noswap_vqrdmlah_s16(int16
35498	35519	int32x2_t __s0_139 = __p0_139; \
35499	35520	int32x2_t __s1_139 = __p1_139; \
35500	35521	int32x2_t __s2_139 = __p2_139; \
35501		int32x2_t __rev0_139; __rev0_139 = __builtin_shufflevector(__s0_139, __s0_139, 1, 0); \
35502		int32x2_t __rev1_139; __rev1_139 = __builtin_shufflevector(__s1_139, __s1_139, 1, 0); \
35503		int32x2_t __rev2_139; __rev2_139 = __builtin_shufflevector(__s2_139, __s2_139, 1, 0); \
	35522	int32x2_t __rev0_139; __rev0_139 = __builtin_shufflevector(__s0_139, __s0_139, __lane_reverse_64_32); \
	35523	int32x2_t __rev1_139; __rev1_139 = __builtin_shufflevector(__s1_139, __s1_139, __lane_reverse_64_32); \
	35524	int32x2_t __rev2_139; __rev2_139 = __builtin_shufflevector(__s2_139, __s2_139, __lane_reverse_64_32); \
35504	35525	__ret_139 = __noswap_vqrdmlah_s32(__rev0_139, __rev1_139, __noswap_splat_lane_s32(__rev2_139, __p3_139)); \
35505		__ret_139 = __builtin_shufflevector(__ret_139, __ret_139, 1, 0); \
	35526	__ret_139 = __builtin_shufflevector(__ret_139, __ret_139, __lane_reverse_64_32); \
35506	35527	__ret_139; \
35507	35528	})
35508	35529	#endif
...	...	@@ -35522,11 +35543,11 @@ __ai __attribute__((target("v8.1a,neon"))) int16x4_t __noswap_vqrdmlah_s16(int16
35522	35543	int16x4_t __s0_141 = __p0_141; \
35523	35544	int16x4_t __s1_141 = __p1_141; \
35524	35545	int16x4_t __s2_141 = __p2_141; \
35525		int16x4_t __rev0_141; __rev0_141 = __builtin_shufflevector(__s0_141, __s0_141, 3, 2, 1, 0); \
35526		int16x4_t __rev1_141; __rev1_141 = __builtin_shufflevector(__s1_141, __s1_141, 3, 2, 1, 0); \
35527		int16x4_t __rev2_141; __rev2_141 = __builtin_shufflevector(__s2_141, __s2_141, 3, 2, 1, 0); \
	35546	int16x4_t __rev0_141; __rev0_141 = __builtin_shufflevector(__s0_141, __s0_141, __lane_reverse_64_16); \
	35547	int16x4_t __rev1_141; __rev1_141 = __builtin_shufflevector(__s1_141, __s1_141, __lane_reverse_64_16); \
	35548	int16x4_t __rev2_141; __rev2_141 = __builtin_shufflevector(__s2_141, __s2_141, __lane_reverse_64_16); \
35528	35549	__ret_141 = __noswap_vqrdmlah_s16(__rev0_141, __rev1_141, __noswap_splat_lane_s16(__rev2_141, __p3_141)); \
35529		__ret_141 = __builtin_shufflevector(__ret_141, __ret_141, 3, 2, 1, 0); \
	35550	__ret_141 = __builtin_shufflevector(__ret_141, __ret_141, __lane_reverse_64_16); \
35530	35551	__ret_141; \
35531	35552	})
35532	35553	#endif
...	...	@@ -35534,22 +35555,22 @@ __ai __attribute__((target("v8.1a,neon"))) int16x4_t __noswap_vqrdmlah_s16(int16
35534	35555	#ifdef __LITTLE_ENDIAN__
35535	35556	__ai __attribute__((target("v8.1a,neon"))) int32x4_t vqrdmlshq_s32(int32x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
35536	35557	int32x4_t __ret;
35537		__ret = (int32x4_t) __builtin_neon_vqrdmlshq_s32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 34);
	35558	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqrdmlshq_s32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 34));
35538	35559	return __ret;
35539	35560	}
35540	35561	#else
35541	35562	__ai __attribute__((target("v8.1a,neon"))) int32x4_t vqrdmlshq_s32(int32x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
35542	35563	int32x4_t __ret;
35543		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35544		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
35545		int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
35546		__ret = (int32x4_t) __builtin_neon_vqrdmlshq_s32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 34);
35547		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	35564	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	35565	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	35566	int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
	35567	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqrdmlshq_s32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 34));
	35568	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
35548	35569	return __ret;
35549	35570	}
35550	35571	__ai __attribute__((target("v8.1a,neon"))) int32x4_t __noswap_vqrdmlshq_s32(int32x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
35551	35572	int32x4_t __ret;
35552		__ret = (int32x4_t) __builtin_neon_vqrdmlshq_s32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 34);
	35573	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqrdmlshq_s32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 34));
35553	35574	return __ret;
35554	35575	}
35555	35576	#endif
...	...	@@ -35557,22 +35578,22 @@ __ai __attribute__((target("v8.1a,neon"))) int32x4_t __noswap_vqrdmlshq_s32(int3
35557	35578	#ifdef __LITTLE_ENDIAN__
35558	35579	__ai __attribute__((target("v8.1a,neon"))) int16x8_t vqrdmlshq_s16(int16x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
35559	35580	int16x8_t __ret;
35560		__ret = (int16x8_t) __builtin_neon_vqrdmlshq_s16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 33);
	35581	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqrdmlshq_s16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 33));
35561	35582	return __ret;
35562	35583	}
35563	35584	#else
35564	35585	__ai __attribute__((target("v8.1a,neon"))) int16x8_t vqrdmlshq_s16(int16x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
35565	35586	int16x8_t __ret;
35566		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
35567		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
35568		int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
35569		__ret = (int16x8_t) __builtin_neon_vqrdmlshq_s16((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 33);
35570		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	35587	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	35588	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	35589	int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
	35590	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqrdmlshq_s16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 33));
	35591	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
35571	35592	return __ret;
35572	35593	}
35573	35594	__ai __attribute__((target("v8.1a,neon"))) int16x8_t __noswap_vqrdmlshq_s16(int16x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
35574	35595	int16x8_t __ret;
35575		__ret = (int16x8_t) __builtin_neon_vqrdmlshq_s16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 33);
	35596	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqrdmlshq_s16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 33));
35576	35597	return __ret;
35577	35598	}
35578	35599	#endif
...	...	@@ -35580,22 +35601,22 @@ __ai __attribute__((target("v8.1a,neon"))) int16x8_t __noswap_vqrdmlshq_s16(int1
35580	35601	#ifdef __LITTLE_ENDIAN__
35581	35602	__ai __attribute__((target("v8.1a,neon"))) int32x2_t vqrdmlsh_s32(int32x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
35582	35603	int32x2_t __ret;
35583		__ret = (int32x2_t) __builtin_neon_vqrdmlsh_s32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 2);
	35604	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqrdmlsh_s32(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 2));
35584	35605	return __ret;
35585	35606	}
35586	35607	#else
35587	35608	__ai __attribute__((target("v8.1a,neon"))) int32x2_t vqrdmlsh_s32(int32x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
35588	35609	int32x2_t __ret;
35589		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
35590		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
35591		int32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
35592		__ret = (int32x2_t) __builtin_neon_vqrdmlsh_s32((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 2);
35593		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	35610	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	35611	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	35612	int32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_32);
	35613	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqrdmlsh_s32(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 2));
	35614	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
35594	35615	return __ret;
35595	35616	}
35596	35617	__ai __attribute__((target("v8.1a,neon"))) int32x2_t __noswap_vqrdmlsh_s32(int32x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
35597	35618	int32x2_t __ret;
35598		__ret = (int32x2_t) __builtin_neon_vqrdmlsh_s32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 2);
	35619	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqrdmlsh_s32(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 2));
35599	35620	return __ret;
35600	35621	}
35601	35622	#endif
...	...	@@ -35603,22 +35624,22 @@ __ai __attribute__((target("v8.1a,neon"))) int32x2_t __noswap_vqrdmlsh_s32(int32
35603	35624	#ifdef __LITTLE_ENDIAN__
35604	35625	__ai __attribute__((target("v8.1a,neon"))) int16x4_t vqrdmlsh_s16(int16x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
35605	35626	int16x4_t __ret;
35606		__ret = (int16x4_t) __builtin_neon_vqrdmlsh_s16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 1);
	35627	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqrdmlsh_s16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 1));
35607	35628	return __ret;
35608	35629	}
35609	35630	#else
35610	35631	__ai __attribute__((target("v8.1a,neon"))) int16x4_t vqrdmlsh_s16(int16x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
35611	35632	int16x4_t __ret;
35612		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35613		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
35614		int16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
35615		__ret = (int16x4_t) __builtin_neon_vqrdmlsh_s16((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 1);
35616		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	35633	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	35634	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	35635	int16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
	35636	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqrdmlsh_s16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 1));
	35637	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
35617	35638	return __ret;
35618	35639	}
35619	35640	__ai __attribute__((target("v8.1a,neon"))) int16x4_t __noswap_vqrdmlsh_s16(int16x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
35620	35641	int16x4_t __ret;
35621		__ret = (int16x4_t) __builtin_neon_vqrdmlsh_s16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 1);
	35642	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqrdmlsh_s16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 1));
35622	35643	return __ret;
35623	35644	}
35624	35645	#endif
...	...	@@ -35638,11 +35659,11 @@ __ai __attribute__((target("v8.1a,neon"))) int16x4_t __noswap_vqrdmlsh_s16(int16
35638	35659	int32x4_t __s0_143 = __p0_143; \
35639	35660	int32x4_t __s1_143 = __p1_143; \
35640	35661	int32x2_t __s2_143 = __p2_143; \
35641		int32x4_t __rev0_143; __rev0_143 = __builtin_shufflevector(__s0_143, __s0_143, 3, 2, 1, 0); \
35642		int32x4_t __rev1_143; __rev1_143 = __builtin_shufflevector(__s1_143, __s1_143, 3, 2, 1, 0); \
35643		int32x2_t __rev2_143; __rev2_143 = __builtin_shufflevector(__s2_143, __s2_143, 1, 0); \
	35662	int32x4_t __rev0_143; __rev0_143 = __builtin_shufflevector(__s0_143, __s0_143, __lane_reverse_128_32); \
	35663	int32x4_t __rev1_143; __rev1_143 = __builtin_shufflevector(__s1_143, __s1_143, __lane_reverse_128_32); \
	35664	int32x2_t __rev2_143; __rev2_143 = __builtin_shufflevector(__s2_143, __s2_143, __lane_reverse_64_32); \
35644	35665	__ret_143 = __noswap_vqrdmlshq_s32(__rev0_143, __rev1_143, __noswap_splatq_lane_s32(__rev2_143, __p3_143)); \
35645		__ret_143 = __builtin_shufflevector(__ret_143, __ret_143, 3, 2, 1, 0); \
	35666	__ret_143 = __builtin_shufflevector(__ret_143, __ret_143, __lane_reverse_128_32); \
35646	35667	__ret_143; \
35647	35668	})
35648	35669	#endif
...	...	@@ -35662,11 +35683,11 @@ __ai __attribute__((target("v8.1a,neon"))) int16x4_t __noswap_vqrdmlsh_s16(int16
35662	35683	int16x8_t __s0_145 = __p0_145; \
35663	35684	int16x8_t __s1_145 = __p1_145; \
35664	35685	int16x4_t __s2_145 = __p2_145; \
35665		int16x8_t __rev0_145; __rev0_145 = __builtin_shufflevector(__s0_145, __s0_145, 7, 6, 5, 4, 3, 2, 1, 0); \
35666		int16x8_t __rev1_145; __rev1_145 = __builtin_shufflevector(__s1_145, __s1_145, 7, 6, 5, 4, 3, 2, 1, 0); \
35667		int16x4_t __rev2_145; __rev2_145 = __builtin_shufflevector(__s2_145, __s2_145, 3, 2, 1, 0); \
	35686	int16x8_t __rev0_145; __rev0_145 = __builtin_shufflevector(__s0_145, __s0_145, __lane_reverse_128_16); \
	35687	int16x8_t __rev1_145; __rev1_145 = __builtin_shufflevector(__s1_145, __s1_145, __lane_reverse_128_16); \
	35688	int16x4_t __rev2_145; __rev2_145 = __builtin_shufflevector(__s2_145, __s2_145, __lane_reverse_64_16); \
35668	35689	__ret_145 = __noswap_vqrdmlshq_s16(__rev0_145, __rev1_145, __noswap_splatq_lane_s16(__rev2_145, __p3_145)); \
35669		__ret_145 = __builtin_shufflevector(__ret_145, __ret_145, 7, 6, 5, 4, 3, 2, 1, 0); \
	35690	__ret_145 = __builtin_shufflevector(__ret_145, __ret_145, __lane_reverse_128_16); \
35670	35691	__ret_145; \
35671	35692	})
35672	35693	#endif
...	...	@@ -35686,11 +35707,11 @@ __ai __attribute__((target("v8.1a,neon"))) int16x4_t __noswap_vqrdmlsh_s16(int16
35686	35707	int32x2_t __s0_147 = __p0_147; \
35687	35708	int32x2_t __s1_147 = __p1_147; \
35688	35709	int32x2_t __s2_147 = __p2_147; \
35689		int32x2_t __rev0_147; __rev0_147 = __builtin_shufflevector(__s0_147, __s0_147, 1, 0); \
35690		int32x2_t __rev1_147; __rev1_147 = __builtin_shufflevector(__s1_147, __s1_147, 1, 0); \
35691		int32x2_t __rev2_147; __rev2_147 = __builtin_shufflevector(__s2_147, __s2_147, 1, 0); \
	35710	int32x2_t __rev0_147; __rev0_147 = __builtin_shufflevector(__s0_147, __s0_147, __lane_reverse_64_32); \
	35711	int32x2_t __rev1_147; __rev1_147 = __builtin_shufflevector(__s1_147, __s1_147, __lane_reverse_64_32); \
	35712	int32x2_t __rev2_147; __rev2_147 = __builtin_shufflevector(__s2_147, __s2_147, __lane_reverse_64_32); \
35692	35713	__ret_147 = __noswap_vqrdmlsh_s32(__rev0_147, __rev1_147, __noswap_splat_lane_s32(__rev2_147, __p3_147)); \
35693		__ret_147 = __builtin_shufflevector(__ret_147, __ret_147, 1, 0); \
	35714	__ret_147 = __builtin_shufflevector(__ret_147, __ret_147, __lane_reverse_64_32); \
35694	35715	__ret_147; \
35695	35716	})
35696	35717	#endif
...	...	@@ -35710,11 +35731,11 @@ __ai __attribute__((target("v8.1a,neon"))) int16x4_t __noswap_vqrdmlsh_s16(int16
35710	35731	int16x4_t __s0_149 = __p0_149; \
35711	35732	int16x4_t __s1_149 = __p1_149; \
35712	35733	int16x4_t __s2_149 = __p2_149; \
35713		int16x4_t __rev0_149; __rev0_149 = __builtin_shufflevector(__s0_149, __s0_149, 3, 2, 1, 0); \
35714		int16x4_t __rev1_149; __rev1_149 = __builtin_shufflevector(__s1_149, __s1_149, 3, 2, 1, 0); \
35715		int16x4_t __rev2_149; __rev2_149 = __builtin_shufflevector(__s2_149, __s2_149, 3, 2, 1, 0); \
	35734	int16x4_t __rev0_149; __rev0_149 = __builtin_shufflevector(__s0_149, __s0_149, __lane_reverse_64_16); \
	35735	int16x4_t __rev1_149; __rev1_149 = __builtin_shufflevector(__s1_149, __s1_149, __lane_reverse_64_16); \
	35736	int16x4_t __rev2_149; __rev2_149 = __builtin_shufflevector(__s2_149, __s2_149, __lane_reverse_64_16); \
35716	35737	__ret_149 = __noswap_vqrdmlsh_s16(__rev0_149, __rev1_149, __noswap_splat_lane_s16(__rev2_149, __p3_149)); \
35717		__ret_149 = __builtin_shufflevector(__ret_149, __ret_149, 3, 2, 1, 0); \
	35738	__ret_149 = __builtin_shufflevector(__ret_149, __ret_149, __lane_reverse_64_16); \
35718	35739	__ret_149; \
35719	35740	})
35720	35741	#endif
...	...	@@ -35722,16 +35743,16 @@ __ai __attribute__((target("v8.1a,neon"))) int16x4_t __noswap_vqrdmlsh_s16(int16
35722	35743	#ifdef __LITTLE_ENDIAN__
35723	35744	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t vcadd_rot270_f16(float16x4_t __p0, float16x4_t __p1) {
35724	35745	float16x4_t __ret;
35725		__ret = (float16x4_t) __builtin_neon_vcadd_rot270_f16((int8x8_t)__p0, (int8x8_t)__p1, 8);
	35746	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vcadd_rot270_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 8));
35726	35747	return __ret;
35727	35748	}
35728	35749	#else
35729	35750	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t vcadd_rot270_f16(float16x4_t __p0, float16x4_t __p1) {
35730	35751	float16x4_t __ret;
35731		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35732		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
35733		__ret = (float16x4_t) __builtin_neon_vcadd_rot270_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 8);
35734		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	35752	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	35753	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	35754	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vcadd_rot270_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 8));
	35755	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
35735	35756	return __ret;
35736	35757	}
35737	35758	#endif
...	...	@@ -35739,16 +35760,16 @@ __ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t vcadd_rot270_f16
35739	35760	#ifdef __LITTLE_ENDIAN__
35740	35761	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t vcadd_rot90_f16(float16x4_t __p0, float16x4_t __p1) {
35741	35762	float16x4_t __ret;
35742		__ret = (float16x4_t) __builtin_neon_vcadd_rot90_f16((int8x8_t)__p0, (int8x8_t)__p1, 8);
	35763	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vcadd_rot90_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 8));
35743	35764	return __ret;
35744	35765	}
35745	35766	#else
35746	35767	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t vcadd_rot90_f16(float16x4_t __p0, float16x4_t __p1) {
35747	35768	float16x4_t __ret;
35748		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35749		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
35750		__ret = (float16x4_t) __builtin_neon_vcadd_rot90_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 8);
35751		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	35769	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	35770	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	35771	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vcadd_rot90_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 8));
	35772	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
35752	35773	return __ret;
35753	35774	}
35754	35775	#endif
...	...	@@ -35756,16 +35777,16 @@ __ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t vcadd_rot90_f16(
35756	35777	#ifdef __LITTLE_ENDIAN__
35757	35778	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x8_t vcaddq_rot270_f16(float16x8_t __p0, float16x8_t __p1) {
35758	35779	float16x8_t __ret;
35759		__ret = (float16x8_t) __builtin_neon_vcaddq_rot270_f16((int8x16_t)__p0, (int8x16_t)__p1, 40);
	35780	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcaddq_rot270_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 40));
35760	35781	return __ret;
35761	35782	}
35762	35783	#else
35763	35784	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x8_t vcaddq_rot270_f16(float16x8_t __p0, float16x8_t __p1) {
35764	35785	float16x8_t __ret;
35765		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
35766		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
35767		__ret = (float16x8_t) __builtin_neon_vcaddq_rot270_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 40);
35768		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	35786	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	35787	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	35788	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcaddq_rot270_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 40));
	35789	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
35769	35790	return __ret;
35770	35791	}
35771	35792	#endif
...	...	@@ -35773,16 +35794,16 @@ __ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x8_t vcaddq_rot270_f1
35773	35794	#ifdef __LITTLE_ENDIAN__
35774	35795	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x8_t vcaddq_rot90_f16(float16x8_t __p0, float16x8_t __p1) {
35775	35796	float16x8_t __ret;
35776		__ret = (float16x8_t) __builtin_neon_vcaddq_rot90_f16((int8x16_t)__p0, (int8x16_t)__p1, 40);
	35797	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcaddq_rot90_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 40));
35777	35798	return __ret;
35778	35799	}
35779	35800	#else
35780	35801	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x8_t vcaddq_rot90_f16(float16x8_t __p0, float16x8_t __p1) {
35781	35802	float16x8_t __ret;
35782		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
35783		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
35784		__ret = (float16x8_t) __builtin_neon_vcaddq_rot90_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 40);
35785		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	35803	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	35804	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	35805	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcaddq_rot90_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 40));
	35806	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
35786	35807	return __ret;
35787	35808	}
35788	35809	#endif
...	...	@@ -35790,22 +35811,22 @@ __ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x8_t vcaddq_rot90_f16
35790	35811	#ifdef __LITTLE_ENDIAN__
35791	35812	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x8_t vcmlaq_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
35792	35813	float16x8_t __ret;
35793		__ret = (float16x8_t) __builtin_neon_vcmlaq_f16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 40);
	35814	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcmlaq_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 40));
35794	35815	return __ret;
35795	35816	}
35796	35817	#else
35797	35818	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x8_t vcmlaq_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
35798	35819	float16x8_t __ret;
35799		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
35800		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
35801		float16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
35802		__ret = (float16x8_t) __builtin_neon_vcmlaq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 40);
35803		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	35820	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	35821	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	35822	float16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
	35823	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcmlaq_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 40));
	35824	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
35804	35825	return __ret;
35805	35826	}
35806	35827	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x8_t __noswap_vcmlaq_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
35807	35828	float16x8_t __ret;
35808		__ret = (float16x8_t) __builtin_neon_vcmlaq_f16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 40);
	35829	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcmlaq_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 40));
35809	35830	return __ret;
35810	35831	}
35811	35832	#endif
...	...	@@ -35813,22 +35834,22 @@ __ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x8_t __noswap_vcmlaq_
35813	35834	#ifdef __LITTLE_ENDIAN__
35814	35835	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t vcmla_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
35815	35836	float16x4_t __ret;
35816		__ret = (float16x4_t) __builtin_neon_vcmla_f16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 8);
	35837	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vcmla_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 8));
35817	35838	return __ret;
35818	35839	}
35819	35840	#else
35820	35841	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t vcmla_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
35821	35842	float16x4_t __ret;
35822		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35823		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
35824		float16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
35825		__ret = (float16x4_t) __builtin_neon_vcmla_f16((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 8);
35826		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	35843	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	35844	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	35845	float16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
	35846	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vcmla_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 8));
	35847	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
35827	35848	return __ret;
35828	35849	}
35829	35850	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t __noswap_vcmla_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
35830	35851	float16x4_t __ret;
35831		__ret = (float16x4_t) __builtin_neon_vcmla_f16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 8);
	35852	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vcmla_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 8));
35832	35853	return __ret;
35833	35854	}
35834	35855	#endif
...	...	@@ -35848,11 +35869,11 @@ __ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t __noswap_vcmla_f
35848	35869	float16x4_t __s0_151 = __p0_151; \
35849	35870	float16x4_t __s1_151 = __p1_151; \
35850	35871	float16x4_t __s2_151 = __p2_151; \
35851		float16x4_t __rev0_151; __rev0_151 = __builtin_shufflevector(__s0_151, __s0_151, 3, 2, 1, 0); \
35852		float16x4_t __rev1_151; __rev1_151 = __builtin_shufflevector(__s1_151, __s1_151, 3, 2, 1, 0); \
35853		float16x4_t __rev2_151; __rev2_151 = __builtin_shufflevector(__s2_151, __s2_151, 3, 2, 1, 0); \
	35872	float16x4_t __rev0_151; __rev0_151 = __builtin_shufflevector(__s0_151, __s0_151, __lane_reverse_64_16); \
	35873	float16x4_t __rev1_151; __rev1_151 = __builtin_shufflevector(__s1_151, __s1_151, __lane_reverse_64_16); \
	35874	float16x4_t __rev2_151; __rev2_151 = __builtin_shufflevector(__s2_151, __s2_151, __lane_reverse_64_16); \
35854	35875	__ret_151 = __noswap_vcmla_f16(__rev0_151, __rev1_151, __builtin_bit_cast(float16x4_t, (uint32x2_t) {__noswap_vget_lane_u32(__builtin_bit_cast(uint32x2_t, __rev2_151), __p3_151), __noswap_vget_lane_u32(__builtin_bit_cast(uint32x2_t, __rev2_151), __p3_151)})); \
35855		__ret_151 = __builtin_shufflevector(__ret_151, __ret_151, 3, 2, 1, 0); \
	35876	__ret_151 = __builtin_shufflevector(__ret_151, __ret_151, __lane_reverse_64_16); \
35856	35877	__ret_151; \
35857	35878	})
35858	35879	#endif
...	...	@@ -35872,11 +35893,11 @@ __ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t __noswap_vcmla_f
35872	35893	float16x8_t __s0_153 = __p0_153; \
35873	35894	float16x8_t __s1_153 = __p1_153; \
35874	35895	float16x4_t __s2_153 = __p2_153; \
35875		float16x8_t __rev0_153; __rev0_153 = __builtin_shufflevector(__s0_153, __s0_153, 7, 6, 5, 4, 3, 2, 1, 0); \
35876		float16x8_t __rev1_153; __rev1_153 = __builtin_shufflevector(__s1_153, __s1_153, 7, 6, 5, 4, 3, 2, 1, 0); \
35877		float16x4_t __rev2_153; __rev2_153 = __builtin_shufflevector(__s2_153, __s2_153, 3, 2, 1, 0); \
	35896	float16x8_t __rev0_153; __rev0_153 = __builtin_shufflevector(__s0_153, __s0_153, __lane_reverse_128_16); \
	35897	float16x8_t __rev1_153; __rev1_153 = __builtin_shufflevector(__s1_153, __s1_153, __lane_reverse_128_16); \
	35898	float16x4_t __rev2_153; __rev2_153 = __builtin_shufflevector(__s2_153, __s2_153, __lane_reverse_64_16); \
35878	35899	__ret_153 = __noswap_vcmlaq_f16(__rev0_153, __rev1_153, __builtin_bit_cast(float16x8_t, (uint32x4_t) {__noswap_vget_lane_u32(__builtin_bit_cast(uint32x2_t, __rev2_153), __p3_153), __noswap_vget_lane_u32(__builtin_bit_cast(uint32x2_t, __rev2_153), __p3_153), __noswap_vget_lane_u32(__builtin_bit_cast(uint32x2_t, __rev2_153), __p3_153), __noswap_vget_lane_u32(__builtin_bit_cast(uint32x2_t, __rev2_153), __p3_153)})); \
35879		__ret_153 = __builtin_shufflevector(__ret_153, __ret_153, 7, 6, 5, 4, 3, 2, 1, 0); \
	35900	__ret_153 = __builtin_shufflevector(__ret_153, __ret_153, __lane_reverse_128_16); \
35880	35901	__ret_153; \
35881	35902	})
35882	35903	#endif
...	...	@@ -35896,11 +35917,11 @@ __ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t __noswap_vcmla_f
35896	35917	float16x4_t __s0_155 = __p0_155; \
35897	35918	float16x4_t __s1_155 = __p1_155; \
35898	35919	float16x8_t __s2_155 = __p2_155; \
35899		float16x4_t __rev0_155; __rev0_155 = __builtin_shufflevector(__s0_155, __s0_155, 3, 2, 1, 0); \
35900		float16x4_t __rev1_155; __rev1_155 = __builtin_shufflevector(__s1_155, __s1_155, 3, 2, 1, 0); \
35901		float16x8_t __rev2_155; __rev2_155 = __builtin_shufflevector(__s2_155, __s2_155, 7, 6, 5, 4, 3, 2, 1, 0); \
	35920	float16x4_t __rev0_155; __rev0_155 = __builtin_shufflevector(__s0_155, __s0_155, __lane_reverse_64_16); \
	35921	float16x4_t __rev1_155; __rev1_155 = __builtin_shufflevector(__s1_155, __s1_155, __lane_reverse_64_16); \
	35922	float16x8_t __rev2_155; __rev2_155 = __builtin_shufflevector(__s2_155, __s2_155, __lane_reverse_128_16); \
35902	35923	__ret_155 = __noswap_vcmla_f16(__rev0_155, __rev1_155, __builtin_bit_cast(float16x4_t, (uint32x2_t) {__noswap_vgetq_lane_u32(__builtin_bit_cast(uint32x4_t, __rev2_155), __p3_155), __noswap_vgetq_lane_u32(__builtin_bit_cast(uint32x4_t, __rev2_155), __p3_155)})); \
35903		__ret_155 = __builtin_shufflevector(__ret_155, __ret_155, 3, 2, 1, 0); \
	35924	__ret_155 = __builtin_shufflevector(__ret_155, __ret_155, __lane_reverse_64_16); \
35904	35925	__ret_155; \
35905	35926	})
35906	35927	#endif
...	...	@@ -35920,11 +35941,11 @@ __ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t __noswap_vcmla_f
35920	35941	float16x8_t __s0_157 = __p0_157; \
35921	35942	float16x8_t __s1_157 = __p1_157; \
35922	35943	float16x8_t __s2_157 = __p2_157; \
35923		float16x8_t __rev0_157; __rev0_157 = __builtin_shufflevector(__s0_157, __s0_157, 7, 6, 5, 4, 3, 2, 1, 0); \
35924		float16x8_t __rev1_157; __rev1_157 = __builtin_shufflevector(__s1_157, __s1_157, 7, 6, 5, 4, 3, 2, 1, 0); \
35925		float16x8_t __rev2_157; __rev2_157 = __builtin_shufflevector(__s2_157, __s2_157, 7, 6, 5, 4, 3, 2, 1, 0); \
	35944	float16x8_t __rev0_157; __rev0_157 = __builtin_shufflevector(__s0_157, __s0_157, __lane_reverse_128_16); \
	35945	float16x8_t __rev1_157; __rev1_157 = __builtin_shufflevector(__s1_157, __s1_157, __lane_reverse_128_16); \
	35946	float16x8_t __rev2_157; __rev2_157 = __builtin_shufflevector(__s2_157, __s2_157, __lane_reverse_128_16); \
35926	35947	__ret_157 = __noswap_vcmlaq_f16(__rev0_157, __rev1_157, __builtin_bit_cast(float16x8_t, (uint32x4_t) {__noswap_vgetq_lane_u32(__builtin_bit_cast(uint32x4_t, __rev2_157), __p3_157), __noswap_vgetq_lane_u32(__builtin_bit_cast(uint32x4_t, __rev2_157), __p3_157), __noswap_vgetq_lane_u32(__builtin_bit_cast(uint32x4_t, __rev2_157), __p3_157), __noswap_vgetq_lane_u32(__builtin_bit_cast(uint32x4_t, __rev2_157), __p3_157)})); \
35927		__ret_157 = __builtin_shufflevector(__ret_157, __ret_157, 7, 6, 5, 4, 3, 2, 1, 0); \
	35948	__ret_157 = __builtin_shufflevector(__ret_157, __ret_157, __lane_reverse_128_16); \
35928	35949	__ret_157; \
35929	35950	})
35930	35951	#endif
...	...	@@ -35932,22 +35953,22 @@ __ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t __noswap_vcmla_f
35932	35953	#ifdef __LITTLE_ENDIAN__
35933	35954	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x8_t vcmlaq_rot180_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
35934	35955	float16x8_t __ret;
35935		__ret = (float16x8_t) __builtin_neon_vcmlaq_rot180_f16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 40);
	35956	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcmlaq_rot180_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 40));
35936	35957	return __ret;
35937	35958	}
35938	35959	#else
35939	35960	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x8_t vcmlaq_rot180_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
35940	35961	float16x8_t __ret;
35941		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
35942		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
35943		float16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
35944		__ret = (float16x8_t) __builtin_neon_vcmlaq_rot180_f16((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 40);
35945		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	35962	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	35963	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	35964	float16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
	35965	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcmlaq_rot180_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 40));
	35966	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
35946	35967	return __ret;
35947	35968	}
35948	35969	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x8_t __noswap_vcmlaq_rot180_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
35949	35970	float16x8_t __ret;
35950		__ret = (float16x8_t) __builtin_neon_vcmlaq_rot180_f16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 40);
	35971	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcmlaq_rot180_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 40));
35951	35972	return __ret;
35952	35973	}
35953	35974	#endif
...	...	@@ -35955,22 +35976,22 @@ __ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x8_t __noswap_vcmlaq_
35955	35976	#ifdef __LITTLE_ENDIAN__
35956	35977	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t vcmla_rot180_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
35957	35978	float16x4_t __ret;
35958		__ret = (float16x4_t) __builtin_neon_vcmla_rot180_f16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 8);
	35979	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vcmla_rot180_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 8));
35959	35980	return __ret;
35960	35981	}
35961	35982	#else
35962	35983	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t vcmla_rot180_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
35963	35984	float16x4_t __ret;
35964		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35965		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
35966		float16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
35967		__ret = (float16x4_t) __builtin_neon_vcmla_rot180_f16((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 8);
35968		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	35985	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	35986	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	35987	float16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
	35988	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vcmla_rot180_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 8));
	35989	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
35969	35990	return __ret;
35970	35991	}
35971	35992	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t __noswap_vcmla_rot180_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
35972	35993	float16x4_t __ret;
35973		__ret = (float16x4_t) __builtin_neon_vcmla_rot180_f16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 8);
	35994	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vcmla_rot180_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 8));
35974	35995	return __ret;
35975	35996	}
35976	35997	#endif
...	...	@@ -35990,11 +36011,11 @@ __ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t __noswap_vcmla_r
35990	36011	float16x4_t __s0_159 = __p0_159; \
35991	36012	float16x4_t __s1_159 = __p1_159; \
35992	36013	float16x4_t __s2_159 = __p2_159; \
35993		float16x4_t __rev0_159; __rev0_159 = __builtin_shufflevector(__s0_159, __s0_159, 3, 2, 1, 0); \
35994		float16x4_t __rev1_159; __rev1_159 = __builtin_shufflevector(__s1_159, __s1_159, 3, 2, 1, 0); \
35995		float16x4_t __rev2_159; __rev2_159 = __builtin_shufflevector(__s2_159, __s2_159, 3, 2, 1, 0); \
	36014	float16x4_t __rev0_159; __rev0_159 = __builtin_shufflevector(__s0_159, __s0_159, __lane_reverse_64_16); \
	36015	float16x4_t __rev1_159; __rev1_159 = __builtin_shufflevector(__s1_159, __s1_159, __lane_reverse_64_16); \
	36016	float16x4_t __rev2_159; __rev2_159 = __builtin_shufflevector(__s2_159, __s2_159, __lane_reverse_64_16); \
35996	36017	__ret_159 = __noswap_vcmla_rot180_f16(__rev0_159, __rev1_159, __builtin_bit_cast(float16x4_t, (uint32x2_t) {__noswap_vget_lane_u32(__builtin_bit_cast(uint32x2_t, __rev2_159), __p3_159), __noswap_vget_lane_u32(__builtin_bit_cast(uint32x2_t, __rev2_159), __p3_159)})); \
35997		__ret_159 = __builtin_shufflevector(__ret_159, __ret_159, 3, 2, 1, 0); \
	36018	__ret_159 = __builtin_shufflevector(__ret_159, __ret_159, __lane_reverse_64_16); \
35998	36019	__ret_159; \
35999	36020	})
36000	36021	#endif
...	...	@@ -36014,11 +36035,11 @@ __ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t __noswap_vcmla_r
36014	36035	float16x8_t __s0_161 = __p0_161; \
36015	36036	float16x8_t __s1_161 = __p1_161; \
36016	36037	float16x4_t __s2_161 = __p2_161; \
36017		float16x8_t __rev0_161; __rev0_161 = __builtin_shufflevector(__s0_161, __s0_161, 7, 6, 5, 4, 3, 2, 1, 0); \
36018		float16x8_t __rev1_161; __rev1_161 = __builtin_shufflevector(__s1_161, __s1_161, 7, 6, 5, 4, 3, 2, 1, 0); \
36019		float16x4_t __rev2_161; __rev2_161 = __builtin_shufflevector(__s2_161, __s2_161, 3, 2, 1, 0); \
	36038	float16x8_t __rev0_161; __rev0_161 = __builtin_shufflevector(__s0_161, __s0_161, __lane_reverse_128_16); \
	36039	float16x8_t __rev1_161; __rev1_161 = __builtin_shufflevector(__s1_161, __s1_161, __lane_reverse_128_16); \
	36040	float16x4_t __rev2_161; __rev2_161 = __builtin_shufflevector(__s2_161, __s2_161, __lane_reverse_64_16); \
36020	36041	__ret_161 = __noswap_vcmlaq_rot180_f16(__rev0_161, __rev1_161, __builtin_bit_cast(float16x8_t, (uint32x4_t) {__noswap_vget_lane_u32(__builtin_bit_cast(uint32x2_t, __rev2_161), __p3_161), __noswap_vget_lane_u32(__builtin_bit_cast(uint32x2_t, __rev2_161), __p3_161), __noswap_vget_lane_u32(__builtin_bit_cast(uint32x2_t, __rev2_161), __p3_161), __noswap_vget_lane_u32(__builtin_bit_cast(uint32x2_t, __rev2_161), __p3_161)})); \
36021		__ret_161 = __builtin_shufflevector(__ret_161, __ret_161, 7, 6, 5, 4, 3, 2, 1, 0); \
	36042	__ret_161 = __builtin_shufflevector(__ret_161, __ret_161, __lane_reverse_128_16); \
36022	36043	__ret_161; \
36023	36044	})
36024	36045	#endif
...	...	@@ -36038,11 +36059,11 @@ __ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t __noswap_vcmla_r
36038	36059	float16x4_t __s0_163 = __p0_163; \
36039	36060	float16x4_t __s1_163 = __p1_163; \
36040	36061	float16x8_t __s2_163 = __p2_163; \
36041		float16x4_t __rev0_163; __rev0_163 = __builtin_shufflevector(__s0_163, __s0_163, 3, 2, 1, 0); \
36042		float16x4_t __rev1_163; __rev1_163 = __builtin_shufflevector(__s1_163, __s1_163, 3, 2, 1, 0); \
36043		float16x8_t __rev2_163; __rev2_163 = __builtin_shufflevector(__s2_163, __s2_163, 7, 6, 5, 4, 3, 2, 1, 0); \
	36062	float16x4_t __rev0_163; __rev0_163 = __builtin_shufflevector(__s0_163, __s0_163, __lane_reverse_64_16); \
	36063	float16x4_t __rev1_163; __rev1_163 = __builtin_shufflevector(__s1_163, __s1_163, __lane_reverse_64_16); \
	36064	float16x8_t __rev2_163; __rev2_163 = __builtin_shufflevector(__s2_163, __s2_163, __lane_reverse_128_16); \
36044	36065	__ret_163 = __noswap_vcmla_rot180_f16(__rev0_163, __rev1_163, __builtin_bit_cast(float16x4_t, (uint32x2_t) {__noswap_vgetq_lane_u32(__builtin_bit_cast(uint32x4_t, __rev2_163), __p3_163), __noswap_vgetq_lane_u32(__builtin_bit_cast(uint32x4_t, __rev2_163), __p3_163)})); \
36045		__ret_163 = __builtin_shufflevector(__ret_163, __ret_163, 3, 2, 1, 0); \
	36066	__ret_163 = __builtin_shufflevector(__ret_163, __ret_163, __lane_reverse_64_16); \
36046	36067	__ret_163; \
36047	36068	})
36048	36069	#endif
...	...	@@ -36062,11 +36083,11 @@ __ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t __noswap_vcmla_r
36062	36083	float16x8_t __s0_165 = __p0_165; \
36063	36084	float16x8_t __s1_165 = __p1_165; \
36064	36085	float16x8_t __s2_165 = __p2_165; \
36065		float16x8_t __rev0_165; __rev0_165 = __builtin_shufflevector(__s0_165, __s0_165, 7, 6, 5, 4, 3, 2, 1, 0); \
36066		float16x8_t __rev1_165; __rev1_165 = __builtin_shufflevector(__s1_165, __s1_165, 7, 6, 5, 4, 3, 2, 1, 0); \
36067		float16x8_t __rev2_165; __rev2_165 = __builtin_shufflevector(__s2_165, __s2_165, 7, 6, 5, 4, 3, 2, 1, 0); \
	36086	float16x8_t __rev0_165; __rev0_165 = __builtin_shufflevector(__s0_165, __s0_165, __lane_reverse_128_16); \
	36087	float16x8_t __rev1_165; __rev1_165 = __builtin_shufflevector(__s1_165, __s1_165, __lane_reverse_128_16); \
	36088	float16x8_t __rev2_165; __rev2_165 = __builtin_shufflevector(__s2_165, __s2_165, __lane_reverse_128_16); \
36068	36089	__ret_165 = __noswap_vcmlaq_rot180_f16(__rev0_165, __rev1_165, __builtin_bit_cast(float16x8_t, (uint32x4_t) {__noswap_vgetq_lane_u32(__builtin_bit_cast(uint32x4_t, __rev2_165), __p3_165), __noswap_vgetq_lane_u32(__builtin_bit_cast(uint32x4_t, __rev2_165), __p3_165), __noswap_vgetq_lane_u32(__builtin_bit_cast(uint32x4_t, __rev2_165), __p3_165), __noswap_vgetq_lane_u32(__builtin_bit_cast(uint32x4_t, __rev2_165), __p3_165)})); \
36069		__ret_165 = __builtin_shufflevector(__ret_165, __ret_165, 7, 6, 5, 4, 3, 2, 1, 0); \
	36090	__ret_165 = __builtin_shufflevector(__ret_165, __ret_165, __lane_reverse_128_16); \
36070	36091	__ret_165; \
36071	36092	})
36072	36093	#endif
...	...	@@ -36074,22 +36095,22 @@ __ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t __noswap_vcmla_r
36074	36095	#ifdef __LITTLE_ENDIAN__
36075	36096	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x8_t vcmlaq_rot270_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
36076	36097	float16x8_t __ret;
36077		__ret = (float16x8_t) __builtin_neon_vcmlaq_rot270_f16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 40);
	36098	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcmlaq_rot270_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 40));
36078	36099	return __ret;
36079	36100	}
36080	36101	#else
36081	36102	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x8_t vcmlaq_rot270_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
36082	36103	float16x8_t __ret;
36083		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
36084		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
36085		float16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
36086		__ret = (float16x8_t) __builtin_neon_vcmlaq_rot270_f16((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 40);
36087		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	36104	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	36105	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	36106	float16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
	36107	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcmlaq_rot270_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 40));
	36108	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
36088	36109	return __ret;
36089	36110	}
36090	36111	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x8_t __noswap_vcmlaq_rot270_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
36091	36112	float16x8_t __ret;
36092		__ret = (float16x8_t) __builtin_neon_vcmlaq_rot270_f16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 40);
	36113	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcmlaq_rot270_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 40));
36093	36114	return __ret;
36094	36115	}
36095	36116	#endif
...	...	@@ -36097,22 +36118,22 @@ __ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x8_t __noswap_vcmlaq_
36097	36118	#ifdef __LITTLE_ENDIAN__
36098	36119	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t vcmla_rot270_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
36099	36120	float16x4_t __ret;
36100		__ret = (float16x4_t) __builtin_neon_vcmla_rot270_f16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 8);
	36121	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vcmla_rot270_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 8));
36101	36122	return __ret;
36102	36123	}
36103	36124	#else
36104	36125	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t vcmla_rot270_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
36105	36126	float16x4_t __ret;
36106		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
36107		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
36108		float16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
36109		__ret = (float16x4_t) __builtin_neon_vcmla_rot270_f16((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 8);
36110		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	36127	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	36128	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	36129	float16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
	36130	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vcmla_rot270_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 8));
	36131	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
36111	36132	return __ret;
36112	36133	}
36113	36134	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t __noswap_vcmla_rot270_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
36114	36135	float16x4_t __ret;
36115		__ret = (float16x4_t) __builtin_neon_vcmla_rot270_f16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 8);
	36136	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vcmla_rot270_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 8));
36116	36137	return __ret;
36117	36138	}
36118	36139	#endif
...	...	@@ -36132,11 +36153,11 @@ __ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t __noswap_vcmla_r
36132	36153	float16x4_t __s0_167 = __p0_167; \
36133	36154	float16x4_t __s1_167 = __p1_167; \
36134	36155	float16x4_t __s2_167 = __p2_167; \
36135		float16x4_t __rev0_167; __rev0_167 = __builtin_shufflevector(__s0_167, __s0_167, 3, 2, 1, 0); \
36136		float16x4_t __rev1_167; __rev1_167 = __builtin_shufflevector(__s1_167, __s1_167, 3, 2, 1, 0); \
36137		float16x4_t __rev2_167; __rev2_167 = __builtin_shufflevector(__s2_167, __s2_167, 3, 2, 1, 0); \
	36156	float16x4_t __rev0_167; __rev0_167 = __builtin_shufflevector(__s0_167, __s0_167, __lane_reverse_64_16); \
	36157	float16x4_t __rev1_167; __rev1_167 = __builtin_shufflevector(__s1_167, __s1_167, __lane_reverse_64_16); \
	36158	float16x4_t __rev2_167; __rev2_167 = __builtin_shufflevector(__s2_167, __s2_167, __lane_reverse_64_16); \
36138	36159	__ret_167 = __noswap_vcmla_rot270_f16(__rev0_167, __rev1_167, __builtin_bit_cast(float16x4_t, (uint32x2_t) {__noswap_vget_lane_u32(__builtin_bit_cast(uint32x2_t, __rev2_167), __p3_167), __noswap_vget_lane_u32(__builtin_bit_cast(uint32x2_t, __rev2_167), __p3_167)})); \
36139		__ret_167 = __builtin_shufflevector(__ret_167, __ret_167, 3, 2, 1, 0); \
	36160	__ret_167 = __builtin_shufflevector(__ret_167, __ret_167, __lane_reverse_64_16); \
36140	36161	__ret_167; \
36141	36162	})
36142	36163	#endif
...	...	@@ -36156,11 +36177,11 @@ __ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t __noswap_vcmla_r
36156	36177	float16x8_t __s0_169 = __p0_169; \
36157	36178	float16x8_t __s1_169 = __p1_169; \
36158	36179	float16x4_t __s2_169 = __p2_169; \
36159		float16x8_t __rev0_169; __rev0_169 = __builtin_shufflevector(__s0_169, __s0_169, 7, 6, 5, 4, 3, 2, 1, 0); \
36160		float16x8_t __rev1_169; __rev1_169 = __builtin_shufflevector(__s1_169, __s1_169, 7, 6, 5, 4, 3, 2, 1, 0); \
36161		float16x4_t __rev2_169; __rev2_169 = __builtin_shufflevector(__s2_169, __s2_169, 3, 2, 1, 0); \
	36180	float16x8_t __rev0_169; __rev0_169 = __builtin_shufflevector(__s0_169, __s0_169, __lane_reverse_128_16); \
	36181	float16x8_t __rev1_169; __rev1_169 = __builtin_shufflevector(__s1_169, __s1_169, __lane_reverse_128_16); \
	36182	float16x4_t __rev2_169; __rev2_169 = __builtin_shufflevector(__s2_169, __s2_169, __lane_reverse_64_16); \
36162	36183	__ret_169 = __noswap_vcmlaq_rot270_f16(__rev0_169, __rev1_169, __builtin_bit_cast(float16x8_t, (uint32x4_t) {__noswap_vget_lane_u32(__builtin_bit_cast(uint32x2_t, __rev2_169), __p3_169), __noswap_vget_lane_u32(__builtin_bit_cast(uint32x2_t, __rev2_169), __p3_169), __noswap_vget_lane_u32(__builtin_bit_cast(uint32x2_t, __rev2_169), __p3_169), __noswap_vget_lane_u32(__builtin_bit_cast(uint32x2_t, __rev2_169), __p3_169)})); \
36163		__ret_169 = __builtin_shufflevector(__ret_169, __ret_169, 7, 6, 5, 4, 3, 2, 1, 0); \
	36184	__ret_169 = __builtin_shufflevector(__ret_169, __ret_169, __lane_reverse_128_16); \
36164	36185	__ret_169; \
36165	36186	})
36166	36187	#endif
...	...	@@ -36180,11 +36201,11 @@ __ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t __noswap_vcmla_r
36180	36201	float16x4_t __s0_171 = __p0_171; \
36181	36202	float16x4_t __s1_171 = __p1_171; \
36182	36203	float16x8_t __s2_171 = __p2_171; \
36183		float16x4_t __rev0_171; __rev0_171 = __builtin_shufflevector(__s0_171, __s0_171, 3, 2, 1, 0); \
36184		float16x4_t __rev1_171; __rev1_171 = __builtin_shufflevector(__s1_171, __s1_171, 3, 2, 1, 0); \
36185		float16x8_t __rev2_171; __rev2_171 = __builtin_shufflevector(__s2_171, __s2_171, 7, 6, 5, 4, 3, 2, 1, 0); \
	36204	float16x4_t __rev0_171; __rev0_171 = __builtin_shufflevector(__s0_171, __s0_171, __lane_reverse_64_16); \
	36205	float16x4_t __rev1_171; __rev1_171 = __builtin_shufflevector(__s1_171, __s1_171, __lane_reverse_64_16); \
	36206	float16x8_t __rev2_171; __rev2_171 = __builtin_shufflevector(__s2_171, __s2_171, __lane_reverse_128_16); \
36186	36207	__ret_171 = __noswap_vcmla_rot270_f16(__rev0_171, __rev1_171, __builtin_bit_cast(float16x4_t, (uint32x2_t) {__noswap_vgetq_lane_u32(__builtin_bit_cast(uint32x4_t, __rev2_171), __p3_171), __noswap_vgetq_lane_u32(__builtin_bit_cast(uint32x4_t, __rev2_171), __p3_171)})); \
36187		__ret_171 = __builtin_shufflevector(__ret_171, __ret_171, 3, 2, 1, 0); \
	36208	__ret_171 = __builtin_shufflevector(__ret_171, __ret_171, __lane_reverse_64_16); \
36188	36209	__ret_171; \
36189	36210	})
36190	36211	#endif
...	...	@@ -36204,11 +36225,11 @@ __ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t __noswap_vcmla_r
36204	36225	float16x8_t __s0_173 = __p0_173; \
36205	36226	float16x8_t __s1_173 = __p1_173; \
36206	36227	float16x8_t __s2_173 = __p2_173; \
36207		float16x8_t __rev0_173; __rev0_173 = __builtin_shufflevector(__s0_173, __s0_173, 7, 6, 5, 4, 3, 2, 1, 0); \
36208		float16x8_t __rev1_173; __rev1_173 = __builtin_shufflevector(__s1_173, __s1_173, 7, 6, 5, 4, 3, 2, 1, 0); \
36209		float16x8_t __rev2_173; __rev2_173 = __builtin_shufflevector(__s2_173, __s2_173, 7, 6, 5, 4, 3, 2, 1, 0); \
	36228	float16x8_t __rev0_173; __rev0_173 = __builtin_shufflevector(__s0_173, __s0_173, __lane_reverse_128_16); \
	36229	float16x8_t __rev1_173; __rev1_173 = __builtin_shufflevector(__s1_173, __s1_173, __lane_reverse_128_16); \
	36230	float16x8_t __rev2_173; __rev2_173 = __builtin_shufflevector(__s2_173, __s2_173, __lane_reverse_128_16); \
36210	36231	__ret_173 = __noswap_vcmlaq_rot270_f16(__rev0_173, __rev1_173, __builtin_bit_cast(float16x8_t, (uint32x4_t) {__noswap_vgetq_lane_u32(__builtin_bit_cast(uint32x4_t, __rev2_173), __p3_173), __noswap_vgetq_lane_u32(__builtin_bit_cast(uint32x4_t, __rev2_173), __p3_173), __noswap_vgetq_lane_u32(__builtin_bit_cast(uint32x4_t, __rev2_173), __p3_173), __noswap_vgetq_lane_u32(__builtin_bit_cast(uint32x4_t, __rev2_173), __p3_173)})); \
36211		__ret_173 = __builtin_shufflevector(__ret_173, __ret_173, 7, 6, 5, 4, 3, 2, 1, 0); \
	36232	__ret_173 = __builtin_shufflevector(__ret_173, __ret_173, __lane_reverse_128_16); \
36212	36233	__ret_173; \
36213	36234	})
36214	36235	#endif
...	...	@@ -36216,22 +36237,22 @@ __ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t __noswap_vcmla_r
36216	36237	#ifdef __LITTLE_ENDIAN__
36217	36238	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x8_t vcmlaq_rot90_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
36218	36239	float16x8_t __ret;
36219		__ret = (float16x8_t) __builtin_neon_vcmlaq_rot90_f16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 40);
	36240	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcmlaq_rot90_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 40));
36220	36241	return __ret;
36221	36242	}
36222	36243	#else
36223	36244	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x8_t vcmlaq_rot90_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
36224	36245	float16x8_t __ret;
36225		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
36226		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
36227		float16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
36228		__ret = (float16x8_t) __builtin_neon_vcmlaq_rot90_f16((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 40);
36229		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	36246	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	36247	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	36248	float16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
	36249	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcmlaq_rot90_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 40));
	36250	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
36230	36251	return __ret;
36231	36252	}
36232	36253	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x8_t __noswap_vcmlaq_rot90_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
36233	36254	float16x8_t __ret;
36234		__ret = (float16x8_t) __builtin_neon_vcmlaq_rot90_f16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 40);
	36255	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcmlaq_rot90_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 40));
36235	36256	return __ret;
36236	36257	}
36237	36258	#endif
...	...	@@ -36239,22 +36260,22 @@ __ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x8_t __noswap_vcmlaq_
36239	36260	#ifdef __LITTLE_ENDIAN__
36240	36261	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t vcmla_rot90_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
36241	36262	float16x4_t __ret;
36242		__ret = (float16x4_t) __builtin_neon_vcmla_rot90_f16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 8);
	36263	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vcmla_rot90_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 8));
36243	36264	return __ret;
36244	36265	}
36245	36266	#else
36246	36267	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t vcmla_rot90_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
36247	36268	float16x4_t __ret;
36248		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
36249		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
36250		float16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
36251		__ret = (float16x4_t) __builtin_neon_vcmla_rot90_f16((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 8);
36252		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	36269	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	36270	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	36271	float16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
	36272	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vcmla_rot90_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 8));
	36273	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
36253	36274	return __ret;
36254	36275	}
36255	36276	__ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t __noswap_vcmla_rot90_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
36256	36277	float16x4_t __ret;
36257		__ret = (float16x4_t) __builtin_neon_vcmla_rot90_f16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 8);
	36278	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vcmla_rot90_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 8));
36258	36279	return __ret;
36259	36280	}
36260	36281	#endif
...	...	@@ -36274,11 +36295,11 @@ __ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t __noswap_vcmla_r
36274	36295	float16x4_t __s0_175 = __p0_175; \
36275	36296	float16x4_t __s1_175 = __p1_175; \
36276	36297	float16x4_t __s2_175 = __p2_175; \
36277		float16x4_t __rev0_175; __rev0_175 = __builtin_shufflevector(__s0_175, __s0_175, 3, 2, 1, 0); \
36278		float16x4_t __rev1_175; __rev1_175 = __builtin_shufflevector(__s1_175, __s1_175, 3, 2, 1, 0); \
36279		float16x4_t __rev2_175; __rev2_175 = __builtin_shufflevector(__s2_175, __s2_175, 3, 2, 1, 0); \
	36298	float16x4_t __rev0_175; __rev0_175 = __builtin_shufflevector(__s0_175, __s0_175, __lane_reverse_64_16); \
	36299	float16x4_t __rev1_175; __rev1_175 = __builtin_shufflevector(__s1_175, __s1_175, __lane_reverse_64_16); \
	36300	float16x4_t __rev2_175; __rev2_175 = __builtin_shufflevector(__s2_175, __s2_175, __lane_reverse_64_16); \
36280	36301	__ret_175 = __noswap_vcmla_rot90_f16(__rev0_175, __rev1_175, __builtin_bit_cast(float16x4_t, (uint32x2_t) {__noswap_vget_lane_u32(__builtin_bit_cast(uint32x2_t, __rev2_175), __p3_175), __noswap_vget_lane_u32(__builtin_bit_cast(uint32x2_t, __rev2_175), __p3_175)})); \
36281		__ret_175 = __builtin_shufflevector(__ret_175, __ret_175, 3, 2, 1, 0); \
	36302	__ret_175 = __builtin_shufflevector(__ret_175, __ret_175, __lane_reverse_64_16); \
36282	36303	__ret_175; \
36283	36304	})
36284	36305	#endif
...	...	@@ -36298,11 +36319,11 @@ __ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t __noswap_vcmla_r
36298	36319	float16x8_t __s0_177 = __p0_177; \
36299	36320	float16x8_t __s1_177 = __p1_177; \
36300	36321	float16x4_t __s2_177 = __p2_177; \
36301		float16x8_t __rev0_177; __rev0_177 = __builtin_shufflevector(__s0_177, __s0_177, 7, 6, 5, 4, 3, 2, 1, 0); \
36302		float16x8_t __rev1_177; __rev1_177 = __builtin_shufflevector(__s1_177, __s1_177, 7, 6, 5, 4, 3, 2, 1, 0); \
36303		float16x4_t __rev2_177; __rev2_177 = __builtin_shufflevector(__s2_177, __s2_177, 3, 2, 1, 0); \
	36322	float16x8_t __rev0_177; __rev0_177 = __builtin_shufflevector(__s0_177, __s0_177, __lane_reverse_128_16); \
	36323	float16x8_t __rev1_177; __rev1_177 = __builtin_shufflevector(__s1_177, __s1_177, __lane_reverse_128_16); \
	36324	float16x4_t __rev2_177; __rev2_177 = __builtin_shufflevector(__s2_177, __s2_177, __lane_reverse_64_16); \
36304	36325	__ret_177 = __noswap_vcmlaq_rot90_f16(__rev0_177, __rev1_177, __builtin_bit_cast(float16x8_t, (uint32x4_t) {__noswap_vget_lane_u32(__builtin_bit_cast(uint32x2_t, __rev2_177), __p3_177), __noswap_vget_lane_u32(__builtin_bit_cast(uint32x2_t, __rev2_177), __p3_177), __noswap_vget_lane_u32(__builtin_bit_cast(uint32x2_t, __rev2_177), __p3_177), __noswap_vget_lane_u32(__builtin_bit_cast(uint32x2_t, __rev2_177), __p3_177)})); \
36305		__ret_177 = __builtin_shufflevector(__ret_177, __ret_177, 7, 6, 5, 4, 3, 2, 1, 0); \
	36326	__ret_177 = __builtin_shufflevector(__ret_177, __ret_177, __lane_reverse_128_16); \
36306	36327	__ret_177; \
36307	36328	})
36308	36329	#endif
...	...	@@ -36322,11 +36343,11 @@ __ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t __noswap_vcmla_r
36322	36343	float16x4_t __s0_179 = __p0_179; \
36323	36344	float16x4_t __s1_179 = __p1_179; \
36324	36345	float16x8_t __s2_179 = __p2_179; \
36325		float16x4_t __rev0_179; __rev0_179 = __builtin_shufflevector(__s0_179, __s0_179, 3, 2, 1, 0); \
36326		float16x4_t __rev1_179; __rev1_179 = __builtin_shufflevector(__s1_179, __s1_179, 3, 2, 1, 0); \
36327		float16x8_t __rev2_179; __rev2_179 = __builtin_shufflevector(__s2_179, __s2_179, 7, 6, 5, 4, 3, 2, 1, 0); \
	36346	float16x4_t __rev0_179; __rev0_179 = __builtin_shufflevector(__s0_179, __s0_179, __lane_reverse_64_16); \
	36347	float16x4_t __rev1_179; __rev1_179 = __builtin_shufflevector(__s1_179, __s1_179, __lane_reverse_64_16); \
	36348	float16x8_t __rev2_179; __rev2_179 = __builtin_shufflevector(__s2_179, __s2_179, __lane_reverse_128_16); \
36328	36349	__ret_179 = __noswap_vcmla_rot90_f16(__rev0_179, __rev1_179, __builtin_bit_cast(float16x4_t, (uint32x2_t) {__noswap_vgetq_lane_u32(__builtin_bit_cast(uint32x4_t, __rev2_179), __p3_179), __noswap_vgetq_lane_u32(__builtin_bit_cast(uint32x4_t, __rev2_179), __p3_179)})); \
36329		__ret_179 = __builtin_shufflevector(__ret_179, __ret_179, 3, 2, 1, 0); \
	36350	__ret_179 = __builtin_shufflevector(__ret_179, __ret_179, __lane_reverse_64_16); \
36330	36351	__ret_179; \
36331	36352	})
36332	36353	#endif
...	...	@@ -36346,11 +36367,11 @@ __ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t __noswap_vcmla_r
36346	36367	float16x8_t __s0_181 = __p0_181; \
36347	36368	float16x8_t __s1_181 = __p1_181; \
36348	36369	float16x8_t __s2_181 = __p2_181; \
36349		float16x8_t __rev0_181; __rev0_181 = __builtin_shufflevector(__s0_181, __s0_181, 7, 6, 5, 4, 3, 2, 1, 0); \
36350		float16x8_t __rev1_181; __rev1_181 = __builtin_shufflevector(__s1_181, __s1_181, 7, 6, 5, 4, 3, 2, 1, 0); \
36351		float16x8_t __rev2_181; __rev2_181 = __builtin_shufflevector(__s2_181, __s2_181, 7, 6, 5, 4, 3, 2, 1, 0); \
	36370	float16x8_t __rev0_181; __rev0_181 = __builtin_shufflevector(__s0_181, __s0_181, __lane_reverse_128_16); \
	36371	float16x8_t __rev1_181; __rev1_181 = __builtin_shufflevector(__s1_181, __s1_181, __lane_reverse_128_16); \
	36372	float16x8_t __rev2_181; __rev2_181 = __builtin_shufflevector(__s2_181, __s2_181, __lane_reverse_128_16); \
36352	36373	__ret_181 = __noswap_vcmlaq_rot90_f16(__rev0_181, __rev1_181, __builtin_bit_cast(float16x8_t, (uint32x4_t) {__noswap_vgetq_lane_u32(__builtin_bit_cast(uint32x4_t, __rev2_181), __p3_181), __noswap_vgetq_lane_u32(__builtin_bit_cast(uint32x4_t, __rev2_181), __p3_181), __noswap_vgetq_lane_u32(__builtin_bit_cast(uint32x4_t, __rev2_181), __p3_181), __noswap_vgetq_lane_u32(__builtin_bit_cast(uint32x4_t, __rev2_181), __p3_181)})); \
36353		__ret_181 = __builtin_shufflevector(__ret_181, __ret_181, 7, 6, 5, 4, 3, 2, 1, 0); \
	36374	__ret_181 = __builtin_shufflevector(__ret_181, __ret_181, __lane_reverse_128_16); \
36354	36375	__ret_181; \
36355	36376	})
36356	36377	#endif
...	...	@@ -36358,16 +36379,16 @@ __ai __attribute__((target("v8.3a,fullfp16,neon"))) float16x4_t __noswap_vcmla_r
36358	36379	#ifdef __LITTLE_ENDIAN__
36359	36380	__ai __attribute__((target("v8.3a,neon"))) float32x2_t vcadd_rot270_f32(float32x2_t __p0, float32x2_t __p1) {
36360	36381	float32x2_t __ret;
36361		__ret = (float32x2_t) __builtin_neon_vcadd_rot270_f32((int8x8_t)__p0, (int8x8_t)__p1, 9);
	36382	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vcadd_rot270_f32(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 9));
36362	36383	return __ret;
36363	36384	}
36364	36385	#else
36365	36386	__ai __attribute__((target("v8.3a,neon"))) float32x2_t vcadd_rot270_f32(float32x2_t __p0, float32x2_t __p1) {
36366	36387	float32x2_t __ret;
36367		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
36368		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
36369		__ret = (float32x2_t) __builtin_neon_vcadd_rot270_f32((int8x8_t)__rev0, (int8x8_t)__rev1, 9);
36370		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	36388	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	36389	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	36390	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vcadd_rot270_f32(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 9));
	36391	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
36371	36392	return __ret;
36372	36393	}
36373	36394	#endif
...	...	@@ -36375,16 +36396,16 @@ __ai __attribute__((target("v8.3a,neon"))) float32x2_t vcadd_rot270_f32(float32x
36375	36396	#ifdef __LITTLE_ENDIAN__
36376	36397	__ai __attribute__((target("v8.3a,neon"))) float32x2_t vcadd_rot90_f32(float32x2_t __p0, float32x2_t __p1) {
36377	36398	float32x2_t __ret;
36378		__ret = (float32x2_t) __builtin_neon_vcadd_rot90_f32((int8x8_t)__p0, (int8x8_t)__p1, 9);
	36399	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vcadd_rot90_f32(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 9));
36379	36400	return __ret;
36380	36401	}
36381	36402	#else
36382	36403	__ai __attribute__((target("v8.3a,neon"))) float32x2_t vcadd_rot90_f32(float32x2_t __p0, float32x2_t __p1) {
36383	36404	float32x2_t __ret;
36384		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
36385		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
36386		__ret = (float32x2_t) __builtin_neon_vcadd_rot90_f32((int8x8_t)__rev0, (int8x8_t)__rev1, 9);
36387		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	36405	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	36406	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	36407	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vcadd_rot90_f32(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 9));
	36408	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
36388	36409	return __ret;
36389	36410	}
36390	36411	#endif
...	...	@@ -36392,16 +36413,16 @@ __ai __attribute__((target("v8.3a,neon"))) float32x2_t vcadd_rot90_f32(float32x2
36392	36413	#ifdef __LITTLE_ENDIAN__
36393	36414	__ai __attribute__((target("v8.3a,neon"))) float32x4_t vcaddq_rot270_f32(float32x4_t __p0, float32x4_t __p1) {
36394	36415	float32x4_t __ret;
36395		__ret = (float32x4_t) __builtin_neon_vcaddq_rot270_f32((int8x16_t)__p0, (int8x16_t)__p1, 41);
	36416	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vcaddq_rot270_f32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 41));
36396	36417	return __ret;
36397	36418	}
36398	36419	#else
36399	36420	__ai __attribute__((target("v8.3a,neon"))) float32x4_t vcaddq_rot270_f32(float32x4_t __p0, float32x4_t __p1) {
36400	36421	float32x4_t __ret;
36401		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
36402		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
36403		__ret = (float32x4_t) __builtin_neon_vcaddq_rot270_f32((int8x16_t)__rev0, (int8x16_t)__rev1, 41);
36404		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	36422	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	36423	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	36424	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vcaddq_rot270_f32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 41));
	36425	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
36405	36426	return __ret;
36406	36427	}
36407	36428	#endif
...	...	@@ -36409,16 +36430,16 @@ __ai __attribute__((target("v8.3a,neon"))) float32x4_t vcaddq_rot270_f32(float32
36409	36430	#ifdef __LITTLE_ENDIAN__
36410	36431	__ai __attribute__((target("v8.3a,neon"))) float32x4_t vcaddq_rot90_f32(float32x4_t __p0, float32x4_t __p1) {
36411	36432	float32x4_t __ret;
36412		__ret = (float32x4_t) __builtin_neon_vcaddq_rot90_f32((int8x16_t)__p0, (int8x16_t)__p1, 41);
	36433	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vcaddq_rot90_f32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 41));
36413	36434	return __ret;
36414	36435	}
36415	36436	#else
36416	36437	__ai __attribute__((target("v8.3a,neon"))) float32x4_t vcaddq_rot90_f32(float32x4_t __p0, float32x4_t __p1) {
36417	36438	float32x4_t __ret;
36418		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
36419		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
36420		__ret = (float32x4_t) __builtin_neon_vcaddq_rot90_f32((int8x16_t)__rev0, (int8x16_t)__rev1, 41);
36421		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	36439	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	36440	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	36441	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vcaddq_rot90_f32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 41));
	36442	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
36422	36443	return __ret;
36423	36444	}
36424	36445	#endif
...	...	@@ -36426,22 +36447,22 @@ __ai __attribute__((target("v8.3a,neon"))) float32x4_t vcaddq_rot90_f32(float32x
36426	36447	#ifdef __LITTLE_ENDIAN__
36427	36448	__ai __attribute__((target("v8.3a,neon"))) float32x4_t vcmlaq_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
36428	36449	float32x4_t __ret;
36429		__ret = (float32x4_t) __builtin_neon_vcmlaq_f32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
	36450	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vcmlaq_f32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 41));
36430	36451	return __ret;
36431	36452	}
36432	36453	#else
36433	36454	__ai __attribute__((target("v8.3a,neon"))) float32x4_t vcmlaq_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
36434	36455	float32x4_t __ret;
36435		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
36436		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
36437		float32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
36438		__ret = (float32x4_t) __builtin_neon_vcmlaq_f32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 41);
36439		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	36456	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	36457	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	36458	float32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
	36459	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vcmlaq_f32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 41));
	36460	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
36440	36461	return __ret;
36441	36462	}
36442	36463	__ai __attribute__((target("v8.3a,neon"))) float32x4_t __noswap_vcmlaq_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
36443	36464	float32x4_t __ret;
36444		__ret = (float32x4_t) __builtin_neon_vcmlaq_f32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
	36465	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vcmlaq_f32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 41));
36445	36466	return __ret;
36446	36467	}
36447	36468	#endif
...	...	@@ -36449,22 +36470,22 @@ __ai __attribute__((target("v8.3a,neon"))) float32x4_t __noswap_vcmlaq_f32(float
36449	36470	#ifdef __LITTLE_ENDIAN__
36450	36471	__ai __attribute__((target("v8.3a,neon"))) float32x2_t vcmla_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
36451	36472	float32x2_t __ret;
36452		__ret = (float32x2_t) __builtin_neon_vcmla_f32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
	36473	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vcmla_f32(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 9));
36453	36474	return __ret;
36454	36475	}
36455	36476	#else
36456	36477	__ai __attribute__((target("v8.3a,neon"))) float32x2_t vcmla_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
36457	36478	float32x2_t __ret;
36458		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
36459		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
36460		float32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
36461		__ret = (float32x2_t) __builtin_neon_vcmla_f32((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 9);
36462		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	36479	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	36480	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	36481	float32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_32);
	36482	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vcmla_f32(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 9));
	36483	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
36463	36484	return __ret;
36464	36485	}
36465	36486	__ai __attribute__((target("v8.3a,neon"))) float32x2_t __noswap_vcmla_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
36466	36487	float32x2_t __ret;
36467		__ret = (float32x2_t) __builtin_neon_vcmla_f32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
	36488	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vcmla_f32(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 9));
36468	36489	return __ret;
36469	36490	}
36470	36491	#endif
...	...	@@ -36484,11 +36505,11 @@ __ai __attribute__((target("v8.3a,neon"))) float32x2_t __noswap_vcmla_f32(float3
36484	36505	float32x2_t __s0_183 = __p0_183; \
36485	36506	float32x2_t __s1_183 = __p1_183; \
36486	36507	float32x2_t __s2_183 = __p2_183; \
36487		float32x2_t __rev0_183; __rev0_183 = __builtin_shufflevector(__s0_183, __s0_183, 1, 0); \
36488		float32x2_t __rev1_183; __rev1_183 = __builtin_shufflevector(__s1_183, __s1_183, 1, 0); \
36489		float32x2_t __rev2_183; __rev2_183 = __builtin_shufflevector(__s2_183, __s2_183, 1, 0); \
	36508	float32x2_t __rev0_183; __rev0_183 = __builtin_shufflevector(__s0_183, __s0_183, __lane_reverse_64_32); \
	36509	float32x2_t __rev1_183; __rev1_183 = __builtin_shufflevector(__s1_183, __s1_183, __lane_reverse_64_32); \
	36510	float32x2_t __rev2_183; __rev2_183 = __builtin_shufflevector(__s2_183, __s2_183, __lane_reverse_64_32); \
36490	36511	__ret_183 = __noswap_vcmla_f32(__rev0_183, __rev1_183, __builtin_bit_cast(float32x2_t, (uint64x1_t) {vget_lane_u64(__builtin_bit_cast(uint64x1_t, __rev2_183), __p3_183)})); \
36491		__ret_183 = __builtin_shufflevector(__ret_183, __ret_183, 1, 0); \
	36512	__ret_183 = __builtin_shufflevector(__ret_183, __ret_183, __lane_reverse_64_32); \
36492	36513	__ret_183; \
36493	36514	})
36494	36515	#endif
...	...	@@ -36508,11 +36529,11 @@ __ai __attribute__((target("v8.3a,neon"))) float32x2_t __noswap_vcmla_f32(float3
36508	36529	float32x4_t __s0_185 = __p0_185; \
36509	36530	float32x4_t __s1_185 = __p1_185; \
36510	36531	float32x2_t __s2_185 = __p2_185; \
36511		float32x4_t __rev0_185; __rev0_185 = __builtin_shufflevector(__s0_185, __s0_185, 3, 2, 1, 0); \
36512		float32x4_t __rev1_185; __rev1_185 = __builtin_shufflevector(__s1_185, __s1_185, 3, 2, 1, 0); \
36513		float32x2_t __rev2_185; __rev2_185 = __builtin_shufflevector(__s2_185, __s2_185, 1, 0); \
	36532	float32x4_t __rev0_185; __rev0_185 = __builtin_shufflevector(__s0_185, __s0_185, __lane_reverse_128_32); \
	36533	float32x4_t __rev1_185; __rev1_185 = __builtin_shufflevector(__s1_185, __s1_185, __lane_reverse_128_32); \
	36534	float32x2_t __rev2_185; __rev2_185 = __builtin_shufflevector(__s2_185, __s2_185, __lane_reverse_64_32); \
36514	36535	__ret_185 = __noswap_vcmlaq_f32(__rev0_185, __rev1_185, __builtin_bit_cast(float32x4_t, (uint64x2_t) {vget_lane_u64(__builtin_bit_cast(uint64x1_t, __rev2_185), __p3_185), vget_lane_u64(__builtin_bit_cast(uint64x1_t, __rev2_185), __p3_185)})); \
36515		__ret_185 = __builtin_shufflevector(__ret_185, __ret_185, 3, 2, 1, 0); \
	36536	__ret_185 = __builtin_shufflevector(__ret_185, __ret_185, __lane_reverse_128_32); \
36516	36537	__ret_185; \
36517	36538	})
36518	36539	#endif
...	...	@@ -36532,11 +36553,11 @@ __ai __attribute__((target("v8.3a,neon"))) float32x2_t __noswap_vcmla_f32(float3
36532	36553	float32x2_t __s0_187 = __p0_187; \
36533	36554	float32x2_t __s1_187 = __p1_187; \
36534	36555	float32x4_t __s2_187 = __p2_187; \
36535		float32x2_t __rev0_187; __rev0_187 = __builtin_shufflevector(__s0_187, __s0_187, 1, 0); \
36536		float32x2_t __rev1_187; __rev1_187 = __builtin_shufflevector(__s1_187, __s1_187, 1, 0); \
36537		float32x4_t __rev2_187; __rev2_187 = __builtin_shufflevector(__s2_187, __s2_187, 3, 2, 1, 0); \
	36556	float32x2_t __rev0_187; __rev0_187 = __builtin_shufflevector(__s0_187, __s0_187, __lane_reverse_64_32); \
	36557	float32x2_t __rev1_187; __rev1_187 = __builtin_shufflevector(__s1_187, __s1_187, __lane_reverse_64_32); \
	36558	float32x4_t __rev2_187; __rev2_187 = __builtin_shufflevector(__s2_187, __s2_187, __lane_reverse_128_32); \
36538	36559	__ret_187 = __noswap_vcmla_f32(__rev0_187, __rev1_187, __builtin_bit_cast(float32x2_t, (uint64x1_t) {__noswap_vgetq_lane_u64(__builtin_bit_cast(uint64x2_t, __rev2_187), __p3_187)})); \
36539		__ret_187 = __builtin_shufflevector(__ret_187, __ret_187, 1, 0); \
	36560	__ret_187 = __builtin_shufflevector(__ret_187, __ret_187, __lane_reverse_64_32); \
36540	36561	__ret_187; \
36541	36562	})
36542	36563	#endif
...	...	@@ -36556,11 +36577,11 @@ __ai __attribute__((target("v8.3a,neon"))) float32x2_t __noswap_vcmla_f32(float3
36556	36577	float32x4_t __s0_189 = __p0_189; \
36557	36578	float32x4_t __s1_189 = __p1_189; \
36558	36579	float32x4_t __s2_189 = __p2_189; \
36559		float32x4_t __rev0_189; __rev0_189 = __builtin_shufflevector(__s0_189, __s0_189, 3, 2, 1, 0); \
36560		float32x4_t __rev1_189; __rev1_189 = __builtin_shufflevector(__s1_189, __s1_189, 3, 2, 1, 0); \
36561		float32x4_t __rev2_189; __rev2_189 = __builtin_shufflevector(__s2_189, __s2_189, 3, 2, 1, 0); \
	36580	float32x4_t __rev0_189; __rev0_189 = __builtin_shufflevector(__s0_189, __s0_189, __lane_reverse_128_32); \
	36581	float32x4_t __rev1_189; __rev1_189 = __builtin_shufflevector(__s1_189, __s1_189, __lane_reverse_128_32); \
	36582	float32x4_t __rev2_189; __rev2_189 = __builtin_shufflevector(__s2_189, __s2_189, __lane_reverse_128_32); \
36562	36583	__ret_189 = __noswap_vcmlaq_f32(__rev0_189, __rev1_189, __builtin_bit_cast(float32x4_t, (uint64x2_t) {__noswap_vgetq_lane_u64(__builtin_bit_cast(uint64x2_t, __rev2_189), __p3_189), __noswap_vgetq_lane_u64(__builtin_bit_cast(uint64x2_t, __rev2_189), __p3_189)})); \
36563		__ret_189 = __builtin_shufflevector(__ret_189, __ret_189, 3, 2, 1, 0); \
	36584	__ret_189 = __builtin_shufflevector(__ret_189, __ret_189, __lane_reverse_128_32); \
36564	36585	__ret_189; \
36565	36586	})
36566	36587	#endif
...	...	@@ -36568,22 +36589,22 @@ __ai __attribute__((target("v8.3a,neon"))) float32x2_t __noswap_vcmla_f32(float3
36568	36589	#ifdef __LITTLE_ENDIAN__
36569	36590	__ai __attribute__((target("v8.3a,neon"))) float32x4_t vcmlaq_rot180_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
36570	36591	float32x4_t __ret;
36571		__ret = (float32x4_t) __builtin_neon_vcmlaq_rot180_f32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
	36592	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vcmlaq_rot180_f32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 41));
36572	36593	return __ret;
36573	36594	}
36574	36595	#else
36575	36596	__ai __attribute__((target("v8.3a,neon"))) float32x4_t vcmlaq_rot180_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
36576	36597	float32x4_t __ret;
36577		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
36578		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
36579		float32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
36580		__ret = (float32x4_t) __builtin_neon_vcmlaq_rot180_f32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 41);
36581		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	36598	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	36599	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	36600	float32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
	36601	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vcmlaq_rot180_f32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 41));
	36602	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
36582	36603	return __ret;
36583	36604	}
36584	36605	__ai __attribute__((target("v8.3a,neon"))) float32x4_t __noswap_vcmlaq_rot180_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
36585	36606	float32x4_t __ret;
36586		__ret = (float32x4_t) __builtin_neon_vcmlaq_rot180_f32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
	36607	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vcmlaq_rot180_f32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 41));
36587	36608	return __ret;
36588	36609	}
36589	36610	#endif
...	...	@@ -36591,22 +36612,22 @@ __ai __attribute__((target("v8.3a,neon"))) float32x4_t __noswap_vcmlaq_rot180_f3
36591	36612	#ifdef __LITTLE_ENDIAN__
36592	36613	__ai __attribute__((target("v8.3a,neon"))) float32x2_t vcmla_rot180_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
36593	36614	float32x2_t __ret;
36594		__ret = (float32x2_t) __builtin_neon_vcmla_rot180_f32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
	36615	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vcmla_rot180_f32(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 9));
36595	36616	return __ret;
36596	36617	}
36597	36618	#else
36598	36619	__ai __attribute__((target("v8.3a,neon"))) float32x2_t vcmla_rot180_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
36599	36620	float32x2_t __ret;
36600		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
36601		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
36602		float32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
36603		__ret = (float32x2_t) __builtin_neon_vcmla_rot180_f32((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 9);
36604		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	36621	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	36622	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	36623	float32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_32);
	36624	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vcmla_rot180_f32(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 9));
	36625	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
36605	36626	return __ret;
36606	36627	}
36607	36628	__ai __attribute__((target("v8.3a,neon"))) float32x2_t __noswap_vcmla_rot180_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
36608	36629	float32x2_t __ret;
36609		__ret = (float32x2_t) __builtin_neon_vcmla_rot180_f32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
	36630	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vcmla_rot180_f32(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 9));
36610	36631	return __ret;
36611	36632	}
36612	36633	#endif
...	...	@@ -36626,11 +36647,11 @@ __ai __attribute__((target("v8.3a,neon"))) float32x2_t __noswap_vcmla_rot180_f32
36626	36647	float32x2_t __s0_191 = __p0_191; \
36627	36648	float32x2_t __s1_191 = __p1_191; \
36628	36649	float32x2_t __s2_191 = __p2_191; \
36629		float32x2_t __rev0_191; __rev0_191 = __builtin_shufflevector(__s0_191, __s0_191, 1, 0); \
36630		float32x2_t __rev1_191; __rev1_191 = __builtin_shufflevector(__s1_191, __s1_191, 1, 0); \
36631		float32x2_t __rev2_191; __rev2_191 = __builtin_shufflevector(__s2_191, __s2_191, 1, 0); \
	36650	float32x2_t __rev0_191; __rev0_191 = __builtin_shufflevector(__s0_191, __s0_191, __lane_reverse_64_32); \
	36651	float32x2_t __rev1_191; __rev1_191 = __builtin_shufflevector(__s1_191, __s1_191, __lane_reverse_64_32); \
	36652	float32x2_t __rev2_191; __rev2_191 = __builtin_shufflevector(__s2_191, __s2_191, __lane_reverse_64_32); \
36632	36653	__ret_191 = __noswap_vcmla_rot180_f32(__rev0_191, __rev1_191, __builtin_bit_cast(float32x2_t, (uint64x1_t) {vget_lane_u64(__builtin_bit_cast(uint64x1_t, __rev2_191), __p3_191)})); \
36633		__ret_191 = __builtin_shufflevector(__ret_191, __ret_191, 1, 0); \
	36654	__ret_191 = __builtin_shufflevector(__ret_191, __ret_191, __lane_reverse_64_32); \
36634	36655	__ret_191; \
36635	36656	})
36636	36657	#endif
...	...	@@ -36650,11 +36671,11 @@ __ai __attribute__((target("v8.3a,neon"))) float32x2_t __noswap_vcmla_rot180_f32
36650	36671	float32x4_t __s0_193 = __p0_193; \
36651	36672	float32x4_t __s1_193 = __p1_193; \
36652	36673	float32x2_t __s2_193 = __p2_193; \
36653		float32x4_t __rev0_193; __rev0_193 = __builtin_shufflevector(__s0_193, __s0_193, 3, 2, 1, 0); \
36654		float32x4_t __rev1_193; __rev1_193 = __builtin_shufflevector(__s1_193, __s1_193, 3, 2, 1, 0); \
36655		float32x2_t __rev2_193; __rev2_193 = __builtin_shufflevector(__s2_193, __s2_193, 1, 0); \
	36674	float32x4_t __rev0_193; __rev0_193 = __builtin_shufflevector(__s0_193, __s0_193, __lane_reverse_128_32); \
	36675	float32x4_t __rev1_193; __rev1_193 = __builtin_shufflevector(__s1_193, __s1_193, __lane_reverse_128_32); \
	36676	float32x2_t __rev2_193; __rev2_193 = __builtin_shufflevector(__s2_193, __s2_193, __lane_reverse_64_32); \
36656	36677	__ret_193 = __noswap_vcmlaq_rot180_f32(__rev0_193, __rev1_193, __builtin_bit_cast(float32x4_t, (uint64x2_t) {vget_lane_u64(__builtin_bit_cast(uint64x1_t, __rev2_193), __p3_193), vget_lane_u64(__builtin_bit_cast(uint64x1_t, __rev2_193), __p3_193)})); \
36657		__ret_193 = __builtin_shufflevector(__ret_193, __ret_193, 3, 2, 1, 0); \
	36678	__ret_193 = __builtin_shufflevector(__ret_193, __ret_193, __lane_reverse_128_32); \
36658	36679	__ret_193; \
36659	36680	})
36660	36681	#endif
...	...	@@ -36674,11 +36695,11 @@ __ai __attribute__((target("v8.3a,neon"))) float32x2_t __noswap_vcmla_rot180_f32
36674	36695	float32x2_t __s0_195 = __p0_195; \
36675	36696	float32x2_t __s1_195 = __p1_195; \
36676	36697	float32x4_t __s2_195 = __p2_195; \
36677		float32x2_t __rev0_195; __rev0_195 = __builtin_shufflevector(__s0_195, __s0_195, 1, 0); \
36678		float32x2_t __rev1_195; __rev1_195 = __builtin_shufflevector(__s1_195, __s1_195, 1, 0); \
36679		float32x4_t __rev2_195; __rev2_195 = __builtin_shufflevector(__s2_195, __s2_195, 3, 2, 1, 0); \
	36698	float32x2_t __rev0_195; __rev0_195 = __builtin_shufflevector(__s0_195, __s0_195, __lane_reverse_64_32); \
	36699	float32x2_t __rev1_195; __rev1_195 = __builtin_shufflevector(__s1_195, __s1_195, __lane_reverse_64_32); \
	36700	float32x4_t __rev2_195; __rev2_195 = __builtin_shufflevector(__s2_195, __s2_195, __lane_reverse_128_32); \
36680	36701	__ret_195 = __noswap_vcmla_rot180_f32(__rev0_195, __rev1_195, __builtin_bit_cast(float32x2_t, (uint64x1_t) {__noswap_vgetq_lane_u64(__builtin_bit_cast(uint64x2_t, __rev2_195), __p3_195)})); \
36681		__ret_195 = __builtin_shufflevector(__ret_195, __ret_195, 1, 0); \
	36702	__ret_195 = __builtin_shufflevector(__ret_195, __ret_195, __lane_reverse_64_32); \
36682	36703	__ret_195; \
36683	36704	})
36684	36705	#endif
...	...	@@ -36698,11 +36719,11 @@ __ai __attribute__((target("v8.3a,neon"))) float32x2_t __noswap_vcmla_rot180_f32
36698	36719	float32x4_t __s0_197 = __p0_197; \
36699	36720	float32x4_t __s1_197 = __p1_197; \
36700	36721	float32x4_t __s2_197 = __p2_197; \
36701		float32x4_t __rev0_197; __rev0_197 = __builtin_shufflevector(__s0_197, __s0_197, 3, 2, 1, 0); \
36702		float32x4_t __rev1_197; __rev1_197 = __builtin_shufflevector(__s1_197, __s1_197, 3, 2, 1, 0); \
36703		float32x4_t __rev2_197; __rev2_197 = __builtin_shufflevector(__s2_197, __s2_197, 3, 2, 1, 0); \
	36722	float32x4_t __rev0_197; __rev0_197 = __builtin_shufflevector(__s0_197, __s0_197, __lane_reverse_128_32); \
	36723	float32x4_t __rev1_197; __rev1_197 = __builtin_shufflevector(__s1_197, __s1_197, __lane_reverse_128_32); \
	36724	float32x4_t __rev2_197; __rev2_197 = __builtin_shufflevector(__s2_197, __s2_197, __lane_reverse_128_32); \
36704	36725	__ret_197 = __noswap_vcmlaq_rot180_f32(__rev0_197, __rev1_197, __builtin_bit_cast(float32x4_t, (uint64x2_t) {__noswap_vgetq_lane_u64(__builtin_bit_cast(uint64x2_t, __rev2_197), __p3_197), __noswap_vgetq_lane_u64(__builtin_bit_cast(uint64x2_t, __rev2_197), __p3_197)})); \
36705		__ret_197 = __builtin_shufflevector(__ret_197, __ret_197, 3, 2, 1, 0); \
	36726	__ret_197 = __builtin_shufflevector(__ret_197, __ret_197, __lane_reverse_128_32); \
36706	36727	__ret_197; \
36707	36728	})
36708	36729	#endif
...	...	@@ -36710,22 +36731,22 @@ __ai __attribute__((target("v8.3a,neon"))) float32x2_t __noswap_vcmla_rot180_f32
36710	36731	#ifdef __LITTLE_ENDIAN__
36711	36732	__ai __attribute__((target("v8.3a,neon"))) float32x4_t vcmlaq_rot270_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
36712	36733	float32x4_t __ret;
36713		__ret = (float32x4_t) __builtin_neon_vcmlaq_rot270_f32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
	36734	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vcmlaq_rot270_f32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 41));
36714	36735	return __ret;
36715	36736	}
36716	36737	#else
36717	36738	__ai __attribute__((target("v8.3a,neon"))) float32x4_t vcmlaq_rot270_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
36718	36739	float32x4_t __ret;
36719		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
36720		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
36721		float32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
36722		__ret = (float32x4_t) __builtin_neon_vcmlaq_rot270_f32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 41);
36723		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	36740	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	36741	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	36742	float32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
	36743	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vcmlaq_rot270_f32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 41));
	36744	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
36724	36745	return __ret;
36725	36746	}
36726	36747	__ai __attribute__((target("v8.3a,neon"))) float32x4_t __noswap_vcmlaq_rot270_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
36727	36748	float32x4_t __ret;
36728		__ret = (float32x4_t) __builtin_neon_vcmlaq_rot270_f32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
	36749	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vcmlaq_rot270_f32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 41));
36729	36750	return __ret;
36730	36751	}
36731	36752	#endif
...	...	@@ -36733,22 +36754,22 @@ __ai __attribute__((target("v8.3a,neon"))) float32x4_t __noswap_vcmlaq_rot270_f3
36733	36754	#ifdef __LITTLE_ENDIAN__
36734	36755	__ai __attribute__((target("v8.3a,neon"))) float32x2_t vcmla_rot270_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
36735	36756	float32x2_t __ret;
36736		__ret = (float32x2_t) __builtin_neon_vcmla_rot270_f32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
	36757	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vcmla_rot270_f32(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 9));
36737	36758	return __ret;
36738	36759	}
36739	36760	#else
36740	36761	__ai __attribute__((target("v8.3a,neon"))) float32x2_t vcmla_rot270_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
36741	36762	float32x2_t __ret;
36742		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
36743		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
36744		float32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
36745		__ret = (float32x2_t) __builtin_neon_vcmla_rot270_f32((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 9);
36746		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	36763	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	36764	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	36765	float32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_32);
	36766	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vcmla_rot270_f32(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 9));
	36767	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
36747	36768	return __ret;
36748	36769	}
36749	36770	__ai __attribute__((target("v8.3a,neon"))) float32x2_t __noswap_vcmla_rot270_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
36750	36771	float32x2_t __ret;
36751		__ret = (float32x2_t) __builtin_neon_vcmla_rot270_f32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
	36772	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vcmla_rot270_f32(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 9));
36752	36773	return __ret;
36753	36774	}
36754	36775	#endif
...	...	@@ -36768,11 +36789,11 @@ __ai __attribute__((target("v8.3a,neon"))) float32x2_t __noswap_vcmla_rot270_f32
36768	36789	float32x2_t __s0_199 = __p0_199; \
36769	36790	float32x2_t __s1_199 = __p1_199; \
36770	36791	float32x2_t __s2_199 = __p2_199; \
36771		float32x2_t __rev0_199; __rev0_199 = __builtin_shufflevector(__s0_199, __s0_199, 1, 0); \
36772		float32x2_t __rev1_199; __rev1_199 = __builtin_shufflevector(__s1_199, __s1_199, 1, 0); \
36773		float32x2_t __rev2_199; __rev2_199 = __builtin_shufflevector(__s2_199, __s2_199, 1, 0); \
	36792	float32x2_t __rev0_199; __rev0_199 = __builtin_shufflevector(__s0_199, __s0_199, __lane_reverse_64_32); \
	36793	float32x2_t __rev1_199; __rev1_199 = __builtin_shufflevector(__s1_199, __s1_199, __lane_reverse_64_32); \
	36794	float32x2_t __rev2_199; __rev2_199 = __builtin_shufflevector(__s2_199, __s2_199, __lane_reverse_64_32); \
36774	36795	__ret_199 = __noswap_vcmla_rot270_f32(__rev0_199, __rev1_199, __builtin_bit_cast(float32x2_t, (uint64x1_t) {vget_lane_u64(__builtin_bit_cast(uint64x1_t, __rev2_199), __p3_199)})); \
36775		__ret_199 = __builtin_shufflevector(__ret_199, __ret_199, 1, 0); \
	36796	__ret_199 = __builtin_shufflevector(__ret_199, __ret_199, __lane_reverse_64_32); \
36776	36797	__ret_199; \
36777	36798	})
36778	36799	#endif
...	...	@@ -36792,11 +36813,11 @@ __ai __attribute__((target("v8.3a,neon"))) float32x2_t __noswap_vcmla_rot270_f32
36792	36813	float32x4_t __s0_201 = __p0_201; \
36793	36814	float32x4_t __s1_201 = __p1_201; \
36794	36815	float32x2_t __s2_201 = __p2_201; \
36795		float32x4_t __rev0_201; __rev0_201 = __builtin_shufflevector(__s0_201, __s0_201, 3, 2, 1, 0); \
36796		float32x4_t __rev1_201; __rev1_201 = __builtin_shufflevector(__s1_201, __s1_201, 3, 2, 1, 0); \
36797		float32x2_t __rev2_201; __rev2_201 = __builtin_shufflevector(__s2_201, __s2_201, 1, 0); \
	36816	float32x4_t __rev0_201; __rev0_201 = __builtin_shufflevector(__s0_201, __s0_201, __lane_reverse_128_32); \
	36817	float32x4_t __rev1_201; __rev1_201 = __builtin_shufflevector(__s1_201, __s1_201, __lane_reverse_128_32); \
	36818	float32x2_t __rev2_201; __rev2_201 = __builtin_shufflevector(__s2_201, __s2_201, __lane_reverse_64_32); \
36798	36819	__ret_201 = __noswap_vcmlaq_rot270_f32(__rev0_201, __rev1_201, __builtin_bit_cast(float32x4_t, (uint64x2_t) {vget_lane_u64(__builtin_bit_cast(uint64x1_t, __rev2_201), __p3_201), vget_lane_u64(__builtin_bit_cast(uint64x1_t, __rev2_201), __p3_201)})); \
36799		__ret_201 = __builtin_shufflevector(__ret_201, __ret_201, 3, 2, 1, 0); \
	36820	__ret_201 = __builtin_shufflevector(__ret_201, __ret_201, __lane_reverse_128_32); \
36800	36821	__ret_201; \
36801	36822	})
36802	36823	#endif
...	...	@@ -36816,11 +36837,11 @@ __ai __attribute__((target("v8.3a,neon"))) float32x2_t __noswap_vcmla_rot270_f32
36816	36837	float32x2_t __s0_203 = __p0_203; \
36817	36838	float32x2_t __s1_203 = __p1_203; \
36818	36839	float32x4_t __s2_203 = __p2_203; \
36819		float32x2_t __rev0_203; __rev0_203 = __builtin_shufflevector(__s0_203, __s0_203, 1, 0); \
36820		float32x2_t __rev1_203; __rev1_203 = __builtin_shufflevector(__s1_203, __s1_203, 1, 0); \
36821		float32x4_t __rev2_203; __rev2_203 = __builtin_shufflevector(__s2_203, __s2_203, 3, 2, 1, 0); \
	36840	float32x2_t __rev0_203; __rev0_203 = __builtin_shufflevector(__s0_203, __s0_203, __lane_reverse_64_32); \
	36841	float32x2_t __rev1_203; __rev1_203 = __builtin_shufflevector(__s1_203, __s1_203, __lane_reverse_64_32); \
	36842	float32x4_t __rev2_203; __rev2_203 = __builtin_shufflevector(__s2_203, __s2_203, __lane_reverse_128_32); \
36822	36843	__ret_203 = __noswap_vcmla_rot270_f32(__rev0_203, __rev1_203, __builtin_bit_cast(float32x2_t, (uint64x1_t) {__noswap_vgetq_lane_u64(__builtin_bit_cast(uint64x2_t, __rev2_203), __p3_203)})); \
36823		__ret_203 = __builtin_shufflevector(__ret_203, __ret_203, 1, 0); \
	36844	__ret_203 = __builtin_shufflevector(__ret_203, __ret_203, __lane_reverse_64_32); \
36824	36845	__ret_203; \
36825	36846	})
36826	36847	#endif
...	...	@@ -36840,11 +36861,11 @@ __ai __attribute__((target("v8.3a,neon"))) float32x2_t __noswap_vcmla_rot270_f32
36840	36861	float32x4_t __s0_205 = __p0_205; \
36841	36862	float32x4_t __s1_205 = __p1_205; \
36842	36863	float32x4_t __s2_205 = __p2_205; \
36843		float32x4_t __rev0_205; __rev0_205 = __builtin_shufflevector(__s0_205, __s0_205, 3, 2, 1, 0); \
36844		float32x4_t __rev1_205; __rev1_205 = __builtin_shufflevector(__s1_205, __s1_205, 3, 2, 1, 0); \
36845		float32x4_t __rev2_205; __rev2_205 = __builtin_shufflevector(__s2_205, __s2_205, 3, 2, 1, 0); \
	36864	float32x4_t __rev0_205; __rev0_205 = __builtin_shufflevector(__s0_205, __s0_205, __lane_reverse_128_32); \
	36865	float32x4_t __rev1_205; __rev1_205 = __builtin_shufflevector(__s1_205, __s1_205, __lane_reverse_128_32); \
	36866	float32x4_t __rev2_205; __rev2_205 = __builtin_shufflevector(__s2_205, __s2_205, __lane_reverse_128_32); \
36846	36867	__ret_205 = __noswap_vcmlaq_rot270_f32(__rev0_205, __rev1_205, __builtin_bit_cast(float32x4_t, (uint64x2_t) {__noswap_vgetq_lane_u64(__builtin_bit_cast(uint64x2_t, __rev2_205), __p3_205), __noswap_vgetq_lane_u64(__builtin_bit_cast(uint64x2_t, __rev2_205), __p3_205)})); \
36847		__ret_205 = __builtin_shufflevector(__ret_205, __ret_205, 3, 2, 1, 0); \
	36868	__ret_205 = __builtin_shufflevector(__ret_205, __ret_205, __lane_reverse_128_32); \
36848	36869	__ret_205; \
36849	36870	})
36850	36871	#endif
...	...	@@ -36852,22 +36873,22 @@ __ai __attribute__((target("v8.3a,neon"))) float32x2_t __noswap_vcmla_rot270_f32
36852	36873	#ifdef __LITTLE_ENDIAN__
36853	36874	__ai __attribute__((target("v8.3a,neon"))) float32x4_t vcmlaq_rot90_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
36854	36875	float32x4_t __ret;
36855		__ret = (float32x4_t) __builtin_neon_vcmlaq_rot90_f32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
	36876	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vcmlaq_rot90_f32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 41));
36856	36877	return __ret;
36857	36878	}
36858	36879	#else
36859	36880	__ai __attribute__((target("v8.3a,neon"))) float32x4_t vcmlaq_rot90_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
36860	36881	float32x4_t __ret;
36861		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
36862		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
36863		float32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
36864		__ret = (float32x4_t) __builtin_neon_vcmlaq_rot90_f32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 41);
36865		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	36882	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	36883	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	36884	float32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
	36885	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vcmlaq_rot90_f32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 41));
	36886	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
36866	36887	return __ret;
36867	36888	}
36868	36889	__ai __attribute__((target("v8.3a,neon"))) float32x4_t __noswap_vcmlaq_rot90_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
36869	36890	float32x4_t __ret;
36870		__ret = (float32x4_t) __builtin_neon_vcmlaq_rot90_f32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
	36891	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vcmlaq_rot90_f32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 41));
36871	36892	return __ret;
36872	36893	}
36873	36894	#endif
...	...	@@ -36875,22 +36896,22 @@ __ai __attribute__((target("v8.3a,neon"))) float32x4_t __noswap_vcmlaq_rot90_f32
36875	36896	#ifdef __LITTLE_ENDIAN__
36876	36897	__ai __attribute__((target("v8.3a,neon"))) float32x2_t vcmla_rot90_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
36877	36898	float32x2_t __ret;
36878		__ret = (float32x2_t) __builtin_neon_vcmla_rot90_f32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
	36899	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vcmla_rot90_f32(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 9));
36879	36900	return __ret;
36880	36901	}
36881	36902	#else
36882	36903	__ai __attribute__((target("v8.3a,neon"))) float32x2_t vcmla_rot90_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
36883	36904	float32x2_t __ret;
36884		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
36885		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
36886		float32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
36887		__ret = (float32x2_t) __builtin_neon_vcmla_rot90_f32((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 9);
36888		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	36905	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	36906	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	36907	float32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_32);
	36908	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vcmla_rot90_f32(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 9));
	36909	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
36889	36910	return __ret;
36890	36911	}
36891	36912	__ai __attribute__((target("v8.3a,neon"))) float32x2_t __noswap_vcmla_rot90_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
36892	36913	float32x2_t __ret;
36893		__ret = (float32x2_t) __builtin_neon_vcmla_rot90_f32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
	36914	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vcmla_rot90_f32(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 9));
36894	36915	return __ret;
36895	36916	}
36896	36917	#endif
...	...	@@ -36910,11 +36931,11 @@ __ai __attribute__((target("v8.3a,neon"))) float32x2_t __noswap_vcmla_rot90_f32(
36910	36931	float32x2_t __s0_207 = __p0_207; \
36911	36932	float32x2_t __s1_207 = __p1_207; \
36912	36933	float32x2_t __s2_207 = __p2_207; \
36913		float32x2_t __rev0_207; __rev0_207 = __builtin_shufflevector(__s0_207, __s0_207, 1, 0); \
36914		float32x2_t __rev1_207; __rev1_207 = __builtin_shufflevector(__s1_207, __s1_207, 1, 0); \
36915		float32x2_t __rev2_207; __rev2_207 = __builtin_shufflevector(__s2_207, __s2_207, 1, 0); \
	36934	float32x2_t __rev0_207; __rev0_207 = __builtin_shufflevector(__s0_207, __s0_207, __lane_reverse_64_32); \
	36935	float32x2_t __rev1_207; __rev1_207 = __builtin_shufflevector(__s1_207, __s1_207, __lane_reverse_64_32); \
	36936	float32x2_t __rev2_207; __rev2_207 = __builtin_shufflevector(__s2_207, __s2_207, __lane_reverse_64_32); \
36916	36937	__ret_207 = __noswap_vcmla_rot90_f32(__rev0_207, __rev1_207, __builtin_bit_cast(float32x2_t, (uint64x1_t) {vget_lane_u64(__builtin_bit_cast(uint64x1_t, __rev2_207), __p3_207)})); \
36917		__ret_207 = __builtin_shufflevector(__ret_207, __ret_207, 1, 0); \
	36938	__ret_207 = __builtin_shufflevector(__ret_207, __ret_207, __lane_reverse_64_32); \
36918	36939	__ret_207; \
36919	36940	})
36920	36941	#endif
...	...	@@ -36934,11 +36955,11 @@ __ai __attribute__((target("v8.3a,neon"))) float32x2_t __noswap_vcmla_rot90_f32(
36934	36955	float32x4_t __s0_209 = __p0_209; \
36935	36956	float32x4_t __s1_209 = __p1_209; \
36936	36957	float32x2_t __s2_209 = __p2_209; \
36937		float32x4_t __rev0_209; __rev0_209 = __builtin_shufflevector(__s0_209, __s0_209, 3, 2, 1, 0); \
36938		float32x4_t __rev1_209; __rev1_209 = __builtin_shufflevector(__s1_209, __s1_209, 3, 2, 1, 0); \
36939		float32x2_t __rev2_209; __rev2_209 = __builtin_shufflevector(__s2_209, __s2_209, 1, 0); \
	36958	float32x4_t __rev0_209; __rev0_209 = __builtin_shufflevector(__s0_209, __s0_209, __lane_reverse_128_32); \
	36959	float32x4_t __rev1_209; __rev1_209 = __builtin_shufflevector(__s1_209, __s1_209, __lane_reverse_128_32); \
	36960	float32x2_t __rev2_209; __rev2_209 = __builtin_shufflevector(__s2_209, __s2_209, __lane_reverse_64_32); \
36940	36961	__ret_209 = __noswap_vcmlaq_rot90_f32(__rev0_209, __rev1_209, __builtin_bit_cast(float32x4_t, (uint64x2_t) {vget_lane_u64(__builtin_bit_cast(uint64x1_t, __rev2_209), __p3_209), vget_lane_u64(__builtin_bit_cast(uint64x1_t, __rev2_209), __p3_209)})); \
36941		__ret_209 = __builtin_shufflevector(__ret_209, __ret_209, 3, 2, 1, 0); \
	36962	__ret_209 = __builtin_shufflevector(__ret_209, __ret_209, __lane_reverse_128_32); \
36942	36963	__ret_209; \
36943	36964	})
36944	36965	#endif
...	...	@@ -36958,11 +36979,11 @@ __ai __attribute__((target("v8.3a,neon"))) float32x2_t __noswap_vcmla_rot90_f32(
36958	36979	float32x2_t __s0_211 = __p0_211; \
36959	36980	float32x2_t __s1_211 = __p1_211; \
36960	36981	float32x4_t __s2_211 = __p2_211; \
36961		float32x2_t __rev0_211; __rev0_211 = __builtin_shufflevector(__s0_211, __s0_211, 1, 0); \
36962		float32x2_t __rev1_211; __rev1_211 = __builtin_shufflevector(__s1_211, __s1_211, 1, 0); \
36963		float32x4_t __rev2_211; __rev2_211 = __builtin_shufflevector(__s2_211, __s2_211, 3, 2, 1, 0); \
	36982	float32x2_t __rev0_211; __rev0_211 = __builtin_shufflevector(__s0_211, __s0_211, __lane_reverse_64_32); \
	36983	float32x2_t __rev1_211; __rev1_211 = __builtin_shufflevector(__s1_211, __s1_211, __lane_reverse_64_32); \
	36984	float32x4_t __rev2_211; __rev2_211 = __builtin_shufflevector(__s2_211, __s2_211, __lane_reverse_128_32); \
36964	36985	__ret_211 = __noswap_vcmla_rot90_f32(__rev0_211, __rev1_211, __builtin_bit_cast(float32x2_t, (uint64x1_t) {__noswap_vgetq_lane_u64(__builtin_bit_cast(uint64x2_t, __rev2_211), __p3_211)})); \
36965		__ret_211 = __builtin_shufflevector(__ret_211, __ret_211, 1, 0); \
	36986	__ret_211 = __builtin_shufflevector(__ret_211, __ret_211, __lane_reverse_64_32); \
36966	36987	__ret_211; \
36967	36988	})
36968	36989	#endif
...	...	@@ -36982,11 +37003,11 @@ __ai __attribute__((target("v8.3a,neon"))) float32x2_t __noswap_vcmla_rot90_f32(
36982	37003	float32x4_t __s0_213 = __p0_213; \
36983	37004	float32x4_t __s1_213 = __p1_213; \
36984	37005	float32x4_t __s2_213 = __p2_213; \
36985		float32x4_t __rev0_213; __rev0_213 = __builtin_shufflevector(__s0_213, __s0_213, 3, 2, 1, 0); \
36986		float32x4_t __rev1_213; __rev1_213 = __builtin_shufflevector(__s1_213, __s1_213, 3, 2, 1, 0); \
36987		float32x4_t __rev2_213; __rev2_213 = __builtin_shufflevector(__s2_213, __s2_213, 3, 2, 1, 0); \
	37006	float32x4_t __rev0_213; __rev0_213 = __builtin_shufflevector(__s0_213, __s0_213, __lane_reverse_128_32); \
	37007	float32x4_t __rev1_213; __rev1_213 = __builtin_shufflevector(__s1_213, __s1_213, __lane_reverse_128_32); \
	37008	float32x4_t __rev2_213; __rev2_213 = __builtin_shufflevector(__s2_213, __s2_213, __lane_reverse_128_32); \
36988	37009	__ret_213 = __noswap_vcmlaq_rot90_f32(__rev0_213, __rev1_213, __builtin_bit_cast(float32x4_t, (uint64x2_t) {__noswap_vgetq_lane_u64(__builtin_bit_cast(uint64x2_t, __rev2_213), __p3_213), __noswap_vgetq_lane_u64(__builtin_bit_cast(uint64x2_t, __rev2_213), __p3_213)})); \
36989		__ret_213 = __builtin_shufflevector(__ret_213, __ret_213, 3, 2, 1, 0); \
	37010	__ret_213 = __builtin_shufflevector(__ret_213, __ret_213, __lane_reverse_128_32); \
36990	37011	__ret_213; \
36991	37012	})
36992	37013	#endif
...	...	@@ -36995,20 +37016,20 @@ __ai __attribute__((target("v8.3a,neon"))) float32x2_t __noswap_vcmla_rot90_f32(
36995	37016	#ifdef __LITTLE_ENDIAN__
36996	37017	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t __a32_vcvt_bf16_f32(float32x4_t __p0) {
36997	37018	bfloat16x4_t __ret;
36998		__ret = (bfloat16x4_t) __builtin_neon___a32_vcvt_bf16_f32((int8x16_t)__p0, 11);
	37019	__ret = __builtin_bit_cast(bfloat16x4_t, __builtin_neon___a32_vcvt_bf16_f32(__builtin_bit_cast(int8x16_t, __p0), 11));
36999	37020	return __ret;
37000	37021	}
37001	37022	#else
37002	37023	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t __a32_vcvt_bf16_f32(float32x4_t __p0) {
37003	37024	bfloat16x4_t __ret;
37004		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
37005		__ret = (bfloat16x4_t) __builtin_neon___a32_vcvt_bf16_f32((int8x16_t)__rev0, 11);
37006		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	37025	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	37026	__ret = __builtin_bit_cast(bfloat16x4_t, __builtin_neon___a32_vcvt_bf16_f32(__builtin_bit_cast(int8x16_t, __rev0), 11));
	37027	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
37007	37028	return __ret;
37008	37029	}
37009	37030	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t __noswap___a32_vcvt_bf16_f32(float32x4_t __p0) {
37010	37031	bfloat16x4_t __ret;
37011		__ret = (bfloat16x4_t) __builtin_neon___a32_vcvt_bf16_f32((int8x16_t)__p0, 11);
	37032	__ret = __builtin_bit_cast(bfloat16x4_t, __builtin_neon___a32_vcvt_bf16_f32(__builtin_bit_cast(int8x16_t, __p0), 11));
37012	37033	return __ret;
37013	37034	}
37014	37035	#endif
...	...	@@ -37022,9 +37043,9 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t vcvt_bf16_f32(float32x4_t
37022	37043	#else
37023	37044	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vcvt_bf16_f32(float32x4_t __p0) {
37024	37045	bfloat16x4_t __ret;
37025		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	37046	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
37026	37047	__ret = __noswap___a32_vcvt_bf16_f32(__rev0);
37027		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	37048	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
37028	37049	return __ret;
37029	37050	}
37030	37051	#endif
...	...	@@ -37038,10 +37059,10 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x8_t vcvtq_high_bf16_f32(bfloa
37038	37059	#else
37039	37060	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vcvtq_high_bf16_f32(bfloat16x8_t __p0, float32x4_t __p1) {
37040	37061	bfloat16x8_t __ret;
37041		bfloat16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
37042		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	37062	bfloat16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	37063	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
37043	37064	__ret = __noswap_vcombine_bf16(__noswap___a32_vcvt_bf16_f32(__rev1), __noswap_vget_low_bf16(__rev0));
37044		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	37065	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
37045	37066	return __ret;
37046	37067	}
37047	37068	#endif
...	...	@@ -37049,277 +37070,277 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x8_t vcvtq_high_bf16_f32(bfloa
37049	37070	#ifdef __LITTLE_ENDIAN__
37050	37071	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vcvtq_low_bf16_f32(float32x4_t __p0) {
37051	37072	bfloat16x8_t __ret;
37052		__ret = vcombine_bf16((bfloat16x4_t)(0ULL), __a32_vcvt_bf16_f32(__p0));
	37073	__ret = vcombine_bf16(__builtin_bit_cast(bfloat16x4_t, 0ULL), __a32_vcvt_bf16_f32(__p0));
37053	37074	return __ret;
37054	37075	}
37055	37076	#else
37056	37077	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vcvtq_low_bf16_f32(float32x4_t __p0) {
37057	37078	bfloat16x8_t __ret;
37058		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
37059		__ret = __noswap_vcombine_bf16((bfloat16x4_t)(0ULL), __noswap___a32_vcvt_bf16_f32(__rev0));
37060		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	37079	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	37080	__ret = __noswap_vcombine_bf16(__builtin_bit_cast(bfloat16x4_t, 0ULL), __noswap___a32_vcvt_bf16_f32(__rev0));
	37081	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
37061	37082	return __ret;
37062	37083	}
37063	37084	#endif
37064	37085
37065	37086	__ai __attribute__((target("bf16,neon"))) poly8x8_t vreinterpret_p8_bf16(bfloat16x4_t __p0) {
37066	37087	poly8x8_t __ret;
37067		__ret = (poly8x8_t)(__p0);
	37088	__ret = __builtin_bit_cast(poly8x8_t, __p0);
37068	37089	return __ret;
37069	37090	}
37070	37091	__ai __attribute__((target("bf16,neon"))) poly64x1_t vreinterpret_p64_bf16(bfloat16x4_t __p0) {
37071	37092	poly64x1_t __ret;
37072		__ret = (poly64x1_t)(__p0);
	37093	__ret = __builtin_bit_cast(poly64x1_t, __p0);
37073	37094	return __ret;
37074	37095	}
37075	37096	__ai __attribute__((target("bf16,neon"))) poly16x4_t vreinterpret_p16_bf16(bfloat16x4_t __p0) {
37076	37097	poly16x4_t __ret;
37077		__ret = (poly16x4_t)(__p0);
	37098	__ret = __builtin_bit_cast(poly16x4_t, __p0);
37078	37099	return __ret;
37079	37100	}
37080	37101	__ai __attribute__((target("bf16,neon"))) poly8x16_t vreinterpretq_p8_bf16(bfloat16x8_t __p0) {
37081	37102	poly8x16_t __ret;
37082		__ret = (poly8x16_t)(__p0);
	37103	__ret = __builtin_bit_cast(poly8x16_t, __p0);
37083	37104	return __ret;
37084	37105	}
37085	37106	__ai __attribute__((target("bf16,neon"))) poly64x2_t vreinterpretq_p64_bf16(bfloat16x8_t __p0) {
37086	37107	poly64x2_t __ret;
37087		__ret = (poly64x2_t)(__p0);
	37108	__ret = __builtin_bit_cast(poly64x2_t, __p0);
37088	37109	return __ret;
37089	37110	}
37090	37111	__ai __attribute__((target("bf16,neon"))) poly16x8_t vreinterpretq_p16_bf16(bfloat16x8_t __p0) {
37091	37112	poly16x8_t __ret;
37092		__ret = (poly16x8_t)(__p0);
	37113	__ret = __builtin_bit_cast(poly16x8_t, __p0);
37093	37114	return __ret;
37094	37115	}
37095	37116	__ai __attribute__((target("bf16,neon"))) uint8x16_t vreinterpretq_u8_bf16(bfloat16x8_t __p0) {
37096	37117	uint8x16_t __ret;
37097		__ret = (uint8x16_t)(__p0);
	37118	__ret = __builtin_bit_cast(uint8x16_t, __p0);
37098	37119	return __ret;
37099	37120	}
37100	37121	__ai __attribute__((target("bf16,neon"))) uint32x4_t vreinterpretq_u32_bf16(bfloat16x8_t __p0) {
37101	37122	uint32x4_t __ret;
37102		__ret = (uint32x4_t)(__p0);
	37123	__ret = __builtin_bit_cast(uint32x4_t, __p0);
37103	37124	return __ret;
37104	37125	}
37105	37126	__ai __attribute__((target("bf16,neon"))) uint64x2_t vreinterpretq_u64_bf16(bfloat16x8_t __p0) {
37106	37127	uint64x2_t __ret;
37107		__ret = (uint64x2_t)(__p0);
	37128	__ret = __builtin_bit_cast(uint64x2_t, __p0);
37108	37129	return __ret;
37109	37130	}
37110	37131	__ai __attribute__((target("bf16,neon"))) uint16x8_t vreinterpretq_u16_bf16(bfloat16x8_t __p0) {
37111	37132	uint16x8_t __ret;
37112		__ret = (uint16x8_t)(__p0);
	37133	__ret = __builtin_bit_cast(uint16x8_t, __p0);
37113	37134	return __ret;
37114	37135	}
37115	37136	__ai __attribute__((target("bf16,neon"))) int8x16_t vreinterpretq_s8_bf16(bfloat16x8_t __p0) {
37116	37137	int8x16_t __ret;
37117		__ret = (int8x16_t)(__p0);
	37138	__ret = __builtin_bit_cast(int8x16_t, __p0);
37118	37139	return __ret;
37119	37140	}
37120	37141	__ai __attribute__((target("bf16,neon"))) float32x4_t vreinterpretq_f32_bf16(bfloat16x8_t __p0) {
37121	37142	float32x4_t __ret;
37122		__ret = (float32x4_t)(__p0);
	37143	__ret = __builtin_bit_cast(float32x4_t, __p0);
37123	37144	return __ret;
37124	37145	}
37125	37146	__ai __attribute__((target("bf16,neon"))) float16x8_t vreinterpretq_f16_bf16(bfloat16x8_t __p0) {
37126	37147	float16x8_t __ret;
37127		__ret = (float16x8_t)(__p0);
	37148	__ret = __builtin_bit_cast(float16x8_t, __p0);
37128	37149	return __ret;
37129	37150	}
37130	37151	__ai __attribute__((target("bf16,neon"))) int32x4_t vreinterpretq_s32_bf16(bfloat16x8_t __p0) {
37131	37152	int32x4_t __ret;
37132		__ret = (int32x4_t)(__p0);
	37153	__ret = __builtin_bit_cast(int32x4_t, __p0);
37133	37154	return __ret;
37134	37155	}
37135	37156	__ai __attribute__((target("bf16,neon"))) int64x2_t vreinterpretq_s64_bf16(bfloat16x8_t __p0) {
37136	37157	int64x2_t __ret;
37137		__ret = (int64x2_t)(__p0);
	37158	__ret = __builtin_bit_cast(int64x2_t, __p0);
37138	37159	return __ret;
37139	37160	}
37140	37161	__ai __attribute__((target("bf16,neon"))) int16x8_t vreinterpretq_s16_bf16(bfloat16x8_t __p0) {
37141	37162	int16x8_t __ret;
37142		__ret = (int16x8_t)(__p0);
	37163	__ret = __builtin_bit_cast(int16x8_t, __p0);
37143	37164	return __ret;
37144	37165	}
37145	37166	__ai __attribute__((target("bf16,neon"))) uint8x8_t vreinterpret_u8_bf16(bfloat16x4_t __p0) {
37146	37167	uint8x8_t __ret;
37147		__ret = (uint8x8_t)(__p0);
	37168	__ret = __builtin_bit_cast(uint8x8_t, __p0);
37148	37169	return __ret;
37149	37170	}
37150	37171	__ai __attribute__((target("bf16,neon"))) uint32x2_t vreinterpret_u32_bf16(bfloat16x4_t __p0) {
37151	37172	uint32x2_t __ret;
37152		__ret = (uint32x2_t)(__p0);
	37173	__ret = __builtin_bit_cast(uint32x2_t, __p0);
37153	37174	return __ret;
37154	37175	}
37155	37176	__ai __attribute__((target("bf16,neon"))) uint64x1_t vreinterpret_u64_bf16(bfloat16x4_t __p0) {
37156	37177	uint64x1_t __ret;
37157		__ret = (uint64x1_t)(__p0);
	37178	__ret = __builtin_bit_cast(uint64x1_t, __p0);
37158	37179	return __ret;
37159	37180	}
37160	37181	__ai __attribute__((target("bf16,neon"))) uint16x4_t vreinterpret_u16_bf16(bfloat16x4_t __p0) {
37161	37182	uint16x4_t __ret;
37162		__ret = (uint16x4_t)(__p0);
	37183	__ret = __builtin_bit_cast(uint16x4_t, __p0);
37163	37184	return __ret;
37164	37185	}
37165	37186	__ai __attribute__((target("bf16,neon"))) int8x8_t vreinterpret_s8_bf16(bfloat16x4_t __p0) {
37166	37187	int8x8_t __ret;
37167		__ret = (int8x8_t)(__p0);
	37188	__ret = __builtin_bit_cast(int8x8_t, __p0);
37168	37189	return __ret;
37169	37190	}
37170	37191	__ai __attribute__((target("bf16,neon"))) float32x2_t vreinterpret_f32_bf16(bfloat16x4_t __p0) {
37171	37192	float32x2_t __ret;
37172		__ret = (float32x2_t)(__p0);
	37193	__ret = __builtin_bit_cast(float32x2_t, __p0);
37173	37194	return __ret;
37174	37195	}
37175	37196	__ai __attribute__((target("bf16,neon"))) float16x4_t vreinterpret_f16_bf16(bfloat16x4_t __p0) {
37176	37197	float16x4_t __ret;
37177		__ret = (float16x4_t)(__p0);
	37198	__ret = __builtin_bit_cast(float16x4_t, __p0);
37178	37199	return __ret;
37179	37200	}
37180	37201	__ai __attribute__((target("bf16,neon"))) int32x2_t vreinterpret_s32_bf16(bfloat16x4_t __p0) {
37181	37202	int32x2_t __ret;
37182		__ret = (int32x2_t)(__p0);
	37203	__ret = __builtin_bit_cast(int32x2_t, __p0);
37183	37204	return __ret;
37184	37205	}
37185	37206	__ai __attribute__((target("bf16,neon"))) int64x1_t vreinterpret_s64_bf16(bfloat16x4_t __p0) {
37186	37207	int64x1_t __ret;
37187		__ret = (int64x1_t)(__p0);
	37208	__ret = __builtin_bit_cast(int64x1_t, __p0);
37188	37209	return __ret;
37189	37210	}
37190	37211	__ai __attribute__((target("bf16,neon"))) int16x4_t vreinterpret_s16_bf16(bfloat16x4_t __p0) {
37191	37212	int16x4_t __ret;
37192		__ret = (int16x4_t)(__p0);
	37213	__ret = __builtin_bit_cast(int16x4_t, __p0);
37193	37214	return __ret;
37194	37215	}
37195	37216	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vreinterpretq_bf16_p8(poly8x16_t __p0) {
37196	37217	bfloat16x8_t __ret;
37197		__ret = (bfloat16x8_t)(__p0);
	37218	__ret = __builtin_bit_cast(bfloat16x8_t, __p0);
37198	37219	return __ret;
37199	37220	}
37200	37221	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vreinterpretq_bf16_p64(poly64x2_t __p0) {
37201	37222	bfloat16x8_t __ret;
37202		__ret = (bfloat16x8_t)(__p0);
	37223	__ret = __builtin_bit_cast(bfloat16x8_t, __p0);
37203	37224	return __ret;
37204	37225	}
37205	37226	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vreinterpretq_bf16_p16(poly16x8_t __p0) {
37206	37227	bfloat16x8_t __ret;
37207		__ret = (bfloat16x8_t)(__p0);
	37228	__ret = __builtin_bit_cast(bfloat16x8_t, __p0);
37208	37229	return __ret;
37209	37230	}
37210	37231	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vreinterpretq_bf16_u8(uint8x16_t __p0) {
37211	37232	bfloat16x8_t __ret;
37212		__ret = (bfloat16x8_t)(__p0);
	37233	__ret = __builtin_bit_cast(bfloat16x8_t, __p0);
37213	37234	return __ret;
37214	37235	}
37215	37236	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vreinterpretq_bf16_u32(uint32x4_t __p0) {
37216	37237	bfloat16x8_t __ret;
37217		__ret = (bfloat16x8_t)(__p0);
	37238	__ret = __builtin_bit_cast(bfloat16x8_t, __p0);
37218	37239	return __ret;
37219	37240	}
37220	37241	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vreinterpretq_bf16_u64(uint64x2_t __p0) {
37221	37242	bfloat16x8_t __ret;
37222		__ret = (bfloat16x8_t)(__p0);
	37243	__ret = __builtin_bit_cast(bfloat16x8_t, __p0);
37223	37244	return __ret;
37224	37245	}
37225	37246	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vreinterpretq_bf16_u16(uint16x8_t __p0) {
37226	37247	bfloat16x8_t __ret;
37227		__ret = (bfloat16x8_t)(__p0);
	37248	__ret = __builtin_bit_cast(bfloat16x8_t, __p0);
37228	37249	return __ret;
37229	37250	}
37230	37251	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vreinterpretq_bf16_s8(int8x16_t __p0) {
37231	37252	bfloat16x8_t __ret;
37232		__ret = (bfloat16x8_t)(__p0);
	37253	__ret = __builtin_bit_cast(bfloat16x8_t, __p0);
37233	37254	return __ret;
37234	37255	}
37235	37256	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vreinterpretq_bf16_f32(float32x4_t __p0) {
37236	37257	bfloat16x8_t __ret;
37237		__ret = (bfloat16x8_t)(__p0);
	37258	__ret = __builtin_bit_cast(bfloat16x8_t, __p0);
37238	37259	return __ret;
37239	37260	}
37240	37261	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vreinterpretq_bf16_f16(float16x8_t __p0) {
37241	37262	bfloat16x8_t __ret;
37242		__ret = (bfloat16x8_t)(__p0);
	37263	__ret = __builtin_bit_cast(bfloat16x8_t, __p0);
37243	37264	return __ret;
37244	37265	}
37245	37266	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vreinterpretq_bf16_s32(int32x4_t __p0) {
37246	37267	bfloat16x8_t __ret;
37247		__ret = (bfloat16x8_t)(__p0);
	37268	__ret = __builtin_bit_cast(bfloat16x8_t, __p0);
37248	37269	return __ret;
37249	37270	}
37250	37271	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vreinterpretq_bf16_s64(int64x2_t __p0) {
37251	37272	bfloat16x8_t __ret;
37252		__ret = (bfloat16x8_t)(__p0);
	37273	__ret = __builtin_bit_cast(bfloat16x8_t, __p0);
37253	37274	return __ret;
37254	37275	}
37255	37276	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vreinterpretq_bf16_s16(int16x8_t __p0) {
37256	37277	bfloat16x8_t __ret;
37257		__ret = (bfloat16x8_t)(__p0);
	37278	__ret = __builtin_bit_cast(bfloat16x8_t, __p0);
37258	37279	return __ret;
37259	37280	}
37260	37281	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_p8(poly8x8_t __p0) {
37261	37282	bfloat16x4_t __ret;
37262		__ret = (bfloat16x4_t)(__p0);
	37283	__ret = __builtin_bit_cast(bfloat16x4_t, __p0);
37263	37284	return __ret;
37264	37285	}
37265	37286	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_p64(poly64x1_t __p0) {
37266	37287	bfloat16x4_t __ret;
37267		__ret = (bfloat16x4_t)(__p0);
	37288	__ret = __builtin_bit_cast(bfloat16x4_t, __p0);
37268	37289	return __ret;
37269	37290	}
37270	37291	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_p16(poly16x4_t __p0) {
37271	37292	bfloat16x4_t __ret;
37272		__ret = (bfloat16x4_t)(__p0);
	37293	__ret = __builtin_bit_cast(bfloat16x4_t, __p0);
37273	37294	return __ret;
37274	37295	}
37275	37296	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_u8(uint8x8_t __p0) {
37276	37297	bfloat16x4_t __ret;
37277		__ret = (bfloat16x4_t)(__p0);
	37298	__ret = __builtin_bit_cast(bfloat16x4_t, __p0);
37278	37299	return __ret;
37279	37300	}
37280	37301	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_u32(uint32x2_t __p0) {
37281	37302	bfloat16x4_t __ret;
37282		__ret = (bfloat16x4_t)(__p0);
	37303	__ret = __builtin_bit_cast(bfloat16x4_t, __p0);
37283	37304	return __ret;
37284	37305	}
37285	37306	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_u64(uint64x1_t __p0) {
37286	37307	bfloat16x4_t __ret;
37287		__ret = (bfloat16x4_t)(__p0);
	37308	__ret = __builtin_bit_cast(bfloat16x4_t, __p0);
37288	37309	return __ret;
37289	37310	}
37290	37311	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_u16(uint16x4_t __p0) {
37291	37312	bfloat16x4_t __ret;
37292		__ret = (bfloat16x4_t)(__p0);
	37313	__ret = __builtin_bit_cast(bfloat16x4_t, __p0);
37293	37314	return __ret;
37294	37315	}
37295	37316	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_s8(int8x8_t __p0) {
37296	37317	bfloat16x4_t __ret;
37297		__ret = (bfloat16x4_t)(__p0);
	37318	__ret = __builtin_bit_cast(bfloat16x4_t, __p0);
37298	37319	return __ret;
37299	37320	}
37300	37321	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_f32(float32x2_t __p0) {
37301	37322	bfloat16x4_t __ret;
37302		__ret = (bfloat16x4_t)(__p0);
	37323	__ret = __builtin_bit_cast(bfloat16x4_t, __p0);
37303	37324	return __ret;
37304	37325	}
37305	37326	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_f16(float16x4_t __p0) {
37306	37327	bfloat16x4_t __ret;
37307		__ret = (bfloat16x4_t)(__p0);
	37328	__ret = __builtin_bit_cast(bfloat16x4_t, __p0);
37308	37329	return __ret;
37309	37330	}
37310	37331	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_s32(int32x2_t __p0) {
37311	37332	bfloat16x4_t __ret;
37312		__ret = (bfloat16x4_t)(__p0);
	37333	__ret = __builtin_bit_cast(bfloat16x4_t, __p0);
37313	37334	return __ret;
37314	37335	}
37315	37336	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_s64(int64x1_t __p0) {
37316	37337	bfloat16x4_t __ret;
37317		__ret = (bfloat16x4_t)(__p0);
	37338	__ret = __builtin_bit_cast(bfloat16x4_t, __p0);
37318	37339	return __ret;
37319	37340	}
37320	37341	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_s16(int16x4_t __p0) {
37321	37342	bfloat16x4_t __ret;
37322		__ret = (bfloat16x4_t)(__p0);
	37343	__ret = __builtin_bit_cast(bfloat16x4_t, __p0);
37323	37344	return __ret;
37324	37345	}
37325	37346	#ifdef __LITTLE_ENDIAN__
...	...	@@ -37335,10 +37356,10 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_s16(int
37335	37356	int32x4_t __ret_215; \
37336	37357	int32x4_t __s0_215 = __p0_215; \
37337	37358	int32x2_t __s1_215 = __p1_215; \
37338		int32x4_t __rev0_215; __rev0_215 = __builtin_shufflevector(__s0_215, __s0_215, 3, 2, 1, 0); \
37339		int32x2_t __rev1_215; __rev1_215 = __builtin_shufflevector(__s1_215, __s1_215, 1, 0); \
	37359	int32x4_t __rev0_215; __rev0_215 = __builtin_shufflevector(__s0_215, __s0_215, __lane_reverse_128_32); \
	37360	int32x2_t __rev1_215; __rev1_215 = __builtin_shufflevector(__s1_215, __s1_215, __lane_reverse_64_32); \
37340	37361	__ret_215 = __noswap_vqdmulhq_s32(__rev0_215, __noswap_splatq_lane_s32(__rev1_215, __p2_215)); \
37341		__ret_215 = __builtin_shufflevector(__ret_215, __ret_215, 3, 2, 1, 0); \
	37362	__ret_215 = __builtin_shufflevector(__ret_215, __ret_215, __lane_reverse_128_32); \
37342	37363	__ret_215; \
37343	37364	})
37344	37365	#endif
...	...	@@ -37356,10 +37377,10 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_s16(int
37356	37377	int16x8_t __ret_217; \
37357	37378	int16x8_t __s0_217 = __p0_217; \
37358	37379	int16x4_t __s1_217 = __p1_217; \
37359		int16x8_t __rev0_217; __rev0_217 = __builtin_shufflevector(__s0_217, __s0_217, 7, 6, 5, 4, 3, 2, 1, 0); \
37360		int16x4_t __rev1_217; __rev1_217 = __builtin_shufflevector(__s1_217, __s1_217, 3, 2, 1, 0); \
	37380	int16x8_t __rev0_217; __rev0_217 = __builtin_shufflevector(__s0_217, __s0_217, __lane_reverse_128_16); \
	37381	int16x4_t __rev1_217; __rev1_217 = __builtin_shufflevector(__s1_217, __s1_217, __lane_reverse_64_16); \
37361	37382	__ret_217 = __noswap_vqdmulhq_s16(__rev0_217, __noswap_splatq_lane_s16(__rev1_217, __p2_217)); \
37362		__ret_217 = __builtin_shufflevector(__ret_217, __ret_217, 7, 6, 5, 4, 3, 2, 1, 0); \
	37383	__ret_217 = __builtin_shufflevector(__ret_217, __ret_217, __lane_reverse_128_16); \
37363	37384	__ret_217; \
37364	37385	})
37365	37386	#endif
...	...	@@ -37377,10 +37398,10 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_s16(int
37377	37398	int32x2_t __ret_219; \
37378	37399	int32x2_t __s0_219 = __p0_219; \
37379	37400	int32x2_t __s1_219 = __p1_219; \
37380		int32x2_t __rev0_219; __rev0_219 = __builtin_shufflevector(__s0_219, __s0_219, 1, 0); \
37381		int32x2_t __rev1_219; __rev1_219 = __builtin_shufflevector(__s1_219, __s1_219, 1, 0); \
	37401	int32x2_t __rev0_219; __rev0_219 = __builtin_shufflevector(__s0_219, __s0_219, __lane_reverse_64_32); \
	37402	int32x2_t __rev1_219; __rev1_219 = __builtin_shufflevector(__s1_219, __s1_219, __lane_reverse_64_32); \
37382	37403	__ret_219 = __noswap_vqdmulh_s32(__rev0_219, __noswap_splat_lane_s32(__rev1_219, __p2_219)); \
37383		__ret_219 = __builtin_shufflevector(__ret_219, __ret_219, 1, 0); \
	37404	__ret_219 = __builtin_shufflevector(__ret_219, __ret_219, __lane_reverse_64_32); \
37384	37405	__ret_219; \
37385	37406	})
37386	37407	#endif
...	...	@@ -37398,10 +37419,10 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_s16(int
37398	37419	int16x4_t __ret_221; \
37399	37420	int16x4_t __s0_221 = __p0_221; \
37400	37421	int16x4_t __s1_221 = __p1_221; \
37401		int16x4_t __rev0_221; __rev0_221 = __builtin_shufflevector(__s0_221, __s0_221, 3, 2, 1, 0); \
37402		int16x4_t __rev1_221; __rev1_221 = __builtin_shufflevector(__s1_221, __s1_221, 3, 2, 1, 0); \
	37422	int16x4_t __rev0_221; __rev0_221 = __builtin_shufflevector(__s0_221, __s0_221, __lane_reverse_64_16); \
	37423	int16x4_t __rev1_221; __rev1_221 = __builtin_shufflevector(__s1_221, __s1_221, __lane_reverse_64_16); \
37403	37424	__ret_221 = __noswap_vqdmulh_s16(__rev0_221, __noswap_splat_lane_s16(__rev1_221, __p2_221)); \
37404		__ret_221 = __builtin_shufflevector(__ret_221, __ret_221, 3, 2, 1, 0); \
	37425	__ret_221 = __builtin_shufflevector(__ret_221, __ret_221, __lane_reverse_64_16); \
37405	37426	__ret_221; \
37406	37427	})
37407	37428	#endif
...	...	@@ -37419,10 +37440,10 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_s16(int
37419	37440	int32x4_t __ret_223; \
37420	37441	int32x4_t __s0_223 = __p0_223; \
37421	37442	int32x2_t __s1_223 = __p1_223; \
37422		int32x4_t __rev0_223; __rev0_223 = __builtin_shufflevector(__s0_223, __s0_223, 3, 2, 1, 0); \
37423		int32x2_t __rev1_223; __rev1_223 = __builtin_shufflevector(__s1_223, __s1_223, 1, 0); \
	37443	int32x4_t __rev0_223; __rev0_223 = __builtin_shufflevector(__s0_223, __s0_223, __lane_reverse_128_32); \
	37444	int32x2_t __rev1_223; __rev1_223 = __builtin_shufflevector(__s1_223, __s1_223, __lane_reverse_64_32); \
37424	37445	__ret_223 = __noswap_vqrdmulhq_s32(__rev0_223, __noswap_splatq_lane_s32(__rev1_223, __p2_223)); \
37425		__ret_223 = __builtin_shufflevector(__ret_223, __ret_223, 3, 2, 1, 0); \
	37446	__ret_223 = __builtin_shufflevector(__ret_223, __ret_223, __lane_reverse_128_32); \
37426	37447	__ret_223; \
37427	37448	})
37428	37449	#endif
...	...	@@ -37440,10 +37461,10 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_s16(int
37440	37461	int16x8_t __ret_225; \
37441	37462	int16x8_t __s0_225 = __p0_225; \
37442	37463	int16x4_t __s1_225 = __p1_225; \
37443		int16x8_t __rev0_225; __rev0_225 = __builtin_shufflevector(__s0_225, __s0_225, 7, 6, 5, 4, 3, 2, 1, 0); \
37444		int16x4_t __rev1_225; __rev1_225 = __builtin_shufflevector(__s1_225, __s1_225, 3, 2, 1, 0); \
	37464	int16x8_t __rev0_225; __rev0_225 = __builtin_shufflevector(__s0_225, __s0_225, __lane_reverse_128_16); \
	37465	int16x4_t __rev1_225; __rev1_225 = __builtin_shufflevector(__s1_225, __s1_225, __lane_reverse_64_16); \
37445	37466	__ret_225 = __noswap_vqrdmulhq_s16(__rev0_225, __noswap_splatq_lane_s16(__rev1_225, __p2_225)); \
37446		__ret_225 = __builtin_shufflevector(__ret_225, __ret_225, 7, 6, 5, 4, 3, 2, 1, 0); \
	37467	__ret_225 = __builtin_shufflevector(__ret_225, __ret_225, __lane_reverse_128_16); \
37447	37468	__ret_225; \
37448	37469	})
37449	37470	#endif
...	...	@@ -37461,10 +37482,10 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_s16(int
37461	37482	int32x2_t __ret_227; \
37462	37483	int32x2_t __s0_227 = __p0_227; \
37463	37484	int32x2_t __s1_227 = __p1_227; \
37464		int32x2_t __rev0_227; __rev0_227 = __builtin_shufflevector(__s0_227, __s0_227, 1, 0); \
37465		int32x2_t __rev1_227; __rev1_227 = __builtin_shufflevector(__s1_227, __s1_227, 1, 0); \
	37485	int32x2_t __rev0_227; __rev0_227 = __builtin_shufflevector(__s0_227, __s0_227, __lane_reverse_64_32); \
	37486	int32x2_t __rev1_227; __rev1_227 = __builtin_shufflevector(__s1_227, __s1_227, __lane_reverse_64_32); \
37466	37487	__ret_227 = __noswap_vqrdmulh_s32(__rev0_227, __noswap_splat_lane_s32(__rev1_227, __p2_227)); \
37467		__ret_227 = __builtin_shufflevector(__ret_227, __ret_227, 1, 0); \
	37488	__ret_227 = __builtin_shufflevector(__ret_227, __ret_227, __lane_reverse_64_32); \
37468	37489	__ret_227; \
37469	37490	})
37470	37491	#endif
...	...	@@ -37482,1332 +37503,1332 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_s16(int
37482	37503	int16x4_t __ret_229; \
37483	37504	int16x4_t __s0_229 = __p0_229; \
37484	37505	int16x4_t __s1_229 = __p1_229; \
37485		int16x4_t __rev0_229; __rev0_229 = __builtin_shufflevector(__s0_229, __s0_229, 3, 2, 1, 0); \
37486		int16x4_t __rev1_229; __rev1_229 = __builtin_shufflevector(__s1_229, __s1_229, 3, 2, 1, 0); \
	37506	int16x4_t __rev0_229; __rev0_229 = __builtin_shufflevector(__s0_229, __s0_229, __lane_reverse_64_16); \
	37507	int16x4_t __rev1_229; __rev1_229 = __builtin_shufflevector(__s1_229, __s1_229, __lane_reverse_64_16); \
37487	37508	__ret_229 = __noswap_vqrdmulh_s16(__rev0_229, __noswap_splat_lane_s16(__rev1_229, __p2_229)); \
37488		__ret_229 = __builtin_shufflevector(__ret_229, __ret_229, 3, 2, 1, 0); \
	37509	__ret_229 = __builtin_shufflevector(__ret_229, __ret_229, __lane_reverse_64_16); \
37489	37510	__ret_229; \
37490	37511	})
37491	37512	#endif
37492	37513
37493	37514	__ai __attribute__((target("neon"))) poly8x8_t vreinterpret_p8_p16(poly16x4_t __p0) {
37494	37515	poly8x8_t __ret;
37495		__ret = (poly8x8_t)(__p0);
	37516	__ret = __builtin_bit_cast(poly8x8_t, __p0);
37496	37517	return __ret;
37497	37518	}
37498	37519	__ai __attribute__((target("neon"))) poly8x8_t vreinterpret_p8_u8(uint8x8_t __p0) {
37499	37520	poly8x8_t __ret;
37500		__ret = (poly8x8_t)(__p0);
	37521	__ret = __builtin_bit_cast(poly8x8_t, __p0);
37501	37522	return __ret;
37502	37523	}
37503	37524	__ai __attribute__((target("neon"))) poly8x8_t vreinterpret_p8_u32(uint32x2_t __p0) {
37504	37525	poly8x8_t __ret;
37505		__ret = (poly8x8_t)(__p0);
	37526	__ret = __builtin_bit_cast(poly8x8_t, __p0);
37506	37527	return __ret;
37507	37528	}
37508	37529	__ai __attribute__((target("neon"))) poly8x8_t vreinterpret_p8_u64(uint64x1_t __p0) {
37509	37530	poly8x8_t __ret;
37510		__ret = (poly8x8_t)(__p0);
	37531	__ret = __builtin_bit_cast(poly8x8_t, __p0);
37511	37532	return __ret;
37512	37533	}
37513	37534	__ai __attribute__((target("neon"))) poly8x8_t vreinterpret_p8_u16(uint16x4_t __p0) {
37514	37535	poly8x8_t __ret;
37515		__ret = (poly8x8_t)(__p0);
	37536	__ret = __builtin_bit_cast(poly8x8_t, __p0);
37516	37537	return __ret;
37517	37538	}
37518	37539	__ai __attribute__((target("neon"))) poly8x8_t vreinterpret_p8_s8(int8x8_t __p0) {
37519	37540	poly8x8_t __ret;
37520		__ret = (poly8x8_t)(__p0);
	37541	__ret = __builtin_bit_cast(poly8x8_t, __p0);
37521	37542	return __ret;
37522	37543	}
37523	37544	__ai __attribute__((target("neon"))) poly8x8_t vreinterpret_p8_f32(float32x2_t __p0) {
37524	37545	poly8x8_t __ret;
37525		__ret = (poly8x8_t)(__p0);
	37546	__ret = __builtin_bit_cast(poly8x8_t, __p0);
37526	37547	return __ret;
37527	37548	}
37528	37549	__ai __attribute__((target("neon"))) poly8x8_t vreinterpret_p8_f16(float16x4_t __p0) {
37529	37550	poly8x8_t __ret;
37530		__ret = (poly8x8_t)(__p0);
	37551	__ret = __builtin_bit_cast(poly8x8_t, __p0);
37531	37552	return __ret;
37532	37553	}
37533	37554	__ai __attribute__((target("neon"))) poly8x8_t vreinterpret_p8_s32(int32x2_t __p0) {
37534	37555	poly8x8_t __ret;
37535		__ret = (poly8x8_t)(__p0);
	37556	__ret = __builtin_bit_cast(poly8x8_t, __p0);
37536	37557	return __ret;
37537	37558	}
37538	37559	__ai __attribute__((target("neon"))) poly8x8_t vreinterpret_p8_s64(int64x1_t __p0) {
37539	37560	poly8x8_t __ret;
37540		__ret = (poly8x8_t)(__p0);
	37561	__ret = __builtin_bit_cast(poly8x8_t, __p0);
37541	37562	return __ret;
37542	37563	}
37543	37564	__ai __attribute__((target("neon"))) poly8x8_t vreinterpret_p8_s16(int16x4_t __p0) {
37544	37565	poly8x8_t __ret;
37545		__ret = (poly8x8_t)(__p0);
	37566	__ret = __builtin_bit_cast(poly8x8_t, __p0);
37546	37567	return __ret;
37547	37568	}
37548	37569	__ai __attribute__((target("neon"))) poly16x4_t vreinterpret_p16_p8(poly8x8_t __p0) {
37549	37570	poly16x4_t __ret;
37550		__ret = (poly16x4_t)(__p0);
	37571	__ret = __builtin_bit_cast(poly16x4_t, __p0);
37551	37572	return __ret;
37552	37573	}
37553	37574	__ai __attribute__((target("neon"))) poly16x4_t vreinterpret_p16_u8(uint8x8_t __p0) {
37554	37575	poly16x4_t __ret;
37555		__ret = (poly16x4_t)(__p0);
	37576	__ret = __builtin_bit_cast(poly16x4_t, __p0);
37556	37577	return __ret;
37557	37578	}
37558	37579	__ai __attribute__((target("neon"))) poly16x4_t vreinterpret_p16_u32(uint32x2_t __p0) {
37559	37580	poly16x4_t __ret;
37560		__ret = (poly16x4_t)(__p0);
	37581	__ret = __builtin_bit_cast(poly16x4_t, __p0);
37561	37582	return __ret;
37562	37583	}
37563	37584	__ai __attribute__((target("neon"))) poly16x4_t vreinterpret_p16_u64(uint64x1_t __p0) {
37564	37585	poly16x4_t __ret;
37565		__ret = (poly16x4_t)(__p0);
	37586	__ret = __builtin_bit_cast(poly16x4_t, __p0);
37566	37587	return __ret;
37567	37588	}
37568	37589	__ai __attribute__((target("neon"))) poly16x4_t vreinterpret_p16_u16(uint16x4_t __p0) {
37569	37590	poly16x4_t __ret;
37570		__ret = (poly16x4_t)(__p0);
	37591	__ret = __builtin_bit_cast(poly16x4_t, __p0);
37571	37592	return __ret;
37572	37593	}
37573	37594	__ai __attribute__((target("neon"))) poly16x4_t vreinterpret_p16_s8(int8x8_t __p0) {
37574	37595	poly16x4_t __ret;
37575		__ret = (poly16x4_t)(__p0);
	37596	__ret = __builtin_bit_cast(poly16x4_t, __p0);
37576	37597	return __ret;
37577	37598	}
37578	37599	__ai __attribute__((target("neon"))) poly16x4_t vreinterpret_p16_f32(float32x2_t __p0) {
37579	37600	poly16x4_t __ret;
37580		__ret = (poly16x4_t)(__p0);
	37601	__ret = __builtin_bit_cast(poly16x4_t, __p0);
37581	37602	return __ret;
37582	37603	}
37583	37604	__ai __attribute__((target("neon"))) poly16x4_t vreinterpret_p16_f16(float16x4_t __p0) {
37584	37605	poly16x4_t __ret;
37585		__ret = (poly16x4_t)(__p0);
	37606	__ret = __builtin_bit_cast(poly16x4_t, __p0);
37586	37607	return __ret;
37587	37608	}
37588	37609	__ai __attribute__((target("neon"))) poly16x4_t vreinterpret_p16_s32(int32x2_t __p0) {
37589	37610	poly16x4_t __ret;
37590		__ret = (poly16x4_t)(__p0);
	37611	__ret = __builtin_bit_cast(poly16x4_t, __p0);
37591	37612	return __ret;
37592	37613	}
37593	37614	__ai __attribute__((target("neon"))) poly16x4_t vreinterpret_p16_s64(int64x1_t __p0) {
37594	37615	poly16x4_t __ret;
37595		__ret = (poly16x4_t)(__p0);
	37616	__ret = __builtin_bit_cast(poly16x4_t, __p0);
37596	37617	return __ret;
37597	37618	}
37598	37619	__ai __attribute__((target("neon"))) poly16x4_t vreinterpret_p16_s16(int16x4_t __p0) {
37599	37620	poly16x4_t __ret;
37600		__ret = (poly16x4_t)(__p0);
	37621	__ret = __builtin_bit_cast(poly16x4_t, __p0);
37601	37622	return __ret;
37602	37623	}
37603	37624	__ai __attribute__((target("neon"))) poly8x16_t vreinterpretq_p8_p16(poly16x8_t __p0) {
37604	37625	poly8x16_t __ret;
37605		__ret = (poly8x16_t)(__p0);
	37626	__ret = __builtin_bit_cast(poly8x16_t, __p0);
37606	37627	return __ret;
37607	37628	}
37608	37629	__ai __attribute__((target("neon"))) poly8x16_t vreinterpretq_p8_u8(uint8x16_t __p0) {
37609	37630	poly8x16_t __ret;
37610		__ret = (poly8x16_t)(__p0);
	37631	__ret = __builtin_bit_cast(poly8x16_t, __p0);
37611	37632	return __ret;
37612	37633	}
37613	37634	__ai __attribute__((target("neon"))) poly8x16_t vreinterpretq_p8_u32(uint32x4_t __p0) {
37614	37635	poly8x16_t __ret;
37615		__ret = (poly8x16_t)(__p0);
	37636	__ret = __builtin_bit_cast(poly8x16_t, __p0);
37616	37637	return __ret;
37617	37638	}
37618	37639	__ai __attribute__((target("neon"))) poly8x16_t vreinterpretq_p8_u64(uint64x2_t __p0) {
37619	37640	poly8x16_t __ret;
37620		__ret = (poly8x16_t)(__p0);
	37641	__ret = __builtin_bit_cast(poly8x16_t, __p0);
37621	37642	return __ret;
37622	37643	}
37623	37644	__ai __attribute__((target("neon"))) poly8x16_t vreinterpretq_p8_u16(uint16x8_t __p0) {
37624	37645	poly8x16_t __ret;
37625		__ret = (poly8x16_t)(__p0);
	37646	__ret = __builtin_bit_cast(poly8x16_t, __p0);
37626	37647	return __ret;
37627	37648	}
37628	37649	__ai __attribute__((target("neon"))) poly8x16_t vreinterpretq_p8_s8(int8x16_t __p0) {
37629	37650	poly8x16_t __ret;
37630		__ret = (poly8x16_t)(__p0);
	37651	__ret = __builtin_bit_cast(poly8x16_t, __p0);
37631	37652	return __ret;
37632	37653	}
37633	37654	__ai __attribute__((target("neon"))) poly8x16_t vreinterpretq_p8_f32(float32x4_t __p0) {
37634	37655	poly8x16_t __ret;
37635		__ret = (poly8x16_t)(__p0);
	37656	__ret = __builtin_bit_cast(poly8x16_t, __p0);
37636	37657	return __ret;
37637	37658	}
37638	37659	__ai __attribute__((target("neon"))) poly8x16_t vreinterpretq_p8_f16(float16x8_t __p0) {
37639	37660	poly8x16_t __ret;
37640		__ret = (poly8x16_t)(__p0);
	37661	__ret = __builtin_bit_cast(poly8x16_t, __p0);
37641	37662	return __ret;
37642	37663	}
37643	37664	__ai __attribute__((target("neon"))) poly8x16_t vreinterpretq_p8_s32(int32x4_t __p0) {
37644	37665	poly8x16_t __ret;
37645		__ret = (poly8x16_t)(__p0);
	37666	__ret = __builtin_bit_cast(poly8x16_t, __p0);
37646	37667	return __ret;
37647	37668	}
37648	37669	__ai __attribute__((target("neon"))) poly8x16_t vreinterpretq_p8_s64(int64x2_t __p0) {
37649	37670	poly8x16_t __ret;
37650		__ret = (poly8x16_t)(__p0);
	37671	__ret = __builtin_bit_cast(poly8x16_t, __p0);
37651	37672	return __ret;
37652	37673	}
37653	37674	__ai __attribute__((target("neon"))) poly8x16_t vreinterpretq_p8_s16(int16x8_t __p0) {
37654	37675	poly8x16_t __ret;
37655		__ret = (poly8x16_t)(__p0);
	37676	__ret = __builtin_bit_cast(poly8x16_t, __p0);
37656	37677	return __ret;
37657	37678	}
37658	37679	__ai __attribute__((target("neon"))) poly16x8_t vreinterpretq_p16_p8(poly8x16_t __p0) {
37659	37680	poly16x8_t __ret;
37660		__ret = (poly16x8_t)(__p0);
	37681	__ret = __builtin_bit_cast(poly16x8_t, __p0);
37661	37682	return __ret;
37662	37683	}
37663	37684	__ai __attribute__((target("neon"))) poly16x8_t vreinterpretq_p16_u8(uint8x16_t __p0) {
37664	37685	poly16x8_t __ret;
37665		__ret = (poly16x8_t)(__p0);
	37686	__ret = __builtin_bit_cast(poly16x8_t, __p0);
37666	37687	return __ret;
37667	37688	}
37668	37689	__ai __attribute__((target("neon"))) poly16x8_t vreinterpretq_p16_u32(uint32x4_t __p0) {
37669	37690	poly16x8_t __ret;
37670		__ret = (poly16x8_t)(__p0);
	37691	__ret = __builtin_bit_cast(poly16x8_t, __p0);
37671	37692	return __ret;
37672	37693	}
37673	37694	__ai __attribute__((target("neon"))) poly16x8_t vreinterpretq_p16_u64(uint64x2_t __p0) {
37674	37695	poly16x8_t __ret;
37675		__ret = (poly16x8_t)(__p0);
	37696	__ret = __builtin_bit_cast(poly16x8_t, __p0);
37676	37697	return __ret;
37677	37698	}
37678	37699	__ai __attribute__((target("neon"))) poly16x8_t vreinterpretq_p16_u16(uint16x8_t __p0) {
37679	37700	poly16x8_t __ret;
37680		__ret = (poly16x8_t)(__p0);
	37701	__ret = __builtin_bit_cast(poly16x8_t, __p0);
37681	37702	return __ret;
37682	37703	}
37683	37704	__ai __attribute__((target("neon"))) poly16x8_t vreinterpretq_p16_s8(int8x16_t __p0) {
37684	37705	poly16x8_t __ret;
37685		__ret = (poly16x8_t)(__p0);
	37706	__ret = __builtin_bit_cast(poly16x8_t, __p0);
37686	37707	return __ret;
37687	37708	}
37688	37709	__ai __attribute__((target("neon"))) poly16x8_t vreinterpretq_p16_f32(float32x4_t __p0) {
37689	37710	poly16x8_t __ret;
37690		__ret = (poly16x8_t)(__p0);
	37711	__ret = __builtin_bit_cast(poly16x8_t, __p0);
37691	37712	return __ret;
37692	37713	}
37693	37714	__ai __attribute__((target("neon"))) poly16x8_t vreinterpretq_p16_f16(float16x8_t __p0) {
37694	37715	poly16x8_t __ret;
37695		__ret = (poly16x8_t)(__p0);
	37716	__ret = __builtin_bit_cast(poly16x8_t, __p0);
37696	37717	return __ret;
37697	37718	}
37698	37719	__ai __attribute__((target("neon"))) poly16x8_t vreinterpretq_p16_s32(int32x4_t __p0) {
37699	37720	poly16x8_t __ret;
37700		__ret = (poly16x8_t)(__p0);
	37721	__ret = __builtin_bit_cast(poly16x8_t, __p0);
37701	37722	return __ret;
37702	37723	}
37703	37724	__ai __attribute__((target("neon"))) poly16x8_t vreinterpretq_p16_s64(int64x2_t __p0) {
37704	37725	poly16x8_t __ret;
37705		__ret = (poly16x8_t)(__p0);
	37726	__ret = __builtin_bit_cast(poly16x8_t, __p0);
37706	37727	return __ret;
37707	37728	}
37708	37729	__ai __attribute__((target("neon"))) poly16x8_t vreinterpretq_p16_s16(int16x8_t __p0) {
37709	37730	poly16x8_t __ret;
37710		__ret = (poly16x8_t)(__p0);
	37731	__ret = __builtin_bit_cast(poly16x8_t, __p0);
37711	37732	return __ret;
37712	37733	}
37713	37734	__ai __attribute__((target("neon"))) uint8x16_t vreinterpretq_u8_p8(poly8x16_t __p0) {
37714	37735	uint8x16_t __ret;
37715		__ret = (uint8x16_t)(__p0);
	37736	__ret = __builtin_bit_cast(uint8x16_t, __p0);
37716	37737	return __ret;
37717	37738	}
37718	37739	__ai __attribute__((target("neon"))) uint8x16_t vreinterpretq_u8_p16(poly16x8_t __p0) {
37719	37740	uint8x16_t __ret;
37720		__ret = (uint8x16_t)(__p0);
	37741	__ret = __builtin_bit_cast(uint8x16_t, __p0);
37721	37742	return __ret;
37722	37743	}
37723	37744	__ai __attribute__((target("neon"))) uint8x16_t vreinterpretq_u8_u32(uint32x4_t __p0) {
37724	37745	uint8x16_t __ret;
37725		__ret = (uint8x16_t)(__p0);
	37746	__ret = __builtin_bit_cast(uint8x16_t, __p0);
37726	37747	return __ret;
37727	37748	}
37728	37749	__ai __attribute__((target("neon"))) uint8x16_t vreinterpretq_u8_u64(uint64x2_t __p0) {
37729	37750	uint8x16_t __ret;
37730		__ret = (uint8x16_t)(__p0);
	37751	__ret = __builtin_bit_cast(uint8x16_t, __p0);
37731	37752	return __ret;
37732	37753	}
37733	37754	__ai __attribute__((target("neon"))) uint8x16_t vreinterpretq_u8_u16(uint16x8_t __p0) {
37734	37755	uint8x16_t __ret;
37735		__ret = (uint8x16_t)(__p0);
	37756	__ret = __builtin_bit_cast(uint8x16_t, __p0);
37736	37757	return __ret;
37737	37758	}
37738	37759	__ai __attribute__((target("neon"))) uint8x16_t vreinterpretq_u8_s8(int8x16_t __p0) {
37739	37760	uint8x16_t __ret;
37740		__ret = (uint8x16_t)(__p0);
	37761	__ret = __builtin_bit_cast(uint8x16_t, __p0);
37741	37762	return __ret;
37742	37763	}
37743	37764	__ai __attribute__((target("neon"))) uint8x16_t vreinterpretq_u8_f32(float32x4_t __p0) {
37744	37765	uint8x16_t __ret;
37745		__ret = (uint8x16_t)(__p0);
	37766	__ret = __builtin_bit_cast(uint8x16_t, __p0);
37746	37767	return __ret;
37747	37768	}
37748	37769	__ai __attribute__((target("neon"))) uint8x16_t vreinterpretq_u8_f16(float16x8_t __p0) {
37749	37770	uint8x16_t __ret;
37750		__ret = (uint8x16_t)(__p0);
	37771	__ret = __builtin_bit_cast(uint8x16_t, __p0);
37751	37772	return __ret;
37752	37773	}
37753	37774	__ai __attribute__((target("neon"))) uint8x16_t vreinterpretq_u8_s32(int32x4_t __p0) {
37754	37775	uint8x16_t __ret;
37755		__ret = (uint8x16_t)(__p0);
	37776	__ret = __builtin_bit_cast(uint8x16_t, __p0);
37756	37777	return __ret;
37757	37778	}
37758	37779	__ai __attribute__((target("neon"))) uint8x16_t vreinterpretq_u8_s64(int64x2_t __p0) {
37759	37780	uint8x16_t __ret;
37760		__ret = (uint8x16_t)(__p0);
	37781	__ret = __builtin_bit_cast(uint8x16_t, __p0);
37761	37782	return __ret;
37762	37783	}
37763	37784	__ai __attribute__((target("neon"))) uint8x16_t vreinterpretq_u8_s16(int16x8_t __p0) {
37764	37785	uint8x16_t __ret;
37765		__ret = (uint8x16_t)(__p0);
	37786	__ret = __builtin_bit_cast(uint8x16_t, __p0);
37766	37787	return __ret;
37767	37788	}
37768	37789	__ai __attribute__((target("neon"))) uint32x4_t vreinterpretq_u32_p8(poly8x16_t __p0) {
37769	37790	uint32x4_t __ret;
37770		__ret = (uint32x4_t)(__p0);
	37791	__ret = __builtin_bit_cast(uint32x4_t, __p0);
37771	37792	return __ret;
37772	37793	}
37773	37794	__ai __attribute__((target("neon"))) uint32x4_t vreinterpretq_u32_p16(poly16x8_t __p0) {
37774	37795	uint32x4_t __ret;
37775		__ret = (uint32x4_t)(__p0);
	37796	__ret = __builtin_bit_cast(uint32x4_t, __p0);
37776	37797	return __ret;
37777	37798	}
37778	37799	__ai __attribute__((target("neon"))) uint32x4_t vreinterpretq_u32_u8(uint8x16_t __p0) {
37779	37800	uint32x4_t __ret;
37780		__ret = (uint32x4_t)(__p0);
	37801	__ret = __builtin_bit_cast(uint32x4_t, __p0);
37781	37802	return __ret;
37782	37803	}
37783	37804	__ai __attribute__((target("neon"))) uint32x4_t vreinterpretq_u32_u64(uint64x2_t __p0) {
37784	37805	uint32x4_t __ret;
37785		__ret = (uint32x4_t)(__p0);
	37806	__ret = __builtin_bit_cast(uint32x4_t, __p0);
37786	37807	return __ret;
37787	37808	}
37788	37809	__ai __attribute__((target("neon"))) uint32x4_t vreinterpretq_u32_u16(uint16x8_t __p0) {
37789	37810	uint32x4_t __ret;
37790		__ret = (uint32x4_t)(__p0);
	37811	__ret = __builtin_bit_cast(uint32x4_t, __p0);
37791	37812	return __ret;
37792	37813	}
37793	37814	__ai __attribute__((target("neon"))) uint32x4_t vreinterpretq_u32_s8(int8x16_t __p0) {
37794	37815	uint32x4_t __ret;
37795		__ret = (uint32x4_t)(__p0);
	37816	__ret = __builtin_bit_cast(uint32x4_t, __p0);
37796	37817	return __ret;
37797	37818	}
37798	37819	__ai __attribute__((target("neon"))) uint32x4_t vreinterpretq_u32_f32(float32x4_t __p0) {
37799	37820	uint32x4_t __ret;
37800		__ret = (uint32x4_t)(__p0);
	37821	__ret = __builtin_bit_cast(uint32x4_t, __p0);
37801	37822	return __ret;
37802	37823	}
37803	37824	__ai __attribute__((target("neon"))) uint32x4_t vreinterpretq_u32_f16(float16x8_t __p0) {
37804	37825	uint32x4_t __ret;
37805		__ret = (uint32x4_t)(__p0);
	37826	__ret = __builtin_bit_cast(uint32x4_t, __p0);
37806	37827	return __ret;
37807	37828	}
37808	37829	__ai __attribute__((target("neon"))) uint32x4_t vreinterpretq_u32_s32(int32x4_t __p0) {
37809	37830	uint32x4_t __ret;
37810		__ret = (uint32x4_t)(__p0);
	37831	__ret = __builtin_bit_cast(uint32x4_t, __p0);
37811	37832	return __ret;
37812	37833	}
37813	37834	__ai __attribute__((target("neon"))) uint32x4_t vreinterpretq_u32_s64(int64x2_t __p0) {
37814	37835	uint32x4_t __ret;
37815		__ret = (uint32x4_t)(__p0);
	37836	__ret = __builtin_bit_cast(uint32x4_t, __p0);
37816	37837	return __ret;
37817	37838	}
37818	37839	__ai __attribute__((target("neon"))) uint32x4_t vreinterpretq_u32_s16(int16x8_t __p0) {
37819	37840	uint32x4_t __ret;
37820		__ret = (uint32x4_t)(__p0);
	37841	__ret = __builtin_bit_cast(uint32x4_t, __p0);
37821	37842	return __ret;
37822	37843	}
37823	37844	__ai __attribute__((target("neon"))) uint64x2_t vreinterpretq_u64_p8(poly8x16_t __p0) {
37824	37845	uint64x2_t __ret;
37825		__ret = (uint64x2_t)(__p0);
	37846	__ret = __builtin_bit_cast(uint64x2_t, __p0);
37826	37847	return __ret;
37827	37848	}
37828	37849	__ai __attribute__((target("neon"))) uint64x2_t vreinterpretq_u64_p16(poly16x8_t __p0) {
37829	37850	uint64x2_t __ret;
37830		__ret = (uint64x2_t)(__p0);
	37851	__ret = __builtin_bit_cast(uint64x2_t, __p0);
37831	37852	return __ret;
37832	37853	}
37833	37854	__ai __attribute__((target("neon"))) uint64x2_t vreinterpretq_u64_u8(uint8x16_t __p0) {
37834	37855	uint64x2_t __ret;
37835		__ret = (uint64x2_t)(__p0);
	37856	__ret = __builtin_bit_cast(uint64x2_t, __p0);
37836	37857	return __ret;
37837	37858	}
37838	37859	__ai __attribute__((target("neon"))) uint64x2_t vreinterpretq_u64_u32(uint32x4_t __p0) {
37839	37860	uint64x2_t __ret;
37840		__ret = (uint64x2_t)(__p0);
	37861	__ret = __builtin_bit_cast(uint64x2_t, __p0);
37841	37862	return __ret;
37842	37863	}
37843	37864	__ai __attribute__((target("neon"))) uint64x2_t vreinterpretq_u64_u16(uint16x8_t __p0) {
37844	37865	uint64x2_t __ret;
37845		__ret = (uint64x2_t)(__p0);
	37866	__ret = __builtin_bit_cast(uint64x2_t, __p0);
37846	37867	return __ret;
37847	37868	}
37848	37869	__ai __attribute__((target("neon"))) uint64x2_t vreinterpretq_u64_s8(int8x16_t __p0) {
37849	37870	uint64x2_t __ret;
37850		__ret = (uint64x2_t)(__p0);
	37871	__ret = __builtin_bit_cast(uint64x2_t, __p0);
37851	37872	return __ret;
37852	37873	}
37853	37874	__ai __attribute__((target("neon"))) uint64x2_t vreinterpretq_u64_f32(float32x4_t __p0) {
37854	37875	uint64x2_t __ret;
37855		__ret = (uint64x2_t)(__p0);
	37876	__ret = __builtin_bit_cast(uint64x2_t, __p0);
37856	37877	return __ret;
37857	37878	}
37858	37879	__ai __attribute__((target("neon"))) uint64x2_t vreinterpretq_u64_f16(float16x8_t __p0) {
37859	37880	uint64x2_t __ret;
37860		__ret = (uint64x2_t)(__p0);
	37881	__ret = __builtin_bit_cast(uint64x2_t, __p0);
37861	37882	return __ret;
37862	37883	}
37863	37884	__ai __attribute__((target("neon"))) uint64x2_t vreinterpretq_u64_s32(int32x4_t __p0) {
37864	37885	uint64x2_t __ret;
37865		__ret = (uint64x2_t)(__p0);
	37886	__ret = __builtin_bit_cast(uint64x2_t, __p0);
37866	37887	return __ret;
37867	37888	}
37868	37889	__ai __attribute__((target("neon"))) uint64x2_t vreinterpretq_u64_s64(int64x2_t __p0) {
37869	37890	uint64x2_t __ret;
37870		__ret = (uint64x2_t)(__p0);
	37891	__ret = __builtin_bit_cast(uint64x2_t, __p0);
37871	37892	return __ret;
37872	37893	}
37873	37894	__ai __attribute__((target("neon"))) uint64x2_t vreinterpretq_u64_s16(int16x8_t __p0) {
37874	37895	uint64x2_t __ret;
37875		__ret = (uint64x2_t)(__p0);
	37896	__ret = __builtin_bit_cast(uint64x2_t, __p0);
37876	37897	return __ret;
37877	37898	}
37878	37899	__ai __attribute__((target("neon"))) uint16x8_t vreinterpretq_u16_p8(poly8x16_t __p0) {
37879	37900	uint16x8_t __ret;
37880		__ret = (uint16x8_t)(__p0);
	37901	__ret = __builtin_bit_cast(uint16x8_t, __p0);
37881	37902	return __ret;
37882	37903	}
37883	37904	__ai __attribute__((target("neon"))) uint16x8_t vreinterpretq_u16_p16(poly16x8_t __p0) {
37884	37905	uint16x8_t __ret;
37885		__ret = (uint16x8_t)(__p0);
	37906	__ret = __builtin_bit_cast(uint16x8_t, __p0);
37886	37907	return __ret;
37887	37908	}
37888	37909	__ai __attribute__((target("neon"))) uint16x8_t vreinterpretq_u16_u8(uint8x16_t __p0) {
37889	37910	uint16x8_t __ret;
37890		__ret = (uint16x8_t)(__p0);
	37911	__ret = __builtin_bit_cast(uint16x8_t, __p0);
37891	37912	return __ret;
37892	37913	}
37893	37914	__ai __attribute__((target("neon"))) uint16x8_t vreinterpretq_u16_u32(uint32x4_t __p0) {
37894	37915	uint16x8_t __ret;
37895		__ret = (uint16x8_t)(__p0);
	37916	__ret = __builtin_bit_cast(uint16x8_t, __p0);
37896	37917	return __ret;
37897	37918	}
37898	37919	__ai __attribute__((target("neon"))) uint16x8_t vreinterpretq_u16_u64(uint64x2_t __p0) {
37899	37920	uint16x8_t __ret;
37900		__ret = (uint16x8_t)(__p0);
	37921	__ret = __builtin_bit_cast(uint16x8_t, __p0);
37901	37922	return __ret;
37902	37923	}
37903	37924	__ai __attribute__((target("neon"))) uint16x8_t vreinterpretq_u16_s8(int8x16_t __p0) {
37904	37925	uint16x8_t __ret;
37905		__ret = (uint16x8_t)(__p0);
	37926	__ret = __builtin_bit_cast(uint16x8_t, __p0);
37906	37927	return __ret;
37907	37928	}
37908	37929	__ai __attribute__((target("neon"))) uint16x8_t vreinterpretq_u16_f32(float32x4_t __p0) {
37909	37930	uint16x8_t __ret;
37910		__ret = (uint16x8_t)(__p0);
	37931	__ret = __builtin_bit_cast(uint16x8_t, __p0);
37911	37932	return __ret;
37912	37933	}
37913	37934	__ai __attribute__((target("neon"))) uint16x8_t vreinterpretq_u16_f16(float16x8_t __p0) {
37914	37935	uint16x8_t __ret;
37915		__ret = (uint16x8_t)(__p0);
	37936	__ret = __builtin_bit_cast(uint16x8_t, __p0);
37916	37937	return __ret;
37917	37938	}
37918	37939	__ai __attribute__((target("neon"))) uint16x8_t vreinterpretq_u16_s32(int32x4_t __p0) {
37919	37940	uint16x8_t __ret;
37920		__ret = (uint16x8_t)(__p0);
	37941	__ret = __builtin_bit_cast(uint16x8_t, __p0);
37921	37942	return __ret;
37922	37943	}
37923	37944	__ai __attribute__((target("neon"))) uint16x8_t vreinterpretq_u16_s64(int64x2_t __p0) {
37924	37945	uint16x8_t __ret;
37925		__ret = (uint16x8_t)(__p0);
	37946	__ret = __builtin_bit_cast(uint16x8_t, __p0);
37926	37947	return __ret;
37927	37948	}
37928	37949	__ai __attribute__((target("neon"))) uint16x8_t vreinterpretq_u16_s16(int16x8_t __p0) {
37929	37950	uint16x8_t __ret;
37930		__ret = (uint16x8_t)(__p0);
	37951	__ret = __builtin_bit_cast(uint16x8_t, __p0);
37931	37952	return __ret;
37932	37953	}
37933	37954	__ai __attribute__((target("neon"))) int8x16_t vreinterpretq_s8_p8(poly8x16_t __p0) {
37934	37955	int8x16_t __ret;
37935		__ret = (int8x16_t)(__p0);
	37956	__ret = __builtin_bit_cast(int8x16_t, __p0);
37936	37957	return __ret;
37937	37958	}
37938	37959	__ai __attribute__((target("neon"))) int8x16_t vreinterpretq_s8_p16(poly16x8_t __p0) {
37939	37960	int8x16_t __ret;
37940		__ret = (int8x16_t)(__p0);
	37961	__ret = __builtin_bit_cast(int8x16_t, __p0);
37941	37962	return __ret;
37942	37963	}
37943	37964	__ai __attribute__((target("neon"))) int8x16_t vreinterpretq_s8_u8(uint8x16_t __p0) {
37944	37965	int8x16_t __ret;
37945		__ret = (int8x16_t)(__p0);
	37966	__ret = __builtin_bit_cast(int8x16_t, __p0);
37946	37967	return __ret;
37947	37968	}
37948	37969	__ai __attribute__((target("neon"))) int8x16_t vreinterpretq_s8_u32(uint32x4_t __p0) {
37949	37970	int8x16_t __ret;
37950		__ret = (int8x16_t)(__p0);
	37971	__ret = __builtin_bit_cast(int8x16_t, __p0);
37951	37972	return __ret;
37952	37973	}
37953	37974	__ai __attribute__((target("neon"))) int8x16_t vreinterpretq_s8_u64(uint64x2_t __p0) {
37954	37975	int8x16_t __ret;
37955		__ret = (int8x16_t)(__p0);
	37976	__ret = __builtin_bit_cast(int8x16_t, __p0);
37956	37977	return __ret;
37957	37978	}
37958	37979	__ai __attribute__((target("neon"))) int8x16_t vreinterpretq_s8_u16(uint16x8_t __p0) {
37959	37980	int8x16_t __ret;
37960		__ret = (int8x16_t)(__p0);
	37981	__ret = __builtin_bit_cast(int8x16_t, __p0);
37961	37982	return __ret;
37962	37983	}
37963	37984	__ai __attribute__((target("neon"))) int8x16_t vreinterpretq_s8_f32(float32x4_t __p0) {
37964	37985	int8x16_t __ret;
37965		__ret = (int8x16_t)(__p0);
	37986	__ret = __builtin_bit_cast(int8x16_t, __p0);
37966	37987	return __ret;
37967	37988	}
37968	37989	__ai __attribute__((target("neon"))) int8x16_t vreinterpretq_s8_f16(float16x8_t __p0) {
37969	37990	int8x16_t __ret;
37970		__ret = (int8x16_t)(__p0);
	37991	__ret = __builtin_bit_cast(int8x16_t, __p0);
37971	37992	return __ret;
37972	37993	}
37973	37994	__ai __attribute__((target("neon"))) int8x16_t vreinterpretq_s8_s32(int32x4_t __p0) {
37974	37995	int8x16_t __ret;
37975		__ret = (int8x16_t)(__p0);
	37996	__ret = __builtin_bit_cast(int8x16_t, __p0);
37976	37997	return __ret;
37977	37998	}
37978	37999	__ai __attribute__((target("neon"))) int8x16_t vreinterpretq_s8_s64(int64x2_t __p0) {
37979	38000	int8x16_t __ret;
37980		__ret = (int8x16_t)(__p0);
	38001	__ret = __builtin_bit_cast(int8x16_t, __p0);
37981	38002	return __ret;
37982	38003	}
37983	38004	__ai __attribute__((target("neon"))) int8x16_t vreinterpretq_s8_s16(int16x8_t __p0) {
37984	38005	int8x16_t __ret;
37985		__ret = (int8x16_t)(__p0);
	38006	__ret = __builtin_bit_cast(int8x16_t, __p0);
37986	38007	return __ret;
37987	38008	}
37988	38009	__ai __attribute__((target("neon"))) float32x4_t vreinterpretq_f32_p8(poly8x16_t __p0) {
37989	38010	float32x4_t __ret;
37990		__ret = (float32x4_t)(__p0);
	38011	__ret = __builtin_bit_cast(float32x4_t, __p0);
37991	38012	return __ret;
37992	38013	}
37993	38014	__ai __attribute__((target("neon"))) float32x4_t vreinterpretq_f32_p16(poly16x8_t __p0) {
37994	38015	float32x4_t __ret;
37995		__ret = (float32x4_t)(__p0);
	38016	__ret = __builtin_bit_cast(float32x4_t, __p0);
37996	38017	return __ret;
37997	38018	}
37998	38019	__ai __attribute__((target("neon"))) float32x4_t vreinterpretq_f32_u8(uint8x16_t __p0) {
37999	38020	float32x4_t __ret;
38000		__ret = (float32x4_t)(__p0);
	38021	__ret = __builtin_bit_cast(float32x4_t, __p0);
38001	38022	return __ret;
38002	38023	}
38003	38024	__ai __attribute__((target("neon"))) float32x4_t vreinterpretq_f32_u32(uint32x4_t __p0) {
38004	38025	float32x4_t __ret;
38005		__ret = (float32x4_t)(__p0);
	38026	__ret = __builtin_bit_cast(float32x4_t, __p0);
38006	38027	return __ret;
38007	38028	}
38008	38029	__ai __attribute__((target("neon"))) float32x4_t vreinterpretq_f32_u64(uint64x2_t __p0) {
38009	38030	float32x4_t __ret;
38010		__ret = (float32x4_t)(__p0);
	38031	__ret = __builtin_bit_cast(float32x4_t, __p0);
38011	38032	return __ret;
38012	38033	}
38013	38034	__ai __attribute__((target("neon"))) float32x4_t vreinterpretq_f32_u16(uint16x8_t __p0) {
38014	38035	float32x4_t __ret;
38015		__ret = (float32x4_t)(__p0);
	38036	__ret = __builtin_bit_cast(float32x4_t, __p0);
38016	38037	return __ret;
38017	38038	}
38018	38039	__ai __attribute__((target("neon"))) float32x4_t vreinterpretq_f32_s8(int8x16_t __p0) {
38019	38040	float32x4_t __ret;
38020		__ret = (float32x4_t)(__p0);
	38041	__ret = __builtin_bit_cast(float32x4_t, __p0);
38021	38042	return __ret;
38022	38043	}
38023	38044	__ai __attribute__((target("neon"))) float32x4_t vreinterpretq_f32_f16(float16x8_t __p0) {
38024	38045	float32x4_t __ret;
38025		__ret = (float32x4_t)(__p0);
	38046	__ret = __builtin_bit_cast(float32x4_t, __p0);
38026	38047	return __ret;
38027	38048	}
38028	38049	__ai __attribute__((target("neon"))) float32x4_t vreinterpretq_f32_s32(int32x4_t __p0) {
38029	38050	float32x4_t __ret;
38030		__ret = (float32x4_t)(__p0);
	38051	__ret = __builtin_bit_cast(float32x4_t, __p0);
38031	38052	return __ret;
38032	38053	}
38033	38054	__ai __attribute__((target("neon"))) float32x4_t vreinterpretq_f32_s64(int64x2_t __p0) {
38034	38055	float32x4_t __ret;
38035		__ret = (float32x4_t)(__p0);
	38056	__ret = __builtin_bit_cast(float32x4_t, __p0);
38036	38057	return __ret;
38037	38058	}
38038	38059	__ai __attribute__((target("neon"))) float32x4_t vreinterpretq_f32_s16(int16x8_t __p0) {
38039	38060	float32x4_t __ret;
38040		__ret = (float32x4_t)(__p0);
	38061	__ret = __builtin_bit_cast(float32x4_t, __p0);
38041	38062	return __ret;
38042	38063	}
38043	38064	__ai __attribute__((target("neon"))) float16x8_t vreinterpretq_f16_p8(poly8x16_t __p0) {
38044	38065	float16x8_t __ret;
38045		__ret = (float16x8_t)(__p0);
	38066	__ret = __builtin_bit_cast(float16x8_t, __p0);
38046	38067	return __ret;
38047	38068	}
38048	38069	__ai __attribute__((target("neon"))) float16x8_t vreinterpretq_f16_p16(poly16x8_t __p0) {
38049	38070	float16x8_t __ret;
38050		__ret = (float16x8_t)(__p0);
	38071	__ret = __builtin_bit_cast(float16x8_t, __p0);
38051	38072	return __ret;
38052	38073	}
38053	38074	__ai __attribute__((target("neon"))) float16x8_t vreinterpretq_f16_u8(uint8x16_t __p0) {
38054	38075	float16x8_t __ret;
38055		__ret = (float16x8_t)(__p0);
	38076	__ret = __builtin_bit_cast(float16x8_t, __p0);
38056	38077	return __ret;
38057	38078	}
38058	38079	__ai __attribute__((target("neon"))) float16x8_t vreinterpretq_f16_u32(uint32x4_t __p0) {
38059	38080	float16x8_t __ret;
38060		__ret = (float16x8_t)(__p0);
	38081	__ret = __builtin_bit_cast(float16x8_t, __p0);
38061	38082	return __ret;
38062	38083	}
38063	38084	__ai __attribute__((target("neon"))) float16x8_t vreinterpretq_f16_u64(uint64x2_t __p0) {
38064	38085	float16x8_t __ret;
38065		__ret = (float16x8_t)(__p0);
	38086	__ret = __builtin_bit_cast(float16x8_t, __p0);
38066	38087	return __ret;
38067	38088	}
38068	38089	__ai __attribute__((target("neon"))) float16x8_t vreinterpretq_f16_u16(uint16x8_t __p0) {
38069	38090	float16x8_t __ret;
38070		__ret = (float16x8_t)(__p0);
	38091	__ret = __builtin_bit_cast(float16x8_t, __p0);
38071	38092	return __ret;
38072	38093	}
38073	38094	__ai __attribute__((target("neon"))) float16x8_t vreinterpretq_f16_s8(int8x16_t __p0) {
38074	38095	float16x8_t __ret;
38075		__ret = (float16x8_t)(__p0);
	38096	__ret = __builtin_bit_cast(float16x8_t, __p0);
38076	38097	return __ret;
38077	38098	}
38078	38099	__ai __attribute__((target("neon"))) float16x8_t vreinterpretq_f16_f32(float32x4_t __p0) {
38079	38100	float16x8_t __ret;
38080		__ret = (float16x8_t)(__p0);
	38101	__ret = __builtin_bit_cast(float16x8_t, __p0);
38081	38102	return __ret;
38082	38103	}
38083	38104	__ai __attribute__((target("neon"))) float16x8_t vreinterpretq_f16_s32(int32x4_t __p0) {
38084	38105	float16x8_t __ret;
38085		__ret = (float16x8_t)(__p0);
	38106	__ret = __builtin_bit_cast(float16x8_t, __p0);
38086	38107	return __ret;
38087	38108	}
38088	38109	__ai __attribute__((target("neon"))) float16x8_t vreinterpretq_f16_s64(int64x2_t __p0) {
38089	38110	float16x8_t __ret;
38090		__ret = (float16x8_t)(__p0);
	38111	__ret = __builtin_bit_cast(float16x8_t, __p0);
38091	38112	return __ret;
38092	38113	}
38093	38114	__ai __attribute__((target("neon"))) float16x8_t vreinterpretq_f16_s16(int16x8_t __p0) {
38094	38115	float16x8_t __ret;
38095		__ret = (float16x8_t)(__p0);
	38116	__ret = __builtin_bit_cast(float16x8_t, __p0);
38096	38117	return __ret;
38097	38118	}
38098	38119	__ai __attribute__((target("neon"))) int32x4_t vreinterpretq_s32_p8(poly8x16_t __p0) {
38099	38120	int32x4_t __ret;
38100		__ret = (int32x4_t)(__p0);
	38121	__ret = __builtin_bit_cast(int32x4_t, __p0);
38101	38122	return __ret;
38102	38123	}
38103	38124	__ai __attribute__((target("neon"))) int32x4_t vreinterpretq_s32_p16(poly16x8_t __p0) {
38104	38125	int32x4_t __ret;
38105		__ret = (int32x4_t)(__p0);
	38126	__ret = __builtin_bit_cast(int32x4_t, __p0);
38106	38127	return __ret;
38107	38128	}
38108	38129	__ai __attribute__((target("neon"))) int32x4_t vreinterpretq_s32_u8(uint8x16_t __p0) {
38109	38130	int32x4_t __ret;
38110		__ret = (int32x4_t)(__p0);
	38131	__ret = __builtin_bit_cast(int32x4_t, __p0);
38111	38132	return __ret;
38112	38133	}
38113	38134	__ai __attribute__((target("neon"))) int32x4_t vreinterpretq_s32_u32(uint32x4_t __p0) {
38114	38135	int32x4_t __ret;
38115		__ret = (int32x4_t)(__p0);
	38136	__ret = __builtin_bit_cast(int32x4_t, __p0);
38116	38137	return __ret;
38117	38138	}
38118	38139	__ai __attribute__((target("neon"))) int32x4_t vreinterpretq_s32_u64(uint64x2_t __p0) {
38119	38140	int32x4_t __ret;
38120		__ret = (int32x4_t)(__p0);
	38141	__ret = __builtin_bit_cast(int32x4_t, __p0);
38121	38142	return __ret;
38122	38143	}
38123	38144	__ai __attribute__((target("neon"))) int32x4_t vreinterpretq_s32_u16(uint16x8_t __p0) {
38124	38145	int32x4_t __ret;
38125		__ret = (int32x4_t)(__p0);
	38146	__ret = __builtin_bit_cast(int32x4_t, __p0);
38126	38147	return __ret;
38127	38148	}
38128	38149	__ai __attribute__((target("neon"))) int32x4_t vreinterpretq_s32_s8(int8x16_t __p0) {
38129	38150	int32x4_t __ret;
38130		__ret = (int32x4_t)(__p0);
	38151	__ret = __builtin_bit_cast(int32x4_t, __p0);
38131	38152	return __ret;
38132	38153	}
38133	38154	__ai __attribute__((target("neon"))) int32x4_t vreinterpretq_s32_f32(float32x4_t __p0) {
38134	38155	int32x4_t __ret;
38135		__ret = (int32x4_t)(__p0);
	38156	__ret = __builtin_bit_cast(int32x4_t, __p0);
38136	38157	return __ret;
38137	38158	}
38138	38159	__ai __attribute__((target("neon"))) int32x4_t vreinterpretq_s32_f16(float16x8_t __p0) {
38139	38160	int32x4_t __ret;
38140		__ret = (int32x4_t)(__p0);
	38161	__ret = __builtin_bit_cast(int32x4_t, __p0);
38141	38162	return __ret;
38142	38163	}
38143	38164	__ai __attribute__((target("neon"))) int32x4_t vreinterpretq_s32_s64(int64x2_t __p0) {
38144	38165	int32x4_t __ret;
38145		__ret = (int32x4_t)(__p0);
	38166	__ret = __builtin_bit_cast(int32x4_t, __p0);
38146	38167	return __ret;
38147	38168	}
38148	38169	__ai __attribute__((target("neon"))) int32x4_t vreinterpretq_s32_s16(int16x8_t __p0) {
38149	38170	int32x4_t __ret;
38150		__ret = (int32x4_t)(__p0);
	38171	__ret = __builtin_bit_cast(int32x4_t, __p0);
38151	38172	return __ret;
38152	38173	}
38153	38174	__ai __attribute__((target("neon"))) int64x2_t vreinterpretq_s64_p8(poly8x16_t __p0) {
38154	38175	int64x2_t __ret;
38155		__ret = (int64x2_t)(__p0);
	38176	__ret = __builtin_bit_cast(int64x2_t, __p0);
38156	38177	return __ret;
38157	38178	}
38158	38179	__ai __attribute__((target("neon"))) int64x2_t vreinterpretq_s64_p16(poly16x8_t __p0) {
38159	38180	int64x2_t __ret;
38160		__ret = (int64x2_t)(__p0);
	38181	__ret = __builtin_bit_cast(int64x2_t, __p0);
38161	38182	return __ret;
38162	38183	}
38163	38184	__ai __attribute__((target("neon"))) int64x2_t vreinterpretq_s64_u8(uint8x16_t __p0) {
38164	38185	int64x2_t __ret;
38165		__ret = (int64x2_t)(__p0);
	38186	__ret = __builtin_bit_cast(int64x2_t, __p0);
38166	38187	return __ret;
38167	38188	}
38168	38189	__ai __attribute__((target("neon"))) int64x2_t vreinterpretq_s64_u32(uint32x4_t __p0) {
38169	38190	int64x2_t __ret;
38170		__ret = (int64x2_t)(__p0);
	38191	__ret = __builtin_bit_cast(int64x2_t, __p0);
38171	38192	return __ret;
38172	38193	}
38173	38194	__ai __attribute__((target("neon"))) int64x2_t vreinterpretq_s64_u64(uint64x2_t __p0) {
38174	38195	int64x2_t __ret;
38175		__ret = (int64x2_t)(__p0);
	38196	__ret = __builtin_bit_cast(int64x2_t, __p0);
38176	38197	return __ret;
38177	38198	}
38178	38199	__ai __attribute__((target("neon"))) int64x2_t vreinterpretq_s64_u16(uint16x8_t __p0) {
38179	38200	int64x2_t __ret;
38180		__ret = (int64x2_t)(__p0);
	38201	__ret = __builtin_bit_cast(int64x2_t, __p0);
38181	38202	return __ret;
38182	38203	}
38183	38204	__ai __attribute__((target("neon"))) int64x2_t vreinterpretq_s64_s8(int8x16_t __p0) {
38184	38205	int64x2_t __ret;
38185		__ret = (int64x2_t)(__p0);
	38206	__ret = __builtin_bit_cast(int64x2_t, __p0);
38186	38207	return __ret;
38187	38208	}
38188	38209	__ai __attribute__((target("neon"))) int64x2_t vreinterpretq_s64_f32(float32x4_t __p0) {
38189	38210	int64x2_t __ret;
38190		__ret = (int64x2_t)(__p0);
	38211	__ret = __builtin_bit_cast(int64x2_t, __p0);
38191	38212	return __ret;
38192	38213	}
38193	38214	__ai __attribute__((target("neon"))) int64x2_t vreinterpretq_s64_f16(float16x8_t __p0) {
38194	38215	int64x2_t __ret;
38195		__ret = (int64x2_t)(__p0);
	38216	__ret = __builtin_bit_cast(int64x2_t, __p0);
38196	38217	return __ret;
38197	38218	}
38198	38219	__ai __attribute__((target("neon"))) int64x2_t vreinterpretq_s64_s32(int32x4_t __p0) {
38199	38220	int64x2_t __ret;
38200		__ret = (int64x2_t)(__p0);
	38221	__ret = __builtin_bit_cast(int64x2_t, __p0);
38201	38222	return __ret;
38202	38223	}
38203	38224	__ai __attribute__((target("neon"))) int64x2_t vreinterpretq_s64_s16(int16x8_t __p0) {
38204	38225	int64x2_t __ret;
38205		__ret = (int64x2_t)(__p0);
	38226	__ret = __builtin_bit_cast(int64x2_t, __p0);
38206	38227	return __ret;
38207	38228	}
38208	38229	__ai __attribute__((target("neon"))) int16x8_t vreinterpretq_s16_p8(poly8x16_t __p0) {
38209	38230	int16x8_t __ret;
38210		__ret = (int16x8_t)(__p0);
	38231	__ret = __builtin_bit_cast(int16x8_t, __p0);
38211	38232	return __ret;
38212	38233	}
38213	38234	__ai __attribute__((target("neon"))) int16x8_t vreinterpretq_s16_p16(poly16x8_t __p0) {
38214	38235	int16x8_t __ret;
38215		__ret = (int16x8_t)(__p0);
	38236	__ret = __builtin_bit_cast(int16x8_t, __p0);
38216	38237	return __ret;
38217	38238	}
38218	38239	__ai __attribute__((target("neon"))) int16x8_t vreinterpretq_s16_u8(uint8x16_t __p0) {
38219	38240	int16x8_t __ret;
38220		__ret = (int16x8_t)(__p0);
	38241	__ret = __builtin_bit_cast(int16x8_t, __p0);
38221	38242	return __ret;
38222	38243	}
38223	38244	__ai __attribute__((target("neon"))) int16x8_t vreinterpretq_s16_u32(uint32x4_t __p0) {
38224	38245	int16x8_t __ret;
38225		__ret = (int16x8_t)(__p0);
	38246	__ret = __builtin_bit_cast(int16x8_t, __p0);
38226	38247	return __ret;
38227	38248	}
38228	38249	__ai __attribute__((target("neon"))) int16x8_t vreinterpretq_s16_u64(uint64x2_t __p0) {
38229	38250	int16x8_t __ret;
38230		__ret = (int16x8_t)(__p0);
	38251	__ret = __builtin_bit_cast(int16x8_t, __p0);
38231	38252	return __ret;
38232	38253	}
38233	38254	__ai __attribute__((target("neon"))) int16x8_t vreinterpretq_s16_u16(uint16x8_t __p0) {
38234	38255	int16x8_t __ret;
38235		__ret = (int16x8_t)(__p0);
	38256	__ret = __builtin_bit_cast(int16x8_t, __p0);
38236	38257	return __ret;
38237	38258	}
38238	38259	__ai __attribute__((target("neon"))) int16x8_t vreinterpretq_s16_s8(int8x16_t __p0) {
38239	38260	int16x8_t __ret;
38240		__ret = (int16x8_t)(__p0);
	38261	__ret = __builtin_bit_cast(int16x8_t, __p0);
38241	38262	return __ret;
38242	38263	}
38243	38264	__ai __attribute__((target("neon"))) int16x8_t vreinterpretq_s16_f32(float32x4_t __p0) {
38244	38265	int16x8_t __ret;
38245		__ret = (int16x8_t)(__p0);
	38266	__ret = __builtin_bit_cast(int16x8_t, __p0);
38246	38267	return __ret;
38247	38268	}
38248	38269	__ai __attribute__((target("neon"))) int16x8_t vreinterpretq_s16_f16(float16x8_t __p0) {
38249	38270	int16x8_t __ret;
38250		__ret = (int16x8_t)(__p0);
	38271	__ret = __builtin_bit_cast(int16x8_t, __p0);
38251	38272	return __ret;
38252	38273	}
38253	38274	__ai __attribute__((target("neon"))) int16x8_t vreinterpretq_s16_s32(int32x4_t __p0) {
38254	38275	int16x8_t __ret;
38255		__ret = (int16x8_t)(__p0);
	38276	__ret = __builtin_bit_cast(int16x8_t, __p0);
38256	38277	return __ret;
38257	38278	}
38258	38279	__ai __attribute__((target("neon"))) int16x8_t vreinterpretq_s16_s64(int64x2_t __p0) {
38259	38280	int16x8_t __ret;
38260		__ret = (int16x8_t)(__p0);
	38281	__ret = __builtin_bit_cast(int16x8_t, __p0);
38261	38282	return __ret;
38262	38283	}
38263	38284	__ai __attribute__((target("neon"))) uint8x8_t vreinterpret_u8_p8(poly8x8_t __p0) {
38264	38285	uint8x8_t __ret;
38265		__ret = (uint8x8_t)(__p0);
	38286	__ret = __builtin_bit_cast(uint8x8_t, __p0);
38266	38287	return __ret;
38267	38288	}
38268	38289	__ai __attribute__((target("neon"))) uint8x8_t vreinterpret_u8_p16(poly16x4_t __p0) {
38269	38290	uint8x8_t __ret;
38270		__ret = (uint8x8_t)(__p0);
	38291	__ret = __builtin_bit_cast(uint8x8_t, __p0);
38271	38292	return __ret;
38272	38293	}
38273	38294	__ai __attribute__((target("neon"))) uint8x8_t vreinterpret_u8_u32(uint32x2_t __p0) {
38274	38295	uint8x8_t __ret;
38275		__ret = (uint8x8_t)(__p0);
	38296	__ret = __builtin_bit_cast(uint8x8_t, __p0);
38276	38297	return __ret;
38277	38298	}
38278	38299	__ai __attribute__((target("neon"))) uint8x8_t vreinterpret_u8_u64(uint64x1_t __p0) {
38279	38300	uint8x8_t __ret;
38280		__ret = (uint8x8_t)(__p0);
	38301	__ret = __builtin_bit_cast(uint8x8_t, __p0);
38281	38302	return __ret;
38282	38303	}
38283	38304	__ai __attribute__((target("neon"))) uint8x8_t vreinterpret_u8_u16(uint16x4_t __p0) {
38284	38305	uint8x8_t __ret;
38285		__ret = (uint8x8_t)(__p0);
	38306	__ret = __builtin_bit_cast(uint8x8_t, __p0);
38286	38307	return __ret;
38287	38308	}
38288	38309	__ai __attribute__((target("neon"))) uint8x8_t vreinterpret_u8_s8(int8x8_t __p0) {
38289	38310	uint8x8_t __ret;
38290		__ret = (uint8x8_t)(__p0);
	38311	__ret = __builtin_bit_cast(uint8x8_t, __p0);
38291	38312	return __ret;
38292	38313	}
38293	38314	__ai __attribute__((target("neon"))) uint8x8_t vreinterpret_u8_f32(float32x2_t __p0) {
38294	38315	uint8x8_t __ret;
38295		__ret = (uint8x8_t)(__p0);
	38316	__ret = __builtin_bit_cast(uint8x8_t, __p0);
38296	38317	return __ret;
38297	38318	}
38298	38319	__ai __attribute__((target("neon"))) uint8x8_t vreinterpret_u8_f16(float16x4_t __p0) {
38299	38320	uint8x8_t __ret;
38300		__ret = (uint8x8_t)(__p0);
	38321	__ret = __builtin_bit_cast(uint8x8_t, __p0);
38301	38322	return __ret;
38302	38323	}
38303	38324	__ai __attribute__((target("neon"))) uint8x8_t vreinterpret_u8_s32(int32x2_t __p0) {
38304	38325	uint8x8_t __ret;
38305		__ret = (uint8x8_t)(__p0);
	38326	__ret = __builtin_bit_cast(uint8x8_t, __p0);
38306	38327	return __ret;
38307	38328	}
38308	38329	__ai __attribute__((target("neon"))) uint8x8_t vreinterpret_u8_s64(int64x1_t __p0) {
38309	38330	uint8x8_t __ret;
38310		__ret = (uint8x8_t)(__p0);
	38331	__ret = __builtin_bit_cast(uint8x8_t, __p0);
38311	38332	return __ret;
38312	38333	}
38313	38334	__ai __attribute__((target("neon"))) uint8x8_t vreinterpret_u8_s16(int16x4_t __p0) {
38314	38335	uint8x8_t __ret;
38315		__ret = (uint8x8_t)(__p0);
	38336	__ret = __builtin_bit_cast(uint8x8_t, __p0);
38316	38337	return __ret;
38317	38338	}
38318	38339	__ai __attribute__((target("neon"))) uint32x2_t vreinterpret_u32_p8(poly8x8_t __p0) {
38319	38340	uint32x2_t __ret;
38320		__ret = (uint32x2_t)(__p0);
	38341	__ret = __builtin_bit_cast(uint32x2_t, __p0);
38321	38342	return __ret;
38322	38343	}
38323	38344	__ai __attribute__((target("neon"))) uint32x2_t vreinterpret_u32_p16(poly16x4_t __p0) {
38324	38345	uint32x2_t __ret;
38325		__ret = (uint32x2_t)(__p0);
	38346	__ret = __builtin_bit_cast(uint32x2_t, __p0);
38326	38347	return __ret;
38327	38348	}
38328	38349	__ai __attribute__((target("neon"))) uint32x2_t vreinterpret_u32_u8(uint8x8_t __p0) {
38329	38350	uint32x2_t __ret;
38330		__ret = (uint32x2_t)(__p0);
	38351	__ret = __builtin_bit_cast(uint32x2_t, __p0);
38331	38352	return __ret;
38332	38353	}
38333	38354	__ai __attribute__((target("neon"))) uint32x2_t vreinterpret_u32_u64(uint64x1_t __p0) {
38334	38355	uint32x2_t __ret;
38335		__ret = (uint32x2_t)(__p0);
	38356	__ret = __builtin_bit_cast(uint32x2_t, __p0);
38336	38357	return __ret;
38337	38358	}
38338	38359	__ai __attribute__((target("neon"))) uint32x2_t vreinterpret_u32_u16(uint16x4_t __p0) {
38339	38360	uint32x2_t __ret;
38340		__ret = (uint32x2_t)(__p0);
	38361	__ret = __builtin_bit_cast(uint32x2_t, __p0);
38341	38362	return __ret;
38342	38363	}
38343	38364	__ai __attribute__((target("neon"))) uint32x2_t vreinterpret_u32_s8(int8x8_t __p0) {
38344	38365	uint32x2_t __ret;
38345		__ret = (uint32x2_t)(__p0);
	38366	__ret = __builtin_bit_cast(uint32x2_t, __p0);
38346	38367	return __ret;
38347	38368	}
38348	38369	__ai __attribute__((target("neon"))) uint32x2_t vreinterpret_u32_f32(float32x2_t __p0) {
38349	38370	uint32x2_t __ret;
38350		__ret = (uint32x2_t)(__p0);
	38371	__ret = __builtin_bit_cast(uint32x2_t, __p0);
38351	38372	return __ret;
38352	38373	}
38353	38374	__ai __attribute__((target("neon"))) uint32x2_t vreinterpret_u32_f16(float16x4_t __p0) {
38354	38375	uint32x2_t __ret;
38355		__ret = (uint32x2_t)(__p0);
	38376	__ret = __builtin_bit_cast(uint32x2_t, __p0);
38356	38377	return __ret;
38357	38378	}
38358	38379	__ai __attribute__((target("neon"))) uint32x2_t vreinterpret_u32_s32(int32x2_t __p0) {
38359	38380	uint32x2_t __ret;
38360		__ret = (uint32x2_t)(__p0);
	38381	__ret = __builtin_bit_cast(uint32x2_t, __p0);
38361	38382	return __ret;
38362	38383	}
38363	38384	__ai __attribute__((target("neon"))) uint32x2_t vreinterpret_u32_s64(int64x1_t __p0) {
38364	38385	uint32x2_t __ret;
38365		__ret = (uint32x2_t)(__p0);
	38386	__ret = __builtin_bit_cast(uint32x2_t, __p0);
38366	38387	return __ret;
38367	38388	}
38368	38389	__ai __attribute__((target("neon"))) uint32x2_t vreinterpret_u32_s16(int16x4_t __p0) {
38369	38390	uint32x2_t __ret;
38370		__ret = (uint32x2_t)(__p0);
	38391	__ret = __builtin_bit_cast(uint32x2_t, __p0);
38371	38392	return __ret;
38372	38393	}
38373	38394	__ai __attribute__((target("neon"))) uint64x1_t vreinterpret_u64_p8(poly8x8_t __p0) {
38374	38395	uint64x1_t __ret;
38375		__ret = (uint64x1_t)(__p0);
	38396	__ret = __builtin_bit_cast(uint64x1_t, __p0);
38376	38397	return __ret;
38377	38398	}
38378	38399	__ai __attribute__((target("neon"))) uint64x1_t vreinterpret_u64_p16(poly16x4_t __p0) {
38379	38400	uint64x1_t __ret;
38380		__ret = (uint64x1_t)(__p0);
	38401	__ret = __builtin_bit_cast(uint64x1_t, __p0);
38381	38402	return __ret;
38382	38403	}
38383	38404	__ai __attribute__((target("neon"))) uint64x1_t vreinterpret_u64_u8(uint8x8_t __p0) {
38384	38405	uint64x1_t __ret;
38385		__ret = (uint64x1_t)(__p0);
	38406	__ret = __builtin_bit_cast(uint64x1_t, __p0);
38386	38407	return __ret;
38387	38408	}
38388	38409	__ai __attribute__((target("neon"))) uint64x1_t vreinterpret_u64_u32(uint32x2_t __p0) {
38389	38410	uint64x1_t __ret;
38390		__ret = (uint64x1_t)(__p0);
	38411	__ret = __builtin_bit_cast(uint64x1_t, __p0);
38391	38412	return __ret;
38392	38413	}
38393	38414	__ai __attribute__((target("neon"))) uint64x1_t vreinterpret_u64_u16(uint16x4_t __p0) {
38394	38415	uint64x1_t __ret;
38395		__ret = (uint64x1_t)(__p0);
	38416	__ret = __builtin_bit_cast(uint64x1_t, __p0);
38396	38417	return __ret;
38397	38418	}
38398	38419	__ai __attribute__((target("neon"))) uint64x1_t vreinterpret_u64_s8(int8x8_t __p0) {
38399	38420	uint64x1_t __ret;
38400		__ret = (uint64x1_t)(__p0);
	38421	__ret = __builtin_bit_cast(uint64x1_t, __p0);
38401	38422	return __ret;
38402	38423	}
38403	38424	__ai __attribute__((target("neon"))) uint64x1_t vreinterpret_u64_f32(float32x2_t __p0) {
38404	38425	uint64x1_t __ret;
38405		__ret = (uint64x1_t)(__p0);
	38426	__ret = __builtin_bit_cast(uint64x1_t, __p0);
38406	38427	return __ret;
38407	38428	}
38408	38429	__ai __attribute__((target("neon"))) uint64x1_t vreinterpret_u64_f16(float16x4_t __p0) {
38409	38430	uint64x1_t __ret;
38410		__ret = (uint64x1_t)(__p0);
	38431	__ret = __builtin_bit_cast(uint64x1_t, __p0);
38411	38432	return __ret;
38412	38433	}
38413	38434	__ai __attribute__((target("neon"))) uint64x1_t vreinterpret_u64_s32(int32x2_t __p0) {
38414	38435	uint64x1_t __ret;
38415		__ret = (uint64x1_t)(__p0);
	38436	__ret = __builtin_bit_cast(uint64x1_t, __p0);
38416	38437	return __ret;
38417	38438	}
38418	38439	__ai __attribute__((target("neon"))) uint64x1_t vreinterpret_u64_s64(int64x1_t __p0) {
38419	38440	uint64x1_t __ret;
38420		__ret = (uint64x1_t)(__p0);
	38441	__ret = __builtin_bit_cast(uint64x1_t, __p0);
38421	38442	return __ret;
38422	38443	}
38423	38444	__ai __attribute__((target("neon"))) uint64x1_t vreinterpret_u64_s16(int16x4_t __p0) {
38424	38445	uint64x1_t __ret;
38425		__ret = (uint64x1_t)(__p0);
	38446	__ret = __builtin_bit_cast(uint64x1_t, __p0);
38426	38447	return __ret;
38427	38448	}
38428	38449	__ai __attribute__((target("neon"))) uint16x4_t vreinterpret_u16_p8(poly8x8_t __p0) {
38429	38450	uint16x4_t __ret;
38430		__ret = (uint16x4_t)(__p0);
	38451	__ret = __builtin_bit_cast(uint16x4_t, __p0);
38431	38452	return __ret;
38432	38453	}
38433	38454	__ai __attribute__((target("neon"))) uint16x4_t vreinterpret_u16_p16(poly16x4_t __p0) {
38434	38455	uint16x4_t __ret;
38435		__ret = (uint16x4_t)(__p0);
	38456	__ret = __builtin_bit_cast(uint16x4_t, __p0);
38436	38457	return __ret;
38437	38458	}
38438	38459	__ai __attribute__((target("neon"))) uint16x4_t vreinterpret_u16_u8(uint8x8_t __p0) {
38439	38460	uint16x4_t __ret;
38440		__ret = (uint16x4_t)(__p0);
	38461	__ret = __builtin_bit_cast(uint16x4_t, __p0);
38441	38462	return __ret;
38442	38463	}
38443	38464	__ai __attribute__((target("neon"))) uint16x4_t vreinterpret_u16_u32(uint32x2_t __p0) {
38444	38465	uint16x4_t __ret;
38445		__ret = (uint16x4_t)(__p0);
	38466	__ret = __builtin_bit_cast(uint16x4_t, __p0);
38446	38467	return __ret;
38447	38468	}
38448	38469	__ai __attribute__((target("neon"))) uint16x4_t vreinterpret_u16_u64(uint64x1_t __p0) {
38449	38470	uint16x4_t __ret;
38450		__ret = (uint16x4_t)(__p0);
	38471	__ret = __builtin_bit_cast(uint16x4_t, __p0);
38451	38472	return __ret;
38452	38473	}
38453	38474	__ai __attribute__((target("neon"))) uint16x4_t vreinterpret_u16_s8(int8x8_t __p0) {
38454	38475	uint16x4_t __ret;
38455		__ret = (uint16x4_t)(__p0);
	38476	__ret = __builtin_bit_cast(uint16x4_t, __p0);
38456	38477	return __ret;
38457	38478	}
38458	38479	__ai __attribute__((target("neon"))) uint16x4_t vreinterpret_u16_f32(float32x2_t __p0) {
38459	38480	uint16x4_t __ret;
38460		__ret = (uint16x4_t)(__p0);
	38481	__ret = __builtin_bit_cast(uint16x4_t, __p0);
38461	38482	return __ret;
38462	38483	}
38463	38484	__ai __attribute__((target("neon"))) uint16x4_t vreinterpret_u16_f16(float16x4_t __p0) {
38464	38485	uint16x4_t __ret;
38465		__ret = (uint16x4_t)(__p0);
	38486	__ret = __builtin_bit_cast(uint16x4_t, __p0);
38466	38487	return __ret;
38467	38488	}
38468	38489	__ai __attribute__((target("neon"))) uint16x4_t vreinterpret_u16_s32(int32x2_t __p0) {
38469	38490	uint16x4_t __ret;
38470		__ret = (uint16x4_t)(__p0);
	38491	__ret = __builtin_bit_cast(uint16x4_t, __p0);
38471	38492	return __ret;
38472	38493	}
38473	38494	__ai __attribute__((target("neon"))) uint16x4_t vreinterpret_u16_s64(int64x1_t __p0) {
38474	38495	uint16x4_t __ret;
38475		__ret = (uint16x4_t)(__p0);
	38496	__ret = __builtin_bit_cast(uint16x4_t, __p0);
38476	38497	return __ret;
38477	38498	}
38478	38499	__ai __attribute__((target("neon"))) uint16x4_t vreinterpret_u16_s16(int16x4_t __p0) {
38479	38500	uint16x4_t __ret;
38480		__ret = (uint16x4_t)(__p0);
	38501	__ret = __builtin_bit_cast(uint16x4_t, __p0);
38481	38502	return __ret;
38482	38503	}
38483	38504	__ai __attribute__((target("neon"))) int8x8_t vreinterpret_s8_p8(poly8x8_t __p0) {
38484	38505	int8x8_t __ret;
38485		__ret = (int8x8_t)(__p0);
	38506	__ret = __builtin_bit_cast(int8x8_t, __p0);
38486	38507	return __ret;
38487	38508	}
38488	38509	__ai __attribute__((target("neon"))) int8x8_t vreinterpret_s8_p16(poly16x4_t __p0) {
38489	38510	int8x8_t __ret;
38490		__ret = (int8x8_t)(__p0);
	38511	__ret = __builtin_bit_cast(int8x8_t, __p0);
38491	38512	return __ret;
38492	38513	}
38493	38514	__ai __attribute__((target("neon"))) int8x8_t vreinterpret_s8_u8(uint8x8_t __p0) {
38494	38515	int8x8_t __ret;
38495		__ret = (int8x8_t)(__p0);
	38516	__ret = __builtin_bit_cast(int8x8_t, __p0);
38496	38517	return __ret;
38497	38518	}
38498	38519	__ai __attribute__((target("neon"))) int8x8_t vreinterpret_s8_u32(uint32x2_t __p0) {
38499	38520	int8x8_t __ret;
38500		__ret = (int8x8_t)(__p0);
	38521	__ret = __builtin_bit_cast(int8x8_t, __p0);
38501	38522	return __ret;
38502	38523	}
38503	38524	__ai __attribute__((target("neon"))) int8x8_t vreinterpret_s8_u64(uint64x1_t __p0) {
38504	38525	int8x8_t __ret;
38505		__ret = (int8x8_t)(__p0);
	38526	__ret = __builtin_bit_cast(int8x8_t, __p0);
38506	38527	return __ret;
38507	38528	}
38508	38529	__ai __attribute__((target("neon"))) int8x8_t vreinterpret_s8_u16(uint16x4_t __p0) {
38509	38530	int8x8_t __ret;
38510		__ret = (int8x8_t)(__p0);
	38531	__ret = __builtin_bit_cast(int8x8_t, __p0);
38511	38532	return __ret;
38512	38533	}
38513	38534	__ai __attribute__((target("neon"))) int8x8_t vreinterpret_s8_f32(float32x2_t __p0) {
38514	38535	int8x8_t __ret;
38515		__ret = (int8x8_t)(__p0);
	38536	__ret = __builtin_bit_cast(int8x8_t, __p0);
38516	38537	return __ret;
38517	38538	}
38518	38539	__ai __attribute__((target("neon"))) int8x8_t vreinterpret_s8_f16(float16x4_t __p0) {
38519	38540	int8x8_t __ret;
38520		__ret = (int8x8_t)(__p0);
	38541	__ret = __builtin_bit_cast(int8x8_t, __p0);
38521	38542	return __ret;
38522	38543	}
38523	38544	__ai __attribute__((target("neon"))) int8x8_t vreinterpret_s8_s32(int32x2_t __p0) {
38524	38545	int8x8_t __ret;
38525		__ret = (int8x8_t)(__p0);
	38546	__ret = __builtin_bit_cast(int8x8_t, __p0);
38526	38547	return __ret;
38527	38548	}
38528	38549	__ai __attribute__((target("neon"))) int8x8_t vreinterpret_s8_s64(int64x1_t __p0) {
38529	38550	int8x8_t __ret;
38530		__ret = (int8x8_t)(__p0);
	38551	__ret = __builtin_bit_cast(int8x8_t, __p0);
38531	38552	return __ret;
38532	38553	}
38533	38554	__ai __attribute__((target("neon"))) int8x8_t vreinterpret_s8_s16(int16x4_t __p0) {
38534	38555	int8x8_t __ret;
38535		__ret = (int8x8_t)(__p0);
	38556	__ret = __builtin_bit_cast(int8x8_t, __p0);
38536	38557	return __ret;
38537	38558	}
38538	38559	__ai __attribute__((target("neon"))) float32x2_t vreinterpret_f32_p8(poly8x8_t __p0) {
38539	38560	float32x2_t __ret;
38540		__ret = (float32x2_t)(__p0);
	38561	__ret = __builtin_bit_cast(float32x2_t, __p0);
38541	38562	return __ret;
38542	38563	}
38543	38564	__ai __attribute__((target("neon"))) float32x2_t vreinterpret_f32_p16(poly16x4_t __p0) {
38544	38565	float32x2_t __ret;
38545		__ret = (float32x2_t)(__p0);
	38566	__ret = __builtin_bit_cast(float32x2_t, __p0);
38546	38567	return __ret;
38547	38568	}
38548	38569	__ai __attribute__((target("neon"))) float32x2_t vreinterpret_f32_u8(uint8x8_t __p0) {
38549	38570	float32x2_t __ret;
38550		__ret = (float32x2_t)(__p0);
	38571	__ret = __builtin_bit_cast(float32x2_t, __p0);
38551	38572	return __ret;
38552	38573	}
38553	38574	__ai __attribute__((target("neon"))) float32x2_t vreinterpret_f32_u32(uint32x2_t __p0) {
38554	38575	float32x2_t __ret;
38555		__ret = (float32x2_t)(__p0);
	38576	__ret = __builtin_bit_cast(float32x2_t, __p0);
38556	38577	return __ret;
38557	38578	}
38558	38579	__ai __attribute__((target("neon"))) float32x2_t vreinterpret_f32_u64(uint64x1_t __p0) {
38559	38580	float32x2_t __ret;
38560		__ret = (float32x2_t)(__p0);
	38581	__ret = __builtin_bit_cast(float32x2_t, __p0);
38561	38582	return __ret;
38562	38583	}
38563	38584	__ai __attribute__((target("neon"))) float32x2_t vreinterpret_f32_u16(uint16x4_t __p0) {
38564	38585	float32x2_t __ret;
38565		__ret = (float32x2_t)(__p0);
	38586	__ret = __builtin_bit_cast(float32x2_t, __p0);
38566	38587	return __ret;
38567	38588	}
38568	38589	__ai __attribute__((target("neon"))) float32x2_t vreinterpret_f32_s8(int8x8_t __p0) {
38569	38590	float32x2_t __ret;
38570		__ret = (float32x2_t)(__p0);
	38591	__ret = __builtin_bit_cast(float32x2_t, __p0);
38571	38592	return __ret;
38572	38593	}
38573	38594	__ai __attribute__((target("neon"))) float32x2_t vreinterpret_f32_f16(float16x4_t __p0) {
38574	38595	float32x2_t __ret;
38575		__ret = (float32x2_t)(__p0);
	38596	__ret = __builtin_bit_cast(float32x2_t, __p0);
38576	38597	return __ret;
38577	38598	}
38578	38599	__ai __attribute__((target("neon"))) float32x2_t vreinterpret_f32_s32(int32x2_t __p0) {
38579	38600	float32x2_t __ret;
38580		__ret = (float32x2_t)(__p0);
	38601	__ret = __builtin_bit_cast(float32x2_t, __p0);
38581	38602	return __ret;
38582	38603	}
38583	38604	__ai __attribute__((target("neon"))) float32x2_t vreinterpret_f32_s64(int64x1_t __p0) {
38584	38605	float32x2_t __ret;
38585		__ret = (float32x2_t)(__p0);
	38606	__ret = __builtin_bit_cast(float32x2_t, __p0);
38586	38607	return __ret;
38587	38608	}
38588	38609	__ai __attribute__((target("neon"))) float32x2_t vreinterpret_f32_s16(int16x4_t __p0) {
38589	38610	float32x2_t __ret;
38590		__ret = (float32x2_t)(__p0);
	38611	__ret = __builtin_bit_cast(float32x2_t, __p0);
38591	38612	return __ret;
38592	38613	}
38593	38614	__ai __attribute__((target("neon"))) float16x4_t vreinterpret_f16_p8(poly8x8_t __p0) {
38594	38615	float16x4_t __ret;
38595		__ret = (float16x4_t)(__p0);
	38616	__ret = __builtin_bit_cast(float16x4_t, __p0);
38596	38617	return __ret;
38597	38618	}
38598	38619	__ai __attribute__((target("neon"))) float16x4_t vreinterpret_f16_p16(poly16x4_t __p0) {
38599	38620	float16x4_t __ret;
38600		__ret = (float16x4_t)(__p0);
	38621	__ret = __builtin_bit_cast(float16x4_t, __p0);
38601	38622	return __ret;
38602	38623	}
38603	38624	__ai __attribute__((target("neon"))) float16x4_t vreinterpret_f16_u8(uint8x8_t __p0) {
38604	38625	float16x4_t __ret;
38605		__ret = (float16x4_t)(__p0);
	38626	__ret = __builtin_bit_cast(float16x4_t, __p0);
38606	38627	return __ret;
38607	38628	}
38608	38629	__ai __attribute__((target("neon"))) float16x4_t vreinterpret_f16_u32(uint32x2_t __p0) {
38609	38630	float16x4_t __ret;
38610		__ret = (float16x4_t)(__p0);
	38631	__ret = __builtin_bit_cast(float16x4_t, __p0);
38611	38632	return __ret;
38612	38633	}
38613	38634	__ai __attribute__((target("neon"))) float16x4_t vreinterpret_f16_u64(uint64x1_t __p0) {
38614	38635	float16x4_t __ret;
38615		__ret = (float16x4_t)(__p0);
	38636	__ret = __builtin_bit_cast(float16x4_t, __p0);
38616	38637	return __ret;
38617	38638	}
38618	38639	__ai __attribute__((target("neon"))) float16x4_t vreinterpret_f16_u16(uint16x4_t __p0) {
38619	38640	float16x4_t __ret;
38620		__ret = (float16x4_t)(__p0);
	38641	__ret = __builtin_bit_cast(float16x4_t, __p0);
38621	38642	return __ret;
38622	38643	}
38623	38644	__ai __attribute__((target("neon"))) float16x4_t vreinterpret_f16_s8(int8x8_t __p0) {
38624	38645	float16x4_t __ret;
38625		__ret = (float16x4_t)(__p0);
	38646	__ret = __builtin_bit_cast(float16x4_t, __p0);
38626	38647	return __ret;
38627	38648	}
38628	38649	__ai __attribute__((target("neon"))) float16x4_t vreinterpret_f16_f32(float32x2_t __p0) {
38629	38650	float16x4_t __ret;
38630		__ret = (float16x4_t)(__p0);
	38651	__ret = __builtin_bit_cast(float16x4_t, __p0);
38631	38652	return __ret;
38632	38653	}
38633	38654	__ai __attribute__((target("neon"))) float16x4_t vreinterpret_f16_s32(int32x2_t __p0) {
38634	38655	float16x4_t __ret;
38635		__ret = (float16x4_t)(__p0);
	38656	__ret = __builtin_bit_cast(float16x4_t, __p0);
38636	38657	return __ret;
38637	38658	}
38638	38659	__ai __attribute__((target("neon"))) float16x4_t vreinterpret_f16_s64(int64x1_t __p0) {
38639	38660	float16x4_t __ret;
38640		__ret = (float16x4_t)(__p0);
	38661	__ret = __builtin_bit_cast(float16x4_t, __p0);
38641	38662	return __ret;
38642	38663	}
38643	38664	__ai __attribute__((target("neon"))) float16x4_t vreinterpret_f16_s16(int16x4_t __p0) {
38644	38665	float16x4_t __ret;
38645		__ret = (float16x4_t)(__p0);
	38666	__ret = __builtin_bit_cast(float16x4_t, __p0);
38646	38667	return __ret;
38647	38668	}
38648	38669	__ai __attribute__((target("neon"))) int32x2_t vreinterpret_s32_p8(poly8x8_t __p0) {
38649	38670	int32x2_t __ret;
38650		__ret = (int32x2_t)(__p0);
	38671	__ret = __builtin_bit_cast(int32x2_t, __p0);
38651	38672	return __ret;
38652	38673	}
38653	38674	__ai __attribute__((target("neon"))) int32x2_t vreinterpret_s32_p16(poly16x4_t __p0) {
38654	38675	int32x2_t __ret;
38655		__ret = (int32x2_t)(__p0);
	38676	__ret = __builtin_bit_cast(int32x2_t, __p0);
38656	38677	return __ret;
38657	38678	}
38658	38679	__ai __attribute__((target("neon"))) int32x2_t vreinterpret_s32_u8(uint8x8_t __p0) {
38659	38680	int32x2_t __ret;
38660		__ret = (int32x2_t)(__p0);
	38681	__ret = __builtin_bit_cast(int32x2_t, __p0);
38661	38682	return __ret;
38662	38683	}
38663	38684	__ai __attribute__((target("neon"))) int32x2_t vreinterpret_s32_u32(uint32x2_t __p0) {
38664	38685	int32x2_t __ret;
38665		__ret = (int32x2_t)(__p0);
	38686	__ret = __builtin_bit_cast(int32x2_t, __p0);
38666	38687	return __ret;
38667	38688	}
38668	38689	__ai __attribute__((target("neon"))) int32x2_t vreinterpret_s32_u64(uint64x1_t __p0) {
38669	38690	int32x2_t __ret;
38670		__ret = (int32x2_t)(__p0);
	38691	__ret = __builtin_bit_cast(int32x2_t, __p0);
38671	38692	return __ret;
38672	38693	}
38673	38694	__ai __attribute__((target("neon"))) int32x2_t vreinterpret_s32_u16(uint16x4_t __p0) {
38674	38695	int32x2_t __ret;
38675		__ret = (int32x2_t)(__p0);
	38696	__ret = __builtin_bit_cast(int32x2_t, __p0);
38676	38697	return __ret;
38677	38698	}
38678	38699	__ai __attribute__((target("neon"))) int32x2_t vreinterpret_s32_s8(int8x8_t __p0) {
38679	38700	int32x2_t __ret;
38680		__ret = (int32x2_t)(__p0);
	38701	__ret = __builtin_bit_cast(int32x2_t, __p0);
38681	38702	return __ret;
38682	38703	}
38683	38704	__ai __attribute__((target("neon"))) int32x2_t vreinterpret_s32_f32(float32x2_t __p0) {
38684	38705	int32x2_t __ret;
38685		__ret = (int32x2_t)(__p0);
	38706	__ret = __builtin_bit_cast(int32x2_t, __p0);
38686	38707	return __ret;
38687	38708	}
38688	38709	__ai __attribute__((target("neon"))) int32x2_t vreinterpret_s32_f16(float16x4_t __p0) {
38689	38710	int32x2_t __ret;
38690		__ret = (int32x2_t)(__p0);
	38711	__ret = __builtin_bit_cast(int32x2_t, __p0);
38691	38712	return __ret;
38692	38713	}
38693	38714	__ai __attribute__((target("neon"))) int32x2_t vreinterpret_s32_s64(int64x1_t __p0) {
38694	38715	int32x2_t __ret;
38695		__ret = (int32x2_t)(__p0);
	38716	__ret = __builtin_bit_cast(int32x2_t, __p0);
38696	38717	return __ret;
38697	38718	}
38698	38719	__ai __attribute__((target("neon"))) int32x2_t vreinterpret_s32_s16(int16x4_t __p0) {
38699	38720	int32x2_t __ret;
38700		__ret = (int32x2_t)(__p0);
	38721	__ret = __builtin_bit_cast(int32x2_t, __p0);
38701	38722	return __ret;
38702	38723	}
38703	38724	__ai __attribute__((target("neon"))) int64x1_t vreinterpret_s64_p8(poly8x8_t __p0) {
38704	38725	int64x1_t __ret;
38705		__ret = (int64x1_t)(__p0);
	38726	__ret = __builtin_bit_cast(int64x1_t, __p0);
38706	38727	return __ret;
38707	38728	}
38708	38729	__ai __attribute__((target("neon"))) int64x1_t vreinterpret_s64_p16(poly16x4_t __p0) {
38709	38730	int64x1_t __ret;
38710		__ret = (int64x1_t)(__p0);
	38731	__ret = __builtin_bit_cast(int64x1_t, __p0);
38711	38732	return __ret;
38712	38733	}
38713	38734	__ai __attribute__((target("neon"))) int64x1_t vreinterpret_s64_u8(uint8x8_t __p0) {
38714	38735	int64x1_t __ret;
38715		__ret = (int64x1_t)(__p0);
	38736	__ret = __builtin_bit_cast(int64x1_t, __p0);
38716	38737	return __ret;
38717	38738	}
38718	38739	__ai __attribute__((target("neon"))) int64x1_t vreinterpret_s64_u32(uint32x2_t __p0) {
38719	38740	int64x1_t __ret;
38720		__ret = (int64x1_t)(__p0);
	38741	__ret = __builtin_bit_cast(int64x1_t, __p0);
38721	38742	return __ret;
38722	38743	}
38723	38744	__ai __attribute__((target("neon"))) int64x1_t vreinterpret_s64_u64(uint64x1_t __p0) {
38724	38745	int64x1_t __ret;
38725		__ret = (int64x1_t)(__p0);
	38746	__ret = __builtin_bit_cast(int64x1_t, __p0);
38726	38747	return __ret;
38727	38748	}
38728	38749	__ai __attribute__((target("neon"))) int64x1_t vreinterpret_s64_u16(uint16x4_t __p0) {
38729	38750	int64x1_t __ret;
38730		__ret = (int64x1_t)(__p0);
	38751	__ret = __builtin_bit_cast(int64x1_t, __p0);
38731	38752	return __ret;
38732	38753	}
38733	38754	__ai __attribute__((target("neon"))) int64x1_t vreinterpret_s64_s8(int8x8_t __p0) {
38734	38755	int64x1_t __ret;
38735		__ret = (int64x1_t)(__p0);
	38756	__ret = __builtin_bit_cast(int64x1_t, __p0);
38736	38757	return __ret;
38737	38758	}
38738	38759	__ai __attribute__((target("neon"))) int64x1_t vreinterpret_s64_f32(float32x2_t __p0) {
38739	38760	int64x1_t __ret;
38740		__ret = (int64x1_t)(__p0);
	38761	__ret = __builtin_bit_cast(int64x1_t, __p0);
38741	38762	return __ret;
38742	38763	}
38743	38764	__ai __attribute__((target("neon"))) int64x1_t vreinterpret_s64_f16(float16x4_t __p0) {
38744	38765	int64x1_t __ret;
38745		__ret = (int64x1_t)(__p0);
	38766	__ret = __builtin_bit_cast(int64x1_t, __p0);
38746	38767	return __ret;
38747	38768	}
38748	38769	__ai __attribute__((target("neon"))) int64x1_t vreinterpret_s64_s32(int32x2_t __p0) {
38749	38770	int64x1_t __ret;
38750		__ret = (int64x1_t)(__p0);
	38771	__ret = __builtin_bit_cast(int64x1_t, __p0);
38751	38772	return __ret;
38752	38773	}
38753	38774	__ai __attribute__((target("neon"))) int64x1_t vreinterpret_s64_s16(int16x4_t __p0) {
38754	38775	int64x1_t __ret;
38755		__ret = (int64x1_t)(__p0);
	38776	__ret = __builtin_bit_cast(int64x1_t, __p0);
38756	38777	return __ret;
38757	38778	}
38758	38779	__ai __attribute__((target("neon"))) int16x4_t vreinterpret_s16_p8(poly8x8_t __p0) {
38759	38780	int16x4_t __ret;
38760		__ret = (int16x4_t)(__p0);
	38781	__ret = __builtin_bit_cast(int16x4_t, __p0);
38761	38782	return __ret;
38762	38783	}
38763	38784	__ai __attribute__((target("neon"))) int16x4_t vreinterpret_s16_p16(poly16x4_t __p0) {
38764	38785	int16x4_t __ret;
38765		__ret = (int16x4_t)(__p0);
	38786	__ret = __builtin_bit_cast(int16x4_t, __p0);
38766	38787	return __ret;
38767	38788	}
38768	38789	__ai __attribute__((target("neon"))) int16x4_t vreinterpret_s16_u8(uint8x8_t __p0) {
38769	38790	int16x4_t __ret;
38770		__ret = (int16x4_t)(__p0);
	38791	__ret = __builtin_bit_cast(int16x4_t, __p0);
38771	38792	return __ret;
38772	38793	}
38773	38794	__ai __attribute__((target("neon"))) int16x4_t vreinterpret_s16_u32(uint32x2_t __p0) {
38774	38795	int16x4_t __ret;
38775		__ret = (int16x4_t)(__p0);
	38796	__ret = __builtin_bit_cast(int16x4_t, __p0);
38776	38797	return __ret;
38777	38798	}
38778	38799	__ai __attribute__((target("neon"))) int16x4_t vreinterpret_s16_u64(uint64x1_t __p0) {
38779	38800	int16x4_t __ret;
38780		__ret = (int16x4_t)(__p0);
	38801	__ret = __builtin_bit_cast(int16x4_t, __p0);
38781	38802	return __ret;
38782	38803	}
38783	38804	__ai __attribute__((target("neon"))) int16x4_t vreinterpret_s16_u16(uint16x4_t __p0) {
38784	38805	int16x4_t __ret;
38785		__ret = (int16x4_t)(__p0);
	38806	__ret = __builtin_bit_cast(int16x4_t, __p0);
38786	38807	return __ret;
38787	38808	}
38788	38809	__ai __attribute__((target("neon"))) int16x4_t vreinterpret_s16_s8(int8x8_t __p0) {
38789	38810	int16x4_t __ret;
38790		__ret = (int16x4_t)(__p0);
	38811	__ret = __builtin_bit_cast(int16x4_t, __p0);
38791	38812	return __ret;
38792	38813	}
38793	38814	__ai __attribute__((target("neon"))) int16x4_t vreinterpret_s16_f32(float32x2_t __p0) {
38794	38815	int16x4_t __ret;
38795		__ret = (int16x4_t)(__p0);
	38816	__ret = __builtin_bit_cast(int16x4_t, __p0);
38796	38817	return __ret;
38797	38818	}
38798	38819	__ai __attribute__((target("neon"))) int16x4_t vreinterpret_s16_f16(float16x4_t __p0) {
38799	38820	int16x4_t __ret;
38800		__ret = (int16x4_t)(__p0);
	38821	__ret = __builtin_bit_cast(int16x4_t, __p0);
38801	38822	return __ret;
38802	38823	}
38803	38824	__ai __attribute__((target("neon"))) int16x4_t vreinterpret_s16_s32(int32x2_t __p0) {
38804	38825	int16x4_t __ret;
38805		__ret = (int16x4_t)(__p0);
	38826	__ret = __builtin_bit_cast(int16x4_t, __p0);
38806	38827	return __ret;
38807	38828	}
38808	38829	__ai __attribute__((target("neon"))) int16x4_t vreinterpret_s16_s64(int64x1_t __p0) {
38809	38830	int16x4_t __ret;
38810		__ret = (int16x4_t)(__p0);
	38831	__ret = __builtin_bit_cast(int16x4_t, __p0);
38811	38832	return __ret;
38812	38833	}
38813	38834	#endif
...	...	@@ -38815,20 +38836,20 @@ __ai __attribute__((target("neon"))) int16x4_t vreinterpret_s16_s64(int64x1_t __
38815	38836	#ifdef __LITTLE_ENDIAN__
38816	38837	__ai __attribute__((target("neon"))) float16x4_t vcvt_f16_f32(float32x4_t __p0) {
38817	38838	float16x4_t __ret;
38818		__ret = (float16x4_t) __builtin_neon_vcvt_f16_f32((int8x16_t)__p0, 41);
	38839	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vcvt_f16_f32(__builtin_bit_cast(int8x16_t, __p0), 41));
38819	38840	return __ret;
38820	38841	}
38821	38842	#else
38822	38843	__ai __attribute__((target("neon"))) float16x4_t vcvt_f16_f32(float32x4_t __p0) {
38823	38844	float16x4_t __ret;
38824		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
38825		__ret = (float16x4_t) __builtin_neon_vcvt_f16_f32((int8x16_t)__rev0, 41);
38826		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	38845	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	38846	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vcvt_f16_f32(__builtin_bit_cast(int8x16_t, __rev0), 41));
	38847	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
38827	38848	return __ret;
38828	38849	}
38829	38850	__ai __attribute__((target("neon"))) float16x4_t __noswap_vcvt_f16_f32(float32x4_t __p0) {
38830	38851	float16x4_t __ret;
38831		__ret = (float16x4_t) __builtin_neon_vcvt_f16_f32((int8x16_t)__p0, 41);
	38852	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vcvt_f16_f32(__builtin_bit_cast(int8x16_t, __p0), 41));
38832	38853	return __ret;
38833	38854	}
38834	38855	#endif
...	...	@@ -38836,20 +38857,20 @@ __ai __attribute__((target("neon"))) float16x4_t __noswap_vcvt_f16_f32(float32x4
38836	38857	#ifdef __LITTLE_ENDIAN__
38837	38858	__ai __attribute__((target("neon"))) float32x4_t vcvt_f32_f16(float16x4_t __p0) {
38838	38859	float32x4_t __ret;
38839		__ret = (float32x4_t) __builtin_neon_vcvt_f32_f16((int8x8_t)__p0, 8);
	38860	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vcvt_f32_f16(__builtin_bit_cast(int8x8_t, __p0), 8));
38840	38861	return __ret;
38841	38862	}
38842	38863	#else
38843	38864	__ai __attribute__((target("neon"))) float32x4_t vcvt_f32_f16(float16x4_t __p0) {
38844	38865	float32x4_t __ret;
38845		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
38846		__ret = (float32x4_t) __builtin_neon_vcvt_f32_f16((int8x8_t)__rev0, 8);
38847		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	38866	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	38867	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vcvt_f32_f16(__builtin_bit_cast(int8x8_t, __rev0), 8));
	38868	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
38848	38869	return __ret;
38849	38870	}
38850	38871	__ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4_t __p0) {
38851	38872	float32x4_t __ret;
38852		__ret = (float32x4_t) __builtin_neon_vcvt_f32_f16((int8x8_t)__p0, 8);
	38873	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vcvt_f32_f16(__builtin_bit_cast(int8x8_t, __p0), 8));
38853	38874	return __ret;
38854	38875	}
38855	38876	#endif
...	...	@@ -38857,14 +38878,14 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
38857	38878	#ifdef __LITTLE_ENDIAN__
38858	38879	#define vld1q_f16(__p0) __extension__ ({ \
38859	38880	float16x8_t __ret; \
38860		__ret = (float16x8_t) __builtin_neon_vld1q_v(__p0, 40); \
	38881	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vld1q_v(__p0, 40)); \
38861	38882	__ret; \
38862	38883	})
38863	38884	#else
38864	38885	#define vld1q_f16(__p0) __extension__ ({ \
38865	38886	float16x8_t __ret; \
38866		__ret = (float16x8_t) __builtin_neon_vld1q_v(__p0, 40); \
38867		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	38887	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vld1q_v(__p0, 40)); \
	38888	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
38868	38889	__ret; \
38869	38890	})
38870	38891	#endif
...	...	@@ -38872,14 +38893,14 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
38872	38893	#ifdef __LITTLE_ENDIAN__
38873	38894	#define vld1_f16(__p0) __extension__ ({ \
38874	38895	float16x4_t __ret; \
38875		__ret = (float16x4_t) __builtin_neon_vld1_v(__p0, 8); \
	38896	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vld1_v(__p0, 8)); \
38876	38897	__ret; \
38877	38898	})
38878	38899	#else
38879	38900	#define vld1_f16(__p0) __extension__ ({ \
38880	38901	float16x4_t __ret; \
38881		__ret = (float16x4_t) __builtin_neon_vld1_v(__p0, 8); \
38882		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	38902	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vld1_v(__p0, 8)); \
	38903	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
38883	38904	__ret; \
38884	38905	})
38885	38906	#endif
...	...	@@ -38887,14 +38908,14 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
38887	38908	#ifdef __LITTLE_ENDIAN__
38888	38909	#define vld1q_dup_f16(__p0) __extension__ ({ \
38889	38910	float16x8_t __ret; \
38890		__ret = (float16x8_t) __builtin_neon_vld1q_dup_v(__p0, 40); \
	38911	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vld1q_dup_v(__p0, 40)); \
38891	38912	__ret; \
38892	38913	})
38893	38914	#else
38894	38915	#define vld1q_dup_f16(__p0) __extension__ ({ \
38895	38916	float16x8_t __ret; \
38896		__ret = (float16x8_t) __builtin_neon_vld1q_dup_v(__p0, 40); \
38897		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	38917	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vld1q_dup_v(__p0, 40)); \
	38918	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
38898	38919	__ret; \
38899	38920	})
38900	38921	#endif
...	...	@@ -38902,14 +38923,14 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
38902	38923	#ifdef __LITTLE_ENDIAN__
38903	38924	#define vld1_dup_f16(__p0) __extension__ ({ \
38904	38925	float16x4_t __ret; \
38905		__ret = (float16x4_t) __builtin_neon_vld1_dup_v(__p0, 8); \
	38926	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vld1_dup_v(__p0, 8)); \
38906	38927	__ret; \
38907	38928	})
38908	38929	#else
38909	38930	#define vld1_dup_f16(__p0) __extension__ ({ \
38910	38931	float16x4_t __ret; \
38911		__ret = (float16x4_t) __builtin_neon_vld1_dup_v(__p0, 8); \
38912		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	38932	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vld1_dup_v(__p0, 8)); \
	38933	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
38913	38934	__ret; \
38914	38935	})
38915	38936	#endif
...	...	@@ -38918,16 +38939,16 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
38918	38939	#define vld1q_lane_f16(__p0, __p1, __p2) __extension__ ({ \
38919	38940	float16x8_t __ret; \
38920	38941	float16x8_t __s1 = __p1; \
38921		__ret = (float16x8_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__s1, __p2, 40); \
	38942	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vld1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 40)); \
38922	38943	__ret; \
38923	38944	})
38924	38945	#else
38925	38946	#define vld1q_lane_f16(__p0, __p1, __p2) __extension__ ({ \
38926	38947	float16x8_t __ret; \
38927	38948	float16x8_t __s1 = __p1; \
38928		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
38929		__ret = (float16x8_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 40); \
38930		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	38949	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	38950	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vld1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 40)); \
	38951	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
38931	38952	__ret; \
38932	38953	})
38933	38954	#endif
...	...	@@ -38936,16 +38957,16 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
38936	38957	#define vld1_lane_f16(__p0, __p1, __p2) __extension__ ({ \
38937	38958	float16x4_t __ret; \
38938	38959	float16x4_t __s1 = __p1; \
38939		__ret = (float16x4_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__s1, __p2, 8); \
	38960	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vld1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 8)); \
38940	38961	__ret; \
38941	38962	})
38942	38963	#else
38943	38964	#define vld1_lane_f16(__p0, __p1, __p2) __extension__ ({ \
38944	38965	float16x4_t __ret; \
38945	38966	float16x4_t __s1 = __p1; \
38946		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
38947		__ret = (float16x4_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__rev1, __p2, 8); \
38948		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	38967	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	38968	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vld1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), __p2, 8)); \
	38969	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
38949	38970	__ret; \
38950	38971	})
38951	38972	#endif
...	...	@@ -38961,8 +38982,8 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
38961	38982	float16x8x2_t __ret; \
38962	38983	__builtin_neon_vld1q_x2_v(&__ret, __p0, 40); \
38963	38984	\
38964		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
38965		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	38985	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	38986	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
38966	38987	__ret; \
38967	38988	})
38968	38989	#endif
...	...	@@ -38978,8 +38999,8 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
38978	38999	float16x4x2_t __ret; \
38979	39000	__builtin_neon_vld1_x2_v(&__ret, __p0, 8); \
38980	39001	\
38981		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
38982		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	39002	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	39003	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
38983	39004	__ret; \
38984	39005	})
38985	39006	#endif
...	...	@@ -38995,9 +39016,9 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
38995	39016	float16x8x3_t __ret; \
38996	39017	__builtin_neon_vld1q_x3_v(&__ret, __p0, 40); \
38997	39018	\
38998		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
38999		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
39000		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	39019	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	39020	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	39021	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
39001	39022	__ret; \
39002	39023	})
39003	39024	#endif
...	...	@@ -39013,9 +39034,9 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
39013	39034	float16x4x3_t __ret; \
39014	39035	__builtin_neon_vld1_x3_v(&__ret, __p0, 8); \
39015	39036	\
39016		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
39017		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
39018		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	39037	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	39038	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	39039	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
39019	39040	__ret; \
39020	39041	})
39021	39042	#endif
...	...	@@ -39031,10 +39052,10 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
39031	39052	float16x8x4_t __ret; \
39032	39053	__builtin_neon_vld1q_x4_v(&__ret, __p0, 40); \
39033	39054	\
39034		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
39035		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
39036		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
39037		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	39055	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	39056	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	39057	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
	39058	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_16); \
39038	39059	__ret; \
39039	39060	})
39040	39061	#endif
...	...	@@ -39050,10 +39071,10 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
39050	39071	float16x4x4_t __ret; \
39051	39072	__builtin_neon_vld1_x4_v(&__ret, __p0, 8); \
39052	39073	\
39053		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
39054		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
39055		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
39056		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	39074	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	39075	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	39076	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
	39077	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_16); \
39057	39078	__ret; \
39058	39079	})
39059	39080	#endif
...	...	@@ -39069,8 +39090,8 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
39069	39090	float16x8x2_t __ret; \
39070	39091	__builtin_neon_vld2q_v(&__ret, __p0, 40); \
39071	39092	\
39072		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
39073		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	39093	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	39094	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
39074	39095	__ret; \
39075	39096	})
39076	39097	#endif
...	...	@@ -39086,8 +39107,8 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
39086	39107	float16x4x2_t __ret; \
39087	39108	__builtin_neon_vld2_v(&__ret, __p0, 8); \
39088	39109	\
39089		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
39090		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	39110	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	39111	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
39091	39112	__ret; \
39092	39113	})
39093	39114	#endif
...	...	@@ -39103,8 +39124,8 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
39103	39124	float16x8x2_t __ret; \
39104	39125	__builtin_neon_vld2q_dup_v(&__ret, __p0, 40); \
39105	39126	\
39106		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
39107		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	39127	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	39128	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
39108	39129	__ret; \
39109	39130	})
39110	39131	#endif
...	...	@@ -39120,8 +39141,8 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
39120	39141	float16x4x2_t __ret; \
39121	39142	__builtin_neon_vld2_dup_v(&__ret, __p0, 8); \
39122	39143	\
39123		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
39124		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	39144	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	39145	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
39125	39146	__ret; \
39126	39147	})
39127	39148	#endif
...	...	@@ -39130,7 +39151,7 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
39130	39151	#define vld2q_lane_f16(__p0, __p1, __p2) __extension__ ({ \
39131	39152	float16x8x2_t __ret; \
39132	39153	float16x8x2_t __s1 = __p1; \
39133		__builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 40); \
	39154	__builtin_neon_vld2q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 40); \
39134	39155	__ret; \
39135	39156	})
39136	39157	#else
...	...	@@ -39138,12 +39159,12 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
39138	39159	float16x8x2_t __ret; \
39139	39160	float16x8x2_t __s1 = __p1; \
39140	39161	float16x8x2_t __rev1; \
39141		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
39142		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
39143		__builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 40); \
	39162	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	39163	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	39164	__builtin_neon_vld2q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 40); \
39144	39165	\
39145		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
39146		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
	39166	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	39167	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
39147	39168	__ret; \
39148	39169	})
39149	39170	#endif
...	...	@@ -39152,7 +39173,7 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
39152	39173	#define vld2_lane_f16(__p0, __p1, __p2) __extension__ ({ \
39153	39174	float16x4x2_t __ret; \
39154	39175	float16x4x2_t __s1 = __p1; \
39155		__builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 8); \
	39176	__builtin_neon_vld2_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 8); \
39156	39177	__ret; \
39157	39178	})
39158	39179	#else
...	...	@@ -39160,12 +39181,12 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
39160	39181	float16x4x2_t __ret; \
39161	39182	float16x4x2_t __s1 = __p1; \
39162	39183	float16x4x2_t __rev1; \
39163		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
39164		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
39165		__builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 8); \
	39184	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	39185	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	39186	__builtin_neon_vld2_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __p2, 8); \
39166	39187	\
39167		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
39168		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
	39188	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	39189	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
39169	39190	__ret; \
39170	39191	})
39171	39192	#endif
...	...	@@ -39181,9 +39202,9 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
39181	39202	float16x8x3_t __ret; \
39182	39203	__builtin_neon_vld3q_v(&__ret, __p0, 40); \
39183	39204	\
39184		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
39185		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
39186		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	39205	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	39206	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	39207	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
39187	39208	__ret; \
39188	39209	})
39189	39210	#endif
...	...	@@ -39199,9 +39220,9 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
39199	39220	float16x4x3_t __ret; \
39200	39221	__builtin_neon_vld3_v(&__ret, __p0, 8); \
39201	39222	\
39202		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
39203		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
39204		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	39223	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	39224	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	39225	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
39205	39226	__ret; \
39206	39227	})
39207	39228	#endif
...	...	@@ -39217,9 +39238,9 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
39217	39238	float16x8x3_t __ret; \
39218	39239	__builtin_neon_vld3q_dup_v(&__ret, __p0, 40); \
39219	39240	\
39220		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
39221		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
39222		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	39241	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	39242	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	39243	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
39223	39244	__ret; \
39224	39245	})
39225	39246	#endif
...	...	@@ -39235,9 +39256,9 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
39235	39256	float16x4x3_t __ret; \
39236	39257	__builtin_neon_vld3_dup_v(&__ret, __p0, 8); \
39237	39258	\
39238		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
39239		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
39240		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	39259	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	39260	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	39261	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
39241	39262	__ret; \
39242	39263	})
39243	39264	#endif
...	...	@@ -39246,7 +39267,7 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
39246	39267	#define vld3q_lane_f16(__p0, __p1, __p2) __extension__ ({ \
39247	39268	float16x8x3_t __ret; \
39248	39269	float16x8x3_t __s1 = __p1; \
39249		__builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 40); \
	39270	__builtin_neon_vld3q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 40); \
39250	39271	__ret; \
39251	39272	})
39252	39273	#else
...	...	@@ -39254,14 +39275,14 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
39254	39275	float16x8x3_t __ret; \
39255	39276	float16x8x3_t __s1 = __p1; \
39256	39277	float16x8x3_t __rev1; \
39257		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
39258		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
39259		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
39260		__builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 40); \
	39278	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	39279	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	39280	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	39281	__builtin_neon_vld3q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 40); \
39261	39282	\
39262		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
39263		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
39264		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
	39283	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	39284	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	39285	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
39265	39286	__ret; \
39266	39287	})
39267	39288	#endif
...	...	@@ -39270,7 +39291,7 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
39270	39291	#define vld3_lane_f16(__p0, __p1, __p2) __extension__ ({ \
39271	39292	float16x4x3_t __ret; \
39272	39293	float16x4x3_t __s1 = __p1; \
39273		__builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 8); \
	39294	__builtin_neon_vld3_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 8); \
39274	39295	__ret; \
39275	39296	})
39276	39297	#else
...	...	@@ -39278,14 +39299,14 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
39278	39299	float16x4x3_t __ret; \
39279	39300	float16x4x3_t __s1 = __p1; \
39280	39301	float16x4x3_t __rev1; \
39281		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
39282		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
39283		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
39284		__builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 8); \
	39302	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	39303	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	39304	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	39305	__builtin_neon_vld3_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __p2, 8); \
39285	39306	\
39286		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
39287		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
39288		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
	39307	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	39308	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	39309	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
39289	39310	__ret; \
39290	39311	})
39291	39312	#endif
...	...	@@ -39301,10 +39322,10 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
39301	39322	float16x8x4_t __ret; \
39302	39323	__builtin_neon_vld4q_v(&__ret, __p0, 40); \
39303	39324	\
39304		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
39305		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
39306		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
39307		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	39325	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	39326	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	39327	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
	39328	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_16); \
39308	39329	__ret; \
39309	39330	})
39310	39331	#endif
...	...	@@ -39320,10 +39341,10 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
39320	39341	float16x4x4_t __ret; \
39321	39342	__builtin_neon_vld4_v(&__ret, __p0, 8); \
39322	39343	\
39323		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
39324		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
39325		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
39326		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	39344	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	39345	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	39346	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
	39347	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_16); \
39327	39348	__ret; \
39328	39349	})
39329	39350	#endif
...	...	@@ -39339,10 +39360,10 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
39339	39360	float16x8x4_t __ret; \
39340	39361	__builtin_neon_vld4q_dup_v(&__ret, __p0, 40); \
39341	39362	\
39342		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
39343		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
39344		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
39345		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	39363	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	39364	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	39365	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
	39366	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_16); \
39346	39367	__ret; \
39347	39368	})
39348	39369	#endif
...	...	@@ -39358,10 +39379,10 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
39358	39379	float16x4x4_t __ret; \
39359	39380	__builtin_neon_vld4_dup_v(&__ret, __p0, 8); \
39360	39381	\
39361		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
39362		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
39363		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
39364		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	39382	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	39383	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	39384	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
	39385	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_16); \
39365	39386	__ret; \
39366	39387	})
39367	39388	#endif
...	...	@@ -39370,7 +39391,7 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
39370	39391	#define vld4q_lane_f16(__p0, __p1, __p2) __extension__ ({ \
39371	39392	float16x8x4_t __ret; \
39372	39393	float16x8x4_t __s1 = __p1; \
39373		__builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 40); \
	39394	__builtin_neon_vld4q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 40); \
39374	39395	__ret; \
39375	39396	})
39376	39397	#else
...	...	@@ -39378,16 +39399,16 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
39378	39399	float16x8x4_t __ret; \
39379	39400	float16x8x4_t __s1 = __p1; \
39380	39401	float16x8x4_t __rev1; \
39381		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
39382		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
39383		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
39384		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
39385		__builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 40); \
	39402	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	39403	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	39404	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	39405	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_16); \
	39406	__builtin_neon_vld4q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 40); \
39386	39407	\
39387		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
39388		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
39389		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
39390		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
	39408	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_16); \
	39409	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_16); \
	39410	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_16); \
	39411	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_16); \
39391	39412	__ret; \
39392	39413	})
39393	39414	#endif
...	...	@@ -39396,7 +39417,7 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
39396	39417	#define vld4_lane_f16(__p0, __p1, __p2) __extension__ ({ \
39397	39418	float16x4x4_t __ret; \
39398	39419	float16x4x4_t __s1 = __p1; \
39399		__builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 8); \
	39420	__builtin_neon_vld4_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 8); \
39400	39421	__ret; \
39401	39422	})
39402	39423	#else
...	...	@@ -39404,16 +39425,16 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
39404	39425	float16x4x4_t __ret; \
39405	39426	float16x4x4_t __s1 = __p1; \
39406	39427	float16x4x4_t __rev1; \
39407		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
39408		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
39409		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
39410		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
39411		__builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 8); \
	39428	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	39429	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	39430	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	39431	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_16); \
	39432	__builtin_neon_vld4_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), __p2, 8); \
39412	39433	\
39413		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
39414		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
39415		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
39416		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
	39434	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_16); \
	39435	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_16); \
	39436	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_16); \
	39437	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_16); \
39417	39438	__ret; \
39418	39439	})
39419	39440	#endif
...	...	@@ -39421,340 +39442,340 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
39421	39442	#ifdef __LITTLE_ENDIAN__
39422	39443	#define vst1q_f16(__p0, __p1) __extension__ ({ \
39423	39444	float16x8_t __s1 = __p1; \
39424		__builtin_neon_vst1q_v(__p0, (int8x16_t)__s1, 40); \
	39445	__builtin_neon_vst1q_v(__p0, __builtin_bit_cast(int8x16_t, __s1), 40); \
39425	39446	})
39426	39447	#else
39427	39448	#define vst1q_f16(__p0, __p1) __extension__ ({ \
39428	39449	float16x8_t __s1 = __p1; \
39429		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
39430		__builtin_neon_vst1q_v(__p0, (int8x16_t)__rev1, 40); \
	39450	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	39451	__builtin_neon_vst1q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), 40); \
39431	39452	})
39432	39453	#endif
39433	39454
39434	39455	#ifdef __LITTLE_ENDIAN__
39435	39456	#define vst1_f16(__p0, __p1) __extension__ ({ \
39436	39457	float16x4_t __s1 = __p1; \
39437		__builtin_neon_vst1_v(__p0, (int8x8_t)__s1, 8); \
	39458	__builtin_neon_vst1_v(__p0, __builtin_bit_cast(int8x8_t, __s1), 8); \
39438	39459	})
39439	39460	#else
39440	39461	#define vst1_f16(__p0, __p1) __extension__ ({ \
39441	39462	float16x4_t __s1 = __p1; \
39442		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
39443		__builtin_neon_vst1_v(__p0, (int8x8_t)__rev1, 8); \
	39463	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	39464	__builtin_neon_vst1_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), 8); \
39444	39465	})
39445	39466	#endif
39446	39467
39447	39468	#ifdef __LITTLE_ENDIAN__
39448	39469	#define vst1q_lane_f16(__p0, __p1, __p2) __extension__ ({ \
39449	39470	float16x8_t __s1 = __p1; \
39450		__builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__s1, __p2, 40); \
	39471	__builtin_neon_vst1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 40); \
39451	39472	})
39452	39473	#else
39453	39474	#define vst1q_lane_f16(__p0, __p1, __p2) __extension__ ({ \
39454	39475	float16x8_t __s1 = __p1; \
39455		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
39456		__builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 40); \
	39476	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	39477	__builtin_neon_vst1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 40); \
39457	39478	})
39458	39479	#endif
39459	39480
39460	39481	#ifdef __LITTLE_ENDIAN__
39461	39482	#define vst1_lane_f16(__p0, __p1, __p2) __extension__ ({ \
39462	39483	float16x4_t __s1 = __p1; \
39463		__builtin_neon_vst1_lane_v(__p0, (int8x8_t)__s1, __p2, 8); \
	39484	__builtin_neon_vst1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 8); \
39464	39485	})
39465	39486	#else
39466	39487	#define vst1_lane_f16(__p0, __p1, __p2) __extension__ ({ \
39467	39488	float16x4_t __s1 = __p1; \
39468		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
39469		__builtin_neon_vst1_lane_v(__p0, (int8x8_t)__rev1, __p2, 8); \
	39489	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	39490	__builtin_neon_vst1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), __p2, 8); \
39470	39491	})
39471	39492	#endif
39472	39493
39473	39494	#ifdef __LITTLE_ENDIAN__
39474	39495	#define vst1q_f16_x2(__p0, __p1) __extension__ ({ \
39475	39496	float16x8x2_t __s1 = __p1; \
39476		__builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 40); \
	39497	__builtin_neon_vst1q_x2_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 40); \
39477	39498	})
39478	39499	#else
39479	39500	#define vst1q_f16_x2(__p0, __p1) __extension__ ({ \
39480	39501	float16x8x2_t __s1 = __p1; \
39481	39502	float16x8x2_t __rev1; \
39482		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
39483		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
39484		__builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 40); \
	39503	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	39504	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	39505	__builtin_neon_vst1q_x2_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 40); \
39485	39506	})
39486	39507	#endif
39487	39508
39488	39509	#ifdef __LITTLE_ENDIAN__
39489	39510	#define vst1_f16_x2(__p0, __p1) __extension__ ({ \
39490	39511	float16x4x2_t __s1 = __p1; \
39491		__builtin_neon_vst1_x2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 8); \
	39512	__builtin_neon_vst1_x2_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 8); \
39492	39513	})
39493	39514	#else
39494	39515	#define vst1_f16_x2(__p0, __p1) __extension__ ({ \
39495	39516	float16x4x2_t __s1 = __p1; \
39496	39517	float16x4x2_t __rev1; \
39497		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
39498		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
39499		__builtin_neon_vst1_x2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 8); \
	39518	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	39519	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	39520	__builtin_neon_vst1_x2_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), 8); \
39500	39521	})
39501	39522	#endif
39502	39523
39503	39524	#ifdef __LITTLE_ENDIAN__
39504	39525	#define vst1q_f16_x3(__p0, __p1) __extension__ ({ \
39505	39526	float16x8x3_t __s1 = __p1; \
39506		__builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 40); \
	39527	__builtin_neon_vst1q_x3_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 40); \
39507	39528	})
39508	39529	#else
39509	39530	#define vst1q_f16_x3(__p0, __p1) __extension__ ({ \
39510	39531	float16x8x3_t __s1 = __p1; \
39511	39532	float16x8x3_t __rev1; \
39512		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
39513		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
39514		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
39515		__builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 40); \
	39533	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	39534	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	39535	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	39536	__builtin_neon_vst1q_x3_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 40); \
39516	39537	})
39517	39538	#endif
39518	39539
39519	39540	#ifdef __LITTLE_ENDIAN__
39520	39541	#define vst1_f16_x3(__p0, __p1) __extension__ ({ \
39521	39542	float16x4x3_t __s1 = __p1; \
39522		__builtin_neon_vst1_x3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 8); \
	39543	__builtin_neon_vst1_x3_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 8); \
39523	39544	})
39524	39545	#else
39525	39546	#define vst1_f16_x3(__p0, __p1) __extension__ ({ \
39526	39547	float16x4x3_t __s1 = __p1; \
39527	39548	float16x4x3_t __rev1; \
39528		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
39529		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
39530		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
39531		__builtin_neon_vst1_x3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 8); \
	39549	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	39550	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	39551	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	39552	__builtin_neon_vst1_x3_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), 8); \
39532	39553	})
39533	39554	#endif
39534	39555
39535	39556	#ifdef __LITTLE_ENDIAN__
39536	39557	#define vst1q_f16_x4(__p0, __p1) __extension__ ({ \
39537	39558	float16x8x4_t __s1 = __p1; \
39538		__builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 40); \
	39559	__builtin_neon_vst1q_x4_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 40); \
39539	39560	})
39540	39561	#else
39541	39562	#define vst1q_f16_x4(__p0, __p1) __extension__ ({ \
39542	39563	float16x8x4_t __s1 = __p1; \
39543	39564	float16x8x4_t __rev1; \
39544		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
39545		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
39546		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
39547		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
39548		__builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 40); \
	39565	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	39566	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	39567	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	39568	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_16); \
	39569	__builtin_neon_vst1q_x4_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 40); \
39549	39570	})
39550	39571	#endif
39551	39572
39552	39573	#ifdef __LITTLE_ENDIAN__
39553	39574	#define vst1_f16_x4(__p0, __p1) __extension__ ({ \
39554	39575	float16x4x4_t __s1 = __p1; \
39555		__builtin_neon_vst1_x4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 8); \
	39576	__builtin_neon_vst1_x4_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 8); \
39556	39577	})
39557	39578	#else
39558	39579	#define vst1_f16_x4(__p0, __p1) __extension__ ({ \
39559	39580	float16x4x4_t __s1 = __p1; \
39560	39581	float16x4x4_t __rev1; \
39561		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
39562		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
39563		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
39564		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
39565		__builtin_neon_vst1_x4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 8); \
	39582	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	39583	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	39584	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	39585	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_16); \
	39586	__builtin_neon_vst1_x4_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), 8); \
39566	39587	})
39567	39588	#endif
39568	39589
39569	39590	#ifdef __LITTLE_ENDIAN__
39570	39591	#define vst2q_f16(__p0, __p1) __extension__ ({ \
39571	39592	float16x8x2_t __s1 = __p1; \
39572		__builtin_neon_vst2q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 40); \
	39593	__builtin_neon_vst2q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 40); \
39573	39594	})
39574	39595	#else
39575	39596	#define vst2q_f16(__p0, __p1) __extension__ ({ \
39576	39597	float16x8x2_t __s1 = __p1; \
39577	39598	float16x8x2_t __rev1; \
39578		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
39579		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
39580		__builtin_neon_vst2q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 40); \
	39599	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	39600	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	39601	__builtin_neon_vst2q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 40); \
39581	39602	})
39582	39603	#endif
39583	39604
39584	39605	#ifdef __LITTLE_ENDIAN__
39585	39606	#define vst2_f16(__p0, __p1) __extension__ ({ \
39586	39607	float16x4x2_t __s1 = __p1; \
39587		__builtin_neon_vst2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 8); \
	39608	__builtin_neon_vst2_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 8); \
39588	39609	})
39589	39610	#else
39590	39611	#define vst2_f16(__p0, __p1) __extension__ ({ \
39591	39612	float16x4x2_t __s1 = __p1; \
39592	39613	float16x4x2_t __rev1; \
39593		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
39594		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
39595		__builtin_neon_vst2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 8); \
	39614	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	39615	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	39616	__builtin_neon_vst2_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), 8); \
39596	39617	})
39597	39618	#endif
39598	39619
39599	39620	#ifdef __LITTLE_ENDIAN__
39600	39621	#define vst2q_lane_f16(__p0, __p1, __p2) __extension__ ({ \
39601	39622	float16x8x2_t __s1 = __p1; \
39602		__builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 40); \
	39623	__builtin_neon_vst2q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 40); \
39603	39624	})
39604	39625	#else
39605	39626	#define vst2q_lane_f16(__p0, __p1, __p2) __extension__ ({ \
39606	39627	float16x8x2_t __s1 = __p1; \
39607	39628	float16x8x2_t __rev1; \
39608		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
39609		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
39610		__builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 40); \
	39629	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	39630	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	39631	__builtin_neon_vst2q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 40); \
39611	39632	})
39612	39633	#endif
39613	39634
39614	39635	#ifdef __LITTLE_ENDIAN__
39615	39636	#define vst2_lane_f16(__p0, __p1, __p2) __extension__ ({ \
39616	39637	float16x4x2_t __s1 = __p1; \
39617		__builtin_neon_vst2_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 8); \
	39638	__builtin_neon_vst2_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 8); \
39618	39639	})
39619	39640	#else
39620	39641	#define vst2_lane_f16(__p0, __p1, __p2) __extension__ ({ \
39621	39642	float16x4x2_t __s1 = __p1; \
39622	39643	float16x4x2_t __rev1; \
39623		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
39624		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
39625		__builtin_neon_vst2_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 8); \
	39644	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	39645	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	39646	__builtin_neon_vst2_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __p2, 8); \
39626	39647	})
39627	39648	#endif
39628	39649
39629	39650	#ifdef __LITTLE_ENDIAN__
39630	39651	#define vst3q_f16(__p0, __p1) __extension__ ({ \
39631	39652	float16x8x3_t __s1 = __p1; \
39632		__builtin_neon_vst3q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 40); \
	39653	__builtin_neon_vst3q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 40); \
39633	39654	})
39634	39655	#else
39635	39656	#define vst3q_f16(__p0, __p1) __extension__ ({ \
39636	39657	float16x8x3_t __s1 = __p1; \
39637	39658	float16x8x3_t __rev1; \
39638		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
39639		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
39640		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
39641		__builtin_neon_vst3q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 40); \
	39659	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	39660	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	39661	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	39662	__builtin_neon_vst3q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 40); \
39642	39663	})
39643	39664	#endif
39644	39665
39645	39666	#ifdef __LITTLE_ENDIAN__
39646	39667	#define vst3_f16(__p0, __p1) __extension__ ({ \
39647	39668	float16x4x3_t __s1 = __p1; \
39648		__builtin_neon_vst3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 8); \
	39669	__builtin_neon_vst3_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 8); \
39649	39670	})
39650	39671	#else
39651	39672	#define vst3_f16(__p0, __p1) __extension__ ({ \
39652	39673	float16x4x3_t __s1 = __p1; \
39653	39674	float16x4x3_t __rev1; \
39654		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
39655		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
39656		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
39657		__builtin_neon_vst3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 8); \
	39675	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	39676	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	39677	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	39678	__builtin_neon_vst3_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), 8); \
39658	39679	})
39659	39680	#endif
39660	39681
39661	39682	#ifdef __LITTLE_ENDIAN__
39662	39683	#define vst3q_lane_f16(__p0, __p1, __p2) __extension__ ({ \
39663	39684	float16x8x3_t __s1 = __p1; \
39664		__builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 40); \
	39685	__builtin_neon_vst3q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 40); \
39665	39686	})
39666	39687	#else
39667	39688	#define vst3q_lane_f16(__p0, __p1, __p2) __extension__ ({ \
39668	39689	float16x8x3_t __s1 = __p1; \
39669	39690	float16x8x3_t __rev1; \
39670		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
39671		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
39672		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
39673		__builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 40); \
	39691	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	39692	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	39693	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	39694	__builtin_neon_vst3q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 40); \
39674	39695	})
39675	39696	#endif
39676	39697
39677	39698	#ifdef __LITTLE_ENDIAN__
39678	39699	#define vst3_lane_f16(__p0, __p1, __p2) __extension__ ({ \
39679	39700	float16x4x3_t __s1 = __p1; \
39680		__builtin_neon_vst3_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 8); \
	39701	__builtin_neon_vst3_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 8); \
39681	39702	})
39682	39703	#else
39683	39704	#define vst3_lane_f16(__p0, __p1, __p2) __extension__ ({ \
39684	39705	float16x4x3_t __s1 = __p1; \
39685	39706	float16x4x3_t __rev1; \
39686		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
39687		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
39688		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
39689		__builtin_neon_vst3_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 8); \
	39707	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	39708	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	39709	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	39710	__builtin_neon_vst3_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __p2, 8); \
39690	39711	})
39691	39712	#endif
39692	39713
39693	39714	#ifdef __LITTLE_ENDIAN__
39694	39715	#define vst4q_f16(__p0, __p1) __extension__ ({ \
39695	39716	float16x8x4_t __s1 = __p1; \
39696		__builtin_neon_vst4q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 40); \
	39717	__builtin_neon_vst4q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 40); \
39697	39718	})
39698	39719	#else
39699	39720	#define vst4q_f16(__p0, __p1) __extension__ ({ \
39700	39721	float16x8x4_t __s1 = __p1; \
39701	39722	float16x8x4_t __rev1; \
39702		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
39703		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
39704		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
39705		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
39706		__builtin_neon_vst4q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 40); \
	39723	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	39724	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	39725	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	39726	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_16); \
	39727	__builtin_neon_vst4q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 40); \
39707	39728	})
39708	39729	#endif
39709	39730
39710	39731	#ifdef __LITTLE_ENDIAN__
39711	39732	#define vst4_f16(__p0, __p1) __extension__ ({ \
39712	39733	float16x4x4_t __s1 = __p1; \
39713		__builtin_neon_vst4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 8); \
	39734	__builtin_neon_vst4_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 8); \
39714	39735	})
39715	39736	#else
39716	39737	#define vst4_f16(__p0, __p1) __extension__ ({ \
39717	39738	float16x4x4_t __s1 = __p1; \
39718	39739	float16x4x4_t __rev1; \
39719		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
39720		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
39721		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
39722		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
39723		__builtin_neon_vst4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 8); \
	39740	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	39741	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	39742	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	39743	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_16); \
	39744	__builtin_neon_vst4_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), 8); \
39724	39745	})
39725	39746	#endif
39726	39747
39727	39748	#ifdef __LITTLE_ENDIAN__
39728	39749	#define vst4q_lane_f16(__p0, __p1, __p2) __extension__ ({ \
39729	39750	float16x8x4_t __s1 = __p1; \
39730		__builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 40); \
	39751	__builtin_neon_vst4q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 40); \
39731	39752	})
39732	39753	#else
39733	39754	#define vst4q_lane_f16(__p0, __p1, __p2) __extension__ ({ \
39734	39755	float16x8x4_t __s1 = __p1; \
39735	39756	float16x8x4_t __rev1; \
39736		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
39737		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
39738		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
39739		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
39740		__builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 40); \
	39757	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_16); \
	39758	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_16); \
	39759	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_16); \
	39760	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_16); \
	39761	__builtin_neon_vst4q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 40); \
39741	39762	})
39742	39763	#endif
39743	39764
39744	39765	#ifdef __LITTLE_ENDIAN__
39745	39766	#define vst4_lane_f16(__p0, __p1, __p2) __extension__ ({ \
39746	39767	float16x4x4_t __s1 = __p1; \
39747		__builtin_neon_vst4_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 8); \
	39768	__builtin_neon_vst4_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 8); \
39748	39769	})
39749	39770	#else
39750	39771	#define vst4_lane_f16(__p0, __p1, __p2) __extension__ ({ \
39751	39772	float16x4x4_t __s1 = __p1; \
39752	39773	float16x4x4_t __rev1; \
39753		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
39754		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
39755		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
39756		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
39757		__builtin_neon_vst4_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 8); \
	39774	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_16); \
	39775	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_16); \
	39776	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_16); \
	39777	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_16); \
	39778	__builtin_neon_vst4_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), __p2, 8); \
39758	39779	})
39759	39780	#endif
39760	39781
...	...	@@ -39763,45 +39784,45 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vcvt_f32_f16(float16x4
39763	39784	#ifdef __LITTLE_ENDIAN__
39764	39785	__ai __attribute__((target("neon"))) float64x2_t vmaxnmq_f64(float64x2_t __p0, float64x2_t __p1) {
39765	39786	float64x2_t __ret;
39766		__ret = (float64x2_t) __builtin_neon_vmaxnmq_v((int8x16_t)__p0, (int8x16_t)__p1, 42);
	39787	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vmaxnmq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 42));
39767	39788	return __ret;
39768	39789	}
39769	39790	#else
39770	39791	__ai __attribute__((target("neon"))) float64x2_t vmaxnmq_f64(float64x2_t __p0, float64x2_t __p1) {
39771	39792	float64x2_t __ret;
39772		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
39773		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
39774		__ret = (float64x2_t) __builtin_neon_vmaxnmq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 42);
39775		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	39793	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	39794	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	39795	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vmaxnmq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 42));
	39796	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
39776	39797	return __ret;
39777	39798	}
39778	39799	#endif
39779	39800
39780	39801	__ai __attribute__((target("neon"))) float64x1_t vmaxnm_f64(float64x1_t __p0, float64x1_t __p1) {
39781	39802	float64x1_t __ret;
39782		__ret = (float64x1_t) __builtin_neon_vmaxnm_v((int8x8_t)__p0, (int8x8_t)__p1, 10);
	39803	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vmaxnm_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 10));
39783	39804	return __ret;
39784	39805	}
39785	39806	#ifdef __LITTLE_ENDIAN__
39786	39807	__ai __attribute__((target("neon"))) float64x2_t vminnmq_f64(float64x2_t __p0, float64x2_t __p1) {
39787	39808	float64x2_t __ret;
39788		__ret = (float64x2_t) __builtin_neon_vminnmq_v((int8x16_t)__p0, (int8x16_t)__p1, 42);
	39809	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vminnmq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 42));
39789	39810	return __ret;
39790	39811	}
39791	39812	#else
39792	39813	__ai __attribute__((target("neon"))) float64x2_t vminnmq_f64(float64x2_t __p0, float64x2_t __p1) {
39793	39814	float64x2_t __ret;
39794		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
39795		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
39796		__ret = (float64x2_t) __builtin_neon_vminnmq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 42);
39797		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	39815	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	39816	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	39817	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vminnmq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 42));
	39818	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
39798	39819	return __ret;
39799	39820	}
39800	39821	#endif
39801	39822
39802	39823	__ai __attribute__((target("neon"))) float64x1_t vminnm_f64(float64x1_t __p0, float64x1_t __p1) {
39803	39824	float64x1_t __ret;
39804		__ret = (float64x1_t) __builtin_neon_vminnm_v((int8x8_t)__p0, (int8x8_t)__p1, 10);
	39825	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vminnm_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 10));
39805	39826	return __ret;
39806	39827	}
39807	39828	#endif
...	...	@@ -39809,148 +39830,148 @@ __ai __attribute__((target("neon"))) float64x1_t vminnm_f64(float64x1_t __p0, fl
39809	39830	#ifdef __LITTLE_ENDIAN__
39810	39831	__ai __attribute__((target("neon"))) float64x2_t vrndq_f64(float64x2_t __p0) {
39811	39832	float64x2_t __ret;
39812		__ret = (float64x2_t) __builtin_neon_vrndq_v((int8x16_t)__p0, 42);
	39833	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vrndq_v(__builtin_bit_cast(int8x16_t, __p0), 42));
39813	39834	return __ret;
39814	39835	}
39815	39836	#else
39816	39837	__ai __attribute__((target("neon"))) float64x2_t vrndq_f64(float64x2_t __p0) {
39817	39838	float64x2_t __ret;
39818		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
39819		__ret = (float64x2_t) __builtin_neon_vrndq_v((int8x16_t)__rev0, 42);
39820		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	39839	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	39840	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vrndq_v(__builtin_bit_cast(int8x16_t, __rev0), 42));
	39841	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
39821	39842	return __ret;
39822	39843	}
39823	39844	#endif
39824	39845
39825	39846	__ai __attribute__((target("neon"))) float64x1_t vrnd_f64(float64x1_t __p0) {
39826	39847	float64x1_t __ret;
39827		__ret = (float64x1_t) __builtin_neon_vrnd_v((int8x8_t)__p0, 10);
	39848	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vrnd_v(__builtin_bit_cast(int8x8_t, __p0), 10));
39828	39849	return __ret;
39829	39850	}
39830	39851	#ifdef __LITTLE_ENDIAN__
39831	39852	__ai __attribute__((target("neon"))) float64x2_t vrndaq_f64(float64x2_t __p0) {
39832	39853	float64x2_t __ret;
39833		__ret = (float64x2_t) __builtin_neon_vrndaq_v((int8x16_t)__p0, 42);
	39854	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vrndaq_v(__builtin_bit_cast(int8x16_t, __p0), 42));
39834	39855	return __ret;
39835	39856	}
39836	39857	#else
39837	39858	__ai __attribute__((target("neon"))) float64x2_t vrndaq_f64(float64x2_t __p0) {
39838	39859	float64x2_t __ret;
39839		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
39840		__ret = (float64x2_t) __builtin_neon_vrndaq_v((int8x16_t)__rev0, 42);
39841		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	39860	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	39861	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vrndaq_v(__builtin_bit_cast(int8x16_t, __rev0), 42));
	39862	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
39842	39863	return __ret;
39843	39864	}
39844	39865	#endif
39845	39866
39846	39867	__ai __attribute__((target("neon"))) float64x1_t vrnda_f64(float64x1_t __p0) {
39847	39868	float64x1_t __ret;
39848		__ret = (float64x1_t) __builtin_neon_vrnda_v((int8x8_t)__p0, 10);
	39869	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vrnda_v(__builtin_bit_cast(int8x8_t, __p0), 10));
39849	39870	return __ret;
39850	39871	}
39851	39872	#ifdef __LITTLE_ENDIAN__
39852	39873	__ai __attribute__((target("neon"))) float64x2_t vrndiq_f64(float64x2_t __p0) {
39853	39874	float64x2_t __ret;
39854		__ret = (float64x2_t) __builtin_neon_vrndiq_v((int8x16_t)__p0, 42);
	39875	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vrndiq_v(__builtin_bit_cast(int8x16_t, __p0), 42));
39855	39876	return __ret;
39856	39877	}
39857	39878	#else
39858	39879	__ai __attribute__((target("neon"))) float64x2_t vrndiq_f64(float64x2_t __p0) {
39859	39880	float64x2_t __ret;
39860		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
39861		__ret = (float64x2_t) __builtin_neon_vrndiq_v((int8x16_t)__rev0, 42);
39862		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	39881	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	39882	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vrndiq_v(__builtin_bit_cast(int8x16_t, __rev0), 42));
	39883	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
39863	39884	return __ret;
39864	39885	}
39865	39886	#endif
39866	39887
39867	39888	__ai __attribute__((target("neon"))) float64x1_t vrndi_f64(float64x1_t __p0) {
39868	39889	float64x1_t __ret;
39869		__ret = (float64x1_t) __builtin_neon_vrndi_v((int8x8_t)__p0, 10);
	39890	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vrndi_v(__builtin_bit_cast(int8x8_t, __p0), 10));
39870	39891	return __ret;
39871	39892	}
39872	39893	#ifdef __LITTLE_ENDIAN__
39873	39894	__ai __attribute__((target("neon"))) float64x2_t vrndmq_f64(float64x2_t __p0) {
39874	39895	float64x2_t __ret;
39875		__ret = (float64x2_t) __builtin_neon_vrndmq_v((int8x16_t)__p0, 42);
	39896	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vrndmq_v(__builtin_bit_cast(int8x16_t, __p0), 42));
39876	39897	return __ret;
39877	39898	}
39878	39899	#else
39879	39900	__ai __attribute__((target("neon"))) float64x2_t vrndmq_f64(float64x2_t __p0) {
39880	39901	float64x2_t __ret;
39881		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
39882		__ret = (float64x2_t) __builtin_neon_vrndmq_v((int8x16_t)__rev0, 42);
39883		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	39902	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	39903	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vrndmq_v(__builtin_bit_cast(int8x16_t, __rev0), 42));
	39904	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
39884	39905	return __ret;
39885	39906	}
39886	39907	#endif
39887	39908
39888	39909	__ai __attribute__((target("neon"))) float64x1_t vrndm_f64(float64x1_t __p0) {
39889	39910	float64x1_t __ret;
39890		__ret = (float64x1_t) __builtin_neon_vrndm_v((int8x8_t)__p0, 10);
	39911	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vrndm_v(__builtin_bit_cast(int8x8_t, __p0), 10));
39891	39912	return __ret;
39892	39913	}
39893	39914	#ifdef __LITTLE_ENDIAN__
39894	39915	__ai __attribute__((target("neon"))) float64x2_t vrndnq_f64(float64x2_t __p0) {
39895	39916	float64x2_t __ret;
39896		__ret = (float64x2_t) __builtin_neon_vrndnq_v((int8x16_t)__p0, 42);
	39917	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vrndnq_v(__builtin_bit_cast(int8x16_t, __p0), 42));
39897	39918	return __ret;
39898	39919	}
39899	39920	#else
39900	39921	__ai __attribute__((target("neon"))) float64x2_t vrndnq_f64(float64x2_t __p0) {
39901	39922	float64x2_t __ret;
39902		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
39903		__ret = (float64x2_t) __builtin_neon_vrndnq_v((int8x16_t)__rev0, 42);
39904		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	39923	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	39924	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vrndnq_v(__builtin_bit_cast(int8x16_t, __rev0), 42));
	39925	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
39905	39926	return __ret;
39906	39927	}
39907	39928	#endif
39908	39929
39909	39930	__ai __attribute__((target("neon"))) float64x1_t vrndn_f64(float64x1_t __p0) {
39910	39931	float64x1_t __ret;
39911		__ret = (float64x1_t) __builtin_neon_vrndn_v((int8x8_t)__p0, 10);
	39932	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vrndn_v(__builtin_bit_cast(int8x8_t, __p0), 10));
39912	39933	return __ret;
39913	39934	}
39914	39935	#ifdef __LITTLE_ENDIAN__
39915	39936	__ai __attribute__((target("neon"))) float64x2_t vrndpq_f64(float64x2_t __p0) {
39916	39937	float64x2_t __ret;
39917		__ret = (float64x2_t) __builtin_neon_vrndpq_v((int8x16_t)__p0, 42);
	39938	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vrndpq_v(__builtin_bit_cast(int8x16_t, __p0), 42));
39918	39939	return __ret;
39919	39940	}
39920	39941	#else
39921	39942	__ai __attribute__((target("neon"))) float64x2_t vrndpq_f64(float64x2_t __p0) {
39922	39943	float64x2_t __ret;
39923		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
39924		__ret = (float64x2_t) __builtin_neon_vrndpq_v((int8x16_t)__rev0, 42);
39925		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	39944	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	39945	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vrndpq_v(__builtin_bit_cast(int8x16_t, __rev0), 42));
	39946	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
39926	39947	return __ret;
39927	39948	}
39928	39949	#endif
39929	39950
39930	39951	__ai __attribute__((target("neon"))) float64x1_t vrndp_f64(float64x1_t __p0) {
39931	39952	float64x1_t __ret;
39932		__ret = (float64x1_t) __builtin_neon_vrndp_v((int8x8_t)__p0, 10);
	39953	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vrndp_v(__builtin_bit_cast(int8x8_t, __p0), 10));
39933	39954	return __ret;
39934	39955	}
39935	39956	#ifdef __LITTLE_ENDIAN__
39936	39957	__ai __attribute__((target("neon"))) float64x2_t vrndxq_f64(float64x2_t __p0) {
39937	39958	float64x2_t __ret;
39938		__ret = (float64x2_t) __builtin_neon_vrndxq_v((int8x16_t)__p0, 42);
	39959	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vrndxq_v(__builtin_bit_cast(int8x16_t, __p0), 42));
39939	39960	return __ret;
39940	39961	}
39941	39962	#else
39942	39963	__ai __attribute__((target("neon"))) float64x2_t vrndxq_f64(float64x2_t __p0) {
39943	39964	float64x2_t __ret;
39944		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
39945		__ret = (float64x2_t) __builtin_neon_vrndxq_v((int8x16_t)__rev0, 42);
39946		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	39965	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	39966	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vrndxq_v(__builtin_bit_cast(int8x16_t, __rev0), 42));
	39967	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
39947	39968	return __ret;
39948	39969	}
39949	39970	#endif
39950	39971
39951	39972	__ai __attribute__((target("neon"))) float64x1_t vrndx_f64(float64x1_t __p0) {
39952	39973	float64x1_t __ret;
39953		__ret = (float64x1_t) __builtin_neon_vrndx_v((int8x8_t)__p0, 10);
	39974	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vrndx_v(__builtin_bit_cast(int8x8_t, __p0), 10));
39954	39975	return __ret;
39955	39976	}
39956	39977	#endif
...	...	@@ -39958,16 +39979,16 @@ __ai __attribute__((target("neon"))) float64x1_t vrndx_f64(float64x1_t __p0) {
39958	39979	#ifdef __LITTLE_ENDIAN__
39959	39980	__ai __attribute__((target("aes,neon"))) uint8x16_t vaesdq_u8(uint8x16_t __p0, uint8x16_t __p1) {
39960	39981	uint8x16_t __ret;
39961		__ret = (uint8x16_t) __builtin_neon_vaesdq_u8((int8x16_t)__p0, (int8x16_t)__p1, 48);
	39982	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vaesdq_u8(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 48));
39962	39983	return __ret;
39963	39984	}
39964	39985	#else
39965	39986	__ai __attribute__((target("aes,neon"))) uint8x16_t vaesdq_u8(uint8x16_t __p0, uint8x16_t __p1) {
39966	39987	uint8x16_t __ret;
39967		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
39968		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
39969		__ret = (uint8x16_t) __builtin_neon_vaesdq_u8((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
39970		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	39988	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	39989	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	39990	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vaesdq_u8(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 48));
	39991	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
39971	39992	return __ret;
39972	39993	}
39973	39994	#endif
...	...	@@ -39975,16 +39996,16 @@ __ai __attribute__((target("aes,neon"))) uint8x16_t vaesdq_u8(uint8x16_t __p0, u
39975	39996	#ifdef __LITTLE_ENDIAN__
39976	39997	__ai __attribute__((target("aes,neon"))) uint8x16_t vaeseq_u8(uint8x16_t __p0, uint8x16_t __p1) {
39977	39998	uint8x16_t __ret;
39978		__ret = (uint8x16_t) __builtin_neon_vaeseq_u8((int8x16_t)__p0, (int8x16_t)__p1, 48);
	39999	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vaeseq_u8(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 48));
39979	40000	return __ret;
39980	40001	}
39981	40002	#else
39982	40003	__ai __attribute__((target("aes,neon"))) uint8x16_t vaeseq_u8(uint8x16_t __p0, uint8x16_t __p1) {
39983	40004	uint8x16_t __ret;
39984		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
39985		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
39986		__ret = (uint8x16_t) __builtin_neon_vaeseq_u8((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
39987		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	40005	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	40006	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	40007	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vaeseq_u8(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 48));
	40008	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
39988	40009	return __ret;
39989	40010	}
39990	40011	#endif
...	...	@@ -39992,15 +40013,15 @@ __ai __attribute__((target("aes,neon"))) uint8x16_t vaeseq_u8(uint8x16_t __p0, u
39992	40013	#ifdef __LITTLE_ENDIAN__
39993	40014	__ai __attribute__((target("aes,neon"))) uint8x16_t vaesimcq_u8(uint8x16_t __p0) {
39994	40015	uint8x16_t __ret;
39995		__ret = (uint8x16_t) __builtin_neon_vaesimcq_u8((int8x16_t)__p0, 48);
	40016	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vaesimcq_u8(__builtin_bit_cast(int8x16_t, __p0), 48));
39996	40017	return __ret;
39997	40018	}
39998	40019	#else
39999	40020	__ai __attribute__((target("aes,neon"))) uint8x16_t vaesimcq_u8(uint8x16_t __p0) {
40000	40021	uint8x16_t __ret;
40001		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
40002		__ret = (uint8x16_t) __builtin_neon_vaesimcq_u8((int8x16_t)__rev0, 48);
40003		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	40022	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	40023	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vaesimcq_u8(__builtin_bit_cast(int8x16_t, __rev0), 48));
	40024	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
40004	40025	return __ret;
40005	40026	}
40006	40027	#endif
...	...	@@ -40008,15 +40029,15 @@ __ai __attribute__((target("aes,neon"))) uint8x16_t vaesimcq_u8(uint8x16_t __p0)
40008	40029	#ifdef __LITTLE_ENDIAN__
40009	40030	__ai __attribute__((target("aes,neon"))) uint8x16_t vaesmcq_u8(uint8x16_t __p0) {
40010	40031	uint8x16_t __ret;
40011		__ret = (uint8x16_t) __builtin_neon_vaesmcq_u8((int8x16_t)__p0, 48);
	40032	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vaesmcq_u8(__builtin_bit_cast(int8x16_t, __p0), 48));
40012	40033	return __ret;
40013	40034	}
40014	40035	#else
40015	40036	__ai __attribute__((target("aes,neon"))) uint8x16_t vaesmcq_u8(uint8x16_t __p0) {
40016	40037	uint8x16_t __ret;
40017		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
40018		__ret = (uint8x16_t) __builtin_neon_vaesmcq_u8((int8x16_t)__rev0, 48);
40019		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	40038	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	40039	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vaesmcq_u8(__builtin_bit_cast(int8x16_t, __rev0), 48));
	40040	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
40020	40041	return __ret;
40021	40042	}
40022	40043	#endif
...	...	@@ -40024,15 +40045,15 @@ __ai __attribute__((target("aes,neon"))) uint8x16_t vaesmcq_u8(uint8x16_t __p0)
40024	40045	#ifdef __LITTLE_ENDIAN__
40025	40046	__ai __attribute__((target("neon"))) int32x4_t vcvtaq_s32_f32(float32x4_t __p0) {
40026	40047	int32x4_t __ret;
40027		__ret = (int32x4_t) __builtin_neon_vcvtaq_s32_v((int8x16_t)__p0, 34);
	40048	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vcvtaq_s32_v(__builtin_bit_cast(int8x16_t, __p0), 34));
40028	40049	return __ret;
40029	40050	}
40030	40051	#else
40031	40052	__ai __attribute__((target("neon"))) int32x4_t vcvtaq_s32_f32(float32x4_t __p0) {
40032	40053	int32x4_t __ret;
40033		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40034		__ret = (int32x4_t) __builtin_neon_vcvtaq_s32_v((int8x16_t)__rev0, 34);
40035		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40054	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	40055	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vcvtaq_s32_v(__builtin_bit_cast(int8x16_t, __rev0), 34));
	40056	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
40036	40057	return __ret;
40037	40058	}
40038	40059	#endif
...	...	@@ -40040,15 +40061,15 @@ __ai __attribute__((target("neon"))) int32x4_t vcvtaq_s32_f32(float32x4_t __p0)
40040	40061	#ifdef __LITTLE_ENDIAN__
40041	40062	__ai __attribute__((target("neon"))) int32x2_t vcvta_s32_f32(float32x2_t __p0) {
40042	40063	int32x2_t __ret;
40043		__ret = (int32x2_t) __builtin_neon_vcvta_s32_v((int8x8_t)__p0, 2);
	40064	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vcvta_s32_v(__builtin_bit_cast(int8x8_t, __p0), 2));
40044	40065	return __ret;
40045	40066	}
40046	40067	#else
40047	40068	__ai __attribute__((target("neon"))) int32x2_t vcvta_s32_f32(float32x2_t __p0) {
40048	40069	int32x2_t __ret;
40049		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
40050		__ret = (int32x2_t) __builtin_neon_vcvta_s32_v((int8x8_t)__rev0, 2);
40051		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	40070	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	40071	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vcvta_s32_v(__builtin_bit_cast(int8x8_t, __rev0), 2));
	40072	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
40052	40073	return __ret;
40053	40074	}
40054	40075	#endif
...	...	@@ -40056,15 +40077,15 @@ __ai __attribute__((target("neon"))) int32x2_t vcvta_s32_f32(float32x2_t __p0) {
40056	40077	#ifdef __LITTLE_ENDIAN__
40057	40078	__ai __attribute__((target("neon"))) uint32x4_t vcvtaq_u32_f32(float32x4_t __p0) {
40058	40079	uint32x4_t __ret;
40059		__ret = (uint32x4_t) __builtin_neon_vcvtaq_u32_v((int8x16_t)__p0, 50);
	40080	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcvtaq_u32_v(__builtin_bit_cast(int8x16_t, __p0), 50));
40060	40081	return __ret;
40061	40082	}
40062	40083	#else
40063	40084	__ai __attribute__((target("neon"))) uint32x4_t vcvtaq_u32_f32(float32x4_t __p0) {
40064	40085	uint32x4_t __ret;
40065		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40066		__ret = (uint32x4_t) __builtin_neon_vcvtaq_u32_v((int8x16_t)__rev0, 50);
40067		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40086	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	40087	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcvtaq_u32_v(__builtin_bit_cast(int8x16_t, __rev0), 50));
	40088	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
40068	40089	return __ret;
40069	40090	}
40070	40091	#endif
...	...	@@ -40072,15 +40093,15 @@ __ai __attribute__((target("neon"))) uint32x4_t vcvtaq_u32_f32(float32x4_t __p0)
40072	40093	#ifdef __LITTLE_ENDIAN__
40073	40094	__ai __attribute__((target("neon"))) uint32x2_t vcvta_u32_f32(float32x2_t __p0) {
40074	40095	uint32x2_t __ret;
40075		__ret = (uint32x2_t) __builtin_neon_vcvta_u32_v((int8x8_t)__p0, 18);
	40096	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcvta_u32_v(__builtin_bit_cast(int8x8_t, __p0), 18));
40076	40097	return __ret;
40077	40098	}
40078	40099	#else
40079	40100	__ai __attribute__((target("neon"))) uint32x2_t vcvta_u32_f32(float32x2_t __p0) {
40080	40101	uint32x2_t __ret;
40081		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
40082		__ret = (uint32x2_t) __builtin_neon_vcvta_u32_v((int8x8_t)__rev0, 18);
40083		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	40102	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	40103	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcvta_u32_v(__builtin_bit_cast(int8x8_t, __rev0), 18));
	40104	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
40084	40105	return __ret;
40085	40106	}
40086	40107	#endif
...	...	@@ -40088,15 +40109,15 @@ __ai __attribute__((target("neon"))) uint32x2_t vcvta_u32_f32(float32x2_t __p0)
40088	40109	#ifdef __LITTLE_ENDIAN__
40089	40110	__ai __attribute__((target("neon"))) int32x4_t vcvtmq_s32_f32(float32x4_t __p0) {
40090	40111	int32x4_t __ret;
40091		__ret = (int32x4_t) __builtin_neon_vcvtmq_s32_v((int8x16_t)__p0, 34);
	40112	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vcvtmq_s32_v(__builtin_bit_cast(int8x16_t, __p0), 34));
40092	40113	return __ret;
40093	40114	}
40094	40115	#else
40095	40116	__ai __attribute__((target("neon"))) int32x4_t vcvtmq_s32_f32(float32x4_t __p0) {
40096	40117	int32x4_t __ret;
40097		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40098		__ret = (int32x4_t) __builtin_neon_vcvtmq_s32_v((int8x16_t)__rev0, 34);
40099		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40118	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	40119	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vcvtmq_s32_v(__builtin_bit_cast(int8x16_t, __rev0), 34));
	40120	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
40100	40121	return __ret;
40101	40122	}
40102	40123	#endif
...	...	@@ -40104,15 +40125,15 @@ __ai __attribute__((target("neon"))) int32x4_t vcvtmq_s32_f32(float32x4_t __p0)
40104	40125	#ifdef __LITTLE_ENDIAN__
40105	40126	__ai __attribute__((target("neon"))) int32x2_t vcvtm_s32_f32(float32x2_t __p0) {
40106	40127	int32x2_t __ret;
40107		__ret = (int32x2_t) __builtin_neon_vcvtm_s32_v((int8x8_t)__p0, 2);
	40128	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vcvtm_s32_v(__builtin_bit_cast(int8x8_t, __p0), 2));
40108	40129	return __ret;
40109	40130	}
40110	40131	#else
40111	40132	__ai __attribute__((target("neon"))) int32x2_t vcvtm_s32_f32(float32x2_t __p0) {
40112	40133	int32x2_t __ret;
40113		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
40114		__ret = (int32x2_t) __builtin_neon_vcvtm_s32_v((int8x8_t)__rev0, 2);
40115		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	40134	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	40135	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vcvtm_s32_v(__builtin_bit_cast(int8x8_t, __rev0), 2));
	40136	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
40116	40137	return __ret;
40117	40138	}
40118	40139	#endif
...	...	@@ -40120,15 +40141,15 @@ __ai __attribute__((target("neon"))) int32x2_t vcvtm_s32_f32(float32x2_t __p0) {
40120	40141	#ifdef __LITTLE_ENDIAN__
40121	40142	__ai __attribute__((target("neon"))) uint32x4_t vcvtmq_u32_f32(float32x4_t __p0) {
40122	40143	uint32x4_t __ret;
40123		__ret = (uint32x4_t) __builtin_neon_vcvtmq_u32_v((int8x16_t)__p0, 50);
	40144	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcvtmq_u32_v(__builtin_bit_cast(int8x16_t, __p0), 50));
40124	40145	return __ret;
40125	40146	}
40126	40147	#else
40127	40148	__ai __attribute__((target("neon"))) uint32x4_t vcvtmq_u32_f32(float32x4_t __p0) {
40128	40149	uint32x4_t __ret;
40129		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40130		__ret = (uint32x4_t) __builtin_neon_vcvtmq_u32_v((int8x16_t)__rev0, 50);
40131		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40150	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	40151	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcvtmq_u32_v(__builtin_bit_cast(int8x16_t, __rev0), 50));
	40152	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
40132	40153	return __ret;
40133	40154	}
40134	40155	#endif
...	...	@@ -40136,15 +40157,15 @@ __ai __attribute__((target("neon"))) uint32x4_t vcvtmq_u32_f32(float32x4_t __p0)
40136	40157	#ifdef __LITTLE_ENDIAN__
40137	40158	__ai __attribute__((target("neon"))) uint32x2_t vcvtm_u32_f32(float32x2_t __p0) {
40138	40159	uint32x2_t __ret;
40139		__ret = (uint32x2_t) __builtin_neon_vcvtm_u32_v((int8x8_t)__p0, 18);
	40160	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcvtm_u32_v(__builtin_bit_cast(int8x8_t, __p0), 18));
40140	40161	return __ret;
40141	40162	}
40142	40163	#else
40143	40164	__ai __attribute__((target("neon"))) uint32x2_t vcvtm_u32_f32(float32x2_t __p0) {
40144	40165	uint32x2_t __ret;
40145		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
40146		__ret = (uint32x2_t) __builtin_neon_vcvtm_u32_v((int8x8_t)__rev0, 18);
40147		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	40166	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	40167	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcvtm_u32_v(__builtin_bit_cast(int8x8_t, __rev0), 18));
	40168	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
40148	40169	return __ret;
40149	40170	}
40150	40171	#endif
...	...	@@ -40152,15 +40173,15 @@ __ai __attribute__((target("neon"))) uint32x2_t vcvtm_u32_f32(float32x2_t __p0)
40152	40173	#ifdef __LITTLE_ENDIAN__
40153	40174	__ai __attribute__((target("neon"))) int32x4_t vcvtnq_s32_f32(float32x4_t __p0) {
40154	40175	int32x4_t __ret;
40155		__ret = (int32x4_t) __builtin_neon_vcvtnq_s32_v((int8x16_t)__p0, 34);
	40176	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vcvtnq_s32_v(__builtin_bit_cast(int8x16_t, __p0), 34));
40156	40177	return __ret;
40157	40178	}
40158	40179	#else
40159	40180	__ai __attribute__((target("neon"))) int32x4_t vcvtnq_s32_f32(float32x4_t __p0) {
40160	40181	int32x4_t __ret;
40161		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40162		__ret = (int32x4_t) __builtin_neon_vcvtnq_s32_v((int8x16_t)__rev0, 34);
40163		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40182	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	40183	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vcvtnq_s32_v(__builtin_bit_cast(int8x16_t, __rev0), 34));
	40184	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
40164	40185	return __ret;
40165	40186	}
40166	40187	#endif
...	...	@@ -40168,15 +40189,15 @@ __ai __attribute__((target("neon"))) int32x4_t vcvtnq_s32_f32(float32x4_t __p0)
40168	40189	#ifdef __LITTLE_ENDIAN__
40169	40190	__ai __attribute__((target("neon"))) int32x2_t vcvtn_s32_f32(float32x2_t __p0) {
40170	40191	int32x2_t __ret;
40171		__ret = (int32x2_t) __builtin_neon_vcvtn_s32_v((int8x8_t)__p0, 2);
	40192	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vcvtn_s32_v(__builtin_bit_cast(int8x8_t, __p0), 2));
40172	40193	return __ret;
40173	40194	}
40174	40195	#else
40175	40196	__ai __attribute__((target("neon"))) int32x2_t vcvtn_s32_f32(float32x2_t __p0) {
40176	40197	int32x2_t __ret;
40177		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
40178		__ret = (int32x2_t) __builtin_neon_vcvtn_s32_v((int8x8_t)__rev0, 2);
40179		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	40198	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	40199	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vcvtn_s32_v(__builtin_bit_cast(int8x8_t, __rev0), 2));
	40200	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
40180	40201	return __ret;
40181	40202	}
40182	40203	#endif
...	...	@@ -40184,15 +40205,15 @@ __ai __attribute__((target("neon"))) int32x2_t vcvtn_s32_f32(float32x2_t __p0) {
40184	40205	#ifdef __LITTLE_ENDIAN__
40185	40206	__ai __attribute__((target("neon"))) uint32x4_t vcvtnq_u32_f32(float32x4_t __p0) {
40186	40207	uint32x4_t __ret;
40187		__ret = (uint32x4_t) __builtin_neon_vcvtnq_u32_v((int8x16_t)__p0, 50);
	40208	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcvtnq_u32_v(__builtin_bit_cast(int8x16_t, __p0), 50));
40188	40209	return __ret;
40189	40210	}
40190	40211	#else
40191	40212	__ai __attribute__((target("neon"))) uint32x4_t vcvtnq_u32_f32(float32x4_t __p0) {
40192	40213	uint32x4_t __ret;
40193		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40194		__ret = (uint32x4_t) __builtin_neon_vcvtnq_u32_v((int8x16_t)__rev0, 50);
40195		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40214	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	40215	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcvtnq_u32_v(__builtin_bit_cast(int8x16_t, __rev0), 50));
	40216	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
40196	40217	return __ret;
40197	40218	}
40198	40219	#endif
...	...	@@ -40200,15 +40221,15 @@ __ai __attribute__((target("neon"))) uint32x4_t vcvtnq_u32_f32(float32x4_t __p0)
40200	40221	#ifdef __LITTLE_ENDIAN__
40201	40222	__ai __attribute__((target("neon"))) uint32x2_t vcvtn_u32_f32(float32x2_t __p0) {
40202	40223	uint32x2_t __ret;
40203		__ret = (uint32x2_t) __builtin_neon_vcvtn_u32_v((int8x8_t)__p0, 18);
	40224	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcvtn_u32_v(__builtin_bit_cast(int8x8_t, __p0), 18));
40204	40225	return __ret;
40205	40226	}
40206	40227	#else
40207	40228	__ai __attribute__((target("neon"))) uint32x2_t vcvtn_u32_f32(float32x2_t __p0) {
40208	40229	uint32x2_t __ret;
40209		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
40210		__ret = (uint32x2_t) __builtin_neon_vcvtn_u32_v((int8x8_t)__rev0, 18);
40211		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	40230	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	40231	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcvtn_u32_v(__builtin_bit_cast(int8x8_t, __rev0), 18));
	40232	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
40212	40233	return __ret;
40213	40234	}
40214	40235	#endif
...	...	@@ -40216,15 +40237,15 @@ __ai __attribute__((target("neon"))) uint32x2_t vcvtn_u32_f32(float32x2_t __p0)
40216	40237	#ifdef __LITTLE_ENDIAN__
40217	40238	__ai __attribute__((target("neon"))) int32x4_t vcvtpq_s32_f32(float32x4_t __p0) {
40218	40239	int32x4_t __ret;
40219		__ret = (int32x4_t) __builtin_neon_vcvtpq_s32_v((int8x16_t)__p0, 34);
	40240	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vcvtpq_s32_v(__builtin_bit_cast(int8x16_t, __p0), 34));
40220	40241	return __ret;
40221	40242	}
40222	40243	#else
40223	40244	__ai __attribute__((target("neon"))) int32x4_t vcvtpq_s32_f32(float32x4_t __p0) {
40224	40245	int32x4_t __ret;
40225		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40226		__ret = (int32x4_t) __builtin_neon_vcvtpq_s32_v((int8x16_t)__rev0, 34);
40227		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40246	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	40247	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vcvtpq_s32_v(__builtin_bit_cast(int8x16_t, __rev0), 34));
	40248	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
40228	40249	return __ret;
40229	40250	}
40230	40251	#endif
...	...	@@ -40232,15 +40253,15 @@ __ai __attribute__((target("neon"))) int32x4_t vcvtpq_s32_f32(float32x4_t __p0)
40232	40253	#ifdef __LITTLE_ENDIAN__
40233	40254	__ai __attribute__((target("neon"))) int32x2_t vcvtp_s32_f32(float32x2_t __p0) {
40234	40255	int32x2_t __ret;
40235		__ret = (int32x2_t) __builtin_neon_vcvtp_s32_v((int8x8_t)__p0, 2);
	40256	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vcvtp_s32_v(__builtin_bit_cast(int8x8_t, __p0), 2));
40236	40257	return __ret;
40237	40258	}
40238	40259	#else
40239	40260	__ai __attribute__((target("neon"))) int32x2_t vcvtp_s32_f32(float32x2_t __p0) {
40240	40261	int32x2_t __ret;
40241		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
40242		__ret = (int32x2_t) __builtin_neon_vcvtp_s32_v((int8x8_t)__rev0, 2);
40243		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	40262	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	40263	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vcvtp_s32_v(__builtin_bit_cast(int8x8_t, __rev0), 2));
	40264	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
40244	40265	return __ret;
40245	40266	}
40246	40267	#endif
...	...	@@ -40248,15 +40269,15 @@ __ai __attribute__((target("neon"))) int32x2_t vcvtp_s32_f32(float32x2_t __p0) {
40248	40269	#ifdef __LITTLE_ENDIAN__
40249	40270	__ai __attribute__((target("neon"))) uint32x4_t vcvtpq_u32_f32(float32x4_t __p0) {
40250	40271	uint32x4_t __ret;
40251		__ret = (uint32x4_t) __builtin_neon_vcvtpq_u32_v((int8x16_t)__p0, 50);
	40272	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcvtpq_u32_v(__builtin_bit_cast(int8x16_t, __p0), 50));
40252	40273	return __ret;
40253	40274	}
40254	40275	#else
40255	40276	__ai __attribute__((target("neon"))) uint32x4_t vcvtpq_u32_f32(float32x4_t __p0) {
40256	40277	uint32x4_t __ret;
40257		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40258		__ret = (uint32x4_t) __builtin_neon_vcvtpq_u32_v((int8x16_t)__rev0, 50);
40259		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40278	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	40279	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcvtpq_u32_v(__builtin_bit_cast(int8x16_t, __rev0), 50));
	40280	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
40260	40281	return __ret;
40261	40282	}
40262	40283	#endif
...	...	@@ -40264,15 +40285,15 @@ __ai __attribute__((target("neon"))) uint32x4_t vcvtpq_u32_f32(float32x4_t __p0)
40264	40285	#ifdef __LITTLE_ENDIAN__
40265	40286	__ai __attribute__((target("neon"))) uint32x2_t vcvtp_u32_f32(float32x2_t __p0) {
40266	40287	uint32x2_t __ret;
40267		__ret = (uint32x2_t) __builtin_neon_vcvtp_u32_v((int8x8_t)__p0, 18);
	40288	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcvtp_u32_v(__builtin_bit_cast(int8x8_t, __p0), 18));
40268	40289	return __ret;
40269	40290	}
40270	40291	#else
40271	40292	__ai __attribute__((target("neon"))) uint32x2_t vcvtp_u32_f32(float32x2_t __p0) {
40272	40293	uint32x2_t __ret;
40273		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
40274		__ret = (uint32x2_t) __builtin_neon_vcvtp_u32_v((int8x8_t)__rev0, 18);
40275		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	40294	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	40295	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcvtp_u32_v(__builtin_bit_cast(int8x8_t, __rev0), 18));
	40296	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
40276	40297	return __ret;
40277	40298	}
40278	40299	#endif
...	...	@@ -40280,38 +40301,38 @@ __ai __attribute__((target("neon"))) uint32x2_t vcvtp_u32_f32(float32x2_t __p0)
40280	40301	#ifdef __LITTLE_ENDIAN__
40281	40302	__ai __attribute__((target("sha2,neon"))) uint32x4_t vsha1cq_u32(uint32x4_t __p0, uint32_t __p1, uint32x4_t __p2) {
40282	40303	uint32x4_t __ret;
40283		__ret = (uint32x4_t) __builtin_neon_vsha1cq_u32(__p0, __p1, __p2);
	40304	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsha1cq_u32(__p0, __p1, __p2));
40284	40305	return __ret;
40285	40306	}
40286	40307	#else
40287	40308	__ai __attribute__((target("sha2,neon"))) uint32x4_t vsha1cq_u32(uint32x4_t __p0, uint32_t __p1, uint32x4_t __p2) {
40288	40309	uint32x4_t __ret;
40289		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40290		uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
40291		__ret = (uint32x4_t) __builtin_neon_vsha1cq_u32(__rev0, __p1, __rev2);
40292		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40310	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	40311	uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
	40312	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsha1cq_u32(__rev0, __p1, __rev2));
	40313	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
40293	40314	return __ret;
40294	40315	}
40295	40316	#endif
40296	40317
40297	40318	__ai __attribute__((target("sha2,neon"))) uint32_t vsha1h_u32(uint32_t __p0) {
40298	40319	uint32_t __ret;
40299		__ret = (uint32_t) __builtin_neon_vsha1h_u32(__p0);
	40320	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vsha1h_u32(__p0));
40300	40321	return __ret;
40301	40322	}
40302	40323	#ifdef __LITTLE_ENDIAN__
40303	40324	__ai __attribute__((target("sha2,neon"))) uint32x4_t vsha1mq_u32(uint32x4_t __p0, uint32_t __p1, uint32x4_t __p2) {
40304	40325	uint32x4_t __ret;
40305		__ret = (uint32x4_t) __builtin_neon_vsha1mq_u32(__p0, __p1, __p2);
	40326	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsha1mq_u32(__p0, __p1, __p2));
40306	40327	return __ret;
40307	40328	}
40308	40329	#else
40309	40330	__ai __attribute__((target("sha2,neon"))) uint32x4_t vsha1mq_u32(uint32x4_t __p0, uint32_t __p1, uint32x4_t __p2) {
40310	40331	uint32x4_t __ret;
40311		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40312		uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
40313		__ret = (uint32x4_t) __builtin_neon_vsha1mq_u32(__rev0, __p1, __rev2);
40314		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40332	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	40333	uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
	40334	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsha1mq_u32(__rev0, __p1, __rev2));
	40335	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
40315	40336	return __ret;
40316	40337	}
40317	40338	#endif
...	...	@@ -40319,16 +40340,16 @@ __ai __attribute__((target("sha2,neon"))) uint32x4_t vsha1mq_u32(uint32x4_t __p0
40319	40340	#ifdef __LITTLE_ENDIAN__
40320	40341	__ai __attribute__((target("sha2,neon"))) uint32x4_t vsha1pq_u32(uint32x4_t __p0, uint32_t __p1, uint32x4_t __p2) {
40321	40342	uint32x4_t __ret;
40322		__ret = (uint32x4_t) __builtin_neon_vsha1pq_u32(__p0, __p1, __p2);
	40343	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsha1pq_u32(__p0, __p1, __p2));
40323	40344	return __ret;
40324	40345	}
40325	40346	#else
40326	40347	__ai __attribute__((target("sha2,neon"))) uint32x4_t vsha1pq_u32(uint32x4_t __p0, uint32_t __p1, uint32x4_t __p2) {
40327	40348	uint32x4_t __ret;
40328		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40329		uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
40330		__ret = (uint32x4_t) __builtin_neon_vsha1pq_u32(__rev0, __p1, __rev2);
40331		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40349	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	40350	uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
	40351	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsha1pq_u32(__rev0, __p1, __rev2));
	40352	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
40332	40353	return __ret;
40333	40354	}
40334	40355	#endif
...	...	@@ -40336,17 +40357,17 @@ __ai __attribute__((target("sha2,neon"))) uint32x4_t vsha1pq_u32(uint32x4_t __p0
40336	40357	#ifdef __LITTLE_ENDIAN__
40337	40358	__ai __attribute__((target("sha2,neon"))) uint32x4_t vsha1su0q_u32(uint32x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
40338	40359	uint32x4_t __ret;
40339		__ret = (uint32x4_t) __builtin_neon_vsha1su0q_u32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 50);
	40360	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsha1su0q_u32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 50));
40340	40361	return __ret;
40341	40362	}
40342	40363	#else
40343	40364	__ai __attribute__((target("sha2,neon"))) uint32x4_t vsha1su0q_u32(uint32x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
40344	40365	uint32x4_t __ret;
40345		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40346		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
40347		uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
40348		__ret = (uint32x4_t) __builtin_neon_vsha1su0q_u32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 50);
40349		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40366	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	40367	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	40368	uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
	40369	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsha1su0q_u32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 50));
	40370	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
40350	40371	return __ret;
40351	40372	}
40352	40373	#endif
...	...	@@ -40354,16 +40375,16 @@ __ai __attribute__((target("sha2,neon"))) uint32x4_t vsha1su0q_u32(uint32x4_t __
40354	40375	#ifdef __LITTLE_ENDIAN__
40355	40376	__ai __attribute__((target("sha2,neon"))) uint32x4_t vsha1su1q_u32(uint32x4_t __p0, uint32x4_t __p1) {
40356	40377	uint32x4_t __ret;
40357		__ret = (uint32x4_t) __builtin_neon_vsha1su1q_u32((int8x16_t)__p0, (int8x16_t)__p1, 50);
	40378	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsha1su1q_u32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50));
40358	40379	return __ret;
40359	40380	}
40360	40381	#else
40361	40382	__ai __attribute__((target("sha2,neon"))) uint32x4_t vsha1su1q_u32(uint32x4_t __p0, uint32x4_t __p1) {
40362	40383	uint32x4_t __ret;
40363		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40364		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
40365		__ret = (uint32x4_t) __builtin_neon_vsha1su1q_u32((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
40366		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40384	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	40385	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	40386	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsha1su1q_u32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 50));
	40387	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
40367	40388	return __ret;
40368	40389	}
40369	40390	#endif
...	...	@@ -40371,17 +40392,17 @@ __ai __attribute__((target("sha2,neon"))) uint32x4_t vsha1su1q_u32(uint32x4_t __
40371	40392	#ifdef __LITTLE_ENDIAN__
40372	40393	__ai __attribute__((target("sha2,neon"))) uint32x4_t vsha256hq_u32(uint32x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
40373	40394	uint32x4_t __ret;
40374		__ret = (uint32x4_t) __builtin_neon_vsha256hq_u32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 50);
	40395	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsha256hq_u32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 50));
40375	40396	return __ret;
40376	40397	}
40377	40398	#else
40378	40399	__ai __attribute__((target("sha2,neon"))) uint32x4_t vsha256hq_u32(uint32x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
40379	40400	uint32x4_t __ret;
40380		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40381		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
40382		uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
40383		__ret = (uint32x4_t) __builtin_neon_vsha256hq_u32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 50);
40384		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40401	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	40402	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	40403	uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
	40404	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsha256hq_u32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 50));
	40405	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
40385	40406	return __ret;
40386	40407	}
40387	40408	#endif
...	...	@@ -40389,17 +40410,17 @@ __ai __attribute__((target("sha2,neon"))) uint32x4_t vsha256hq_u32(uint32x4_t __
40389	40410	#ifdef __LITTLE_ENDIAN__
40390	40411	__ai __attribute__((target("sha2,neon"))) uint32x4_t vsha256h2q_u32(uint32x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
40391	40412	uint32x4_t __ret;
40392		__ret = (uint32x4_t) __builtin_neon_vsha256h2q_u32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 50);
	40413	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsha256h2q_u32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 50));
40393	40414	return __ret;
40394	40415	}
40395	40416	#else
40396	40417	__ai __attribute__((target("sha2,neon"))) uint32x4_t vsha256h2q_u32(uint32x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
40397	40418	uint32x4_t __ret;
40398		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40399		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
40400		uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
40401		__ret = (uint32x4_t) __builtin_neon_vsha256h2q_u32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 50);
40402		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40419	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	40420	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	40421	uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
	40422	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsha256h2q_u32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 50));
	40423	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
40403	40424	return __ret;
40404	40425	}
40405	40426	#endif
...	...	@@ -40407,16 +40428,16 @@ __ai __attribute__((target("sha2,neon"))) uint32x4_t vsha256h2q_u32(uint32x4_t _
40407	40428	#ifdef __LITTLE_ENDIAN__
40408	40429	__ai __attribute__((target("sha2,neon"))) uint32x4_t vsha256su0q_u32(uint32x4_t __p0, uint32x4_t __p1) {
40409	40430	uint32x4_t __ret;
40410		__ret = (uint32x4_t) __builtin_neon_vsha256su0q_u32((int8x16_t)__p0, (int8x16_t)__p1, 50);
	40431	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsha256su0q_u32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50));
40411	40432	return __ret;
40412	40433	}
40413	40434	#else
40414	40435	__ai __attribute__((target("sha2,neon"))) uint32x4_t vsha256su0q_u32(uint32x4_t __p0, uint32x4_t __p1) {
40415	40436	uint32x4_t __ret;
40416		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40417		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
40418		__ret = (uint32x4_t) __builtin_neon_vsha256su0q_u32((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
40419		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40437	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	40438	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	40439	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsha256su0q_u32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 50));
	40440	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
40420	40441	return __ret;
40421	40442	}
40422	40443	#endif
...	...	@@ -40424,17 +40445,17 @@ __ai __attribute__((target("sha2,neon"))) uint32x4_t vsha256su0q_u32(uint32x4_t
40424	40445	#ifdef __LITTLE_ENDIAN__
40425	40446	__ai __attribute__((target("sha2,neon"))) uint32x4_t vsha256su1q_u32(uint32x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
40426	40447	uint32x4_t __ret;
40427		__ret = (uint32x4_t) __builtin_neon_vsha256su1q_u32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 50);
	40448	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsha256su1q_u32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 50));
40428	40449	return __ret;
40429	40450	}
40430	40451	#else
40431	40452	__ai __attribute__((target("sha2,neon"))) uint32x4_t vsha256su1q_u32(uint32x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
40432	40453	uint32x4_t __ret;
40433		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40434		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
40435		uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
40436		__ret = (uint32x4_t) __builtin_neon_vsha256su1q_u32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 50);
40437		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40454	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	40455	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	40456	uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
	40457	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsha256su1q_u32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 50));
	40458	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
40438	40459	return __ret;
40439	40460	}
40440	40461	#endif
...	...	@@ -40444,15 +40465,15 @@ __ai __attribute__((target("sha2,neon"))) uint32x4_t vsha256su1q_u32(uint32x4_t
40444	40465	#ifdef __LITTLE_ENDIAN__
40445	40466	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vrndq_f16(float16x8_t __p0) {
40446	40467	float16x8_t __ret;
40447		__ret = (float16x8_t) __builtin_neon_vrndq_f16((int8x16_t)__p0, 40);
	40468	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vrndq_f16(__builtin_bit_cast(int8x16_t, __p0), 40));
40448	40469	return __ret;
40449	40470	}
40450	40471	#else
40451	40472	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vrndq_f16(float16x8_t __p0) {
40452	40473	float16x8_t __ret;
40453		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
40454		__ret = (float16x8_t) __builtin_neon_vrndq_f16((int8x16_t)__rev0, 40);
40455		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	40474	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	40475	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vrndq_f16(__builtin_bit_cast(int8x16_t, __rev0), 40));
	40476	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
40456	40477	return __ret;
40457	40478	}
40458	40479	#endif
...	...	@@ -40460,15 +40481,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vrndq_f16(float16x8_t
40460	40481	#ifdef __LITTLE_ENDIAN__
40461	40482	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vrnd_f16(float16x4_t __p0) {
40462	40483	float16x4_t __ret;
40463		__ret = (float16x4_t) __builtin_neon_vrnd_f16((int8x8_t)__p0, 8);
	40484	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vrnd_f16(__builtin_bit_cast(int8x8_t, __p0), 8));
40464	40485	return __ret;
40465	40486	}
40466	40487	#else
40467	40488	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vrnd_f16(float16x4_t __p0) {
40468	40489	float16x4_t __ret;
40469		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40470		__ret = (float16x4_t) __builtin_neon_vrnd_f16((int8x8_t)__rev0, 8);
40471		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40490	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	40491	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vrnd_f16(__builtin_bit_cast(int8x8_t, __rev0), 8));
	40492	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
40472	40493	return __ret;
40473	40494	}
40474	40495	#endif
...	...	@@ -40476,15 +40497,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vrnd_f16(float16x4_t _
40476	40497	#ifdef __LITTLE_ENDIAN__
40477	40498	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vrndaq_f16(float16x8_t __p0) {
40478	40499	float16x8_t __ret;
40479		__ret = (float16x8_t) __builtin_neon_vrndaq_f16((int8x16_t)__p0, 40);
	40500	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vrndaq_f16(__builtin_bit_cast(int8x16_t, __p0), 40));
40480	40501	return __ret;
40481	40502	}
40482	40503	#else
40483	40504	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vrndaq_f16(float16x8_t __p0) {
40484	40505	float16x8_t __ret;
40485		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
40486		__ret = (float16x8_t) __builtin_neon_vrndaq_f16((int8x16_t)__rev0, 40);
40487		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	40506	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	40507	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vrndaq_f16(__builtin_bit_cast(int8x16_t, __rev0), 40));
	40508	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
40488	40509	return __ret;
40489	40510	}
40490	40511	#endif
...	...	@@ -40492,15 +40513,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vrndaq_f16(float16x8_t
40492	40513	#ifdef __LITTLE_ENDIAN__
40493	40514	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vrnda_f16(float16x4_t __p0) {
40494	40515	float16x4_t __ret;
40495		__ret = (float16x4_t) __builtin_neon_vrnda_f16((int8x8_t)__p0, 8);
	40516	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vrnda_f16(__builtin_bit_cast(int8x8_t, __p0), 8));
40496	40517	return __ret;
40497	40518	}
40498	40519	#else
40499	40520	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vrnda_f16(float16x4_t __p0) {
40500	40521	float16x4_t __ret;
40501		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40502		__ret = (float16x4_t) __builtin_neon_vrnda_f16((int8x8_t)__rev0, 8);
40503		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40522	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	40523	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vrnda_f16(__builtin_bit_cast(int8x8_t, __rev0), 8));
	40524	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
40504	40525	return __ret;
40505	40526	}
40506	40527	#endif
...	...	@@ -40508,15 +40529,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vrnda_f16(float16x4_t
40508	40529	#ifdef __LITTLE_ENDIAN__
40509	40530	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vrndmq_f16(float16x8_t __p0) {
40510	40531	float16x8_t __ret;
40511		__ret = (float16x8_t) __builtin_neon_vrndmq_f16((int8x16_t)__p0, 40);
	40532	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vrndmq_f16(__builtin_bit_cast(int8x16_t, __p0), 40));
40512	40533	return __ret;
40513	40534	}
40514	40535	#else
40515	40536	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vrndmq_f16(float16x8_t __p0) {
40516	40537	float16x8_t __ret;
40517		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
40518		__ret = (float16x8_t) __builtin_neon_vrndmq_f16((int8x16_t)__rev0, 40);
40519		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	40538	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	40539	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vrndmq_f16(__builtin_bit_cast(int8x16_t, __rev0), 40));
	40540	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
40520	40541	return __ret;
40521	40542	}
40522	40543	#endif
...	...	@@ -40524,15 +40545,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vrndmq_f16(float16x8_t
40524	40545	#ifdef __LITTLE_ENDIAN__
40525	40546	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vrndm_f16(float16x4_t __p0) {
40526	40547	float16x4_t __ret;
40527		__ret = (float16x4_t) __builtin_neon_vrndm_f16((int8x8_t)__p0, 8);
	40548	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vrndm_f16(__builtin_bit_cast(int8x8_t, __p0), 8));
40528	40549	return __ret;
40529	40550	}
40530	40551	#else
40531	40552	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vrndm_f16(float16x4_t __p0) {
40532	40553	float16x4_t __ret;
40533		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40534		__ret = (float16x4_t) __builtin_neon_vrndm_f16((int8x8_t)__rev0, 8);
40535		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40554	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	40555	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vrndm_f16(__builtin_bit_cast(int8x8_t, __rev0), 8));
	40556	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
40536	40557	return __ret;
40537	40558	}
40538	40559	#endif
...	...	@@ -40540,15 +40561,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vrndm_f16(float16x4_t
40540	40561	#ifdef __LITTLE_ENDIAN__
40541	40562	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vrndnq_f16(float16x8_t __p0) {
40542	40563	float16x8_t __ret;
40543		__ret = (float16x8_t) __builtin_neon_vrndnq_f16((int8x16_t)__p0, 40);
	40564	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vrndnq_f16(__builtin_bit_cast(int8x16_t, __p0), 40));
40544	40565	return __ret;
40545	40566	}
40546	40567	#else
40547	40568	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vrndnq_f16(float16x8_t __p0) {
40548	40569	float16x8_t __ret;
40549		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
40550		__ret = (float16x8_t) __builtin_neon_vrndnq_f16((int8x16_t)__rev0, 40);
40551		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	40570	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	40571	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vrndnq_f16(__builtin_bit_cast(int8x16_t, __rev0), 40));
	40572	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
40552	40573	return __ret;
40553	40574	}
40554	40575	#endif
...	...	@@ -40556,15 +40577,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vrndnq_f16(float16x8_t
40556	40577	#ifdef __LITTLE_ENDIAN__
40557	40578	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vrndn_f16(float16x4_t __p0) {
40558	40579	float16x4_t __ret;
40559		__ret = (float16x4_t) __builtin_neon_vrndn_f16((int8x8_t)__p0, 8);
	40580	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vrndn_f16(__builtin_bit_cast(int8x8_t, __p0), 8));
40560	40581	return __ret;
40561	40582	}
40562	40583	#else
40563	40584	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vrndn_f16(float16x4_t __p0) {
40564	40585	float16x4_t __ret;
40565		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40566		__ret = (float16x4_t) __builtin_neon_vrndn_f16((int8x8_t)__rev0, 8);
40567		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40586	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	40587	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vrndn_f16(__builtin_bit_cast(int8x8_t, __rev0), 8));
	40588	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
40568	40589	return __ret;
40569	40590	}
40570	40591	#endif
...	...	@@ -40572,15 +40593,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vrndn_f16(float16x4_t
40572	40593	#ifdef __LITTLE_ENDIAN__
40573	40594	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vrndpq_f16(float16x8_t __p0) {
40574	40595	float16x8_t __ret;
40575		__ret = (float16x8_t) __builtin_neon_vrndpq_f16((int8x16_t)__p0, 40);
	40596	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vrndpq_f16(__builtin_bit_cast(int8x16_t, __p0), 40));
40576	40597	return __ret;
40577	40598	}
40578	40599	#else
40579	40600	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vrndpq_f16(float16x8_t __p0) {
40580	40601	float16x8_t __ret;
40581		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
40582		__ret = (float16x8_t) __builtin_neon_vrndpq_f16((int8x16_t)__rev0, 40);
40583		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	40602	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	40603	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vrndpq_f16(__builtin_bit_cast(int8x16_t, __rev0), 40));
	40604	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
40584	40605	return __ret;
40585	40606	}
40586	40607	#endif
...	...	@@ -40588,15 +40609,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vrndpq_f16(float16x8_t
40588	40609	#ifdef __LITTLE_ENDIAN__
40589	40610	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vrndp_f16(float16x4_t __p0) {
40590	40611	float16x4_t __ret;
40591		__ret = (float16x4_t) __builtin_neon_vrndp_f16((int8x8_t)__p0, 8);
	40612	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vrndp_f16(__builtin_bit_cast(int8x8_t, __p0), 8));
40592	40613	return __ret;
40593	40614	}
40594	40615	#else
40595	40616	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vrndp_f16(float16x4_t __p0) {
40596	40617	float16x4_t __ret;
40597		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40598		__ret = (float16x4_t) __builtin_neon_vrndp_f16((int8x8_t)__rev0, 8);
40599		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40618	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	40619	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vrndp_f16(__builtin_bit_cast(int8x8_t, __rev0), 8));
	40620	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
40600	40621	return __ret;
40601	40622	}
40602	40623	#endif
...	...	@@ -40604,15 +40625,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vrndp_f16(float16x4_t
40604	40625	#ifdef __LITTLE_ENDIAN__
40605	40626	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vrndxq_f16(float16x8_t __p0) {
40606	40627	float16x8_t __ret;
40607		__ret = (float16x8_t) __builtin_neon_vrndxq_f16((int8x16_t)__p0, 40);
	40628	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vrndxq_f16(__builtin_bit_cast(int8x16_t, __p0), 40));
40608	40629	return __ret;
40609	40630	}
40610	40631	#else
40611	40632	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vrndxq_f16(float16x8_t __p0) {
40612	40633	float16x8_t __ret;
40613		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
40614		__ret = (float16x8_t) __builtin_neon_vrndxq_f16((int8x16_t)__rev0, 40);
40615		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	40634	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	40635	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vrndxq_f16(__builtin_bit_cast(int8x16_t, __rev0), 40));
	40636	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
40616	40637	return __ret;
40617	40638	}
40618	40639	#endif
...	...	@@ -40620,15 +40641,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vrndxq_f16(float16x8_t
40620	40641	#ifdef __LITTLE_ENDIAN__
40621	40642	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vrndx_f16(float16x4_t __p0) {
40622	40643	float16x4_t __ret;
40623		__ret = (float16x4_t) __builtin_neon_vrndx_f16((int8x8_t)__p0, 8);
	40644	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vrndx_f16(__builtin_bit_cast(int8x8_t, __p0), 8));
40624	40645	return __ret;
40625	40646	}
40626	40647	#else
40627	40648	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vrndx_f16(float16x4_t __p0) {
40628	40649	float16x4_t __ret;
40629		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40630		__ret = (float16x4_t) __builtin_neon_vrndx_f16((int8x8_t)__rev0, 8);
40631		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40650	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	40651	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vrndx_f16(__builtin_bit_cast(int8x8_t, __rev0), 8));
	40652	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
40632	40653	return __ret;
40633	40654	}
40634	40655	#endif
...	...	@@ -40636,15 +40657,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vrndx_f16(float16x4_t
40636	40657	#ifdef __LITTLE_ENDIAN__
40637	40658	__ai __attribute__((target("neon"))) float32x4_t vrndq_f32(float32x4_t __p0) {
40638	40659	float32x4_t __ret;
40639		__ret = (float32x4_t) __builtin_neon_vrndq_v((int8x16_t)__p0, 41);
	40660	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vrndq_v(__builtin_bit_cast(int8x16_t, __p0), 41));
40640	40661	return __ret;
40641	40662	}
40642	40663	#else
40643	40664	__ai __attribute__((target("neon"))) float32x4_t vrndq_f32(float32x4_t __p0) {
40644	40665	float32x4_t __ret;
40645		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40646		__ret = (float32x4_t) __builtin_neon_vrndq_v((int8x16_t)__rev0, 41);
40647		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40666	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	40667	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vrndq_v(__builtin_bit_cast(int8x16_t, __rev0), 41));
	40668	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
40648	40669	return __ret;
40649	40670	}
40650	40671	#endif
...	...	@@ -40652,15 +40673,15 @@ __ai __attribute__((target("neon"))) float32x4_t vrndq_f32(float32x4_t __p0) {
40652	40673	#ifdef __LITTLE_ENDIAN__
40653	40674	__ai __attribute__((target("neon"))) float32x2_t vrnd_f32(float32x2_t __p0) {
40654	40675	float32x2_t __ret;
40655		__ret = (float32x2_t) __builtin_neon_vrnd_v((int8x8_t)__p0, 9);
	40676	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vrnd_v(__builtin_bit_cast(int8x8_t, __p0), 9));
40656	40677	return __ret;
40657	40678	}
40658	40679	#else
40659	40680	__ai __attribute__((target("neon"))) float32x2_t vrnd_f32(float32x2_t __p0) {
40660	40681	float32x2_t __ret;
40661		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
40662		__ret = (float32x2_t) __builtin_neon_vrnd_v((int8x8_t)__rev0, 9);
40663		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	40682	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	40683	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vrnd_v(__builtin_bit_cast(int8x8_t, __rev0), 9));
	40684	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
40664	40685	return __ret;
40665	40686	}
40666	40687	#endif
...	...	@@ -40668,15 +40689,15 @@ __ai __attribute__((target("neon"))) float32x2_t vrnd_f32(float32x2_t __p0) {
40668	40689	#ifdef __LITTLE_ENDIAN__
40669	40690	__ai __attribute__((target("neon"))) float32x4_t vrndaq_f32(float32x4_t __p0) {
40670	40691	float32x4_t __ret;
40671		__ret = (float32x4_t) __builtin_neon_vrndaq_v((int8x16_t)__p0, 41);
	40692	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vrndaq_v(__builtin_bit_cast(int8x16_t, __p0), 41));
40672	40693	return __ret;
40673	40694	}
40674	40695	#else
40675	40696	__ai __attribute__((target("neon"))) float32x4_t vrndaq_f32(float32x4_t __p0) {
40676	40697	float32x4_t __ret;
40677		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40678		__ret = (float32x4_t) __builtin_neon_vrndaq_v((int8x16_t)__rev0, 41);
40679		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40698	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	40699	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vrndaq_v(__builtin_bit_cast(int8x16_t, __rev0), 41));
	40700	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
40680	40701	return __ret;
40681	40702	}
40682	40703	#endif
...	...	@@ -40684,15 +40705,15 @@ __ai __attribute__((target("neon"))) float32x4_t vrndaq_f32(float32x4_t __p0) {
40684	40705	#ifdef __LITTLE_ENDIAN__
40685	40706	__ai __attribute__((target("neon"))) float32x2_t vrnda_f32(float32x2_t __p0) {
40686	40707	float32x2_t __ret;
40687		__ret = (float32x2_t) __builtin_neon_vrnda_v((int8x8_t)__p0, 9);
	40708	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vrnda_v(__builtin_bit_cast(int8x8_t, __p0), 9));
40688	40709	return __ret;
40689	40710	}
40690	40711	#else
40691	40712	__ai __attribute__((target("neon"))) float32x2_t vrnda_f32(float32x2_t __p0) {
40692	40713	float32x2_t __ret;
40693		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
40694		__ret = (float32x2_t) __builtin_neon_vrnda_v((int8x8_t)__rev0, 9);
40695		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	40714	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	40715	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vrnda_v(__builtin_bit_cast(int8x8_t, __rev0), 9));
	40716	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
40696	40717	return __ret;
40697	40718	}
40698	40719	#endif
...	...	@@ -40700,15 +40721,15 @@ __ai __attribute__((target("neon"))) float32x2_t vrnda_f32(float32x2_t __p0) {
40700	40721	#ifdef __LITTLE_ENDIAN__
40701	40722	__ai __attribute__((target("neon"))) float32x4_t vrndiq_f32(float32x4_t __p0) {
40702	40723	float32x4_t __ret;
40703		__ret = (float32x4_t) __builtin_neon_vrndiq_v((int8x16_t)__p0, 41);
	40724	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vrndiq_v(__builtin_bit_cast(int8x16_t, __p0), 41));
40704	40725	return __ret;
40705	40726	}
40706	40727	#else
40707	40728	__ai __attribute__((target("neon"))) float32x4_t vrndiq_f32(float32x4_t __p0) {
40708	40729	float32x4_t __ret;
40709		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40710		__ret = (float32x4_t) __builtin_neon_vrndiq_v((int8x16_t)__rev0, 41);
40711		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40730	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	40731	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vrndiq_v(__builtin_bit_cast(int8x16_t, __rev0), 41));
	40732	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
40712	40733	return __ret;
40713	40734	}
40714	40735	#endif
...	...	@@ -40716,15 +40737,15 @@ __ai __attribute__((target("neon"))) float32x4_t vrndiq_f32(float32x4_t __p0) {
40716	40737	#ifdef __LITTLE_ENDIAN__
40717	40738	__ai __attribute__((target("neon"))) float32x2_t vrndi_f32(float32x2_t __p0) {
40718	40739	float32x2_t __ret;
40719		__ret = (float32x2_t) __builtin_neon_vrndi_v((int8x8_t)__p0, 9);
	40740	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vrndi_v(__builtin_bit_cast(int8x8_t, __p0), 9));
40720	40741	return __ret;
40721	40742	}
40722	40743	#else
40723	40744	__ai __attribute__((target("neon"))) float32x2_t vrndi_f32(float32x2_t __p0) {
40724	40745	float32x2_t __ret;
40725		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
40726		__ret = (float32x2_t) __builtin_neon_vrndi_v((int8x8_t)__rev0, 9);
40727		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	40746	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	40747	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vrndi_v(__builtin_bit_cast(int8x8_t, __rev0), 9));
	40748	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
40728	40749	return __ret;
40729	40750	}
40730	40751	#endif
...	...	@@ -40732,15 +40753,15 @@ __ai __attribute__((target("neon"))) float32x2_t vrndi_f32(float32x2_t __p0) {
40732	40753	#ifdef __LITTLE_ENDIAN__
40733	40754	__ai __attribute__((target("neon"))) float32x4_t vrndmq_f32(float32x4_t __p0) {
40734	40755	float32x4_t __ret;
40735		__ret = (float32x4_t) __builtin_neon_vrndmq_v((int8x16_t)__p0, 41);
	40756	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vrndmq_v(__builtin_bit_cast(int8x16_t, __p0), 41));
40736	40757	return __ret;
40737	40758	}
40738	40759	#else
40739	40760	__ai __attribute__((target("neon"))) float32x4_t vrndmq_f32(float32x4_t __p0) {
40740	40761	float32x4_t __ret;
40741		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40742		__ret = (float32x4_t) __builtin_neon_vrndmq_v((int8x16_t)__rev0, 41);
40743		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40762	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	40763	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vrndmq_v(__builtin_bit_cast(int8x16_t, __rev0), 41));
	40764	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
40744	40765	return __ret;
40745	40766	}
40746	40767	#endif
...	...	@@ -40748,15 +40769,15 @@ __ai __attribute__((target("neon"))) float32x4_t vrndmq_f32(float32x4_t __p0) {
40748	40769	#ifdef __LITTLE_ENDIAN__
40749	40770	__ai __attribute__((target("neon"))) float32x2_t vrndm_f32(float32x2_t __p0) {
40750	40771	float32x2_t __ret;
40751		__ret = (float32x2_t) __builtin_neon_vrndm_v((int8x8_t)__p0, 9);
	40772	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vrndm_v(__builtin_bit_cast(int8x8_t, __p0), 9));
40752	40773	return __ret;
40753	40774	}
40754	40775	#else
40755	40776	__ai __attribute__((target("neon"))) float32x2_t vrndm_f32(float32x2_t __p0) {
40756	40777	float32x2_t __ret;
40757		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
40758		__ret = (float32x2_t) __builtin_neon_vrndm_v((int8x8_t)__rev0, 9);
40759		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	40778	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	40779	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vrndm_v(__builtin_bit_cast(int8x8_t, __rev0), 9));
	40780	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
40760	40781	return __ret;
40761	40782	}
40762	40783	#endif
...	...	@@ -40764,15 +40785,15 @@ __ai __attribute__((target("neon"))) float32x2_t vrndm_f32(float32x2_t __p0) {
40764	40785	#ifdef __LITTLE_ENDIAN__
40765	40786	__ai __attribute__((target("neon"))) float32x4_t vrndnq_f32(float32x4_t __p0) {
40766	40787	float32x4_t __ret;
40767		__ret = (float32x4_t) __builtin_neon_vrndnq_v((int8x16_t)__p0, 41);
	40788	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vrndnq_v(__builtin_bit_cast(int8x16_t, __p0), 41));
40768	40789	return __ret;
40769	40790	}
40770	40791	#else
40771	40792	__ai __attribute__((target("neon"))) float32x4_t vrndnq_f32(float32x4_t __p0) {
40772	40793	float32x4_t __ret;
40773		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40774		__ret = (float32x4_t) __builtin_neon_vrndnq_v((int8x16_t)__rev0, 41);
40775		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40794	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	40795	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vrndnq_v(__builtin_bit_cast(int8x16_t, __rev0), 41));
	40796	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
40776	40797	return __ret;
40777	40798	}
40778	40799	#endif
...	...	@@ -40780,36 +40801,36 @@ __ai __attribute__((target("neon"))) float32x4_t vrndnq_f32(float32x4_t __p0) {
40780	40801	#ifdef __LITTLE_ENDIAN__
40781	40802	__ai __attribute__((target("neon"))) float32x2_t vrndn_f32(float32x2_t __p0) {
40782	40803	float32x2_t __ret;
40783		__ret = (float32x2_t) __builtin_neon_vrndn_v((int8x8_t)__p0, 9);
	40804	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vrndn_v(__builtin_bit_cast(int8x8_t, __p0), 9));
40784	40805	return __ret;
40785	40806	}
40786	40807	#else
40787	40808	__ai __attribute__((target("neon"))) float32x2_t vrndn_f32(float32x2_t __p0) {
40788	40809	float32x2_t __ret;
40789		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
40790		__ret = (float32x2_t) __builtin_neon_vrndn_v((int8x8_t)__rev0, 9);
40791		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	40810	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	40811	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vrndn_v(__builtin_bit_cast(int8x8_t, __rev0), 9));
	40812	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
40792	40813	return __ret;
40793	40814	}
40794	40815	#endif
40795	40816
40796	40817	__ai __attribute__((target("neon"))) float32_t vrndns_f32(float32_t __p0) {
40797	40818	float32_t __ret;
40798		__ret = (float32_t) __builtin_neon_vrndns_f32(__p0);
	40819	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vrndns_f32(__p0));
40799	40820	return __ret;
40800	40821	}
40801	40822	#ifdef __LITTLE_ENDIAN__
40802	40823	__ai __attribute__((target("neon"))) float32x4_t vrndpq_f32(float32x4_t __p0) {
40803	40824	float32x4_t __ret;
40804		__ret = (float32x4_t) __builtin_neon_vrndpq_v((int8x16_t)__p0, 41);
	40825	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vrndpq_v(__builtin_bit_cast(int8x16_t, __p0), 41));
40805	40826	return __ret;
40806	40827	}
40807	40828	#else
40808	40829	__ai __attribute__((target("neon"))) float32x4_t vrndpq_f32(float32x4_t __p0) {
40809	40830	float32x4_t __ret;
40810		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40811		__ret = (float32x4_t) __builtin_neon_vrndpq_v((int8x16_t)__rev0, 41);
40812		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40831	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	40832	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vrndpq_v(__builtin_bit_cast(int8x16_t, __rev0), 41));
	40833	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
40813	40834	return __ret;
40814	40835	}
40815	40836	#endif
...	...	@@ -40817,15 +40838,15 @@ __ai __attribute__((target("neon"))) float32x4_t vrndpq_f32(float32x4_t __p0) {
40817	40838	#ifdef __LITTLE_ENDIAN__
40818	40839	__ai __attribute__((target("neon"))) float32x2_t vrndp_f32(float32x2_t __p0) {
40819	40840	float32x2_t __ret;
40820		__ret = (float32x2_t) __builtin_neon_vrndp_v((int8x8_t)__p0, 9);
	40841	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vrndp_v(__builtin_bit_cast(int8x8_t, __p0), 9));
40821	40842	return __ret;
40822	40843	}
40823	40844	#else
40824	40845	__ai __attribute__((target("neon"))) float32x2_t vrndp_f32(float32x2_t __p0) {
40825	40846	float32x2_t __ret;
40826		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
40827		__ret = (float32x2_t) __builtin_neon_vrndp_v((int8x8_t)__rev0, 9);
40828		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	40847	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	40848	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vrndp_v(__builtin_bit_cast(int8x8_t, __rev0), 9));
	40849	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
40829	40850	return __ret;
40830	40851	}
40831	40852	#endif
...	...	@@ -40833,15 +40854,15 @@ __ai __attribute__((target("neon"))) float32x2_t vrndp_f32(float32x2_t __p0) {
40833	40854	#ifdef __LITTLE_ENDIAN__
40834	40855	__ai __attribute__((target("neon"))) float32x4_t vrndxq_f32(float32x4_t __p0) {
40835	40856	float32x4_t __ret;
40836		__ret = (float32x4_t) __builtin_neon_vrndxq_v((int8x16_t)__p0, 41);
	40857	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vrndxq_v(__builtin_bit_cast(int8x16_t, __p0), 41));
40837	40858	return __ret;
40838	40859	}
40839	40860	#else
40840	40861	__ai __attribute__((target("neon"))) float32x4_t vrndxq_f32(float32x4_t __p0) {
40841	40862	float32x4_t __ret;
40842		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40843		__ret = (float32x4_t) __builtin_neon_vrndxq_v((int8x16_t)__rev0, 41);
40844		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40863	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	40864	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vrndxq_v(__builtin_bit_cast(int8x16_t, __rev0), 41));
	40865	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
40845	40866	return __ret;
40846	40867	}
40847	40868	#endif
...	...	@@ -40849,15 +40870,15 @@ __ai __attribute__((target("neon"))) float32x4_t vrndxq_f32(float32x4_t __p0) {
40849	40870	#ifdef __LITTLE_ENDIAN__
40850	40871	__ai __attribute__((target("neon"))) float32x2_t vrndx_f32(float32x2_t __p0) {
40851	40872	float32x2_t __ret;
40852		__ret = (float32x2_t) __builtin_neon_vrndx_v((int8x8_t)__p0, 9);
	40873	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vrndx_v(__builtin_bit_cast(int8x8_t, __p0), 9));
40853	40874	return __ret;
40854	40875	}
40855	40876	#else
40856	40877	__ai __attribute__((target("neon"))) float32x2_t vrndx_f32(float32x2_t __p0) {
40857	40878	float32x2_t __ret;
40858		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
40859		__ret = (float32x2_t) __builtin_neon_vrndx_v((int8x8_t)__rev0, 9);
40860		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	40879	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	40880	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vrndx_v(__builtin_bit_cast(int8x8_t, __rev0), 9));
	40881	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
40861	40882	return __ret;
40862	40883	}
40863	40884	#endif
...	...	@@ -40867,16 +40888,16 @@ __ai __attribute__((target("neon"))) float32x2_t vrndx_f32(float32x2_t __p0) {
40867	40888	#ifdef __LITTLE_ENDIAN__
40868	40889	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vmaxnmq_f16(float16x8_t __p0, float16x8_t __p1) {
40869	40890	float16x8_t __ret;
40870		__ret = (float16x8_t) __builtin_neon_vmaxnmq_f16((int8x16_t)__p0, (int8x16_t)__p1, 40);
	40891	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vmaxnmq_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 40));
40871	40892	return __ret;
40872	40893	}
40873	40894	#else
40874	40895	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vmaxnmq_f16(float16x8_t __p0, float16x8_t __p1) {
40875	40896	float16x8_t __ret;
40876		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
40877		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
40878		__ret = (float16x8_t) __builtin_neon_vmaxnmq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 40);
40879		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	40897	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	40898	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	40899	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vmaxnmq_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 40));
	40900	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
40880	40901	return __ret;
40881	40902	}
40882	40903	#endif
...	...	@@ -40884,16 +40905,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vmaxnmq_f16(float16x8_
40884	40905	#ifdef __LITTLE_ENDIAN__
40885	40906	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vmaxnm_f16(float16x4_t __p0, float16x4_t __p1) {
40886	40907	float16x4_t __ret;
40887		__ret = (float16x4_t) __builtin_neon_vmaxnm_f16((int8x8_t)__p0, (int8x8_t)__p1, 8);
	40908	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vmaxnm_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 8));
40888	40909	return __ret;
40889	40910	}
40890	40911	#else
40891	40912	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vmaxnm_f16(float16x4_t __p0, float16x4_t __p1) {
40892	40913	float16x4_t __ret;
40893		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40894		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
40895		__ret = (float16x4_t) __builtin_neon_vmaxnm_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 8);
40896		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40914	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	40915	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	40916	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vmaxnm_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 8));
	40917	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
40897	40918	return __ret;
40898	40919	}
40899	40920	#endif
...	...	@@ -40901,16 +40922,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vmaxnm_f16(float16x4_t
40901	40922	#ifdef __LITTLE_ENDIAN__
40902	40923	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vminnmq_f16(float16x8_t __p0, float16x8_t __p1) {
40903	40924	float16x8_t __ret;
40904		__ret = (float16x8_t) __builtin_neon_vminnmq_f16((int8x16_t)__p0, (int8x16_t)__p1, 40);
	40925	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vminnmq_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 40));
40905	40926	return __ret;
40906	40927	}
40907	40928	#else
40908	40929	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vminnmq_f16(float16x8_t __p0, float16x8_t __p1) {
40909	40930	float16x8_t __ret;
40910		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
40911		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
40912		__ret = (float16x8_t) __builtin_neon_vminnmq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 40);
40913		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	40931	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	40932	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	40933	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vminnmq_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 40));
	40934	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
40914	40935	return __ret;
40915	40936	}
40916	40937	#endif
...	...	@@ -40918,16 +40939,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vminnmq_f16(float16x8_
40918	40939	#ifdef __LITTLE_ENDIAN__
40919	40940	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vminnm_f16(float16x4_t __p0, float16x4_t __p1) {
40920	40941	float16x4_t __ret;
40921		__ret = (float16x4_t) __builtin_neon_vminnm_f16((int8x8_t)__p0, (int8x8_t)__p1, 8);
	40942	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vminnm_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 8));
40922	40943	return __ret;
40923	40944	}
40924	40945	#else
40925	40946	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vminnm_f16(float16x4_t __p0, float16x4_t __p1) {
40926	40947	float16x4_t __ret;
40927		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40928		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
40929		__ret = (float16x4_t) __builtin_neon_vminnm_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 8);
40930		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40948	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	40949	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	40950	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vminnm_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 8));
	40951	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
40931	40952	return __ret;
40932	40953	}
40933	40954	#endif
...	...	@@ -40935,16 +40956,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vminnm_f16(float16x4_t
40935	40956	#ifdef __LITTLE_ENDIAN__
40936	40957	__ai __attribute__((target("neon"))) float32x4_t vmaxnmq_f32(float32x4_t __p0, float32x4_t __p1) {
40937	40958	float32x4_t __ret;
40938		__ret = (float32x4_t) __builtin_neon_vmaxnmq_v((int8x16_t)__p0, (int8x16_t)__p1, 41);
	40959	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmaxnmq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 41));
40939	40960	return __ret;
40940	40961	}
40941	40962	#else
40942	40963	__ai __attribute__((target("neon"))) float32x4_t vmaxnmq_f32(float32x4_t __p0, float32x4_t __p1) {
40943	40964	float32x4_t __ret;
40944		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40945		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
40946		__ret = (float32x4_t) __builtin_neon_vmaxnmq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 41);
40947		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40965	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	40966	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	40967	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmaxnmq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 41));
	40968	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
40948	40969	return __ret;
40949	40970	}
40950	40971	#endif
...	...	@@ -40952,16 +40973,16 @@ __ai __attribute__((target("neon"))) float32x4_t vmaxnmq_f32(float32x4_t __p0, f
40952	40973	#ifdef __LITTLE_ENDIAN__
40953	40974	__ai __attribute__((target("neon"))) float32x2_t vmaxnm_f32(float32x2_t __p0, float32x2_t __p1) {
40954	40975	float32x2_t __ret;
40955		__ret = (float32x2_t) __builtin_neon_vmaxnm_v((int8x8_t)__p0, (int8x8_t)__p1, 9);
	40976	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vmaxnm_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 9));
40956	40977	return __ret;
40957	40978	}
40958	40979	#else
40959	40980	__ai __attribute__((target("neon"))) float32x2_t vmaxnm_f32(float32x2_t __p0, float32x2_t __p1) {
40960	40981	float32x2_t __ret;
40961		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
40962		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
40963		__ret = (float32x2_t) __builtin_neon_vmaxnm_v((int8x8_t)__rev0, (int8x8_t)__rev1, 9);
40964		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	40982	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	40983	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	40984	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vmaxnm_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 9));
	40985	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
40965	40986	return __ret;
40966	40987	}
40967	40988	#endif
...	...	@@ -40969,16 +40990,16 @@ __ai __attribute__((target("neon"))) float32x2_t vmaxnm_f32(float32x2_t __p0, fl
40969	40990	#ifdef __LITTLE_ENDIAN__
40970	40991	__ai __attribute__((target("neon"))) float32x4_t vminnmq_f32(float32x4_t __p0, float32x4_t __p1) {
40971	40992	float32x4_t __ret;
40972		__ret = (float32x4_t) __builtin_neon_vminnmq_v((int8x16_t)__p0, (int8x16_t)__p1, 41);
	40993	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vminnmq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 41));
40973	40994	return __ret;
40974	40995	}
40975	40996	#else
40976	40997	__ai __attribute__((target("neon"))) float32x4_t vminnmq_f32(float32x4_t __p0, float32x4_t __p1) {
40977	40998	float32x4_t __ret;
40978		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
40979		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
40980		__ret = (float32x4_t) __builtin_neon_vminnmq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 41);
40981		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	40999	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	41000	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	41001	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vminnmq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 41));
	41002	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
40982	41003	return __ret;
40983	41004	}
40984	41005	#endif
...	...	@@ -40986,16 +41007,16 @@ __ai __attribute__((target("neon"))) float32x4_t vminnmq_f32(float32x4_t __p0, f
40986	41007	#ifdef __LITTLE_ENDIAN__
40987	41008	__ai __attribute__((target("neon"))) float32x2_t vminnm_f32(float32x2_t __p0, float32x2_t __p1) {
40988	41009	float32x2_t __ret;
40989		__ret = (float32x2_t) __builtin_neon_vminnm_v((int8x8_t)__p0, (int8x8_t)__p1, 9);
	41010	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vminnm_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 9));
40990	41011	return __ret;
40991	41012	}
40992	41013	#else
40993	41014	__ai __attribute__((target("neon"))) float32x2_t vminnm_f32(float32x2_t __p0, float32x2_t __p1) {
40994	41015	float32x2_t __ret;
40995		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
40996		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
40997		__ret = (float32x2_t) __builtin_neon_vminnm_v((int8x8_t)__rev0, (int8x8_t)__rev1, 9);
40998		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	41016	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	41017	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	41018	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vminnm_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 9));
	41019	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
40999	41020	return __ret;
41000	41021	}
41001	41022	#endif
...	...	@@ -41005,22 +41026,22 @@ __ai __attribute__((target("neon"))) float32x2_t vminnm_f32(float32x2_t __p0, fl
41005	41026	#ifdef __LITTLE_ENDIAN__
41006	41027	__ai __attribute__((target("neon"))) float32x4_t vfmaq_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
41007	41028	float32x4_t __ret;
41008		__ret = (float32x4_t) __builtin_neon_vfmaq_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
	41029	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vfmaq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 41));
41009	41030	return __ret;
41010	41031	}
41011	41032	#else
41012	41033	__ai __attribute__((target("neon"))) float32x4_t vfmaq_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
41013	41034	float32x4_t __ret;
41014		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
41015		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
41016		float32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
41017		__ret = (float32x4_t) __builtin_neon_vfmaq_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 41);
41018		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	41035	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	41036	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	41037	float32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
	41038	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vfmaq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 41));
	41039	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
41019	41040	return __ret;
41020	41041	}
41021	41042	__ai __attribute__((target("neon"))) float32x4_t __noswap_vfmaq_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
41022	41043	float32x4_t __ret;
41023		__ret = (float32x4_t) __builtin_neon_vfmaq_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
	41044	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vfmaq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 41));
41024	41045	return __ret;
41025	41046	}
41026	41047	#endif
...	...	@@ -41028,22 +41049,22 @@ __ai __attribute__((target("neon"))) float32x4_t __noswap_vfmaq_f32(float32x4_t
41028	41049	#ifdef __LITTLE_ENDIAN__
41029	41050	__ai __attribute__((target("neon"))) float32x2_t vfma_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
41030	41051	float32x2_t __ret;
41031		__ret = (float32x2_t) __builtin_neon_vfma_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
	41052	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vfma_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 9));
41032	41053	return __ret;
41033	41054	}
41034	41055	#else
41035	41056	__ai __attribute__((target("neon"))) float32x2_t vfma_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
41036	41057	float32x2_t __ret;
41037		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
41038		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
41039		float32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
41040		__ret = (float32x2_t) __builtin_neon_vfma_v((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 9);
41041		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	41058	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	41059	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	41060	float32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_32);
	41061	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vfma_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 9));
	41062	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
41042	41063	return __ret;
41043	41064	}
41044	41065	__ai __attribute__((target("neon"))) float32x2_t __noswap_vfma_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
41045	41066	float32x2_t __ret;
41046		__ret = (float32x2_t) __builtin_neon_vfma_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
	41067	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vfma_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 9));
41047	41068	return __ret;
41048	41069	}
41049	41070	#endif
...	...	@@ -41057,10 +41078,10 @@ __ai __attribute__((target("neon"))) float32x4_t vfmaq_n_f32(float32x4_t __p0, f
41057	41078	#else
41058	41079	__ai __attribute__((target("neon"))) float32x4_t vfmaq_n_f32(float32x4_t __p0, float32x4_t __p1, float32_t __p2) {
41059	41080	float32x4_t __ret;
41060		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
41061		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	41081	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	41082	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
41062	41083	__ret = __noswap_vfmaq_f32(__rev0, __rev1, (float32x4_t) {__p2, __p2, __p2, __p2});
41063		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	41084	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
41064	41085	return __ret;
41065	41086	}
41066	41087	#endif
...	...	@@ -41074,10 +41095,10 @@ __ai __attribute__((target("neon"))) float32x2_t vfma_n_f32(float32x2_t __p0, fl
41074	41095	#else
41075	41096	__ai __attribute__((target("neon"))) float32x2_t vfma_n_f32(float32x2_t __p0, float32x2_t __p1, float32_t __p2) {
41076	41097	float32x2_t __ret;
41077		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
41078		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	41098	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	41099	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
41079	41100	__ret = __noswap_vfma_f32(__rev0, __rev1, (float32x2_t) {__p2, __p2});
41080		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	41101	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
41081	41102	return __ret;
41082	41103	}
41083	41104	#endif
...	...	@@ -41091,11 +41112,11 @@ __ai __attribute__((target("neon"))) float32x4_t vfmsq_f32(float32x4_t __p0, flo
41091	41112	#else
41092	41113	__ai __attribute__((target("neon"))) float32x4_t vfmsq_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
41093	41114	float32x4_t __ret;
41094		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
41095		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
41096		float32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	41115	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	41116	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	41117	float32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
41097	41118	__ret = __noswap_vfmaq_f32(__rev0, -__rev1, __rev2);
41098		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	41119	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
41099	41120	return __ret;
41100	41121	}
41101	41122	#endif
...	...	@@ -41109,11 +41130,11 @@ __ai __attribute__((target("neon"))) float32x2_t vfms_f32(float32x2_t __p0, floa
41109	41130	#else
41110	41131	__ai __attribute__((target("neon"))) float32x2_t vfms_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
41111	41132	float32x2_t __ret;
41112		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
41113		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
41114		float32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
	41133	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	41134	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	41135	float32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_32);
41115	41136	__ret = __noswap_vfma_f32(__rev0, -__rev1, __rev2);
41116		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	41137	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
41117	41138	return __ret;
41118	41139	}
41119	41140	#endif
...	...	@@ -41123,15 +41144,15 @@ __ai __attribute__((target("neon"))) float32x2_t vfms_f32(float32x2_t __p0, floa
41123	41144	#ifdef __LITTLE_ENDIAN__
41124	41145	__ai __attribute__((target("fp8,neon"))) bfloat16x8_t vcvt1_bf16_mf8_fpm(mfloat8x8_t __p0, fpm_t __p1) {
41125	41146	bfloat16x8_t __ret;
41126		__ret = (bfloat16x8_t) __builtin_neon_vcvt1_bf16_mf8_fpm(__p0, __p1);
	41147	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vcvt1_bf16_mf8_fpm(__p0, __p1));
41127	41148	return __ret;
41128	41149	}
41129	41150	#else
41130	41151	__ai __attribute__((target("fp8,neon"))) bfloat16x8_t vcvt1_bf16_mf8_fpm(mfloat8x8_t __p0, fpm_t __p1) {
41131	41152	bfloat16x8_t __ret;
41132		mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
41133		__ret = (bfloat16x8_t) __builtin_neon_vcvt1_bf16_mf8_fpm(__rev0, __p1);
41134		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	41153	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	41154	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vcvt1_bf16_mf8_fpm(__rev0, __p1));
	41155	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
41135	41156	return __ret;
41136	41157	}
41137	41158	#endif
...	...	@@ -41139,15 +41160,15 @@ __ai __attribute__((target("fp8,neon"))) bfloat16x8_t vcvt1_bf16_mf8_fpm(mfloat8
41139	41160	#ifdef __LITTLE_ENDIAN__
41140	41161	__ai __attribute__((target("fp8,neon"))) float16x8_t vcvt1_f16_mf8_fpm(mfloat8x8_t __p0, fpm_t __p1) {
41141	41162	float16x8_t __ret;
41142		__ret = (float16x8_t) __builtin_neon_vcvt1_f16_mf8_fpm(__p0, __p1);
	41163	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcvt1_f16_mf8_fpm(__p0, __p1));
41143	41164	return __ret;
41144	41165	}
41145	41166	#else
41146	41167	__ai __attribute__((target("fp8,neon"))) float16x8_t vcvt1_f16_mf8_fpm(mfloat8x8_t __p0, fpm_t __p1) {
41147	41168	float16x8_t __ret;
41148		mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
41149		__ret = (float16x8_t) __builtin_neon_vcvt1_f16_mf8_fpm(__rev0, __p1);
41150		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	41169	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	41170	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcvt1_f16_mf8_fpm(__rev0, __p1));
	41171	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
41151	41172	return __ret;
41152	41173	}
41153	41174	#endif
...	...	@@ -41155,15 +41176,15 @@ __ai __attribute__((target("fp8,neon"))) float16x8_t vcvt1_f16_mf8_fpm(mfloat8x8
41155	41176	#ifdef __LITTLE_ENDIAN__
41156	41177	__ai __attribute__((target("fp8,neon"))) bfloat16x8_t vcvt1_high_bf16_mf8_fpm(mfloat8x16_t __p0, fpm_t __p1) {
41157	41178	bfloat16x8_t __ret;
41158		__ret = (bfloat16x8_t) __builtin_neon_vcvt1_high_bf16_mf8_fpm(__p0, __p1);
	41179	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vcvt1_high_bf16_mf8_fpm(__p0, __p1));
41159	41180	return __ret;
41160	41181	}
41161	41182	#else
41162	41183	__ai __attribute__((target("fp8,neon"))) bfloat16x8_t vcvt1_high_bf16_mf8_fpm(mfloat8x16_t __p0, fpm_t __p1) {
41163	41184	bfloat16x8_t __ret;
41164		mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
41165		__ret = (bfloat16x8_t) __builtin_neon_vcvt1_high_bf16_mf8_fpm(__rev0, __p1);
41166		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	41185	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	41186	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vcvt1_high_bf16_mf8_fpm(__rev0, __p1));
	41187	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
41167	41188	return __ret;
41168	41189	}
41169	41190	#endif
...	...	@@ -41171,15 +41192,15 @@ __ai __attribute__((target("fp8,neon"))) bfloat16x8_t vcvt1_high_bf16_mf8_fpm(mf
41171	41192	#ifdef __LITTLE_ENDIAN__
41172	41193	__ai __attribute__((target("fp8,neon"))) float16x8_t vcvt1_high_f16_mf8_fpm(mfloat8x16_t __p0, fpm_t __p1) {
41173	41194	float16x8_t __ret;
41174		__ret = (float16x8_t) __builtin_neon_vcvt1_high_f16_mf8_fpm(__p0, __p1);
	41195	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcvt1_high_f16_mf8_fpm(__p0, __p1));
41175	41196	return __ret;
41176	41197	}
41177	41198	#else
41178	41199	__ai __attribute__((target("fp8,neon"))) float16x8_t vcvt1_high_f16_mf8_fpm(mfloat8x16_t __p0, fpm_t __p1) {
41179	41200	float16x8_t __ret;
41180		mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
41181		__ret = (float16x8_t) __builtin_neon_vcvt1_high_f16_mf8_fpm(__rev0, __p1);
41182		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	41201	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	41202	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcvt1_high_f16_mf8_fpm(__rev0, __p1));
	41203	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
41183	41204	return __ret;
41184	41205	}
41185	41206	#endif
...	...	@@ -41187,15 +41208,15 @@ __ai __attribute__((target("fp8,neon"))) float16x8_t vcvt1_high_f16_mf8_fpm(mflo
41187	41208	#ifdef __LITTLE_ENDIAN__
41188	41209	__ai __attribute__((target("fp8,neon"))) bfloat16x8_t vcvt1_low_bf16_mf8_fpm(mfloat8x16_t __p0, fpm_t __p1) {
41189	41210	bfloat16x8_t __ret;
41190		__ret = (bfloat16x8_t) __builtin_neon_vcvt1_low_bf16_mf8_fpm(__p0, __p1);
	41211	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vcvt1_low_bf16_mf8_fpm(__p0, __p1));
41191	41212	return __ret;
41192	41213	}
41193	41214	#else
41194	41215	__ai __attribute__((target("fp8,neon"))) bfloat16x8_t vcvt1_low_bf16_mf8_fpm(mfloat8x16_t __p0, fpm_t __p1) {
41195	41216	bfloat16x8_t __ret;
41196		mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
41197		__ret = (bfloat16x8_t) __builtin_neon_vcvt1_low_bf16_mf8_fpm(__rev0, __p1);
41198		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	41217	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	41218	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vcvt1_low_bf16_mf8_fpm(__rev0, __p1));
	41219	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
41199	41220	return __ret;
41200	41221	}
41201	41222	#endif
...	...	@@ -41203,15 +41224,15 @@ __ai __attribute__((target("fp8,neon"))) bfloat16x8_t vcvt1_low_bf16_mf8_fpm(mfl
41203	41224	#ifdef __LITTLE_ENDIAN__
41204	41225	__ai __attribute__((target("fp8,neon"))) float16x8_t vcvt1_low_f16_mf8_fpm(mfloat8x16_t __p0, fpm_t __p1) {
41205	41226	float16x8_t __ret;
41206		__ret = (float16x8_t) __builtin_neon_vcvt1_low_f16_mf8_fpm(__p0, __p1);
	41227	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcvt1_low_f16_mf8_fpm(__p0, __p1));
41207	41228	return __ret;
41208	41229	}
41209	41230	#else
41210	41231	__ai __attribute__((target("fp8,neon"))) float16x8_t vcvt1_low_f16_mf8_fpm(mfloat8x16_t __p0, fpm_t __p1) {
41211	41232	float16x8_t __ret;
41212		mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
41213		__ret = (float16x8_t) __builtin_neon_vcvt1_low_f16_mf8_fpm(__rev0, __p1);
41214		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	41233	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	41234	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcvt1_low_f16_mf8_fpm(__rev0, __p1));
	41235	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
41215	41236	return __ret;
41216	41237	}
41217	41238	#endif
...	...	@@ -41219,15 +41240,15 @@ __ai __attribute__((target("fp8,neon"))) float16x8_t vcvt1_low_f16_mf8_fpm(mfloa
41219	41240	#ifdef __LITTLE_ENDIAN__
41220	41241	__ai __attribute__((target("fp8,neon"))) bfloat16x8_t vcvt2_bf16_mf8_fpm(mfloat8x8_t __p0, fpm_t __p1) {
41221	41242	bfloat16x8_t __ret;
41222		__ret = (bfloat16x8_t) __builtin_neon_vcvt2_bf16_mf8_fpm(__p0, __p1);
	41243	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vcvt2_bf16_mf8_fpm(__p0, __p1));
41223	41244	return __ret;
41224	41245	}
41225	41246	#else
41226	41247	__ai __attribute__((target("fp8,neon"))) bfloat16x8_t vcvt2_bf16_mf8_fpm(mfloat8x8_t __p0, fpm_t __p1) {
41227	41248	bfloat16x8_t __ret;
41228		mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
41229		__ret = (bfloat16x8_t) __builtin_neon_vcvt2_bf16_mf8_fpm(__rev0, __p1);
41230		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	41249	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	41250	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vcvt2_bf16_mf8_fpm(__rev0, __p1));
	41251	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
41231	41252	return __ret;
41232	41253	}
41233	41254	#endif
...	...	@@ -41235,15 +41256,15 @@ __ai __attribute__((target("fp8,neon"))) bfloat16x8_t vcvt2_bf16_mf8_fpm(mfloat8
41235	41256	#ifdef __LITTLE_ENDIAN__
41236	41257	__ai __attribute__((target("fp8,neon"))) float16x8_t vcvt2_f16_mf8_fpm(mfloat8x8_t __p0, fpm_t __p1) {
41237	41258	float16x8_t __ret;
41238		__ret = (float16x8_t) __builtin_neon_vcvt2_f16_mf8_fpm(__p0, __p1);
	41259	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcvt2_f16_mf8_fpm(__p0, __p1));
41239	41260	return __ret;
41240	41261	}
41241	41262	#else
41242	41263	__ai __attribute__((target("fp8,neon"))) float16x8_t vcvt2_f16_mf8_fpm(mfloat8x8_t __p0, fpm_t __p1) {
41243	41264	float16x8_t __ret;
41244		mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
41245		__ret = (float16x8_t) __builtin_neon_vcvt2_f16_mf8_fpm(__rev0, __p1);
41246		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	41265	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	41266	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcvt2_f16_mf8_fpm(__rev0, __p1));
	41267	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
41247	41268	return __ret;
41248	41269	}
41249	41270	#endif
...	...	@@ -41251,15 +41272,15 @@ __ai __attribute__((target("fp8,neon"))) float16x8_t vcvt2_f16_mf8_fpm(mfloat8x8
41251	41272	#ifdef __LITTLE_ENDIAN__
41252	41273	__ai __attribute__((target("fp8,neon"))) bfloat16x8_t vcvt2_high_bf16_mf8_fpm(mfloat8x16_t __p0, fpm_t __p1) {
41253	41274	bfloat16x8_t __ret;
41254		__ret = (bfloat16x8_t) __builtin_neon_vcvt2_high_bf16_mf8_fpm(__p0, __p1);
	41275	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vcvt2_high_bf16_mf8_fpm(__p0, __p1));
41255	41276	return __ret;
41256	41277	}
41257	41278	#else
41258	41279	__ai __attribute__((target("fp8,neon"))) bfloat16x8_t vcvt2_high_bf16_mf8_fpm(mfloat8x16_t __p0, fpm_t __p1) {
41259	41280	bfloat16x8_t __ret;
41260		mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
41261		__ret = (bfloat16x8_t) __builtin_neon_vcvt2_high_bf16_mf8_fpm(__rev0, __p1);
41262		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	41281	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	41282	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vcvt2_high_bf16_mf8_fpm(__rev0, __p1));
	41283	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
41263	41284	return __ret;
41264	41285	}
41265	41286	#endif
...	...	@@ -41267,15 +41288,15 @@ __ai __attribute__((target("fp8,neon"))) bfloat16x8_t vcvt2_high_bf16_mf8_fpm(mf
41267	41288	#ifdef __LITTLE_ENDIAN__
41268	41289	__ai __attribute__((target("fp8,neon"))) float16x8_t vcvt2_high_f16_mf8_fpm(mfloat8x16_t __p0, fpm_t __p1) {
41269	41290	float16x8_t __ret;
41270		__ret = (float16x8_t) __builtin_neon_vcvt2_high_f16_mf8_fpm(__p0, __p1);
	41291	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcvt2_high_f16_mf8_fpm(__p0, __p1));
41271	41292	return __ret;
41272	41293	}
41273	41294	#else
41274	41295	__ai __attribute__((target("fp8,neon"))) float16x8_t vcvt2_high_f16_mf8_fpm(mfloat8x16_t __p0, fpm_t __p1) {
41275	41296	float16x8_t __ret;
41276		mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
41277		__ret = (float16x8_t) __builtin_neon_vcvt2_high_f16_mf8_fpm(__rev0, __p1);
41278		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	41297	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	41298	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcvt2_high_f16_mf8_fpm(__rev0, __p1));
	41299	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
41279	41300	return __ret;
41280	41301	}
41281	41302	#endif
...	...	@@ -41283,15 +41304,15 @@ __ai __attribute__((target("fp8,neon"))) float16x8_t vcvt2_high_f16_mf8_fpm(mflo
41283	41304	#ifdef __LITTLE_ENDIAN__
41284	41305	__ai __attribute__((target("fp8,neon"))) bfloat16x8_t vcvt2_low_bf16_mf8_fpm(mfloat8x16_t __p0, fpm_t __p1) {
41285	41306	bfloat16x8_t __ret;
41286		__ret = (bfloat16x8_t) __builtin_neon_vcvt2_low_bf16_mf8_fpm(__p0, __p1);
	41307	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vcvt2_low_bf16_mf8_fpm(__p0, __p1));
41287	41308	return __ret;
41288	41309	}
41289	41310	#else
41290	41311	__ai __attribute__((target("fp8,neon"))) bfloat16x8_t vcvt2_low_bf16_mf8_fpm(mfloat8x16_t __p0, fpm_t __p1) {
41291	41312	bfloat16x8_t __ret;
41292		mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
41293		__ret = (bfloat16x8_t) __builtin_neon_vcvt2_low_bf16_mf8_fpm(__rev0, __p1);
41294		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	41313	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	41314	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vcvt2_low_bf16_mf8_fpm(__rev0, __p1));
	41315	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
41295	41316	return __ret;
41296	41317	}
41297	41318	#endif
...	...	@@ -41299,15 +41320,15 @@ __ai __attribute__((target("fp8,neon"))) bfloat16x8_t vcvt2_low_bf16_mf8_fpm(mfl
41299	41320	#ifdef __LITTLE_ENDIAN__
41300	41321	__ai __attribute__((target("fp8,neon"))) float16x8_t vcvt2_low_f16_mf8_fpm(mfloat8x16_t __p0, fpm_t __p1) {
41301	41322	float16x8_t __ret;
41302		__ret = (float16x8_t) __builtin_neon_vcvt2_low_f16_mf8_fpm(__p0, __p1);
	41323	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcvt2_low_f16_mf8_fpm(__p0, __p1));
41303	41324	return __ret;
41304	41325	}
41305	41326	#else
41306	41327	__ai __attribute__((target("fp8,neon"))) float16x8_t vcvt2_low_f16_mf8_fpm(mfloat8x16_t __p0, fpm_t __p1) {
41307	41328	float16x8_t __ret;
41308		mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
41309		__ret = (float16x8_t) __builtin_neon_vcvt2_low_f16_mf8_fpm(__rev0, __p1);
41310		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	41329	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	41330	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vcvt2_low_f16_mf8_fpm(__rev0, __p1));
	41331	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
41311	41332	return __ret;
41312	41333	}
41313	41334	#endif
...	...	@@ -41315,17 +41336,17 @@ __ai __attribute__((target("fp8,neon"))) float16x8_t vcvt2_low_f16_mf8_fpm(mfloa
41315	41336	#ifdef __LITTLE_ENDIAN__
41316	41337	__ai __attribute__((target("fp8,neon"))) mfloat8x16_t vcvt_high_mf8_f32_fpm(mfloat8x8_t __p0, float32x4_t __p1, float32x4_t __p2, fpm_t __p3) {
41317	41338	mfloat8x16_t __ret;
41318		__ret = (mfloat8x16_t) __builtin_neon_vcvt_high_mf8_f32_fpm(__p0, __p1, __p2, __p3);
	41339	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vcvt_high_mf8_f32_fpm(__p0, __p1, __p2, __p3));
41319	41340	return __ret;
41320	41341	}
41321	41342	#else
41322	41343	__ai __attribute__((target("fp8,neon"))) mfloat8x16_t vcvt_high_mf8_f32_fpm(mfloat8x8_t __p0, float32x4_t __p1, float32x4_t __p2, fpm_t __p3) {
41323	41344	mfloat8x16_t __ret;
41324		mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
41325		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
41326		float32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
41327		__ret = (mfloat8x16_t) __builtin_neon_vcvt_high_mf8_f32_fpm(__rev0, __rev1, __rev2, __p3);
41328		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	41345	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	41346	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	41347	float32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
	41348	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vcvt_high_mf8_f32_fpm(__rev0, __rev1, __rev2, __p3));
	41349	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
41329	41350	return __ret;
41330	41351	}
41331	41352	#endif
...	...	@@ -41333,16 +41354,16 @@ __ai __attribute__((target("fp8,neon"))) mfloat8x16_t vcvt_high_mf8_f32_fpm(mflo
41333	41354	#ifdef __LITTLE_ENDIAN__
41334	41355	__ai __attribute__((target("fp8,neon"))) mfloat8x16_t vcvtq_mf8_f16_fpm(float16x8_t __p0, float16x8_t __p1, fpm_t __p2) {
41335	41356	mfloat8x16_t __ret;
41336		__ret = (mfloat8x16_t) __builtin_neon_vcvtq_mf8_f16_fpm((int8x16_t)__p0, (int8x16_t)__p1, __p2);
	41357	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vcvtq_mf8_f16_fpm(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __p2));
41337	41358	return __ret;
41338	41359	}
41339	41360	#else
41340	41361	__ai __attribute__((target("fp8,neon"))) mfloat8x16_t vcvtq_mf8_f16_fpm(float16x8_t __p0, float16x8_t __p1, fpm_t __p2) {
41341	41362	mfloat8x16_t __ret;
41342		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
41343		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
41344		__ret = (mfloat8x16_t) __builtin_neon_vcvtq_mf8_f16_fpm((int8x16_t)__rev0, (int8x16_t)__rev1, __p2);
41345		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	41363	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	41364	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	41365	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vcvtq_mf8_f16_fpm(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2));
	41366	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
41346	41367	return __ret;
41347	41368	}
41348	41369	#endif
...	...	@@ -41350,16 +41371,16 @@ __ai __attribute__((target("fp8,neon"))) mfloat8x16_t vcvtq_mf8_f16_fpm(float16x
41350	41371	#ifdef __LITTLE_ENDIAN__
41351	41372	__ai __attribute__((target("fp8,neon"))) mfloat8x8_t vcvt_mf8_f16_fpm(float16x4_t __p0, float16x4_t __p1, fpm_t __p2) {
41352	41373	mfloat8x8_t __ret;
41353		__ret = (mfloat8x8_t) __builtin_neon_vcvt_mf8_f16_fpm((int8x8_t)__p0, (int8x8_t)__p1, __p2);
	41374	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vcvt_mf8_f16_fpm(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __p2));
41354	41375	return __ret;
41355	41376	}
41356	41377	#else
41357	41378	__ai __attribute__((target("fp8,neon"))) mfloat8x8_t vcvt_mf8_f16_fpm(float16x4_t __p0, float16x4_t __p1, fpm_t __p2) {
41358	41379	mfloat8x8_t __ret;
41359		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
41360		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
41361		__ret = (mfloat8x8_t) __builtin_neon_vcvt_mf8_f16_fpm((int8x8_t)__rev0, (int8x8_t)__rev1, __p2);
41362		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	41380	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	41381	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	41382	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vcvt_mf8_f16_fpm(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2));
	41383	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
41363	41384	return __ret;
41364	41385	}
41365	41386	#endif
...	...	@@ -41367,16 +41388,16 @@ __ai __attribute__((target("fp8,neon"))) mfloat8x8_t vcvt_mf8_f16_fpm(float16x4_
41367	41388	#ifdef __LITTLE_ENDIAN__
41368	41389	__ai __attribute__((target("fp8,neon"))) mfloat8x8_t vcvt_mf8_f32_fpm(float32x4_t __p0, float32x4_t __p1, fpm_t __p2) {
41369	41390	mfloat8x8_t __ret;
41370		__ret = (mfloat8x8_t) __builtin_neon_vcvt_mf8_f32_fpm(__p0, __p1, __p2);
	41391	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vcvt_mf8_f32_fpm(__p0, __p1, __p2));
41371	41392	return __ret;
41372	41393	}
41373	41394	#else
41374	41395	__ai __attribute__((target("fp8,neon"))) mfloat8x8_t vcvt_mf8_f32_fpm(float32x4_t __p0, float32x4_t __p1, fpm_t __p2) {
41375	41396	mfloat8x8_t __ret;
41376		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
41377		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
41378		__ret = (mfloat8x8_t) __builtin_neon_vcvt_mf8_f32_fpm(__rev0, __rev1, __p2);
41379		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	41397	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	41398	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	41399	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vcvt_mf8_f32_fpm(__rev0, __rev1, __p2));
	41400	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
41380	41401	return __ret;
41381	41402	}
41382	41403	#endif
...	...	@@ -41384,16 +41405,16 @@ __ai __attribute__((target("fp8,neon"))) mfloat8x8_t vcvt_mf8_f32_fpm(float32x4_
41384	41405	#ifdef __LITTLE_ENDIAN__
41385	41406	__ai __attribute__((target("fp8,neon"))) float32x2_t vscale_f32(float32x2_t __p0, int32x2_t __p1) {
41386	41407	float32x2_t __ret;
41387		__ret = (float32x2_t) __builtin_neon_vscale_f32((int8x8_t)__p0, (int8x8_t)__p1, 9);
	41408	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vscale_f32(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 9));
41388	41409	return __ret;
41389	41410	}
41390	41411	#else
41391	41412	__ai __attribute__((target("fp8,neon"))) float32x2_t vscale_f32(float32x2_t __p0, int32x2_t __p1) {
41392	41413	float32x2_t __ret;
41393		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
41394		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
41395		__ret = (float32x2_t) __builtin_neon_vscale_f32((int8x8_t)__rev0, (int8x8_t)__rev1, 9);
41396		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	41414	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	41415	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	41416	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vscale_f32(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 9));
	41417	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
41397	41418	return __ret;
41398	41419	}
41399	41420	#endif
...	...	@@ -41401,16 +41422,16 @@ __ai __attribute__((target("fp8,neon"))) float32x2_t vscale_f32(float32x2_t __p0
41401	41422	#ifdef __LITTLE_ENDIAN__
41402	41423	__ai __attribute__((target("fp8,neon"))) float16x4_t vscale_f16(float16x4_t __p0, int16x4_t __p1) {
41403	41424	float16x4_t __ret;
41404		__ret = (float16x4_t) __builtin_neon_vscale_f16((int8x8_t)__p0, (int8x8_t)__p1, 8);
	41425	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vscale_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 8));
41405	41426	return __ret;
41406	41427	}
41407	41428	#else
41408	41429	__ai __attribute__((target("fp8,neon"))) float16x4_t vscale_f16(float16x4_t __p0, int16x4_t __p1) {
41409	41430	float16x4_t __ret;
41410		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
41411		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
41412		__ret = (float16x4_t) __builtin_neon_vscale_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 8);
41413		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	41431	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	41432	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	41433	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vscale_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 8));
	41434	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
41414	41435	return __ret;
41415	41436	}
41416	41437	#endif
...	...	@@ -41418,16 +41439,16 @@ __ai __attribute__((target("fp8,neon"))) float16x4_t vscale_f16(float16x4_t __p0
41418	41439	#ifdef __LITTLE_ENDIAN__
41419	41440	__ai __attribute__((target("fp8,neon"))) float64x2_t vscaleq_f64(float64x2_t __p0, int64x2_t __p1) {
41420	41441	float64x2_t __ret;
41421		__ret = (float64x2_t) __builtin_neon_vscaleq_f64((int8x16_t)__p0, (int8x16_t)__p1, 42);
	41442	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vscaleq_f64(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 42));
41422	41443	return __ret;
41423	41444	}
41424	41445	#else
41425	41446	__ai __attribute__((target("fp8,neon"))) float64x2_t vscaleq_f64(float64x2_t __p0, int64x2_t __p1) {
41426	41447	float64x2_t __ret;
41427		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
41428		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
41429		__ret = (float64x2_t) __builtin_neon_vscaleq_f64((int8x16_t)__rev0, (int8x16_t)__rev1, 42);
41430		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	41448	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	41449	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	41450	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vscaleq_f64(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 42));
	41451	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
41431	41452	return __ret;
41432	41453	}
41433	41454	#endif
...	...	@@ -41435,16 +41456,16 @@ __ai __attribute__((target("fp8,neon"))) float64x2_t vscaleq_f64(float64x2_t __p
41435	41456	#ifdef __LITTLE_ENDIAN__
41436	41457	__ai __attribute__((target("fp8,neon"))) float32x4_t vscaleq_f32(float32x4_t __p0, int32x4_t __p1) {
41437	41458	float32x4_t __ret;
41438		__ret = (float32x4_t) __builtin_neon_vscaleq_f32((int8x16_t)__p0, (int8x16_t)__p1, 41);
	41459	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vscaleq_f32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 41));
41439	41460	return __ret;
41440	41461	}
41441	41462	#else
41442	41463	__ai __attribute__((target("fp8,neon"))) float32x4_t vscaleq_f32(float32x4_t __p0, int32x4_t __p1) {
41443	41464	float32x4_t __ret;
41444		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
41445		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
41446		__ret = (float32x4_t) __builtin_neon_vscaleq_f32((int8x16_t)__rev0, (int8x16_t)__rev1, 41);
41447		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	41465	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	41466	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	41467	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vscaleq_f32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 41));
	41468	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
41448	41469	return __ret;
41449	41470	}
41450	41471	#endif
...	...	@@ -41452,16 +41473,16 @@ __ai __attribute__((target("fp8,neon"))) float32x4_t vscaleq_f32(float32x4_t __p
41452	41473	#ifdef __LITTLE_ENDIAN__
41453	41474	__ai __attribute__((target("fp8,neon"))) float16x8_t vscaleq_f16(float16x8_t __p0, int16x8_t __p1) {
41454	41475	float16x8_t __ret;
41455		__ret = (float16x8_t) __builtin_neon_vscaleq_f16((int8x16_t)__p0, (int8x16_t)__p1, 40);
	41476	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vscaleq_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 40));
41456	41477	return __ret;
41457	41478	}
41458	41479	#else
41459	41480	__ai __attribute__((target("fp8,neon"))) float16x8_t vscaleq_f16(float16x8_t __p0, int16x8_t __p1) {
41460	41481	float16x8_t __ret;
41461		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
41462		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
41463		__ret = (float16x8_t) __builtin_neon_vscaleq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 40);
41464		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	41482	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	41483	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	41484	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vscaleq_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 40));
	41485	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
41465	41486	return __ret;
41466	41487	}
41467	41488	#endif
...	...	@@ -41469,17 +41490,17 @@ __ai __attribute__((target("fp8,neon"))) float16x8_t vscaleq_f16(float16x8_t __p
41469	41490	#ifdef __LITTLE_ENDIAN__
41470	41491	__ai __attribute__((target("fp8dot2,neon"))) float16x8_t vdotq_f16_mf8_fpm(float16x8_t __p0, mfloat8x16_t __p1, mfloat8x16_t __p2, fpm_t __p3) {
41471	41492	float16x8_t __ret;
41472		__ret = (float16x8_t) __builtin_neon_vdotq_f16_mf8_fpm((int8x16_t)__p0, __p1, __p2, __p3);
	41493	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vdotq_f16_mf8_fpm(__builtin_bit_cast(int8x16_t, __p0), __p1, __p2, __p3));
41473	41494	return __ret;
41474	41495	}
41475	41496	#else
41476	41497	__ai __attribute__((target("fp8dot2,neon"))) float16x8_t vdotq_f16_mf8_fpm(float16x8_t __p0, mfloat8x16_t __p1, mfloat8x16_t __p2, fpm_t __p3) {
41477	41498	float16x8_t __ret;
41478		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
41479		mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
41480		mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
41481		__ret = (float16x8_t) __builtin_neon_vdotq_f16_mf8_fpm((int8x16_t)__rev0, __rev1, __rev2, __p3);
41482		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	41499	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	41500	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	41501	mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	41502	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vdotq_f16_mf8_fpm(__builtin_bit_cast(int8x16_t, __rev0), __rev1, __rev2, __p3));
	41503	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
41483	41504	return __ret;
41484	41505	}
41485	41506	#endif
...	...	@@ -41487,17 +41508,17 @@ __ai __attribute__((target("fp8dot2,neon"))) float16x8_t vdotq_f16_mf8_fpm(float
41487	41508	#ifdef __LITTLE_ENDIAN__
41488	41509	__ai __attribute__((target("fp8dot2,neon"))) float16x4_t vdot_f16_mf8_fpm(float16x4_t __p0, mfloat8x8_t __p1, mfloat8x8_t __p2, fpm_t __p3) {
41489	41510	float16x4_t __ret;
41490		__ret = (float16x4_t) __builtin_neon_vdot_f16_mf8_fpm((int8x8_t)__p0, __p1, __p2, __p3);
	41511	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vdot_f16_mf8_fpm(__builtin_bit_cast(int8x8_t, __p0), __p1, __p2, __p3));
41491	41512	return __ret;
41492	41513	}
41493	41514	#else
41494	41515	__ai __attribute__((target("fp8dot2,neon"))) float16x4_t vdot_f16_mf8_fpm(float16x4_t __p0, mfloat8x8_t __p1, mfloat8x8_t __p2, fpm_t __p3) {
41495	41516	float16x4_t __ret;
41496		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
41497		mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
41498		mfloat8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
41499		__ret = (float16x4_t) __builtin_neon_vdot_f16_mf8_fpm((int8x8_t)__rev0, __rev1, __rev2, __p3);
41500		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	41517	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	41518	mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	41519	mfloat8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	41520	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vdot_f16_mf8_fpm(__builtin_bit_cast(int8x8_t, __rev0), __rev1, __rev2, __p3));
	41521	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
41501	41522	return __ret;
41502	41523	}
41503	41524	#endif
...	...	@@ -41509,7 +41530,7 @@ __ai __attribute__((target("fp8dot2,neon"))) float16x4_t vdot_f16_mf8_fpm(float1
41509	41530	mfloat8x16_t __s1 = __p1; \
41510	41531	mfloat8x8_t __s2 = __p2; \
41511	41532	fpm_t __s4 = __p4; \
41512		__ret = (float16x8_t) __builtin_neon_vdotq_lane_f16_mf8_fpm((int8x16_t)__s0, __s1, __s2, __p3, __s4); \
	41533	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vdotq_lane_f16_mf8_fpm(__builtin_bit_cast(int8x16_t, __s0), __s1, __s2, __p3, __s4)); \
41513	41534	__ret; \
41514	41535	})
41515	41536	#else
...	...	@@ -41519,11 +41540,11 @@ __ai __attribute__((target("fp8dot2,neon"))) float16x4_t vdot_f16_mf8_fpm(float1
41519	41540	mfloat8x16_t __s1 = __p1; \
41520	41541	mfloat8x8_t __s2 = __p2; \
41521	41542	fpm_t __s4 = __p4; \
41522		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
41523		mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
41524		mfloat8x8_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 7, 6, 5, 4, 3, 2, 1, 0); \
41525		__ret = (float16x8_t) __builtin_neon_vdotq_lane_f16_mf8_fpm((int8x16_t)__rev0, __rev1, __rev2, __p3, __s4); \
41526		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	41543	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	41544	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	41545	mfloat8x8_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_64_8); \
	41546	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vdotq_lane_f16_mf8_fpm(__builtin_bit_cast(int8x16_t, __rev0), __rev1, __rev2, __p3, __s4)); \
	41547	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
41527	41548	__ret; \
41528	41549	})
41529	41550	#endif
...	...	@@ -41535,7 +41556,7 @@ __ai __attribute__((target("fp8dot2,neon"))) float16x4_t vdot_f16_mf8_fpm(float1
41535	41556	mfloat8x8_t __s1 = __p1; \
41536	41557	mfloat8x8_t __s2 = __p2; \
41537	41558	fpm_t __s4 = __p4; \
41538		__ret = (float16x4_t) __builtin_neon_vdot_lane_f16_mf8_fpm((int8x8_t)__s0, __s1, __s2, __p3, __s4); \
	41559	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vdot_lane_f16_mf8_fpm(__builtin_bit_cast(int8x8_t, __s0), __s1, __s2, __p3, __s4)); \
41539	41560	__ret; \
41540	41561	})
41541	41562	#else
...	...	@@ -41545,11 +41566,11 @@ __ai __attribute__((target("fp8dot2,neon"))) float16x4_t vdot_f16_mf8_fpm(float1
41545	41566	mfloat8x8_t __s1 = __p1; \
41546	41567	mfloat8x8_t __s2 = __p2; \
41547	41568	fpm_t __s4 = __p4; \
41548		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
41549		mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
41550		mfloat8x8_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 7, 6, 5, 4, 3, 2, 1, 0); \
41551		__ret = (float16x4_t) __builtin_neon_vdot_lane_f16_mf8_fpm((int8x8_t)__rev0, __rev1, __rev2, __p3, __s4); \
41552		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	41569	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	41570	mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	41571	mfloat8x8_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_64_8); \
	41572	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vdot_lane_f16_mf8_fpm(__builtin_bit_cast(int8x8_t, __rev0), __rev1, __rev2, __p3, __s4)); \
	41573	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
41553	41574	__ret; \
41554	41575	})
41555	41576	#endif
...	...	@@ -41561,7 +41582,7 @@ __ai __attribute__((target("fp8dot2,neon"))) float16x4_t vdot_f16_mf8_fpm(float1
41561	41582	mfloat8x16_t __s1 = __p1; \
41562	41583	mfloat8x16_t __s2 = __p2; \
41563	41584	fpm_t __s4 = __p4; \
41564		__ret = (float16x8_t) __builtin_neon_vdotq_laneq_f16_mf8_fpm((int8x16_t)__s0, __s1, __s2, __p3, __s4); \
	41585	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vdotq_laneq_f16_mf8_fpm(__builtin_bit_cast(int8x16_t, __s0), __s1, __s2, __p3, __s4)); \
41565	41586	__ret; \
41566	41587	})
41567	41588	#else
...	...	@@ -41571,11 +41592,11 @@ __ai __attribute__((target("fp8dot2,neon"))) float16x4_t vdot_f16_mf8_fpm(float1
41571	41592	mfloat8x16_t __s1 = __p1; \
41572	41593	mfloat8x16_t __s2 = __p2; \
41573	41594	fpm_t __s4 = __p4; \
41574		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
41575		mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
41576		mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
41577		__ret = (float16x8_t) __builtin_neon_vdotq_laneq_f16_mf8_fpm((int8x16_t)__rev0, __rev1, __rev2, __p3, __s4); \
41578		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	41595	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	41596	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	41597	mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_128_8); \
	41598	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vdotq_laneq_f16_mf8_fpm(__builtin_bit_cast(int8x16_t, __rev0), __rev1, __rev2, __p3, __s4)); \
	41599	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
41579	41600	__ret; \
41580	41601	})
41581	41602	#endif
...	...	@@ -41587,7 +41608,7 @@ __ai __attribute__((target("fp8dot2,neon"))) float16x4_t vdot_f16_mf8_fpm(float1
41587	41608	mfloat8x8_t __s1 = __p1; \
41588	41609	mfloat8x16_t __s2 = __p2; \
41589	41610	fpm_t __s4 = __p4; \
41590		__ret = (float16x4_t) __builtin_neon_vdot_laneq_f16_mf8_fpm((int8x8_t)__s0, __s1, __s2, __p3, __s4); \
	41611	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vdot_laneq_f16_mf8_fpm(__builtin_bit_cast(int8x8_t, __s0), __s1, __s2, __p3, __s4)); \
41591	41612	__ret; \
41592	41613	})
41593	41614	#else
...	...	@@ -41597,11 +41618,11 @@ __ai __attribute__((target("fp8dot2,neon"))) float16x4_t vdot_f16_mf8_fpm(float1
41597	41618	mfloat8x8_t __s1 = __p1; \
41598	41619	mfloat8x16_t __s2 = __p2; \
41599	41620	fpm_t __s4 = __p4; \
41600		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
41601		mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
41602		mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
41603		__ret = (float16x4_t) __builtin_neon_vdot_laneq_f16_mf8_fpm((int8x8_t)__rev0, __rev1, __rev2, __p3, __s4); \
41604		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	41621	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	41622	mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	41623	mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_128_8); \
	41624	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vdot_laneq_f16_mf8_fpm(__builtin_bit_cast(int8x8_t, __rev0), __rev1, __rev2, __p3, __s4)); \
	41625	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
41605	41626	__ret; \
41606	41627	})
41607	41628	#endif
...	...	@@ -41609,17 +41630,17 @@ __ai __attribute__((target("fp8dot2,neon"))) float16x4_t vdot_f16_mf8_fpm(float1
41609	41630	#ifdef __LITTLE_ENDIAN__
41610	41631	__ai __attribute__((target("fp8dot4,neon"))) float32x4_t vdotq_f32_mf8_fpm(float32x4_t __p0, mfloat8x16_t __p1, mfloat8x16_t __p2, fpm_t __p3) {
41611	41632	float32x4_t __ret;
41612		__ret = (float32x4_t) __builtin_neon_vdotq_f32_mf8_fpm(__p0, __p1, __p2, __p3);
	41633	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vdotq_f32_mf8_fpm(__p0, __p1, __p2, __p3));
41613	41634	return __ret;
41614	41635	}
41615	41636	#else
41616	41637	__ai __attribute__((target("fp8dot4,neon"))) float32x4_t vdotq_f32_mf8_fpm(float32x4_t __p0, mfloat8x16_t __p1, mfloat8x16_t __p2, fpm_t __p3) {
41617	41638	float32x4_t __ret;
41618		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
41619		mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
41620		mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
41621		__ret = (float32x4_t) __builtin_neon_vdotq_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3);
41622		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	41639	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	41640	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	41641	mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	41642	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vdotq_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3));
	41643	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
41623	41644	return __ret;
41624	41645	}
41625	41646	#endif
...	...	@@ -41627,17 +41648,17 @@ __ai __attribute__((target("fp8dot4,neon"))) float32x4_t vdotq_f32_mf8_fpm(float
41627	41648	#ifdef __LITTLE_ENDIAN__
41628	41649	__ai __attribute__((target("fp8dot4,neon"))) float32x2_t vdot_f32_mf8_fpm(float32x2_t __p0, mfloat8x8_t __p1, mfloat8x8_t __p2, fpm_t __p3) {
41629	41650	float32x2_t __ret;
41630		__ret = (float32x2_t) __builtin_neon_vdot_f32_mf8_fpm(__p0, __p1, __p2, __p3);
	41651	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vdot_f32_mf8_fpm(__p0, __p1, __p2, __p3));
41631	41652	return __ret;
41632	41653	}
41633	41654	#else
41634	41655	__ai __attribute__((target("fp8dot4,neon"))) float32x2_t vdot_f32_mf8_fpm(float32x2_t __p0, mfloat8x8_t __p1, mfloat8x8_t __p2, fpm_t __p3) {
41635	41656	float32x2_t __ret;
41636		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
41637		mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
41638		mfloat8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
41639		__ret = (float32x2_t) __builtin_neon_vdot_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3);
41640		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	41657	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	41658	mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	41659	mfloat8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	41660	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vdot_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3));
	41661	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
41641	41662	return __ret;
41642	41663	}
41643	41664	#endif
...	...	@@ -41649,7 +41670,7 @@ __ai __attribute__((target("fp8dot4,neon"))) float32x2_t vdot_f32_mf8_fpm(float3
41649	41670	mfloat8x16_t __s1 = __p1; \
41650	41671	mfloat8x8_t __s2 = __p2; \
41651	41672	fpm_t __s4 = __p4; \
41652		__ret = (float32x4_t) __builtin_neon_vdotq_lane_f32_mf8_fpm(__s0, __s1, __s2, __p3, __s4); \
	41673	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vdotq_lane_f32_mf8_fpm(__s0, __s1, __s2, __p3, __s4)); \
41653	41674	__ret; \
41654	41675	})
41655	41676	#else
...	...	@@ -41659,11 +41680,11 @@ __ai __attribute__((target("fp8dot4,neon"))) float32x2_t vdot_f32_mf8_fpm(float3
41659	41680	mfloat8x16_t __s1 = __p1; \
41660	41681	mfloat8x8_t __s2 = __p2; \
41661	41682	fpm_t __s4 = __p4; \
41662		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
41663		mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
41664		mfloat8x8_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 7, 6, 5, 4, 3, 2, 1, 0); \
41665		__ret = (float32x4_t) __builtin_neon_vdotq_lane_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3, __s4); \
41666		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	41683	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	41684	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	41685	mfloat8x8_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_64_8); \
	41686	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vdotq_lane_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3, __s4)); \
	41687	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
41667	41688	__ret; \
41668	41689	})
41669	41690	#endif
...	...	@@ -41675,7 +41696,7 @@ __ai __attribute__((target("fp8dot4,neon"))) float32x2_t vdot_f32_mf8_fpm(float3
41675	41696	mfloat8x8_t __s1 = __p1; \
41676	41697	mfloat8x8_t __s2 = __p2; \
41677	41698	fpm_t __s4 = __p4; \
41678		__ret = (float32x2_t) __builtin_neon_vdot_lane_f32_mf8_fpm(__s0, __s1, __s2, __p3, __s4); \
	41699	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vdot_lane_f32_mf8_fpm(__s0, __s1, __s2, __p3, __s4)); \
41679	41700	__ret; \
41680	41701	})
41681	41702	#else
...	...	@@ -41685,11 +41706,11 @@ __ai __attribute__((target("fp8dot4,neon"))) float32x2_t vdot_f32_mf8_fpm(float3
41685	41706	mfloat8x8_t __s1 = __p1; \
41686	41707	mfloat8x8_t __s2 = __p2; \
41687	41708	fpm_t __s4 = __p4; \
41688		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
41689		mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
41690		mfloat8x8_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 7, 6, 5, 4, 3, 2, 1, 0); \
41691		__ret = (float32x2_t) __builtin_neon_vdot_lane_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3, __s4); \
41692		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	41709	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	41710	mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	41711	mfloat8x8_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_64_8); \
	41712	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vdot_lane_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3, __s4)); \
	41713	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
41693	41714	__ret; \
41694	41715	})
41695	41716	#endif
...	...	@@ -41701,7 +41722,7 @@ __ai __attribute__((target("fp8dot4,neon"))) float32x2_t vdot_f32_mf8_fpm(float3
41701	41722	mfloat8x16_t __s1 = __p1; \
41702	41723	mfloat8x16_t __s2 = __p2; \
41703	41724	fpm_t __s4 = __p4; \
41704		__ret = (float32x4_t) __builtin_neon_vdotq_laneq_f32_mf8_fpm(__s0, __s1, __s2, __p3, __s4); \
	41725	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vdotq_laneq_f32_mf8_fpm(__s0, __s1, __s2, __p3, __s4)); \
41705	41726	__ret; \
41706	41727	})
41707	41728	#else
...	...	@@ -41711,11 +41732,11 @@ __ai __attribute__((target("fp8dot4,neon"))) float32x2_t vdot_f32_mf8_fpm(float3
41711	41732	mfloat8x16_t __s1 = __p1; \
41712	41733	mfloat8x16_t __s2 = __p2; \
41713	41734	fpm_t __s4 = __p4; \
41714		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
41715		mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
41716		mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
41717		__ret = (float32x4_t) __builtin_neon_vdotq_laneq_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3, __s4); \
41718		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	41735	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	41736	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	41737	mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_128_8); \
	41738	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vdotq_laneq_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3, __s4)); \
	41739	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
41719	41740	__ret; \
41720	41741	})
41721	41742	#endif
...	...	@@ -41727,7 +41748,7 @@ __ai __attribute__((target("fp8dot4,neon"))) float32x2_t vdot_f32_mf8_fpm(float3
41727	41748	mfloat8x8_t __s1 = __p1; \
41728	41749	mfloat8x16_t __s2 = __p2; \
41729	41750	fpm_t __s4 = __p4; \
41730		__ret = (float32x2_t) __builtin_neon_vdot_laneq_f32_mf8_fpm(__s0, __s1, __s2, __p3, __s4); \
	41751	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vdot_laneq_f32_mf8_fpm(__s0, __s1, __s2, __p3, __s4)); \
41731	41752	__ret; \
41732	41753	})
41733	41754	#else
...	...	@@ -41737,11 +41758,11 @@ __ai __attribute__((target("fp8dot4,neon"))) float32x2_t vdot_f32_mf8_fpm(float3
41737	41758	mfloat8x8_t __s1 = __p1; \
41738	41759	mfloat8x16_t __s2 = __p2; \
41739	41760	fpm_t __s4 = __p4; \
41740		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
41741		mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
41742		mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
41743		__ret = (float32x2_t) __builtin_neon_vdot_laneq_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3, __s4); \
41744		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	41761	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	41762	mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	41763	mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_128_8); \
	41764	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vdot_laneq_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3, __s4)); \
	41765	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
41745	41766	__ret; \
41746	41767	})
41747	41768	#endif
...	...	@@ -41749,17 +41770,17 @@ __ai __attribute__((target("fp8dot4,neon"))) float32x2_t vdot_f32_mf8_fpm(float3
41749	41770	#ifdef __LITTLE_ENDIAN__
41750	41771	__ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlalbq_f16_mf8_fpm(float16x8_t __p0, mfloat8x16_t __p1, mfloat8x16_t __p2, fpm_t __p3) {
41751	41772	float16x8_t __ret;
41752		__ret = (float16x8_t) __builtin_neon_vmlalbq_f16_mf8_fpm((int8x16_t)__p0, __p1, __p2, __p3);
	41773	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vmlalbq_f16_mf8_fpm(__builtin_bit_cast(int8x16_t, __p0), __p1, __p2, __p3));
41753	41774	return __ret;
41754	41775	}
41755	41776	#else
41756	41777	__ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlalbq_f16_mf8_fpm(float16x8_t __p0, mfloat8x16_t __p1, mfloat8x16_t __p2, fpm_t __p3) {
41757	41778	float16x8_t __ret;
41758		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
41759		mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
41760		mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
41761		__ret = (float16x8_t) __builtin_neon_vmlalbq_f16_mf8_fpm((int8x16_t)__rev0, __rev1, __rev2, __p3);
41762		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	41779	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	41780	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	41781	mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	41782	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vmlalbq_f16_mf8_fpm(__builtin_bit_cast(int8x16_t, __rev0), __rev1, __rev2, __p3));
	41783	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
41763	41784	return __ret;
41764	41785	}
41765	41786	#endif
...	...	@@ -41771,7 +41792,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlalbq_f16_mf8_fpm(floa
41771	41792	mfloat8x16_t __s1 = __p1; \
41772	41793	mfloat8x8_t __s2 = __p2; \
41773	41794	fpm_t __s4 = __p4; \
41774		__ret = (float16x8_t) __builtin_neon_vmlalbq_lane_f16_mf8_fpm((int8x16_t)__s0, __s1, __s2, __p3, __s4); \
	41795	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vmlalbq_lane_f16_mf8_fpm(__builtin_bit_cast(int8x16_t, __s0), __s1, __s2, __p3, __s4)); \
41775	41796	__ret; \
41776	41797	})
41777	41798	#else
...	...	@@ -41781,11 +41802,11 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlalbq_f16_mf8_fpm(floa
41781	41802	mfloat8x16_t __s1 = __p1; \
41782	41803	mfloat8x8_t __s2 = __p2; \
41783	41804	fpm_t __s4 = __p4; \
41784		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
41785		mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
41786		mfloat8x8_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 7, 6, 5, 4, 3, 2, 1, 0); \
41787		__ret = (float16x8_t) __builtin_neon_vmlalbq_lane_f16_mf8_fpm((int8x16_t)__rev0, __rev1, __rev2, __p3, __s4); \
41788		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	41805	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	41806	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	41807	mfloat8x8_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_64_8); \
	41808	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vmlalbq_lane_f16_mf8_fpm(__builtin_bit_cast(int8x16_t, __rev0), __rev1, __rev2, __p3, __s4)); \
	41809	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
41789	41810	__ret; \
41790	41811	})
41791	41812	#endif
...	...	@@ -41797,7 +41818,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlalbq_f16_mf8_fpm(floa
41797	41818	mfloat8x16_t __s1 = __p1; \
41798	41819	mfloat8x16_t __s2 = __p2; \
41799	41820	fpm_t __s4 = __p4; \
41800		__ret = (float16x8_t) __builtin_neon_vmlalbq_laneq_f16_mf8_fpm((int8x16_t)__s0, __s1, __s2, __p3, __s4); \
	41821	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vmlalbq_laneq_f16_mf8_fpm(__builtin_bit_cast(int8x16_t, __s0), __s1, __s2, __p3, __s4)); \
41801	41822	__ret; \
41802	41823	})
41803	41824	#else
...	...	@@ -41807,11 +41828,11 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlalbq_f16_mf8_fpm(floa
41807	41828	mfloat8x16_t __s1 = __p1; \
41808	41829	mfloat8x16_t __s2 = __p2; \
41809	41830	fpm_t __s4 = __p4; \
41810		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
41811		mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
41812		mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
41813		__ret = (float16x8_t) __builtin_neon_vmlalbq_laneq_f16_mf8_fpm((int8x16_t)__rev0, __rev1, __rev2, __p3, __s4); \
41814		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	41831	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	41832	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	41833	mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_128_8); \
	41834	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vmlalbq_laneq_f16_mf8_fpm(__builtin_bit_cast(int8x16_t, __rev0), __rev1, __rev2, __p3, __s4)); \
	41835	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
41815	41836	__ret; \
41816	41837	})
41817	41838	#endif
...	...	@@ -41819,17 +41840,17 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlalbq_f16_mf8_fpm(floa
41819	41840	#ifdef __LITTLE_ENDIAN__
41820	41841	__ai __attribute__((target("fp8fma,neon"))) float32x4_t vmlallbbq_f32_mf8_fpm(float32x4_t __p0, mfloat8x16_t __p1, mfloat8x16_t __p2, fpm_t __p3) {
41821	41842	float32x4_t __ret;
41822		__ret = (float32x4_t) __builtin_neon_vmlallbbq_f32_mf8_fpm(__p0, __p1, __p2, __p3);
	41843	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmlallbbq_f32_mf8_fpm(__p0, __p1, __p2, __p3));
41823	41844	return __ret;
41824	41845	}
41825	41846	#else
41826	41847	__ai __attribute__((target("fp8fma,neon"))) float32x4_t vmlallbbq_f32_mf8_fpm(float32x4_t __p0, mfloat8x16_t __p1, mfloat8x16_t __p2, fpm_t __p3) {
41827	41848	float32x4_t __ret;
41828		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
41829		mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
41830		mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
41831		__ret = (float32x4_t) __builtin_neon_vmlallbbq_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3);
41832		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	41849	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	41850	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	41851	mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	41852	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmlallbbq_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3));
	41853	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
41833	41854	return __ret;
41834	41855	}
41835	41856	#endif
...	...	@@ -41841,7 +41862,7 @@ __ai __attribute__((target("fp8fma,neon"))) float32x4_t vmlallbbq_f32_mf8_fpm(fl
41841	41862	mfloat8x16_t __s1 = __p1; \
41842	41863	mfloat8x8_t __s2 = __p2; \
41843	41864	fpm_t __s4 = __p4; \
41844		__ret = (float32x4_t) __builtin_neon_vmlallbbq_lane_f32_mf8_fpm(__s0, __s1, __s2, __p3, __s4); \
	41865	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmlallbbq_lane_f32_mf8_fpm(__s0, __s1, __s2, __p3, __s4)); \
41845	41866	__ret; \
41846	41867	})
41847	41868	#else
...	...	@@ -41851,11 +41872,11 @@ __ai __attribute__((target("fp8fma,neon"))) float32x4_t vmlallbbq_f32_mf8_fpm(fl
41851	41872	mfloat8x16_t __s1 = __p1; \
41852	41873	mfloat8x8_t __s2 = __p2; \
41853	41874	fpm_t __s4 = __p4; \
41854		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
41855		mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
41856		mfloat8x8_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 7, 6, 5, 4, 3, 2, 1, 0); \
41857		__ret = (float32x4_t) __builtin_neon_vmlallbbq_lane_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3, __s4); \
41858		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	41875	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	41876	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	41877	mfloat8x8_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_64_8); \
	41878	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmlallbbq_lane_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3, __s4)); \
	41879	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
41859	41880	__ret; \
41860	41881	})
41861	41882	#endif
...	...	@@ -41867,7 +41888,7 @@ __ai __attribute__((target("fp8fma,neon"))) float32x4_t vmlallbbq_f32_mf8_fpm(fl
41867	41888	mfloat8x16_t __s1 = __p1; \
41868	41889	mfloat8x16_t __s2 = __p2; \
41869	41890	fpm_t __s4 = __p4; \
41870		__ret = (float32x4_t) __builtin_neon_vmlallbbq_laneq_f32_mf8_fpm(__s0, __s1, __s2, __p3, __s4); \
	41891	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmlallbbq_laneq_f32_mf8_fpm(__s0, __s1, __s2, __p3, __s4)); \
41871	41892	__ret; \
41872	41893	})
41873	41894	#else
...	...	@@ -41877,11 +41898,11 @@ __ai __attribute__((target("fp8fma,neon"))) float32x4_t vmlallbbq_f32_mf8_fpm(fl
41877	41898	mfloat8x16_t __s1 = __p1; \
41878	41899	mfloat8x16_t __s2 = __p2; \
41879	41900	fpm_t __s4 = __p4; \
41880		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
41881		mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
41882		mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
41883		__ret = (float32x4_t) __builtin_neon_vmlallbbq_laneq_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3, __s4); \
41884		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	41901	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	41902	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	41903	mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_128_8); \
	41904	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmlallbbq_laneq_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3, __s4)); \
	41905	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
41885	41906	__ret; \
41886	41907	})
41887	41908	#endif
...	...	@@ -41889,17 +41910,17 @@ __ai __attribute__((target("fp8fma,neon"))) float32x4_t vmlallbbq_f32_mf8_fpm(fl
41889	41910	#ifdef __LITTLE_ENDIAN__
41890	41911	__ai __attribute__((target("fp8fma,neon"))) float32x4_t vmlallbtq_f32_mf8_fpm(float32x4_t __p0, mfloat8x16_t __p1, mfloat8x16_t __p2, fpm_t __p3) {
41891	41912	float32x4_t __ret;
41892		__ret = (float32x4_t) __builtin_neon_vmlallbtq_f32_mf8_fpm(__p0, __p1, __p2, __p3);
	41913	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmlallbtq_f32_mf8_fpm(__p0, __p1, __p2, __p3));
41893	41914	return __ret;
41894	41915	}
41895	41916	#else
41896	41917	__ai __attribute__((target("fp8fma,neon"))) float32x4_t vmlallbtq_f32_mf8_fpm(float32x4_t __p0, mfloat8x16_t __p1, mfloat8x16_t __p2, fpm_t __p3) {
41897	41918	float32x4_t __ret;
41898		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
41899		mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
41900		mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
41901		__ret = (float32x4_t) __builtin_neon_vmlallbtq_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3);
41902		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	41919	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	41920	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	41921	mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	41922	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmlallbtq_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3));
	41923	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
41903	41924	return __ret;
41904	41925	}
41905	41926	#endif
...	...	@@ -41911,7 +41932,7 @@ __ai __attribute__((target("fp8fma,neon"))) float32x4_t vmlallbtq_f32_mf8_fpm(fl
41911	41932	mfloat8x16_t __s1 = __p1; \
41912	41933	mfloat8x8_t __s2 = __p2; \
41913	41934	fpm_t __s4 = __p4; \
41914		__ret = (float32x4_t) __builtin_neon_vmlallbtq_lane_f32_mf8_fpm(__s0, __s1, __s2, __p3, __s4); \
	41935	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmlallbtq_lane_f32_mf8_fpm(__s0, __s1, __s2, __p3, __s4)); \
41915	41936	__ret; \
41916	41937	})
41917	41938	#else
...	...	@@ -41921,11 +41942,11 @@ __ai __attribute__((target("fp8fma,neon"))) float32x4_t vmlallbtq_f32_mf8_fpm(fl
41921	41942	mfloat8x16_t __s1 = __p1; \
41922	41943	mfloat8x8_t __s2 = __p2; \
41923	41944	fpm_t __s4 = __p4; \
41924		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
41925		mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
41926		mfloat8x8_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 7, 6, 5, 4, 3, 2, 1, 0); \
41927		__ret = (float32x4_t) __builtin_neon_vmlallbtq_lane_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3, __s4); \
41928		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	41945	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	41946	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	41947	mfloat8x8_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_64_8); \
	41948	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmlallbtq_lane_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3, __s4)); \
	41949	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
41929	41950	__ret; \
41930	41951	})
41931	41952	#endif
...	...	@@ -41937,7 +41958,7 @@ __ai __attribute__((target("fp8fma,neon"))) float32x4_t vmlallbtq_f32_mf8_fpm(fl
41937	41958	mfloat8x16_t __s1 = __p1; \
41938	41959	mfloat8x16_t __s2 = __p2; \
41939	41960	fpm_t __s4 = __p4; \
41940		__ret = (float32x4_t) __builtin_neon_vmlallbtq_laneq_f32_mf8_fpm(__s0, __s1, __s2, __p3, __s4); \
	41961	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmlallbtq_laneq_f32_mf8_fpm(__s0, __s1, __s2, __p3, __s4)); \
41941	41962	__ret; \
41942	41963	})
41943	41964	#else
...	...	@@ -41947,11 +41968,11 @@ __ai __attribute__((target("fp8fma,neon"))) float32x4_t vmlallbtq_f32_mf8_fpm(fl
41947	41968	mfloat8x16_t __s1 = __p1; \
41948	41969	mfloat8x16_t __s2 = __p2; \
41949	41970	fpm_t __s4 = __p4; \
41950		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
41951		mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
41952		mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
41953		__ret = (float32x4_t) __builtin_neon_vmlallbtq_laneq_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3, __s4); \
41954		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	41971	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	41972	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	41973	mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_128_8); \
	41974	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmlallbtq_laneq_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3, __s4)); \
	41975	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
41955	41976	__ret; \
41956	41977	})
41957	41978	#endif
...	...	@@ -41959,17 +41980,17 @@ __ai __attribute__((target("fp8fma,neon"))) float32x4_t vmlallbtq_f32_mf8_fpm(fl
41959	41980	#ifdef __LITTLE_ENDIAN__
41960	41981	__ai __attribute__((target("fp8fma,neon"))) float32x4_t vmlalltbq_f32_mf8_fpm(float32x4_t __p0, mfloat8x16_t __p1, mfloat8x16_t __p2, fpm_t __p3) {
41961	41982	float32x4_t __ret;
41962		__ret = (float32x4_t) __builtin_neon_vmlalltbq_f32_mf8_fpm(__p0, __p1, __p2, __p3);
	41983	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmlalltbq_f32_mf8_fpm(__p0, __p1, __p2, __p3));
41963	41984	return __ret;
41964	41985	}
41965	41986	#else
41966	41987	__ai __attribute__((target("fp8fma,neon"))) float32x4_t vmlalltbq_f32_mf8_fpm(float32x4_t __p0, mfloat8x16_t __p1, mfloat8x16_t __p2, fpm_t __p3) {
41967	41988	float32x4_t __ret;
41968		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
41969		mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
41970		mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
41971		__ret = (float32x4_t) __builtin_neon_vmlalltbq_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3);
41972		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	41989	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	41990	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	41991	mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	41992	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmlalltbq_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3));
	41993	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
41973	41994	return __ret;
41974	41995	}
41975	41996	#endif
...	...	@@ -41981,7 +42002,7 @@ __ai __attribute__((target("fp8fma,neon"))) float32x4_t vmlalltbq_f32_mf8_fpm(fl
41981	42002	mfloat8x16_t __s1 = __p1; \
41982	42003	mfloat8x8_t __s2 = __p2; \
41983	42004	fpm_t __s4 = __p4; \
41984		__ret = (float32x4_t) __builtin_neon_vmlalltbq_lane_f32_mf8_fpm(__s0, __s1, __s2, __p3, __s4); \
	42005	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmlalltbq_lane_f32_mf8_fpm(__s0, __s1, __s2, __p3, __s4)); \
41985	42006	__ret; \
41986	42007	})
41987	42008	#else
...	...	@@ -41991,11 +42012,11 @@ __ai __attribute__((target("fp8fma,neon"))) float32x4_t vmlalltbq_f32_mf8_fpm(fl
41991	42012	mfloat8x16_t __s1 = __p1; \
41992	42013	mfloat8x8_t __s2 = __p2; \
41993	42014	fpm_t __s4 = __p4; \
41994		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
41995		mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
41996		mfloat8x8_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 7, 6, 5, 4, 3, 2, 1, 0); \
41997		__ret = (float32x4_t) __builtin_neon_vmlalltbq_lane_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3, __s4); \
41998		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	42015	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	42016	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	42017	mfloat8x8_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_64_8); \
	42018	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmlalltbq_lane_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3, __s4)); \
	42019	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
41999	42020	__ret; \
42000	42021	})
42001	42022	#endif
...	...	@@ -42007,7 +42028,7 @@ __ai __attribute__((target("fp8fma,neon"))) float32x4_t vmlalltbq_f32_mf8_fpm(fl
42007	42028	mfloat8x16_t __s1 = __p1; \
42008	42029	mfloat8x16_t __s2 = __p2; \
42009	42030	fpm_t __s4 = __p4; \
42010		__ret = (float32x4_t) __builtin_neon_vmlalltbq_laneq_f32_mf8_fpm(__s0, __s1, __s2, __p3, __s4); \
	42031	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmlalltbq_laneq_f32_mf8_fpm(__s0, __s1, __s2, __p3, __s4)); \
42011	42032	__ret; \
42012	42033	})
42013	42034	#else
...	...	@@ -42017,11 +42038,11 @@ __ai __attribute__((target("fp8fma,neon"))) float32x4_t vmlalltbq_f32_mf8_fpm(fl
42017	42038	mfloat8x16_t __s1 = __p1; \
42018	42039	mfloat8x16_t __s2 = __p2; \
42019	42040	fpm_t __s4 = __p4; \
42020		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
42021		mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42022		mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42023		__ret = (float32x4_t) __builtin_neon_vmlalltbq_laneq_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3, __s4); \
42024		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	42041	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	42042	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	42043	mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_128_8); \
	42044	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmlalltbq_laneq_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3, __s4)); \
	42045	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
42025	42046	__ret; \
42026	42047	})
42027	42048	#endif
...	...	@@ -42029,17 +42050,17 @@ __ai __attribute__((target("fp8fma,neon"))) float32x4_t vmlalltbq_f32_mf8_fpm(fl
42029	42050	#ifdef __LITTLE_ENDIAN__
42030	42051	__ai __attribute__((target("fp8fma,neon"))) float32x4_t vmlallttq_f32_mf8_fpm(float32x4_t __p0, mfloat8x16_t __p1, mfloat8x16_t __p2, fpm_t __p3) {
42031	42052	float32x4_t __ret;
42032		__ret = (float32x4_t) __builtin_neon_vmlallttq_f32_mf8_fpm(__p0, __p1, __p2, __p3);
	42053	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmlallttq_f32_mf8_fpm(__p0, __p1, __p2, __p3));
42033	42054	return __ret;
42034	42055	}
42035	42056	#else
42036	42057	__ai __attribute__((target("fp8fma,neon"))) float32x4_t vmlallttq_f32_mf8_fpm(float32x4_t __p0, mfloat8x16_t __p1, mfloat8x16_t __p2, fpm_t __p3) {
42037	42058	float32x4_t __ret;
42038		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
42039		mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
42040		mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
42041		__ret = (float32x4_t) __builtin_neon_vmlallttq_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3);
42042		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	42059	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	42060	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	42061	mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	42062	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmlallttq_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3));
	42063	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
42043	42064	return __ret;
42044	42065	}
42045	42066	#endif
...	...	@@ -42051,7 +42072,7 @@ __ai __attribute__((target("fp8fma,neon"))) float32x4_t vmlallttq_f32_mf8_fpm(fl
42051	42072	mfloat8x16_t __s1 = __p1; \
42052	42073	mfloat8x8_t __s2 = __p2; \
42053	42074	fpm_t __s4 = __p4; \
42054		__ret = (float32x4_t) __builtin_neon_vmlallttq_lane_f32_mf8_fpm(__s0, __s1, __s2, __p3, __s4); \
	42075	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmlallttq_lane_f32_mf8_fpm(__s0, __s1, __s2, __p3, __s4)); \
42055	42076	__ret; \
42056	42077	})
42057	42078	#else
...	...	@@ -42061,11 +42082,11 @@ __ai __attribute__((target("fp8fma,neon"))) float32x4_t vmlallttq_f32_mf8_fpm(fl
42061	42082	mfloat8x16_t __s1 = __p1; \
42062	42083	mfloat8x8_t __s2 = __p2; \
42063	42084	fpm_t __s4 = __p4; \
42064		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
42065		mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42066		mfloat8x8_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 7, 6, 5, 4, 3, 2, 1, 0); \
42067		__ret = (float32x4_t) __builtin_neon_vmlallttq_lane_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3, __s4); \
42068		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	42085	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	42086	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	42087	mfloat8x8_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_64_8); \
	42088	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmlallttq_lane_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3, __s4)); \
	42089	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
42069	42090	__ret; \
42070	42091	})
42071	42092	#endif
...	...	@@ -42077,7 +42098,7 @@ __ai __attribute__((target("fp8fma,neon"))) float32x4_t vmlallttq_f32_mf8_fpm(fl
42077	42098	mfloat8x16_t __s1 = __p1; \
42078	42099	mfloat8x16_t __s2 = __p2; \
42079	42100	fpm_t __s4 = __p4; \
42080		__ret = (float32x4_t) __builtin_neon_vmlallttq_laneq_f32_mf8_fpm(__s0, __s1, __s2, __p3, __s4); \
	42101	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmlallttq_laneq_f32_mf8_fpm(__s0, __s1, __s2, __p3, __s4)); \
42081	42102	__ret; \
42082	42103	})
42083	42104	#else
...	...	@@ -42087,11 +42108,11 @@ __ai __attribute__((target("fp8fma,neon"))) float32x4_t vmlallttq_f32_mf8_fpm(fl
42087	42108	mfloat8x16_t __s1 = __p1; \
42088	42109	mfloat8x16_t __s2 = __p2; \
42089	42110	fpm_t __s4 = __p4; \
42090		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
42091		mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42092		mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42093		__ret = (float32x4_t) __builtin_neon_vmlallttq_laneq_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3, __s4); \
42094		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	42111	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	42112	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	42113	mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_128_8); \
	42114	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmlallttq_laneq_f32_mf8_fpm(__rev0, __rev1, __rev2, __p3, __s4)); \
	42115	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
42095	42116	__ret; \
42096	42117	})
42097	42118	#endif
...	...	@@ -42099,17 +42120,17 @@ __ai __attribute__((target("fp8fma,neon"))) float32x4_t vmlallttq_f32_mf8_fpm(fl
42099	42120	#ifdef __LITTLE_ENDIAN__
42100	42121	__ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(float16x8_t __p0, mfloat8x16_t __p1, mfloat8x16_t __p2, fpm_t __p3) {
42101	42122	float16x8_t __ret;
42102		__ret = (float16x8_t) __builtin_neon_vmlaltq_f16_mf8_fpm((int8x16_t)__p0, __p1, __p2, __p3);
	42123	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vmlaltq_f16_mf8_fpm(__builtin_bit_cast(int8x16_t, __p0), __p1, __p2, __p3));
42103	42124	return __ret;
42104	42125	}
42105	42126	#else
42106	42127	__ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(float16x8_t __p0, mfloat8x16_t __p1, mfloat8x16_t __p2, fpm_t __p3) {
42107	42128	float16x8_t __ret;
42108		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
42109		mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
42110		mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
42111		__ret = (float16x8_t) __builtin_neon_vmlaltq_f16_mf8_fpm((int8x16_t)__rev0, __rev1, __rev2, __p3);
42112		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	42129	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	42130	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	42131	mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	42132	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vmlaltq_f16_mf8_fpm(__builtin_bit_cast(int8x16_t, __rev0), __rev1, __rev2, __p3));
	42133	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
42113	42134	return __ret;
42114	42135	}
42115	42136	#endif
...	...	@@ -42121,7 +42142,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42121	42142	mfloat8x16_t __s1 = __p1; \
42122	42143	mfloat8x8_t __s2 = __p2; \
42123	42144	fpm_t __s4 = __p4; \
42124		__ret = (float16x8_t) __builtin_neon_vmlaltq_lane_f16_mf8_fpm((int8x16_t)__s0, __s1, __s2, __p3, __s4); \
	42145	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vmlaltq_lane_f16_mf8_fpm(__builtin_bit_cast(int8x16_t, __s0), __s1, __s2, __p3, __s4)); \
42125	42146	__ret; \
42126	42147	})
42127	42148	#else
...	...	@@ -42131,11 +42152,11 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42131	42152	mfloat8x16_t __s1 = __p1; \
42132	42153	mfloat8x8_t __s2 = __p2; \
42133	42154	fpm_t __s4 = __p4; \
42134		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
42135		mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42136		mfloat8x8_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 7, 6, 5, 4, 3, 2, 1, 0); \
42137		__ret = (float16x8_t) __builtin_neon_vmlaltq_lane_f16_mf8_fpm((int8x16_t)__rev0, __rev1, __rev2, __p3, __s4); \
42138		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	42155	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	42156	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	42157	mfloat8x8_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_64_8); \
	42158	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vmlaltq_lane_f16_mf8_fpm(__builtin_bit_cast(int8x16_t, __rev0), __rev1, __rev2, __p3, __s4)); \
	42159	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
42139	42160	__ret; \
42140	42161	})
42141	42162	#endif
...	...	@@ -42147,7 +42168,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42147	42168	mfloat8x16_t __s1 = __p1; \
42148	42169	mfloat8x16_t __s2 = __p2; \
42149	42170	fpm_t __s4 = __p4; \
42150		__ret = (float16x8_t) __builtin_neon_vmlaltq_laneq_f16_mf8_fpm((int8x16_t)__s0, __s1, __s2, __p3, __s4); \
	42171	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vmlaltq_laneq_f16_mf8_fpm(__builtin_bit_cast(int8x16_t, __s0), __s1, __s2, __p3, __s4)); \
42151	42172	__ret; \
42152	42173	})
42153	42174	#else
...	...	@@ -42157,11 +42178,11 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42157	42178	mfloat8x16_t __s1 = __p1; \
42158	42179	mfloat8x16_t __s2 = __p2; \
42159	42180	fpm_t __s4 = __p4; \
42160		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
42161		mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42162		mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42163		__ret = (float16x8_t) __builtin_neon_vmlaltq_laneq_f16_mf8_fpm((int8x16_t)__rev0, __rev1, __rev2, __p3, __s4); \
42164		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	42181	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	42182	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	42183	mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_128_8); \
	42184	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vmlaltq_laneq_f16_mf8_fpm(__builtin_bit_cast(int8x16_t, __rev0), __rev1, __rev2, __p3, __s4)); \
	42185	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
42165	42186	__ret; \
42166	42187	})
42167	42188	#endif
...	...	@@ -42171,7 +42192,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42171	42192	poly8x16_t __ret; \
42172	42193	poly8x8_t __s0 = __p0; \
42173	42194	uint8x8_t __s1 = __p1; \
42174		__ret = (poly8x16_t) __builtin_neon_vluti2_lane_p8((int8x8_t)__s0, (int8x8_t)__s1, __p2, 36); \
	42195	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vluti2_lane_p8(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 36)); \
42175	42196	__ret; \
42176	42197	})
42177	42198	#else
...	...	@@ -42179,10 +42200,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42179	42200	poly8x16_t __ret; \
42180	42201	poly8x8_t __s0 = __p0; \
42181	42202	uint8x8_t __s1 = __p1; \
42182		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
42183		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
42184		__ret = (poly8x16_t) __builtin_neon_vluti2_lane_p8((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 36); \
42185		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	42203	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	42204	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	42205	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vluti2_lane_p8(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 36)); \
	42206	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
42186	42207	__ret; \
42187	42208	})
42188	42209	#endif
...	...	@@ -42192,7 +42213,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42192	42213	poly8x16_t __ret; \
42193	42214	poly8x16_t __s0 = __p0; \
42194	42215	uint8x8_t __s1 = __p1; \
42195		__ret = (poly8x16_t) __builtin_neon_vluti2q_lane_p8((int8x16_t)__s0, (int8x8_t)__s1, __p2, 36); \
	42216	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vluti2q_lane_p8(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 36)); \
42196	42217	__ret; \
42197	42218	})
42198	42219	#else
...	...	@@ -42200,10 +42221,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42200	42221	poly8x16_t __ret; \
42201	42222	poly8x16_t __s0 = __p0; \
42202	42223	uint8x8_t __s1 = __p1; \
42203		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42204		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
42205		__ret = (poly8x16_t) __builtin_neon_vluti2q_lane_p8((int8x16_t)__rev0, (int8x8_t)__rev1, __p2, 36); \
42206		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	42224	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	42225	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	42226	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vluti2q_lane_p8(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 36)); \
	42227	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
42207	42228	__ret; \
42208	42229	})
42209	42230	#endif
...	...	@@ -42213,7 +42234,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42213	42234	uint8x16_t __ret; \
42214	42235	uint8x16_t __s0 = __p0; \
42215	42236	uint8x8_t __s1 = __p1; \
42216		__ret = (uint8x16_t) __builtin_neon_vluti2q_lane_u8((int8x16_t)__s0, (int8x8_t)__s1, __p2, 48); \
	42237	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vluti2q_lane_u8(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 48)); \
42217	42238	__ret; \
42218	42239	})
42219	42240	#else
...	...	@@ -42221,10 +42242,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42221	42242	uint8x16_t __ret; \
42222	42243	uint8x16_t __s0 = __p0; \
42223	42244	uint8x8_t __s1 = __p1; \
42224		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42225		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
42226		__ret = (uint8x16_t) __builtin_neon_vluti2q_lane_u8((int8x16_t)__rev0, (int8x8_t)__rev1, __p2, 48); \
42227		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	42245	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	42246	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	42247	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vluti2q_lane_u8(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 48)); \
	42248	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
42228	42249	__ret; \
42229	42250	})
42230	42251	#endif
...	...	@@ -42234,7 +42255,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42234	42255	int8x16_t __ret; \
42235	42256	int8x16_t __s0 = __p0; \
42236	42257	uint8x8_t __s1 = __p1; \
42237		__ret = (int8x16_t) __builtin_neon_vluti2q_lane_s8((int8x16_t)__s0, (int8x8_t)__s1, __p2, 32); \
	42258	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vluti2q_lane_s8(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 32)); \
42238	42259	__ret; \
42239	42260	})
42240	42261	#else
...	...	@@ -42242,10 +42263,31 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42242	42263	int8x16_t __ret; \
42243	42264	int8x16_t __s0 = __p0; \
42244	42265	uint8x8_t __s1 = __p1; \
42245		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42246		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
42247		__ret = (int8x16_t) __builtin_neon_vluti2q_lane_s8((int8x16_t)__rev0, (int8x8_t)__rev1, __p2, 32); \
42248		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	42266	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	42267	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	42268	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vluti2q_lane_s8(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 32)); \
	42269	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
	42270	__ret; \
	42271	})
	42272	#endif
	42273
	42274	#ifdef __LITTLE_ENDIAN__
	42275	#define vluti2q_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	42276	mfloat8x16_t __ret; \
	42277	mfloat8x16_t __s0 = __p0; \
	42278	uint8x8_t __s1 = __p1; \
	42279	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vluti2q_lane_mf8(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 44)); \
	42280	__ret; \
	42281	})
	42282	#else
	42283	#define vluti2q_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	42284	mfloat8x16_t __ret; \
	42285	mfloat8x16_t __s0 = __p0; \
	42286	uint8x8_t __s1 = __p1; \
	42287	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	42288	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	42289	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vluti2q_lane_mf8(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 44)); \
	42290	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
42249	42291	__ret; \
42250	42292	})
42251	42293	#endif
...	...	@@ -42255,7 +42297,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42255	42297	uint8x16_t __ret; \
42256	42298	uint8x8_t __s0 = __p0; \
42257	42299	uint8x8_t __s1 = __p1; \
42258		__ret = (uint8x16_t) __builtin_neon_vluti2_lane_u8((int8x8_t)__s0, (int8x8_t)__s1, __p2, 48); \
	42300	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vluti2_lane_u8(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 48)); \
42259	42301	__ret; \
42260	42302	})
42261	42303	#else
...	...	@@ -42263,10 +42305,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42263	42305	uint8x16_t __ret; \
42264	42306	uint8x8_t __s0 = __p0; \
42265	42307	uint8x8_t __s1 = __p1; \
42266		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
42267		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
42268		__ret = (uint8x16_t) __builtin_neon_vluti2_lane_u8((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 48); \
42269		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	42308	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	42309	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	42310	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vluti2_lane_u8(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 48)); \
	42311	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
42270	42312	__ret; \
42271	42313	})
42272	42314	#endif
...	...	@@ -42276,7 +42318,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42276	42318	int8x16_t __ret; \
42277	42319	int8x8_t __s0 = __p0; \
42278	42320	uint8x8_t __s1 = __p1; \
42279		__ret = (int8x16_t) __builtin_neon_vluti2_lane_s8((int8x8_t)__s0, (int8x8_t)__s1, __p2, 32); \
	42321	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vluti2_lane_s8(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 32)); \
42280	42322	__ret; \
42281	42323	})
42282	42324	#else
...	...	@@ -42284,10 +42326,31 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42284	42326	int8x16_t __ret; \
42285	42327	int8x8_t __s0 = __p0; \
42286	42328	uint8x8_t __s1 = __p1; \
42287		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
42288		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
42289		__ret = (int8x16_t) __builtin_neon_vluti2_lane_s8((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 32); \
42290		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	42329	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	42330	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	42331	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vluti2_lane_s8(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 32)); \
	42332	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
	42333	__ret; \
	42334	})
	42335	#endif
	42336
	42337	#ifdef __LITTLE_ENDIAN__
	42338	#define vluti2_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	42339	mfloat8x16_t __ret; \
	42340	mfloat8x8_t __s0 = __p0; \
	42341	uint8x8_t __s1 = __p1; \
	42342	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vluti2_lane_mf8(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 44)); \
	42343	__ret; \
	42344	})
	42345	#else
	42346	#define vluti2_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	42347	mfloat8x16_t __ret; \
	42348	mfloat8x8_t __s0 = __p0; \
	42349	uint8x8_t __s1 = __p1; \
	42350	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	42351	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	42352	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vluti2_lane_mf8(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 44)); \
	42353	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
42291	42354	__ret; \
42292	42355	})
42293	42356	#endif
...	...	@@ -42297,7 +42360,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42297	42360	poly16x8_t __ret; \
42298	42361	poly16x4_t __s0 = __p0; \
42299	42362	uint8x8_t __s1 = __p1; \
42300		__ret = (poly16x8_t) __builtin_neon_vluti2_lane_p16((int8x8_t)__s0, (int8x8_t)__s1, __p2, 37); \
	42363	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vluti2_lane_p16(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 37)); \
42301	42364	__ret; \
42302	42365	})
42303	42366	#else
...	...	@@ -42305,10 +42368,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42305	42368	poly16x8_t __ret; \
42306	42369	poly16x4_t __s0 = __p0; \
42307	42370	uint8x8_t __s1 = __p1; \
42308		poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
42309		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
42310		__ret = (poly16x8_t) __builtin_neon_vluti2_lane_p16((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 37); \
42311		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	42371	poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	42372	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	42373	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vluti2_lane_p16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 37)); \
	42374	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
42312	42375	__ret; \
42313	42376	})
42314	42377	#endif
...	...	@@ -42318,7 +42381,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42318	42381	poly16x8_t __ret; \
42319	42382	poly16x8_t __s0 = __p0; \
42320	42383	uint8x8_t __s1 = __p1; \
42321		__ret = (poly16x8_t) __builtin_neon_vluti2q_lane_p16((int8x16_t)__s0, (int8x8_t)__s1, __p2, 37); \
	42384	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vluti2q_lane_p16(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 37)); \
42322	42385	__ret; \
42323	42386	})
42324	42387	#else
...	...	@@ -42326,10 +42389,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42326	42389	poly16x8_t __ret; \
42327	42390	poly16x8_t __s0 = __p0; \
42328	42391	uint8x8_t __s1 = __p1; \
42329		poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
42330		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
42331		__ret = (poly16x8_t) __builtin_neon_vluti2q_lane_p16((int8x16_t)__rev0, (int8x8_t)__rev1, __p2, 37); \
42332		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	42392	poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	42393	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	42394	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vluti2q_lane_p16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 37)); \
	42395	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
42333	42396	__ret; \
42334	42397	})
42335	42398	#endif
...	...	@@ -42339,7 +42402,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42339	42402	uint16x8_t __ret; \
42340	42403	uint16x8_t __s0 = __p0; \
42341	42404	uint8x8_t __s1 = __p1; \
42342		__ret = (uint16x8_t) __builtin_neon_vluti2q_lane_u16((int8x16_t)__s0, (int8x8_t)__s1, __p2, 49); \
	42405	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vluti2q_lane_u16(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 49)); \
42343	42406	__ret; \
42344	42407	})
42345	42408	#else
...	...	@@ -42347,10 +42410,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42347	42410	uint16x8_t __ret; \
42348	42411	uint16x8_t __s0 = __p0; \
42349	42412	uint8x8_t __s1 = __p1; \
42350		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
42351		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
42352		__ret = (uint16x8_t) __builtin_neon_vluti2q_lane_u16((int8x16_t)__rev0, (int8x8_t)__rev1, __p2, 49); \
42353		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	42413	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	42414	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	42415	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vluti2q_lane_u16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 49)); \
	42416	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
42354	42417	__ret; \
42355	42418	})
42356	42419	#endif
...	...	@@ -42360,7 +42423,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42360	42423	float16x8_t __ret; \
42361	42424	float16x8_t __s0 = __p0; \
42362	42425	uint8x8_t __s1 = __p1; \
42363		__ret = (float16x8_t) __builtin_neon_vluti2q_lane_f16((int8x16_t)__s0, (int8x8_t)__s1, __p2, 40); \
	42426	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vluti2q_lane_f16(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 40)); \
42364	42427	__ret; \
42365	42428	})
42366	42429	#else
...	...	@@ -42368,10 +42431,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42368	42431	float16x8_t __ret; \
42369	42432	float16x8_t __s0 = __p0; \
42370	42433	uint8x8_t __s1 = __p1; \
42371		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
42372		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
42373		__ret = (float16x8_t) __builtin_neon_vluti2q_lane_f16((int8x16_t)__rev0, (int8x8_t)__rev1, __p2, 40); \
42374		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	42434	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	42435	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	42436	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vluti2q_lane_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 40)); \
	42437	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
42375	42438	__ret; \
42376	42439	})
42377	42440	#endif
...	...	@@ -42381,7 +42444,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42381	42444	int16x8_t __ret; \
42382	42445	int16x8_t __s0 = __p0; \
42383	42446	uint8x8_t __s1 = __p1; \
42384		__ret = (int16x8_t) __builtin_neon_vluti2q_lane_s16((int8x16_t)__s0, (int8x8_t)__s1, __p2, 33); \
	42447	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vluti2q_lane_s16(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 33)); \
42385	42448	__ret; \
42386	42449	})
42387	42450	#else
...	...	@@ -42389,10 +42452,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42389	42452	int16x8_t __ret; \
42390	42453	int16x8_t __s0 = __p0; \
42391	42454	uint8x8_t __s1 = __p1; \
42392		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
42393		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
42394		__ret = (int16x8_t) __builtin_neon_vluti2q_lane_s16((int8x16_t)__rev0, (int8x8_t)__rev1, __p2, 33); \
42395		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	42455	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	42456	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	42457	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vluti2q_lane_s16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 33)); \
	42458	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
42396	42459	__ret; \
42397	42460	})
42398	42461	#endif
...	...	@@ -42402,7 +42465,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42402	42465	uint16x8_t __ret; \
42403	42466	uint16x4_t __s0 = __p0; \
42404	42467	uint8x8_t __s1 = __p1; \
42405		__ret = (uint16x8_t) __builtin_neon_vluti2_lane_u16((int8x8_t)__s0, (int8x8_t)__s1, __p2, 49); \
	42468	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vluti2_lane_u16(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 49)); \
42406	42469	__ret; \
42407	42470	})
42408	42471	#else
...	...	@@ -42410,10 +42473,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42410	42473	uint16x8_t __ret; \
42411	42474	uint16x4_t __s0 = __p0; \
42412	42475	uint8x8_t __s1 = __p1; \
42413		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
42414		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
42415		__ret = (uint16x8_t) __builtin_neon_vluti2_lane_u16((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 49); \
42416		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	42476	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	42477	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	42478	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vluti2_lane_u16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 49)); \
	42479	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
42417	42480	__ret; \
42418	42481	})
42419	42482	#endif
...	...	@@ -42423,7 +42486,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42423	42486	float16x8_t __ret; \
42424	42487	float16x4_t __s0 = __p0; \
42425	42488	uint8x8_t __s1 = __p1; \
42426		__ret = (float16x8_t) __builtin_neon_vluti2_lane_f16((int8x8_t)__s0, (int8x8_t)__s1, __p2, 40); \
	42489	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vluti2_lane_f16(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 40)); \
42427	42490	__ret; \
42428	42491	})
42429	42492	#else
...	...	@@ -42431,10 +42494,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42431	42494	float16x8_t __ret; \
42432	42495	float16x4_t __s0 = __p0; \
42433	42496	uint8x8_t __s1 = __p1; \
42434		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
42435		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
42436		__ret = (float16x8_t) __builtin_neon_vluti2_lane_f16((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 40); \
42437		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	42497	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	42498	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	42499	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vluti2_lane_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 40)); \
	42500	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
42438	42501	__ret; \
42439	42502	})
42440	42503	#endif
...	...	@@ -42444,7 +42507,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42444	42507	int16x8_t __ret; \
42445	42508	int16x4_t __s0 = __p0; \
42446	42509	uint8x8_t __s1 = __p1; \
42447		__ret = (int16x8_t) __builtin_neon_vluti2_lane_s16((int8x8_t)__s0, (int8x8_t)__s1, __p2, 33); \
	42510	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vluti2_lane_s16(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 33)); \
42448	42511	__ret; \
42449	42512	})
42450	42513	#else
...	...	@@ -42452,10 +42515,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42452	42515	int16x8_t __ret; \
42453	42516	int16x4_t __s0 = __p0; \
42454	42517	uint8x8_t __s1 = __p1; \
42455		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
42456		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
42457		__ret = (int16x8_t) __builtin_neon_vluti2_lane_s16((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 33); \
42458		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	42518	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	42519	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	42520	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vluti2_lane_s16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 33)); \
	42521	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
42459	42522	__ret; \
42460	42523	})
42461	42524	#endif
...	...	@@ -42465,7 +42528,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42465	42528	poly8x16_t __ret; \
42466	42529	poly8x8_t __s0 = __p0; \
42467	42530	uint8x16_t __s1 = __p1; \
42468		__ret = (poly8x16_t) __builtin_neon_vluti2_laneq_p8((int8x8_t)__s0, (int8x16_t)__s1, __p2, 36); \
	42531	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vluti2_laneq_p8(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 36)); \
42469	42532	__ret; \
42470	42533	})
42471	42534	#else
...	...	@@ -42473,10 +42536,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42473	42536	poly8x16_t __ret; \
42474	42537	poly8x8_t __s0 = __p0; \
42475	42538	uint8x16_t __s1 = __p1; \
42476		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
42477		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42478		__ret = (poly8x16_t) __builtin_neon_vluti2_laneq_p8((int8x8_t)__rev0, (int8x16_t)__rev1, __p2, 36); \
42479		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	42539	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	42540	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	42541	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vluti2_laneq_p8(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 36)); \
	42542	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
42480	42543	__ret; \
42481	42544	})
42482	42545	#endif
...	...	@@ -42486,7 +42549,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42486	42549	poly8x16_t __ret; \
42487	42550	poly8x16_t __s0 = __p0; \
42488	42551	uint8x16_t __s1 = __p1; \
42489		__ret = (poly8x16_t) __builtin_neon_vluti2q_laneq_p8((int8x16_t)__s0, (int8x16_t)__s1, __p2, 36); \
	42552	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vluti2q_laneq_p8(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 36)); \
42490	42553	__ret; \
42491	42554	})
42492	42555	#else
...	...	@@ -42494,10 +42557,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42494	42557	poly8x16_t __ret; \
42495	42558	poly8x16_t __s0 = __p0; \
42496	42559	uint8x16_t __s1 = __p1; \
42497		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42498		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42499		__ret = (poly8x16_t) __builtin_neon_vluti2q_laneq_p8((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 36); \
42500		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	42560	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	42561	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	42562	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vluti2q_laneq_p8(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 36)); \
	42563	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
42501	42564	__ret; \
42502	42565	})
42503	42566	#endif
...	...	@@ -42507,7 +42570,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42507	42570	uint8x16_t __ret; \
42508	42571	uint8x16_t __s0 = __p0; \
42509	42572	uint8x16_t __s1 = __p1; \
42510		__ret = (uint8x16_t) __builtin_neon_vluti2q_laneq_u8((int8x16_t)__s0, (int8x16_t)__s1, __p2, 48); \
	42573	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vluti2q_laneq_u8(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 48)); \
42511	42574	__ret; \
42512	42575	})
42513	42576	#else
...	...	@@ -42515,10 +42578,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42515	42578	uint8x16_t __ret; \
42516	42579	uint8x16_t __s0 = __p0; \
42517	42580	uint8x16_t __s1 = __p1; \
42518		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42519		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42520		__ret = (uint8x16_t) __builtin_neon_vluti2q_laneq_u8((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 48); \
42521		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	42581	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	42582	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	42583	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vluti2q_laneq_u8(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 48)); \
	42584	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
42522	42585	__ret; \
42523	42586	})
42524	42587	#endif
...	...	@@ -42528,7 +42591,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42528	42591	int8x16_t __ret; \
42529	42592	int8x16_t __s0 = __p0; \
42530	42593	uint8x16_t __s1 = __p1; \
42531		__ret = (int8x16_t) __builtin_neon_vluti2q_laneq_s8((int8x16_t)__s0, (int8x16_t)__s1, __p2, 32); \
	42594	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vluti2q_laneq_s8(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 32)); \
42532	42595	__ret; \
42533	42596	})
42534	42597	#else
...	...	@@ -42536,10 +42599,31 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42536	42599	int8x16_t __ret; \
42537	42600	int8x16_t __s0 = __p0; \
42538	42601	uint8x16_t __s1 = __p1; \
42539		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42540		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42541		__ret = (int8x16_t) __builtin_neon_vluti2q_laneq_s8((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 32); \
42542		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	42602	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	42603	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	42604	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vluti2q_laneq_s8(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 32)); \
	42605	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
	42606	__ret; \
	42607	})
	42608	#endif
	42609
	42610	#ifdef __LITTLE_ENDIAN__
	42611	#define vluti2q_laneq_mf8(__p0, __p1, __p2) __extension__ ({ \
	42612	mfloat8x16_t __ret; \
	42613	mfloat8x16_t __s0 = __p0; \
	42614	uint8x16_t __s1 = __p1; \
	42615	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vluti2q_laneq_mf8(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 44)); \
	42616	__ret; \
	42617	})
	42618	#else
	42619	#define vluti2q_laneq_mf8(__p0, __p1, __p2) __extension__ ({ \
	42620	mfloat8x16_t __ret; \
	42621	mfloat8x16_t __s0 = __p0; \
	42622	uint8x16_t __s1 = __p1; \
	42623	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	42624	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	42625	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vluti2q_laneq_mf8(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 44)); \
	42626	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
42543	42627	__ret; \
42544	42628	})
42545	42629	#endif
...	...	@@ -42549,7 +42633,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42549	42633	uint8x16_t __ret; \
42550	42634	uint8x8_t __s0 = __p0; \
42551	42635	uint8x16_t __s1 = __p1; \
42552		__ret = (uint8x16_t) __builtin_neon_vluti2_laneq_u8((int8x8_t)__s0, (int8x16_t)__s1, __p2, 48); \
	42636	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vluti2_laneq_u8(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 48)); \
42553	42637	__ret; \
42554	42638	})
42555	42639	#else
...	...	@@ -42557,10 +42641,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42557	42641	uint8x16_t __ret; \
42558	42642	uint8x8_t __s0 = __p0; \
42559	42643	uint8x16_t __s1 = __p1; \
42560		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
42561		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42562		__ret = (uint8x16_t) __builtin_neon_vluti2_laneq_u8((int8x8_t)__rev0, (int8x16_t)__rev1, __p2, 48); \
42563		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	42644	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	42645	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	42646	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vluti2_laneq_u8(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 48)); \
	42647	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
42564	42648	__ret; \
42565	42649	})
42566	42650	#endif
...	...	@@ -42570,7 +42654,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42570	42654	int8x16_t __ret; \
42571	42655	int8x8_t __s0 = __p0; \
42572	42656	uint8x16_t __s1 = __p1; \
42573		__ret = (int8x16_t) __builtin_neon_vluti2_laneq_s8((int8x8_t)__s0, (int8x16_t)__s1, __p2, 32); \
	42657	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vluti2_laneq_s8(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 32)); \
42574	42658	__ret; \
42575	42659	})
42576	42660	#else
...	...	@@ -42578,10 +42662,31 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42578	42662	int8x16_t __ret; \
42579	42663	int8x8_t __s0 = __p0; \
42580	42664	uint8x16_t __s1 = __p1; \
42581		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
42582		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42583		__ret = (int8x16_t) __builtin_neon_vluti2_laneq_s8((int8x8_t)__rev0, (int8x16_t)__rev1, __p2, 32); \
42584		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	42665	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	42666	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	42667	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vluti2_laneq_s8(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 32)); \
	42668	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
	42669	__ret; \
	42670	})
	42671	#endif
	42672
	42673	#ifdef __LITTLE_ENDIAN__
	42674	#define vluti2_laneq_mf8(__p0, __p1, __p2) __extension__ ({ \
	42675	mfloat8x16_t __ret; \
	42676	mfloat8x8_t __s0 = __p0; \
	42677	uint8x16_t __s1 = __p1; \
	42678	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vluti2_laneq_mf8(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 44)); \
	42679	__ret; \
	42680	})
	42681	#else
	42682	#define vluti2_laneq_mf8(__p0, __p1, __p2) __extension__ ({ \
	42683	mfloat8x16_t __ret; \
	42684	mfloat8x8_t __s0 = __p0; \
	42685	uint8x16_t __s1 = __p1; \
	42686	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	42687	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	42688	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vluti2_laneq_mf8(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 44)); \
	42689	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
42585	42690	__ret; \
42586	42691	})
42587	42692	#endif
...	...	@@ -42591,7 +42696,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42591	42696	poly16x8_t __ret; \
42592	42697	poly16x4_t __s0 = __p0; \
42593	42698	uint8x16_t __s1 = __p1; \
42594		__ret = (poly16x8_t) __builtin_neon_vluti2_laneq_p16((int8x8_t)__s0, (int8x16_t)__s1, __p2, 37); \
	42699	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vluti2_laneq_p16(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 37)); \
42595	42700	__ret; \
42596	42701	})
42597	42702	#else
...	...	@@ -42599,10 +42704,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42599	42704	poly16x8_t __ret; \
42600	42705	poly16x4_t __s0 = __p0; \
42601	42706	uint8x16_t __s1 = __p1; \
42602		poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
42603		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42604		__ret = (poly16x8_t) __builtin_neon_vluti2_laneq_p16((int8x8_t)__rev0, (int8x16_t)__rev1, __p2, 37); \
42605		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	42707	poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	42708	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	42709	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vluti2_laneq_p16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 37)); \
	42710	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
42606	42711	__ret; \
42607	42712	})
42608	42713	#endif
...	...	@@ -42612,7 +42717,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42612	42717	poly16x8_t __ret; \
42613	42718	poly16x8_t __s0 = __p0; \
42614	42719	uint8x16_t __s1 = __p1; \
42615		__ret = (poly16x8_t) __builtin_neon_vluti2q_laneq_p16((int8x16_t)__s0, (int8x16_t)__s1, __p2, 37); \
	42720	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vluti2q_laneq_p16(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 37)); \
42616	42721	__ret; \
42617	42722	})
42618	42723	#else
...	...	@@ -42620,10 +42725,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42620	42725	poly16x8_t __ret; \
42621	42726	poly16x8_t __s0 = __p0; \
42622	42727	uint8x16_t __s1 = __p1; \
42623		poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
42624		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42625		__ret = (poly16x8_t) __builtin_neon_vluti2q_laneq_p16((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 37); \
42626		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	42728	poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	42729	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	42730	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vluti2q_laneq_p16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 37)); \
	42731	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
42627	42732	__ret; \
42628	42733	})
42629	42734	#endif
...	...	@@ -42633,7 +42738,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42633	42738	uint16x8_t __ret; \
42634	42739	uint16x8_t __s0 = __p0; \
42635	42740	uint8x16_t __s1 = __p1; \
42636		__ret = (uint16x8_t) __builtin_neon_vluti2q_laneq_u16((int8x16_t)__s0, (int8x16_t)__s1, __p2, 49); \
	42741	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vluti2q_laneq_u16(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 49)); \
42637	42742	__ret; \
42638	42743	})
42639	42744	#else
...	...	@@ -42641,10 +42746,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42641	42746	uint16x8_t __ret; \
42642	42747	uint16x8_t __s0 = __p0; \
42643	42748	uint8x16_t __s1 = __p1; \
42644		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
42645		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42646		__ret = (uint16x8_t) __builtin_neon_vluti2q_laneq_u16((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 49); \
42647		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	42749	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	42750	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	42751	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vluti2q_laneq_u16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 49)); \
	42752	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
42648	42753	__ret; \
42649	42754	})
42650	42755	#endif
...	...	@@ -42654,7 +42759,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42654	42759	float16x8_t __ret; \
42655	42760	float16x8_t __s0 = __p0; \
42656	42761	uint8x16_t __s1 = __p1; \
42657		__ret = (float16x8_t) __builtin_neon_vluti2q_laneq_f16((int8x16_t)__s0, (int8x16_t)__s1, __p2, 40); \
	42762	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vluti2q_laneq_f16(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 40)); \
42658	42763	__ret; \
42659	42764	})
42660	42765	#else
...	...	@@ -42662,10 +42767,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42662	42767	float16x8_t __ret; \
42663	42768	float16x8_t __s0 = __p0; \
42664	42769	uint8x16_t __s1 = __p1; \
42665		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
42666		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42667		__ret = (float16x8_t) __builtin_neon_vluti2q_laneq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 40); \
42668		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	42770	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	42771	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	42772	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vluti2q_laneq_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 40)); \
	42773	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
42669	42774	__ret; \
42670	42775	})
42671	42776	#endif
...	...	@@ -42675,7 +42780,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42675	42780	int16x8_t __ret; \
42676	42781	int16x8_t __s0 = __p0; \
42677	42782	uint8x16_t __s1 = __p1; \
42678		__ret = (int16x8_t) __builtin_neon_vluti2q_laneq_s16((int8x16_t)__s0, (int8x16_t)__s1, __p2, 33); \
	42783	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vluti2q_laneq_s16(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 33)); \
42679	42784	__ret; \
42680	42785	})
42681	42786	#else
...	...	@@ -42683,10 +42788,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42683	42788	int16x8_t __ret; \
42684	42789	int16x8_t __s0 = __p0; \
42685	42790	uint8x16_t __s1 = __p1; \
42686		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
42687		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42688		__ret = (int16x8_t) __builtin_neon_vluti2q_laneq_s16((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 33); \
42689		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	42791	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	42792	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	42793	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vluti2q_laneq_s16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 33)); \
	42794	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
42690	42795	__ret; \
42691	42796	})
42692	42797	#endif
...	...	@@ -42696,7 +42801,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42696	42801	uint16x8_t __ret; \
42697	42802	uint16x4_t __s0 = __p0; \
42698	42803	uint8x16_t __s1 = __p1; \
42699		__ret = (uint16x8_t) __builtin_neon_vluti2_laneq_u16((int8x8_t)__s0, (int8x16_t)__s1, __p2, 49); \
	42804	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vluti2_laneq_u16(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 49)); \
42700	42805	__ret; \
42701	42806	})
42702	42807	#else
...	...	@@ -42704,10 +42809,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42704	42809	uint16x8_t __ret; \
42705	42810	uint16x4_t __s0 = __p0; \
42706	42811	uint8x16_t __s1 = __p1; \
42707		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
42708		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42709		__ret = (uint16x8_t) __builtin_neon_vluti2_laneq_u16((int8x8_t)__rev0, (int8x16_t)__rev1, __p2, 49); \
42710		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	42812	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	42813	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	42814	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vluti2_laneq_u16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 49)); \
	42815	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
42711	42816	__ret; \
42712	42817	})
42713	42818	#endif
...	...	@@ -42717,7 +42822,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42717	42822	float16x8_t __ret; \
42718	42823	float16x4_t __s0 = __p0; \
42719	42824	uint8x16_t __s1 = __p1; \
42720		__ret = (float16x8_t) __builtin_neon_vluti2_laneq_f16((int8x8_t)__s0, (int8x16_t)__s1, __p2, 40); \
	42825	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vluti2_laneq_f16(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 40)); \
42721	42826	__ret; \
42722	42827	})
42723	42828	#else
...	...	@@ -42725,10 +42830,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42725	42830	float16x8_t __ret; \
42726	42831	float16x4_t __s0 = __p0; \
42727	42832	uint8x16_t __s1 = __p1; \
42728		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
42729		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42730		__ret = (float16x8_t) __builtin_neon_vluti2_laneq_f16((int8x8_t)__rev0, (int8x16_t)__rev1, __p2, 40); \
42731		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	42833	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	42834	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	42835	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vluti2_laneq_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 40)); \
	42836	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
42732	42837	__ret; \
42733	42838	})
42734	42839	#endif
...	...	@@ -42738,7 +42843,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42738	42843	int16x8_t __ret; \
42739	42844	int16x4_t __s0 = __p0; \
42740	42845	uint8x16_t __s1 = __p1; \
42741		__ret = (int16x8_t) __builtin_neon_vluti2_laneq_s16((int8x8_t)__s0, (int8x16_t)__s1, __p2, 33); \
	42846	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vluti2_laneq_s16(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 33)); \
42742	42847	__ret; \
42743	42848	})
42744	42849	#else
...	...	@@ -42746,10 +42851,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42746	42851	int16x8_t __ret; \
42747	42852	int16x4_t __s0 = __p0; \
42748	42853	uint8x16_t __s1 = __p1; \
42749		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
42750		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42751		__ret = (int16x8_t) __builtin_neon_vluti2_laneq_s16((int8x8_t)__rev0, (int8x16_t)__rev1, __p2, 33); \
42752		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	42854	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	42855	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	42856	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vluti2_laneq_s16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 33)); \
	42857	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
42753	42858	__ret; \
42754	42859	})
42755	42860	#endif
...	...	@@ -42759,7 +42864,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42759	42864	poly8x16_t __ret; \
42760	42865	poly8x16_t __s0 = __p0; \
42761	42866	uint8x8_t __s1 = __p1; \
42762		__ret = (poly8x16_t) __builtin_neon_vluti4q_lane_p8((int8x16_t)__s0, (int8x8_t)__s1, __p2, 36); \
	42867	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vluti4q_lane_p8(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 36)); \
42763	42868	__ret; \
42764	42869	})
42765	42870	#else
...	...	@@ -42767,10 +42872,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42767	42872	poly8x16_t __ret; \
42768	42873	poly8x16_t __s0 = __p0; \
42769	42874	uint8x8_t __s1 = __p1; \
42770		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42771		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
42772		__ret = (poly8x16_t) __builtin_neon_vluti4q_lane_p8((int8x16_t)__rev0, (int8x8_t)__rev1, __p2, 36); \
42773		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	42875	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	42876	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	42877	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vluti4q_lane_p8(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 36)); \
	42878	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
42774	42879	__ret; \
42775	42880	})
42776	42881	#endif
...	...	@@ -42780,7 +42885,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42780	42885	uint8x16_t __ret; \
42781	42886	uint8x16_t __s0 = __p0; \
42782	42887	uint8x8_t __s1 = __p1; \
42783		__ret = (uint8x16_t) __builtin_neon_vluti4q_lane_u8((int8x16_t)__s0, (int8x8_t)__s1, __p2, 48); \
	42888	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vluti4q_lane_u8(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 48)); \
42784	42889	__ret; \
42785	42890	})
42786	42891	#else
...	...	@@ -42788,10 +42893,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42788	42893	uint8x16_t __ret; \
42789	42894	uint8x16_t __s0 = __p0; \
42790	42895	uint8x8_t __s1 = __p1; \
42791		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42792		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
42793		__ret = (uint8x16_t) __builtin_neon_vluti4q_lane_u8((int8x16_t)__rev0, (int8x8_t)__rev1, __p2, 48); \
42794		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	42896	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	42897	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	42898	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vluti4q_lane_u8(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 48)); \
	42899	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
42795	42900	__ret; \
42796	42901	})
42797	42902	#endif
...	...	@@ -42801,7 +42906,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42801	42906	int8x16_t __ret; \
42802	42907	int8x16_t __s0 = __p0; \
42803	42908	uint8x8_t __s1 = __p1; \
42804		__ret = (int8x16_t) __builtin_neon_vluti4q_lane_s8((int8x16_t)__s0, (int8x8_t)__s1, __p2, 32); \
	42909	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vluti4q_lane_s8(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 32)); \
42805	42910	__ret; \
42806	42911	})
42807	42912	#else
...	...	@@ -42809,10 +42914,31 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42809	42914	int8x16_t __ret; \
42810	42915	int8x16_t __s0 = __p0; \
42811	42916	uint8x8_t __s1 = __p1; \
42812		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42813		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
42814		__ret = (int8x16_t) __builtin_neon_vluti4q_lane_s8((int8x16_t)__rev0, (int8x8_t)__rev1, __p2, 32); \
42815		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	42917	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	42918	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	42919	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vluti4q_lane_s8(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 32)); \
	42920	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
	42921	__ret; \
	42922	})
	42923	#endif
	42924
	42925	#ifdef __LITTLE_ENDIAN__
	42926	#define vluti4q_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	42927	mfloat8x16_t __ret; \
	42928	mfloat8x16_t __s0 = __p0; \
	42929	uint8x8_t __s1 = __p1; \
	42930	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vluti4q_lane_mf8(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 44)); \
	42931	__ret; \
	42932	})
	42933	#else
	42934	#define vluti4q_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	42935	mfloat8x16_t __ret; \
	42936	mfloat8x16_t __s0 = __p0; \
	42937	uint8x8_t __s1 = __p1; \
	42938	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	42939	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	42940	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vluti4q_lane_mf8(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 44)); \
	42941	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
42816	42942	__ret; \
42817	42943	})
42818	42944	#endif
...	...	@@ -42822,7 +42948,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42822	42948	poly16x8_t __ret; \
42823	42949	poly16x8x2_t __s0 = __p0; \
42824	42950	uint8x8_t __s1 = __p1; \
42825		__ret = (poly16x8_t) __builtin_neon_vluti4q_lane_p16_x2((int8x16_t)__s0.val[0], (int8x16_t)__s0.val[1], (int8x8_t)__s1, __p2, 37); \
	42951	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vluti4q_lane_p16_x2(__builtin_bit_cast(int8x16_t, __s0.val[0]), __builtin_bit_cast(int8x16_t, __s0.val[1]), __builtin_bit_cast(int8x8_t, __s1), __p2, 37)); \
42826	42952	__ret; \
42827	42953	})
42828	42954	#else
...	...	@@ -42831,11 +42957,11 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42831	42957	poly16x8x2_t __s0 = __p0; \
42832	42958	uint8x8_t __s1 = __p1; \
42833	42959	poly16x8x2_t __rev0; \
42834		__rev0.val[0] = __builtin_shufflevector(__s0.val[0], __s0.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
42835		__rev0.val[1] = __builtin_shufflevector(__s0.val[1], __s0.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
42836		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
42837		__ret = (poly16x8_t) __builtin_neon_vluti4q_lane_p16_x2((int8x16_t)__rev0.val[0], (int8x16_t)__rev0.val[1], (int8x8_t)__rev1, __p2, 37); \
42838		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	42960	__rev0.val[0] = __builtin_shufflevector(__s0.val[0], __s0.val[0], __lane_reverse_128_16); \
	42961	__rev0.val[1] = __builtin_shufflevector(__s0.val[1], __s0.val[1], __lane_reverse_128_16); \
	42962	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	42963	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vluti4q_lane_p16_x2(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x8_t, __rev1), __p2, 37)); \
	42964	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
42839	42965	__ret; \
42840	42966	})
42841	42967	#endif
...	...	@@ -42845,7 +42971,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42845	42971	uint16x8_t __ret; \
42846	42972	uint16x8x2_t __s0 = __p0; \
42847	42973	uint8x8_t __s1 = __p1; \
42848		__ret = (uint16x8_t) __builtin_neon_vluti4q_lane_u16_x2((int8x16_t)__s0.val[0], (int8x16_t)__s0.val[1], (int8x8_t)__s1, __p2, 49); \
	42974	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vluti4q_lane_u16_x2(__builtin_bit_cast(int8x16_t, __s0.val[0]), __builtin_bit_cast(int8x16_t, __s0.val[1]), __builtin_bit_cast(int8x8_t, __s1), __p2, 49)); \
42849	42975	__ret; \
42850	42976	})
42851	42977	#else
...	...	@@ -42854,11 +42980,11 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42854	42980	uint16x8x2_t __s0 = __p0; \
42855	42981	uint8x8_t __s1 = __p1; \
42856	42982	uint16x8x2_t __rev0; \
42857		__rev0.val[0] = __builtin_shufflevector(__s0.val[0], __s0.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
42858		__rev0.val[1] = __builtin_shufflevector(__s0.val[1], __s0.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
42859		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
42860		__ret = (uint16x8_t) __builtin_neon_vluti4q_lane_u16_x2((int8x16_t)__rev0.val[0], (int8x16_t)__rev0.val[1], (int8x8_t)__rev1, __p2, 49); \
42861		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	42983	__rev0.val[0] = __builtin_shufflevector(__s0.val[0], __s0.val[0], __lane_reverse_128_16); \
	42984	__rev0.val[1] = __builtin_shufflevector(__s0.val[1], __s0.val[1], __lane_reverse_128_16); \
	42985	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	42986	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vluti4q_lane_u16_x2(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x8_t, __rev1), __p2, 49)); \
	42987	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
42862	42988	__ret; \
42863	42989	})
42864	42990	#endif
...	...	@@ -42868,7 +42994,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42868	42994	float16x8_t __ret; \
42869	42995	float16x8x2_t __s0 = __p0; \
42870	42996	uint8x8_t __s1 = __p1; \
42871		__ret = (float16x8_t) __builtin_neon_vluti4q_lane_f16_x2((int8x16_t)__s0.val[0], (int8x16_t)__s0.val[1], (int8x8_t)__s1, __p2, 40); \
	42997	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vluti4q_lane_f16_x2(__builtin_bit_cast(int8x16_t, __s0.val[0]), __builtin_bit_cast(int8x16_t, __s0.val[1]), __builtin_bit_cast(int8x8_t, __s1), __p2, 40)); \
42872	42998	__ret; \
42873	42999	})
42874	43000	#else
...	...	@@ -42877,11 +43003,11 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42877	43003	float16x8x2_t __s0 = __p0; \
42878	43004	uint8x8_t __s1 = __p1; \
42879	43005	float16x8x2_t __rev0; \
42880		__rev0.val[0] = __builtin_shufflevector(__s0.val[0], __s0.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
42881		__rev0.val[1] = __builtin_shufflevector(__s0.val[1], __s0.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
42882		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
42883		__ret = (float16x8_t) __builtin_neon_vluti4q_lane_f16_x2((int8x16_t)__rev0.val[0], (int8x16_t)__rev0.val[1], (int8x8_t)__rev1, __p2, 40); \
42884		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	43006	__rev0.val[0] = __builtin_shufflevector(__s0.val[0], __s0.val[0], __lane_reverse_128_16); \
	43007	__rev0.val[1] = __builtin_shufflevector(__s0.val[1], __s0.val[1], __lane_reverse_128_16); \
	43008	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	43009	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vluti4q_lane_f16_x2(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x8_t, __rev1), __p2, 40)); \
	43010	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
42885	43011	__ret; \
42886	43012	})
42887	43013	#endif
...	...	@@ -42891,7 +43017,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42891	43017	int16x8_t __ret; \
42892	43018	int16x8x2_t __s0 = __p0; \
42893	43019	uint8x8_t __s1 = __p1; \
42894		__ret = (int16x8_t) __builtin_neon_vluti4q_lane_s16_x2((int8x16_t)__s0.val[0], (int8x16_t)__s0.val[1], (int8x8_t)__s1, __p2, 33); \
	43020	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vluti4q_lane_s16_x2(__builtin_bit_cast(int8x16_t, __s0.val[0]), __builtin_bit_cast(int8x16_t, __s0.val[1]), __builtin_bit_cast(int8x8_t, __s1), __p2, 33)); \
42895	43021	__ret; \
42896	43022	})
42897	43023	#else
...	...	@@ -42900,11 +43026,11 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42900	43026	int16x8x2_t __s0 = __p0; \
42901	43027	uint8x8_t __s1 = __p1; \
42902	43028	int16x8x2_t __rev0; \
42903		__rev0.val[0] = __builtin_shufflevector(__s0.val[0], __s0.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
42904		__rev0.val[1] = __builtin_shufflevector(__s0.val[1], __s0.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
42905		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
42906		__ret = (int16x8_t) __builtin_neon_vluti4q_lane_s16_x2((int8x16_t)__rev0.val[0], (int8x16_t)__rev0.val[1], (int8x8_t)__rev1, __p2, 33); \
42907		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	43029	__rev0.val[0] = __builtin_shufflevector(__s0.val[0], __s0.val[0], __lane_reverse_128_16); \
	43030	__rev0.val[1] = __builtin_shufflevector(__s0.val[1], __s0.val[1], __lane_reverse_128_16); \
	43031	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	43032	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vluti4q_lane_s16_x2(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x8_t, __rev1), __p2, 33)); \
	43033	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
42908	43034	__ret; \
42909	43035	})
42910	43036	#endif
...	...	@@ -42914,7 +43040,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42914	43040	poly8x16_t __ret; \
42915	43041	poly8x16_t __s0 = __p0; \
42916	43042	uint8x16_t __s1 = __p1; \
42917		__ret = (poly8x16_t) __builtin_neon_vluti4q_laneq_p8((int8x16_t)__s0, (int8x16_t)__s1, __p2, 36); \
	43043	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vluti4q_laneq_p8(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 36)); \
42918	43044	__ret; \
42919	43045	})
42920	43046	#else
...	...	@@ -42922,10 +43048,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42922	43048	poly8x16_t __ret; \
42923	43049	poly8x16_t __s0 = __p0; \
42924	43050	uint8x16_t __s1 = __p1; \
42925		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42926		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42927		__ret = (poly8x16_t) __builtin_neon_vluti4q_laneq_p8((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 36); \
42928		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	43051	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	43052	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	43053	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vluti4q_laneq_p8(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 36)); \
	43054	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
42929	43055	__ret; \
42930	43056	})
42931	43057	#endif
...	...	@@ -42935,7 +43061,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42935	43061	uint8x16_t __ret; \
42936	43062	uint8x16_t __s0 = __p0; \
42937	43063	uint8x16_t __s1 = __p1; \
42938		__ret = (uint8x16_t) __builtin_neon_vluti4q_laneq_u8((int8x16_t)__s0, (int8x16_t)__s1, __p2, 48); \
	43064	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vluti4q_laneq_u8(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 48)); \
42939	43065	__ret; \
42940	43066	})
42941	43067	#else
...	...	@@ -42943,10 +43069,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42943	43069	uint8x16_t __ret; \
42944	43070	uint8x16_t __s0 = __p0; \
42945	43071	uint8x16_t __s1 = __p1; \
42946		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42947		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42948		__ret = (uint8x16_t) __builtin_neon_vluti4q_laneq_u8((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 48); \
42949		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	43072	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	43073	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	43074	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vluti4q_laneq_u8(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 48)); \
	43075	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
42950	43076	__ret; \
42951	43077	})
42952	43078	#endif
...	...	@@ -42956,7 +43082,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42956	43082	int8x16_t __ret; \
42957	43083	int8x16_t __s0 = __p0; \
42958	43084	uint8x16_t __s1 = __p1; \
42959		__ret = (int8x16_t) __builtin_neon_vluti4q_laneq_s8((int8x16_t)__s0, (int8x16_t)__s1, __p2, 32); \
	43085	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vluti4q_laneq_s8(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 32)); \
42960	43086	__ret; \
42961	43087	})
42962	43088	#else
...	...	@@ -42964,10 +43090,31 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42964	43090	int8x16_t __ret; \
42965	43091	int8x16_t __s0 = __p0; \
42966	43092	uint8x16_t __s1 = __p1; \
42967		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42968		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42969		__ret = (int8x16_t) __builtin_neon_vluti4q_laneq_s8((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 32); \
42970		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	43093	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	43094	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	43095	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vluti4q_laneq_s8(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 32)); \
	43096	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
	43097	__ret; \
	43098	})
	43099	#endif
	43100
	43101	#ifdef __LITTLE_ENDIAN__
	43102	#define vluti4q_laneq_mf8(__p0, __p1, __p2) __extension__ ({ \
	43103	mfloat8x16_t __ret; \
	43104	mfloat8x16_t __s0 = __p0; \
	43105	uint8x16_t __s1 = __p1; \
	43106	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vluti4q_laneq_mf8(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 44)); \
	43107	__ret; \
	43108	})
	43109	#else
	43110	#define vluti4q_laneq_mf8(__p0, __p1, __p2) __extension__ ({ \
	43111	mfloat8x16_t __ret; \
	43112	mfloat8x16_t __s0 = __p0; \
	43113	uint8x16_t __s1 = __p1; \
	43114	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	43115	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	43116	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vluti4q_laneq_mf8(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 44)); \
	43117	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
42971	43118	__ret; \
42972	43119	})
42973	43120	#endif
...	...	@@ -42977,7 +43124,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42977	43124	poly16x8_t __ret; \
42978	43125	poly16x8x2_t __s0 = __p0; \
42979	43126	uint8x16_t __s1 = __p1; \
42980		__ret = (poly16x8_t) __builtin_neon_vluti4q_laneq_p16_x2((int8x16_t)__s0.val[0], (int8x16_t)__s0.val[1], (int8x16_t)__s1, __p2, 37); \
	43127	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vluti4q_laneq_p16_x2(__builtin_bit_cast(int8x16_t, __s0.val[0]), __builtin_bit_cast(int8x16_t, __s0.val[1]), __builtin_bit_cast(int8x16_t, __s1), __p2, 37)); \
42981	43128	__ret; \
42982	43129	})
42983	43130	#else
...	...	@@ -42986,11 +43133,11 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
42986	43133	poly16x8x2_t __s0 = __p0; \
42987	43134	uint8x16_t __s1 = __p1; \
42988	43135	poly16x8x2_t __rev0; \
42989		__rev0.val[0] = __builtin_shufflevector(__s0.val[0], __s0.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
42990		__rev0.val[1] = __builtin_shufflevector(__s0.val[1], __s0.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
42991		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
42992		__ret = (poly16x8_t) __builtin_neon_vluti4q_laneq_p16_x2((int8x16_t)__rev0.val[0], (int8x16_t)__rev0.val[1], (int8x16_t)__rev1, __p2, 37); \
42993		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	43136	__rev0.val[0] = __builtin_shufflevector(__s0.val[0], __s0.val[0], __lane_reverse_128_16); \
	43137	__rev0.val[1] = __builtin_shufflevector(__s0.val[1], __s0.val[1], __lane_reverse_128_16); \
	43138	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	43139	__ret = __builtin_bit_cast(poly16x8_t, __builtin_neon_vluti4q_laneq_p16_x2(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x16_t, __rev1), __p2, 37)); \
	43140	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
42994	43141	__ret; \
42995	43142	})
42996	43143	#endif
...	...	@@ -43000,7 +43147,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
43000	43147	uint16x8_t __ret; \
43001	43148	uint16x8x2_t __s0 = __p0; \
43002	43149	uint8x16_t __s1 = __p1; \
43003		__ret = (uint16x8_t) __builtin_neon_vluti4q_laneq_u16_x2((int8x16_t)__s0.val[0], (int8x16_t)__s0.val[1], (int8x16_t)__s1, __p2, 49); \
	43150	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vluti4q_laneq_u16_x2(__builtin_bit_cast(int8x16_t, __s0.val[0]), __builtin_bit_cast(int8x16_t, __s0.val[1]), __builtin_bit_cast(int8x16_t, __s1), __p2, 49)); \
43004	43151	__ret; \
43005	43152	})
43006	43153	#else
...	...	@@ -43009,11 +43156,11 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
43009	43156	uint16x8x2_t __s0 = __p0; \
43010	43157	uint8x16_t __s1 = __p1; \
43011	43158	uint16x8x2_t __rev0; \
43012		__rev0.val[0] = __builtin_shufflevector(__s0.val[0], __s0.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
43013		__rev0.val[1] = __builtin_shufflevector(__s0.val[1], __s0.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
43014		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
43015		__ret = (uint16x8_t) __builtin_neon_vluti4q_laneq_u16_x2((int8x16_t)__rev0.val[0], (int8x16_t)__rev0.val[1], (int8x16_t)__rev1, __p2, 49); \
43016		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	43159	__rev0.val[0] = __builtin_shufflevector(__s0.val[0], __s0.val[0], __lane_reverse_128_16); \
	43160	__rev0.val[1] = __builtin_shufflevector(__s0.val[1], __s0.val[1], __lane_reverse_128_16); \
	43161	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	43162	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vluti4q_laneq_u16_x2(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x16_t, __rev1), __p2, 49)); \
	43163	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
43017	43164	__ret; \
43018	43165	})
43019	43166	#endif
...	...	@@ -43023,7 +43170,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
43023	43170	float16x8_t __ret; \
43024	43171	float16x8x2_t __s0 = __p0; \
43025	43172	uint8x16_t __s1 = __p1; \
43026		__ret = (float16x8_t) __builtin_neon_vluti4q_laneq_f16_x2((int8x16_t)__s0.val[0], (int8x16_t)__s0.val[1], (int8x16_t)__s1, __p2, 40); \
	43173	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vluti4q_laneq_f16_x2(__builtin_bit_cast(int8x16_t, __s0.val[0]), __builtin_bit_cast(int8x16_t, __s0.val[1]), __builtin_bit_cast(int8x16_t, __s1), __p2, 40)); \
43027	43174	__ret; \
43028	43175	})
43029	43176	#else
...	...	@@ -43032,11 +43179,11 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
43032	43179	float16x8x2_t __s0 = __p0; \
43033	43180	uint8x16_t __s1 = __p1; \
43034	43181	float16x8x2_t __rev0; \
43035		__rev0.val[0] = __builtin_shufflevector(__s0.val[0], __s0.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
43036		__rev0.val[1] = __builtin_shufflevector(__s0.val[1], __s0.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
43037		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
43038		__ret = (float16x8_t) __builtin_neon_vluti4q_laneq_f16_x2((int8x16_t)__rev0.val[0], (int8x16_t)__rev0.val[1], (int8x16_t)__rev1, __p2, 40); \
43039		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	43182	__rev0.val[0] = __builtin_shufflevector(__s0.val[0], __s0.val[0], __lane_reverse_128_16); \
	43183	__rev0.val[1] = __builtin_shufflevector(__s0.val[1], __s0.val[1], __lane_reverse_128_16); \
	43184	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	43185	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vluti4q_laneq_f16_x2(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x16_t, __rev1), __p2, 40)); \
	43186	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
43040	43187	__ret; \
43041	43188	})
43042	43189	#endif
...	...	@@ -43046,7 +43193,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
43046	43193	int16x8_t __ret; \
43047	43194	int16x8x2_t __s0 = __p0; \
43048	43195	uint8x16_t __s1 = __p1; \
43049		__ret = (int16x8_t) __builtin_neon_vluti4q_laneq_s16_x2((int8x16_t)__s0.val[0], (int8x16_t)__s0.val[1], (int8x16_t)__s1, __p2, 33); \
	43196	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vluti4q_laneq_s16_x2(__builtin_bit_cast(int8x16_t, __s0.val[0]), __builtin_bit_cast(int8x16_t, __s0.val[1]), __builtin_bit_cast(int8x16_t, __s1), __p2, 33)); \
43050	43197	__ret; \
43051	43198	})
43052	43199	#else
...	...	@@ -43055,11 +43202,11 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
43055	43202	int16x8x2_t __s0 = __p0; \
43056	43203	uint8x16_t __s1 = __p1; \
43057	43204	int16x8x2_t __rev0; \
43058		__rev0.val[0] = __builtin_shufflevector(__s0.val[0], __s0.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
43059		__rev0.val[1] = __builtin_shufflevector(__s0.val[1], __s0.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
43060		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
43061		__ret = (int16x8_t) __builtin_neon_vluti4q_laneq_s16_x2((int8x16_t)__rev0.val[0], (int8x16_t)__rev0.val[1], (int8x16_t)__rev1, __p2, 33); \
43062		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	43205	__rev0.val[0] = __builtin_shufflevector(__s0.val[0], __s0.val[0], __lane_reverse_128_16); \
	43206	__rev0.val[1] = __builtin_shufflevector(__s0.val[1], __s0.val[1], __lane_reverse_128_16); \
	43207	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	43208	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vluti4q_laneq_s16_x2(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x16_t, __rev1), __p2, 33)); \
	43209	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
43063	43210	__ret; \
43064	43211	})
43065	43212	#endif
...	...	@@ -43069,7 +43216,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
43069	43216	bfloat16x8_t __ret; \
43070	43217	bfloat16x8_t __s0 = __p0; \
43071	43218	uint8x8_t __s1 = __p1; \
43072		__ret = (bfloat16x8_t) __builtin_neon_vluti2q_lane_bf16((int8x16_t)__s0, (int8x8_t)__s1, __p2, 43); \
	43219	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vluti2q_lane_bf16(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 43)); \
43073	43220	__ret; \
43074	43221	})
43075	43222	#else
...	...	@@ -43077,10 +43224,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
43077	43224	bfloat16x8_t __ret; \
43078	43225	bfloat16x8_t __s0 = __p0; \
43079	43226	uint8x8_t __s1 = __p1; \
43080		bfloat16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
43081		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
43082		__ret = (bfloat16x8_t) __builtin_neon_vluti2q_lane_bf16((int8x16_t)__rev0, (int8x8_t)__rev1, __p2, 43); \
43083		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	43227	bfloat16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	43228	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	43229	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vluti2q_lane_bf16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 43)); \
	43230	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
43084	43231	__ret; \
43085	43232	})
43086	43233	#endif
...	...	@@ -43090,7 +43237,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
43090	43237	bfloat16x8_t __ret; \
43091	43238	bfloat16x4_t __s0 = __p0; \
43092	43239	uint8x8_t __s1 = __p1; \
43093		__ret = (bfloat16x8_t) __builtin_neon_vluti2_lane_bf16((int8x8_t)__s0, (int8x8_t)__s1, __p2, 43); \
	43240	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vluti2_lane_bf16(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 43)); \
43094	43241	__ret; \
43095	43242	})
43096	43243	#else
...	...	@@ -43098,10 +43245,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
43098	43245	bfloat16x8_t __ret; \
43099	43246	bfloat16x4_t __s0 = __p0; \
43100	43247	uint8x8_t __s1 = __p1; \
43101		bfloat16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
43102		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
43103		__ret = (bfloat16x8_t) __builtin_neon_vluti2_lane_bf16((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 43); \
43104		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	43248	bfloat16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	43249	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	43250	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vluti2_lane_bf16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 43)); \
	43251	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
43105	43252	__ret; \
43106	43253	})
43107	43254	#endif
...	...	@@ -43111,7 +43258,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
43111	43258	bfloat16x8_t __ret; \
43112	43259	bfloat16x8_t __s0 = __p0; \
43113	43260	uint8x16_t __s1 = __p1; \
43114		__ret = (bfloat16x8_t) __builtin_neon_vluti2q_laneq_bf16((int8x16_t)__s0, (int8x16_t)__s1, __p2, 43); \
	43261	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vluti2q_laneq_bf16(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 43)); \
43115	43262	__ret; \
43116	43263	})
43117	43264	#else
...	...	@@ -43119,10 +43266,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
43119	43266	bfloat16x8_t __ret; \
43120	43267	bfloat16x8_t __s0 = __p0; \
43121	43268	uint8x16_t __s1 = __p1; \
43122		bfloat16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
43123		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
43124		__ret = (bfloat16x8_t) __builtin_neon_vluti2q_laneq_bf16((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 43); \
43125		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	43269	bfloat16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	43270	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	43271	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vluti2q_laneq_bf16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 43)); \
	43272	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
43126	43273	__ret; \
43127	43274	})
43128	43275	#endif
...	...	@@ -43132,7 +43279,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
43132	43279	bfloat16x8_t __ret; \
43133	43280	bfloat16x4_t __s0 = __p0; \
43134	43281	uint8x16_t __s1 = __p1; \
43135		__ret = (bfloat16x8_t) __builtin_neon_vluti2_laneq_bf16((int8x8_t)__s0, (int8x16_t)__s1, __p2, 43); \
	43282	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vluti2_laneq_bf16(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 43)); \
43136	43283	__ret; \
43137	43284	})
43138	43285	#else
...	...	@@ -43140,10 +43287,10 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
43140	43287	bfloat16x8_t __ret; \
43141	43288	bfloat16x4_t __s0 = __p0; \
43142	43289	uint8x16_t __s1 = __p1; \
43143		bfloat16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
43144		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
43145		__ret = (bfloat16x8_t) __builtin_neon_vluti2_laneq_bf16((int8x8_t)__rev0, (int8x16_t)__rev1, __p2, 43); \
43146		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	43290	bfloat16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	43291	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	43292	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vluti2_laneq_bf16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 43)); \
	43293	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
43147	43294	__ret; \
43148	43295	})
43149	43296	#endif
...	...	@@ -43153,7 +43300,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
43153	43300	bfloat16x8_t __ret; \
43154	43301	bfloat16x8x2_t __s0 = __p0; \
43155	43302	uint8x8_t __s1 = __p1; \
43156		__ret = (bfloat16x8_t) __builtin_neon_vluti4q_lane_bf16_x2((int8x16_t)__s0.val[0], (int8x16_t)__s0.val[1], (int8x8_t)__s1, __p2, 43); \
	43303	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vluti4q_lane_bf16_x2(__builtin_bit_cast(int8x16_t, __s0.val[0]), __builtin_bit_cast(int8x16_t, __s0.val[1]), __builtin_bit_cast(int8x8_t, __s1), __p2, 43)); \
43157	43304	__ret; \
43158	43305	})
43159	43306	#else
...	...	@@ -43162,11 +43309,11 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
43162	43309	bfloat16x8x2_t __s0 = __p0; \
43163	43310	uint8x8_t __s1 = __p1; \
43164	43311	bfloat16x8x2_t __rev0; \
43165		__rev0.val[0] = __builtin_shufflevector(__s0.val[0], __s0.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
43166		__rev0.val[1] = __builtin_shufflevector(__s0.val[1], __s0.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
43167		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
43168		__ret = (bfloat16x8_t) __builtin_neon_vluti4q_lane_bf16_x2((int8x16_t)__rev0.val[0], (int8x16_t)__rev0.val[1], (int8x8_t)__rev1, __p2, 43); \
43169		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	43312	__rev0.val[0] = __builtin_shufflevector(__s0.val[0], __s0.val[0], __lane_reverse_128_16); \
	43313	__rev0.val[1] = __builtin_shufflevector(__s0.val[1], __s0.val[1], __lane_reverse_128_16); \
	43314	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	43315	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vluti4q_lane_bf16_x2(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x8_t, __rev1), __p2, 43)); \
	43316	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
43170	43317	__ret; \
43171	43318	})
43172	43319	#endif
...	...	@@ -43176,7 +43323,7 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
43176	43323	bfloat16x8_t __ret; \
43177	43324	bfloat16x8x2_t __s0 = __p0; \
43178	43325	uint8x16_t __s1 = __p1; \
43179		__ret = (bfloat16x8_t) __builtin_neon_vluti4q_laneq_bf16_x2((int8x16_t)__s0.val[0], (int8x16_t)__s0.val[1], (int8x16_t)__s1, __p2, 43); \
	43326	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vluti4q_laneq_bf16_x2(__builtin_bit_cast(int8x16_t, __s0.val[0]), __builtin_bit_cast(int8x16_t, __s0.val[1]), __builtin_bit_cast(int8x16_t, __s1), __p2, 43)); \
43180	43327	__ret; \
43181	43328	})
43182	43329	#else
...	...	@@ -43185,28 +43332,821 @@ __ai __attribute__((target("fp8fma,neon"))) float16x8_t vmlaltq_f16_mf8_fpm(floa
43185	43332	bfloat16x8x2_t __s0 = __p0; \
43186	43333	uint8x16_t __s1 = __p1; \
43187	43334	bfloat16x8x2_t __rev0; \
43188		__rev0.val[0] = __builtin_shufflevector(__s0.val[0], __s0.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
43189		__rev0.val[1] = __builtin_shufflevector(__s0.val[1], __s0.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
43190		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
43191		__ret = (bfloat16x8_t) __builtin_neon_vluti4q_laneq_bf16_x2((int8x16_t)__rev0.val[0], (int8x16_t)__rev0.val[1], (int8x16_t)__rev1, __p2, 43); \
43192		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	43335	__rev0.val[0] = __builtin_shufflevector(__s0.val[0], __s0.val[0], __lane_reverse_128_16); \
	43336	__rev0.val[1] = __builtin_shufflevector(__s0.val[1], __s0.val[1], __lane_reverse_128_16); \
	43337	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	43338	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vluti4q_laneq_bf16_x2(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x16_t, __rev1), __p2, 43)); \
	43339	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
	43340	__ret; \
	43341	})
	43342	#endif
	43343
	43344	#ifdef __LITTLE_ENDIAN__
	43345	#define splatq_lane_mf8(__p0, __p1) __extension__ ({ \
	43346	mfloat8x16_t __ret; \
	43347	mfloat8x8_t __s0 = __p0; \
	43348	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 12)); \
	43349	__ret; \
	43350	})
	43351	#else
	43352	#define splatq_lane_mf8(__p0, __p1) __extension__ ({ \
	43353	mfloat8x16_t __ret; \
	43354	mfloat8x8_t __s0 = __p0; \
	43355	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	43356	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 12)); \
	43357	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
	43358	__ret; \
	43359	})
	43360	#define __noswap_splatq_lane_mf8(__p0, __p1) __extension__ ({ \
	43361	mfloat8x16_t __ret; \
	43362	mfloat8x8_t __s0 = __p0; \
	43363	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_splatq_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 12)); \
	43364	__ret; \
	43365	})
	43366	#endif
	43367
	43368	#ifdef __LITTLE_ENDIAN__
	43369	#define splat_lane_mf8(__p0, __p1) __extension__ ({ \
	43370	mfloat8x8_t __ret; \
	43371	mfloat8x8_t __s0 = __p0; \
	43372	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 12)); \
	43373	__ret; \
	43374	})
	43375	#else
	43376	#define splat_lane_mf8(__p0, __p1) __extension__ ({ \
	43377	mfloat8x8_t __ret; \
	43378	mfloat8x8_t __s0 = __p0; \
	43379	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	43380	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __rev0), __p1, 12)); \
	43381	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
	43382	__ret; \
	43383	})
	43384	#define __noswap_splat_lane_mf8(__p0, __p1) __extension__ ({ \
	43385	mfloat8x8_t __ret; \
	43386	mfloat8x8_t __s0 = __p0; \
	43387	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_splat_lane_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 12)); \
	43388	__ret; \
	43389	})
	43390	#endif
	43391
	43392	#ifdef __LITTLE_ENDIAN__
	43393	#define splatq_laneq_mf8(__p0, __p1) __extension__ ({ \
	43394	mfloat8x16_t __ret; \
	43395	mfloat8x16_t __s0 = __p0; \
	43396	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 44)); \
	43397	__ret; \
	43398	})
	43399	#else
	43400	#define splatq_laneq_mf8(__p0, __p1) __extension__ ({ \
	43401	mfloat8x16_t __ret; \
	43402	mfloat8x16_t __s0 = __p0; \
	43403	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	43404	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 44)); \
	43405	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
	43406	__ret; \
	43407	})
	43408	#define __noswap_splatq_laneq_mf8(__p0, __p1) __extension__ ({ \
	43409	mfloat8x16_t __ret; \
	43410	mfloat8x16_t __s0 = __p0; \
	43411	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_splatq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 44)); \
	43412	__ret; \
	43413	})
	43414	#endif
	43415
	43416	#ifdef __LITTLE_ENDIAN__
	43417	#define splat_laneq_mf8(__p0, __p1) __extension__ ({ \
	43418	mfloat8x8_t __ret; \
	43419	mfloat8x16_t __s0 = __p0; \
	43420	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 44)); \
	43421	__ret; \
	43422	})
	43423	#else
	43424	#define splat_laneq_mf8(__p0, __p1) __extension__ ({ \
	43425	mfloat8x8_t __ret; \
	43426	mfloat8x16_t __s0 = __p0; \
	43427	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	43428	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 44)); \
	43429	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
	43430	__ret; \
	43431	})
	43432	#define __noswap_splat_laneq_mf8(__p0, __p1) __extension__ ({ \
	43433	mfloat8x8_t __ret; \
	43434	mfloat8x16_t __s0 = __p0; \
	43435	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_splat_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 44)); \
	43436	__ret; \
	43437	})
	43438	#endif
	43439
	43440	#ifdef __LITTLE_ENDIAN__
	43441	__ai __attribute__((target("neon"))) mfloat8x16_t vbslq_mf8(uint8x16_t __p0, mfloat8x16_t __p1, mfloat8x16_t __p2) {
	43442	mfloat8x16_t __ret;
	43443	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vbslq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 44));
	43444	return __ret;
	43445	}
	43446	#else
	43447	__ai __attribute__((target("neon"))) mfloat8x16_t vbslq_mf8(uint8x16_t __p0, mfloat8x16_t __p1, mfloat8x16_t __p2) {
	43448	mfloat8x16_t __ret;
	43449	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	43450	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	43451	mfloat8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	43452	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vbslq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 44));
	43453	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
	43454	return __ret;
	43455	}
	43456	#endif
	43457
	43458	#ifdef __LITTLE_ENDIAN__
	43459	__ai __attribute__((target("neon"))) mfloat8x8_t vbsl_mf8(uint8x8_t __p0, mfloat8x8_t __p1, mfloat8x8_t __p2) {
	43460	mfloat8x8_t __ret;
	43461	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vbsl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 12));
	43462	return __ret;
	43463	}
	43464	#else
	43465	__ai __attribute__((target("neon"))) mfloat8x8_t vbsl_mf8(uint8x8_t __p0, mfloat8x8_t __p1, mfloat8x8_t __p2) {
	43466	mfloat8x8_t __ret;
	43467	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	43468	mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	43469	mfloat8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	43470	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vbsl_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 12));
	43471	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
	43472	return __ret;
	43473	}
	43474	#endif
	43475
	43476	#ifdef __LITTLE_ENDIAN__
	43477	__ai __attribute__((target("neon"))) mfloat8x16_t vcombine_mf8(mfloat8x8_t __p0, mfloat8x8_t __p1) {
	43478	mfloat8x16_t __ret;
	43479	__ret = __builtin_shufflevector(__p0, __p1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15);
	43480	return __ret;
	43481	}
	43482	#else
	43483	__ai __attribute__((target("neon"))) mfloat8x16_t vcombine_mf8(mfloat8x8_t __p0, mfloat8x8_t __p1) {
	43484	mfloat8x16_t __ret;
	43485	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	43486	mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	43487	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15);
	43488	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
	43489	return __ret;
	43490	}
	43491	#endif
	43492
	43493	#define vcreate_mf8(__p0) __extension__ ({ \
	43494	mfloat8x8_t __ret; \
	43495	uint64_t __promote = __p0; \
	43496	__ret = __builtin_bit_cast(mfloat8x8_t, __promote); \
	43497	__ret; \
	43498	})
	43499	#ifdef __LITTLE_ENDIAN__
	43500	#define vdupq_lane_mf8(__p0_230, __p1_230) __extension__ ({ \
	43501	mfloat8x16_t __ret_230; \
	43502	mfloat8x8_t __s0_230 = __p0_230; \
	43503	__ret_230 = splatq_lane_mf8(__s0_230, __p1_230); \
	43504	__ret_230; \
	43505	})
	43506	#else
	43507	#define vdupq_lane_mf8(__p0_231, __p1_231) __extension__ ({ \
	43508	mfloat8x16_t __ret_231; \
	43509	mfloat8x8_t __s0_231 = __p0_231; \
	43510	mfloat8x8_t __rev0_231; __rev0_231 = __builtin_shufflevector(__s0_231, __s0_231, __lane_reverse_64_8); \
	43511	__ret_231 = __noswap_splatq_lane_mf8(__rev0_231, __p1_231); \
	43512	__ret_231 = __builtin_shufflevector(__ret_231, __ret_231, __lane_reverse_128_8); \
	43513	__ret_231; \
	43514	})
	43515	#endif
	43516
	43517	#ifdef __LITTLE_ENDIAN__
	43518	#define vdup_lane_mf8(__p0_232, __p1_232) __extension__ ({ \
	43519	mfloat8x8_t __ret_232; \
	43520	mfloat8x8_t __s0_232 = __p0_232; \
	43521	__ret_232 = splat_lane_mf8(__s0_232, __p1_232); \
	43522	__ret_232; \
	43523	})
	43524	#else
	43525	#define vdup_lane_mf8(__p0_233, __p1_233) __extension__ ({ \
	43526	mfloat8x8_t __ret_233; \
	43527	mfloat8x8_t __s0_233 = __p0_233; \
	43528	mfloat8x8_t __rev0_233; __rev0_233 = __builtin_shufflevector(__s0_233, __s0_233, __lane_reverse_64_8); \
	43529	__ret_233 = __noswap_splat_lane_mf8(__rev0_233, __p1_233); \
	43530	__ret_233 = __builtin_shufflevector(__ret_233, __ret_233, __lane_reverse_64_8); \
	43531	__ret_233; \
	43532	})
	43533	#endif
	43534
	43535	#ifdef __LITTLE_ENDIAN__
	43536	__ai __attribute__((target("neon"))) mfloat8x16_t vdupq_n_mf8(mfloat8_t __p0) {
	43537	mfloat8x16_t __ret;
	43538	__ret = (mfloat8x16_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
	43539	return __ret;
	43540	}
	43541	#else
	43542	__ai __attribute__((target("neon"))) mfloat8x16_t vdupq_n_mf8(mfloat8_t __p0) {
	43543	mfloat8x16_t __ret;
	43544	__ret = (mfloat8x16_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
	43545	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
	43546	return __ret;
	43547	}
	43548	#endif
	43549
	43550	#ifdef __LITTLE_ENDIAN__
	43551	__ai __attribute__((target("neon"))) mfloat8x8_t vdup_n_mf8(mfloat8_t __p0) {
	43552	mfloat8x8_t __ret;
	43553	__ret = (mfloat8x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
	43554	return __ret;
	43555	}
	43556	#else
	43557	__ai __attribute__((target("neon"))) mfloat8x8_t vdup_n_mf8(mfloat8_t __p0) {
	43558	mfloat8x8_t __ret;
	43559	__ret = (mfloat8x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
	43560	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
	43561	return __ret;
	43562	}
	43563	#endif
	43564
	43565	#ifdef __LITTLE_ENDIAN__
	43566	#define vextq_mf8(__p0, __p1, __p2) __extension__ ({ \
	43567	mfloat8x16_t __ret; \
	43568	mfloat8x16_t __s0 = __p0; \
	43569	mfloat8x16_t __s1 = __p1; \
	43570	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vextq_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 44)); \
	43571	__ret; \
	43572	})
	43573	#else
	43574	#define vextq_mf8(__p0, __p1, __p2) __extension__ ({ \
	43575	mfloat8x16_t __ret; \
	43576	mfloat8x16_t __s0 = __p0; \
	43577	mfloat8x16_t __s1 = __p1; \
	43578	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	43579	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	43580	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vextq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 44)); \
	43581	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
	43582	__ret; \
	43583	})
	43584	#endif
	43585
	43586	#ifdef __LITTLE_ENDIAN__
	43587	#define vext_mf8(__p0, __p1, __p2) __extension__ ({ \
	43588	mfloat8x8_t __ret; \
	43589	mfloat8x8_t __s0 = __p0; \
	43590	mfloat8x8_t __s1 = __p1; \
	43591	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vext_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 12)); \
	43592	__ret; \
	43593	})
	43594	#else
	43595	#define vext_mf8(__p0, __p1, __p2) __extension__ ({ \
	43596	mfloat8x8_t __ret; \
	43597	mfloat8x8_t __s0 = __p0; \
	43598	mfloat8x8_t __s1 = __p1; \
	43599	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	43600	mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	43601	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vext_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 12)); \
	43602	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
	43603	__ret; \
	43604	})
	43605	#endif
	43606
	43607	#ifdef __LITTLE_ENDIAN__
	43608	__ai __attribute__((target("neon"))) mfloat8x8_t vget_high_mf8(mfloat8x16_t __p0) {
	43609	mfloat8x8_t __ret;
	43610	__ret = __builtin_shufflevector(__p0, __p0, 8, 9, 10, 11, 12, 13, 14, 15);
	43611	return __ret;
	43612	}
	43613	#else
	43614	__ai __attribute__((target("neon"))) mfloat8x8_t vget_high_mf8(mfloat8x16_t __p0) {
	43615	mfloat8x8_t __ret;
	43616	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	43617	__ret = __builtin_shufflevector(__rev0, __rev0, 8, 9, 10, 11, 12, 13, 14, 15);
	43618	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
	43619	return __ret;
	43620	}
	43621	#endif
	43622
	43623	#ifdef __LITTLE_ENDIAN__
	43624	#define vgetq_lane_mf8(__p0, __p1) __extension__ ({ \
	43625	mfloat8_t __ret; \
	43626	mfloat8x16_t __s0 = __p0; \
	43627	__ret = __builtin_bit_cast(mfloat8_t, __builtin_neon_vgetq_lane_mf8(__s0, __p1)); \
43193	43628	__ret; \
43194	43629	})
	43630	#else
	43631	#define vgetq_lane_mf8(__p0, __p1) __extension__ ({ \
	43632	mfloat8_t __ret; \
	43633	mfloat8x16_t __s0 = __p0; \
	43634	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	43635	__ret = __builtin_bit_cast(mfloat8_t, __builtin_neon_vgetq_lane_mf8(__rev0, __p1)); \
	43636	__ret; \
	43637	})
	43638	#define __noswap_vgetq_lane_mf8(__p0, __p1) __extension__ ({ \
	43639	mfloat8_t __ret; \
	43640	mfloat8x16_t __s0 = __p0; \
	43641	__ret = __builtin_bit_cast(mfloat8_t, __builtin_neon_vgetq_lane_mf8(__s0, __p1)); \
	43642	__ret; \
	43643	})
	43644	#endif
	43645
	43646	#ifdef __LITTLE_ENDIAN__
	43647	#define vget_lane_mf8(__p0, __p1) __extension__ ({ \
	43648	mfloat8_t __ret; \
	43649	mfloat8x8_t __s0 = __p0; \
	43650	__ret = __builtin_bit_cast(mfloat8_t, __builtin_neon_vget_lane_mf8(__s0, __p1)); \
	43651	__ret; \
	43652	})
	43653	#else
	43654	#define vget_lane_mf8(__p0, __p1) __extension__ ({ \
	43655	mfloat8_t __ret; \
	43656	mfloat8x8_t __s0 = __p0; \
	43657	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	43658	__ret = __builtin_bit_cast(mfloat8_t, __builtin_neon_vget_lane_mf8(__rev0, __p1)); \
	43659	__ret; \
	43660	})
	43661	#define __noswap_vget_lane_mf8(__p0, __p1) __extension__ ({ \
	43662	mfloat8_t __ret; \
	43663	mfloat8x8_t __s0 = __p0; \
	43664	__ret = __builtin_bit_cast(mfloat8_t, __builtin_neon_vget_lane_mf8(__s0, __p1)); \
	43665	__ret; \
	43666	})
	43667	#endif
	43668
	43669	#ifdef __LITTLE_ENDIAN__
	43670	__ai __attribute__((target("neon"))) mfloat8x8_t vget_low_mf8(mfloat8x16_t __p0) {
	43671	mfloat8x8_t __ret;
	43672	__ret = __builtin_shufflevector(__p0, __p0, 0, 1, 2, 3, 4, 5, 6, 7);
	43673	return __ret;
	43674	}
	43675	#else
	43676	__ai __attribute__((target("neon"))) mfloat8x8_t vget_low_mf8(mfloat8x16_t __p0) {
	43677	mfloat8x8_t __ret;
	43678	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	43679	__ret = __builtin_shufflevector(__rev0, __rev0, 0, 1, 2, 3, 4, 5, 6, 7);
	43680	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
	43681	return __ret;
	43682	}
	43683	#endif
	43684
	43685	#ifdef __LITTLE_ENDIAN__
	43686	__ai __attribute__((target("neon"))) mfloat8x16_t vmovq_n_mf8(mfloat8_t __p0) {
	43687	mfloat8x16_t __ret;
	43688	__ret = (mfloat8x16_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
	43689	return __ret;
	43690	}
	43691	#else
	43692	__ai __attribute__((target("neon"))) mfloat8x16_t vmovq_n_mf8(mfloat8_t __p0) {
	43693	mfloat8x16_t __ret;
	43694	__ret = (mfloat8x16_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
	43695	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
	43696	return __ret;
	43697	}
	43698	#endif
	43699
	43700	#ifdef __LITTLE_ENDIAN__
	43701	__ai __attribute__((target("neon"))) mfloat8x8_t vmov_n_mf8(mfloat8_t __p0) {
	43702	mfloat8x8_t __ret;
	43703	__ret = (mfloat8x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
	43704	return __ret;
	43705	}
	43706	#else
	43707	__ai __attribute__((target("neon"))) mfloat8x8_t vmov_n_mf8(mfloat8_t __p0) {
	43708	mfloat8x8_t __ret;
	43709	__ret = (mfloat8x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
	43710	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
	43711	return __ret;
	43712	}
	43713	#endif
	43714
	43715	#ifdef __LITTLE_ENDIAN__
	43716	__ai __attribute__((target("neon"))) mfloat8x16_t vrev16q_mf8(mfloat8x16_t __p0) {
	43717	mfloat8x16_t __ret;
	43718	__ret = __builtin_shufflevector(__p0, __p0, 1, 0, 3, 2, 5, 4, 7, 6, 9, 8, 11, 10, 13, 12, 15, 14);
	43719	return __ret;
	43720	}
	43721	#else
	43722	__ai __attribute__((target("neon"))) mfloat8x16_t vrev16q_mf8(mfloat8x16_t __p0) {
	43723	mfloat8x16_t __ret;
	43724	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	43725	__ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2, 5, 4, 7, 6, 9, 8, 11, 10, 13, 12, 15, 14);
	43726	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
	43727	return __ret;
	43728	}
	43729	#endif
	43730
	43731	#ifdef __LITTLE_ENDIAN__
	43732	__ai __attribute__((target("neon"))) mfloat8x8_t vrev16_mf8(mfloat8x8_t __p0) {
	43733	mfloat8x8_t __ret;
	43734	__ret = __builtin_shufflevector(__p0, __p0, 1, 0, 3, 2, 5, 4, 7, 6);
	43735	return __ret;
	43736	}
	43737	#else
	43738	__ai __attribute__((target("neon"))) mfloat8x8_t vrev16_mf8(mfloat8x8_t __p0) {
	43739	mfloat8x8_t __ret;
	43740	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	43741	__ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2, 5, 4, 7, 6);
	43742	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
	43743	return __ret;
	43744	}
	43745	#endif
	43746
	43747	#ifdef __LITTLE_ENDIAN__
	43748	__ai __attribute__((target("neon"))) mfloat8x16_t vrev32q_mf8(mfloat8x16_t __p0) {
	43749	mfloat8x16_t __ret;
	43750	__ret = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0, 7, 6, 5, 4, 11, 10, 9, 8, 15, 14, 13, 12);
	43751	return __ret;
	43752	}
	43753	#else
	43754	__ai __attribute__((target("neon"))) mfloat8x16_t vrev32q_mf8(mfloat8x16_t __p0) {
	43755	mfloat8x16_t __ret;
	43756	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	43757	__ret = __builtin_shufflevector(__rev0, __rev0, 3, 2, 1, 0, 7, 6, 5, 4, 11, 10, 9, 8, 15, 14, 13, 12);
	43758	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
	43759	return __ret;
	43760	}
	43761	#endif
	43762
	43763	#ifdef __LITTLE_ENDIAN__
	43764	__ai __attribute__((target("neon"))) mfloat8x8_t vrev32_mf8(mfloat8x8_t __p0) {
	43765	mfloat8x8_t __ret;
	43766	__ret = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0, 7, 6, 5, 4);
	43767	return __ret;
	43768	}
	43769	#else
	43770	__ai __attribute__((target("neon"))) mfloat8x8_t vrev32_mf8(mfloat8x8_t __p0) {
	43771	mfloat8x8_t __ret;
	43772	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	43773	__ret = __builtin_shufflevector(__rev0, __rev0, 3, 2, 1, 0, 7, 6, 5, 4);
	43774	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
	43775	return __ret;
	43776	}
	43777	#endif
	43778
	43779	#ifdef __LITTLE_ENDIAN__
	43780	__ai __attribute__((target("neon"))) mfloat8x16_t vrev64q_mf8(mfloat8x16_t __p0) {
	43781	mfloat8x16_t __ret;
	43782	__ret = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0, 15, 14, 13, 12, 11, 10, 9, 8);
	43783	return __ret;
	43784	}
	43785	#else
	43786	__ai __attribute__((target("neon"))) mfloat8x16_t vrev64q_mf8(mfloat8x16_t __p0) {
	43787	mfloat8x16_t __ret;
	43788	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	43789	__ret = __builtin_shufflevector(__rev0, __rev0, 7, 6, 5, 4, 3, 2, 1, 0, 15, 14, 13, 12, 11, 10, 9, 8);
	43790	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
	43791	return __ret;
	43792	}
	43793	#endif
	43794
	43795	#ifdef __LITTLE_ENDIAN__
	43796	__ai __attribute__((target("neon"))) mfloat8x8_t vrev64_mf8(mfloat8x8_t __p0) {
	43797	mfloat8x8_t __ret;
	43798	__ret = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	43799	return __ret;
	43800	}
	43801	#else
	43802	__ai __attribute__((target("neon"))) mfloat8x8_t vrev64_mf8(mfloat8x8_t __p0) {
	43803	mfloat8x8_t __ret;
	43804	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	43805	__ret = __builtin_shufflevector(__rev0, __rev0, 7, 6, 5, 4, 3, 2, 1, 0);
	43806	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
	43807	return __ret;
	43808	}
	43809	#endif
	43810
	43811	#ifdef __LITTLE_ENDIAN__
	43812	#define vsetq_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	43813	mfloat8x16_t __ret; \
	43814	mfloat8_t __s0 = __p0; \
	43815	mfloat8x16_t __s1 = __p1; \
	43816	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vsetq_lane_mf8(__s0, __s1, __p2)); \
	43817	__ret; \
	43818	})
	43819	#else
	43820	#define vsetq_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	43821	mfloat8x16_t __ret; \
	43822	mfloat8_t __s0 = __p0; \
	43823	mfloat8x16_t __s1 = __p1; \
	43824	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	43825	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vsetq_lane_mf8(__s0, __rev1, __p2)); \
	43826	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
	43827	__ret; \
	43828	})
	43829	#define __noswap_vsetq_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	43830	mfloat8x16_t __ret; \
	43831	mfloat8_t __s0 = __p0; \
	43832	mfloat8x16_t __s1 = __p1; \
	43833	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vsetq_lane_mf8(__s0, __s1, __p2)); \
	43834	__ret; \
	43835	})
	43836	#endif
	43837
	43838	#ifdef __LITTLE_ENDIAN__
	43839	#define vset_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	43840	mfloat8x8_t __ret; \
	43841	mfloat8_t __s0 = __p0; \
	43842	mfloat8x8_t __s1 = __p1; \
	43843	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vset_lane_mf8(__s0, __s1, __p2)); \
	43844	__ret; \
	43845	})
	43846	#else
	43847	#define vset_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	43848	mfloat8x8_t __ret; \
	43849	mfloat8_t __s0 = __p0; \
	43850	mfloat8x8_t __s1 = __p1; \
	43851	mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	43852	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vset_lane_mf8(__s0, __rev1, __p2)); \
	43853	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
	43854	__ret; \
	43855	})
	43856	#define __noswap_vset_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	43857	mfloat8x8_t __ret; \
	43858	mfloat8_t __s0 = __p0; \
	43859	mfloat8x8_t __s1 = __p1; \
	43860	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vset_lane_mf8(__s0, __s1, __p2)); \
	43861	__ret; \
	43862	})
	43863	#endif
	43864
	43865	#ifdef __LITTLE_ENDIAN__
	43866	__ai __attribute__((target("neon"))) mfloat8x8_t vtbl1_mf8(mfloat8x8_t __p0, mfloat8x8_t __p1) {
	43867	mfloat8x8_t __ret;
	43868	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vtbl1_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 12));
	43869	return __ret;
	43870	}
	43871	#else
	43872	__ai __attribute__((target("neon"))) mfloat8x8_t vtbl1_mf8(mfloat8x8_t __p0, mfloat8x8_t __p1) {
	43873	mfloat8x8_t __ret;
	43874	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	43875	mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	43876	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vtbl1_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 12));
	43877	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
	43878	return __ret;
	43879	}
	43880	#endif
	43881
	43882	#ifdef __LITTLE_ENDIAN__
	43883	__ai __attribute__((target("neon"))) mfloat8x8_t vtbl2_mf8(mfloat8x8x2_t __p0, mfloat8x8_t __p1) {
	43884	mfloat8x8_t __ret;
	43885	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vtbl2_v(__builtin_bit_cast(int8x8_t, __p0.val[0]), __builtin_bit_cast(int8x8_t, __p0.val[1]), __builtin_bit_cast(int8x8_t, __p1), 12));
	43886	return __ret;
	43887	}
	43888	#else
	43889	__ai __attribute__((target("neon"))) mfloat8x8_t vtbl2_mf8(mfloat8x8x2_t __p0, mfloat8x8_t __p1) {
	43890	mfloat8x8_t __ret;
	43891	mfloat8x8x2_t __rev0;
	43892	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_64_8);
	43893	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_64_8);
	43894	mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	43895	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vtbl2_v(__builtin_bit_cast(int8x8_t, __rev0.val[0]), __builtin_bit_cast(int8x8_t, __rev0.val[1]), __builtin_bit_cast(int8x8_t, __rev1), 12));
	43896	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
	43897	return __ret;
	43898	}
	43899	#endif
	43900
	43901	#ifdef __LITTLE_ENDIAN__
	43902	__ai __attribute__((target("neon"))) mfloat8x8_t vtbl3_mf8(mfloat8x8x3_t __p0, mfloat8x8_t __p1) {
	43903	mfloat8x8_t __ret;
	43904	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vtbl3_v(__builtin_bit_cast(int8x8_t, __p0.val[0]), __builtin_bit_cast(int8x8_t, __p0.val[1]), __builtin_bit_cast(int8x8_t, __p0.val[2]), __builtin_bit_cast(int8x8_t, __p1), 12));
	43905	return __ret;
	43906	}
	43907	#else
	43908	__ai __attribute__((target("neon"))) mfloat8x8_t vtbl3_mf8(mfloat8x8x3_t __p0, mfloat8x8_t __p1) {
	43909	mfloat8x8_t __ret;
	43910	mfloat8x8x3_t __rev0;
	43911	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_64_8);
	43912	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_64_8);
	43913	__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], __lane_reverse_64_8);
	43914	mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	43915	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vtbl3_v(__builtin_bit_cast(int8x8_t, __rev0.val[0]), __builtin_bit_cast(int8x8_t, __rev0.val[1]), __builtin_bit_cast(int8x8_t, __rev0.val[2]), __builtin_bit_cast(int8x8_t, __rev1), 12));
	43916	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
	43917	return __ret;
	43918	}
	43919	#endif
	43920
	43921	#ifdef __LITTLE_ENDIAN__
	43922	__ai __attribute__((target("neon"))) mfloat8x8_t vtbl4_mf8(mfloat8x8x4_t __p0, mfloat8x8_t __p1) {
	43923	mfloat8x8_t __ret;
	43924	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vtbl4_v(__builtin_bit_cast(int8x8_t, __p0.val[0]), __builtin_bit_cast(int8x8_t, __p0.val[1]), __builtin_bit_cast(int8x8_t, __p0.val[2]), __builtin_bit_cast(int8x8_t, __p0.val[3]), __builtin_bit_cast(int8x8_t, __p1), 12));
	43925	return __ret;
	43926	}
	43927	#else
	43928	__ai __attribute__((target("neon"))) mfloat8x8_t vtbl4_mf8(mfloat8x8x4_t __p0, mfloat8x8_t __p1) {
	43929	mfloat8x8_t __ret;
	43930	mfloat8x8x4_t __rev0;
	43931	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_64_8);
	43932	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_64_8);
	43933	__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], __lane_reverse_64_8);
	43934	__rev0.val[3] = __builtin_shufflevector(__p0.val[3], __p0.val[3], __lane_reverse_64_8);
	43935	mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	43936	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vtbl4_v(__builtin_bit_cast(int8x8_t, __rev0.val[0]), __builtin_bit_cast(int8x8_t, __rev0.val[1]), __builtin_bit_cast(int8x8_t, __rev0.val[2]), __builtin_bit_cast(int8x8_t, __rev0.val[3]), __builtin_bit_cast(int8x8_t, __rev1), 12));
	43937	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
	43938	return __ret;
	43939	}
	43940	#endif
	43941
	43942	#ifdef __LITTLE_ENDIAN__
	43943	__ai __attribute__((target("neon"))) mfloat8x8_t vtbx1_mf8(mfloat8x8_t __p0, mfloat8x8_t __p1, mfloat8x8_t __p2) {
	43944	mfloat8x8_t __ret;
	43945	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vtbx1_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 12));
	43946	return __ret;
	43947	}
	43948	#else
	43949	__ai __attribute__((target("neon"))) mfloat8x8_t vtbx1_mf8(mfloat8x8_t __p0, mfloat8x8_t __p1, mfloat8x8_t __p2) {
	43950	mfloat8x8_t __ret;
	43951	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	43952	mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	43953	mfloat8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	43954	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vtbx1_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 12));
	43955	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
	43956	return __ret;
	43957	}
	43958	#endif
	43959
	43960	#ifdef __LITTLE_ENDIAN__
	43961	__ai __attribute__((target("neon"))) mfloat8x8_t vtbx2_mf8(mfloat8x8_t __p0, mfloat8x8x2_t __p1, mfloat8x8_t __p2) {
	43962	mfloat8x8_t __ret;
	43963	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vtbx2_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1.val[0]), __builtin_bit_cast(int8x8_t, __p1.val[1]), __builtin_bit_cast(int8x8_t, __p2), 12));
	43964	return __ret;
	43965	}
	43966	#else
	43967	__ai __attribute__((target("neon"))) mfloat8x8_t vtbx2_mf8(mfloat8x8_t __p0, mfloat8x8x2_t __p1, mfloat8x8_t __p2) {
	43968	mfloat8x8_t __ret;
	43969	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	43970	mfloat8x8x2_t __rev1;
	43971	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_64_8);
	43972	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_64_8);
	43973	mfloat8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	43974	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vtbx2_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev2), 12));
	43975	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
	43976	return __ret;
	43977	}
	43978	#endif
	43979
	43980	#ifdef __LITTLE_ENDIAN__
	43981	__ai __attribute__((target("neon"))) mfloat8x8_t vtbx3_mf8(mfloat8x8_t __p0, mfloat8x8x3_t __p1, mfloat8x8_t __p2) {
	43982	mfloat8x8_t __ret;
	43983	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vtbx3_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1.val[0]), __builtin_bit_cast(int8x8_t, __p1.val[1]), __builtin_bit_cast(int8x8_t, __p1.val[2]), __builtin_bit_cast(int8x8_t, __p2), 12));
	43984	return __ret;
	43985	}
	43986	#else
	43987	__ai __attribute__((target("neon"))) mfloat8x8_t vtbx3_mf8(mfloat8x8_t __p0, mfloat8x8x3_t __p1, mfloat8x8_t __p2) {
	43988	mfloat8x8_t __ret;
	43989	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	43990	mfloat8x8x3_t __rev1;
	43991	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_64_8);
	43992	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_64_8);
	43993	__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], __lane_reverse_64_8);
	43994	mfloat8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	43995	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vtbx3_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev2), 12));
	43996	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
	43997	return __ret;
	43998	}
	43999	#endif
	44000
	44001	#ifdef __LITTLE_ENDIAN__
	44002	__ai __attribute__((target("neon"))) mfloat8x8_t vtbx4_mf8(mfloat8x8_t __p0, mfloat8x8x4_t __p1, mfloat8x8_t __p2) {
	44003	mfloat8x8_t __ret;
	44004	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vtbx4_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1.val[0]), __builtin_bit_cast(int8x8_t, __p1.val[1]), __builtin_bit_cast(int8x8_t, __p1.val[2]), __builtin_bit_cast(int8x8_t, __p1.val[3]), __builtin_bit_cast(int8x8_t, __p2), 12));
	44005	return __ret;
	44006	}
	44007	#else
	44008	__ai __attribute__((target("neon"))) mfloat8x8_t vtbx4_mf8(mfloat8x8_t __p0, mfloat8x8x4_t __p1, mfloat8x8_t __p2) {
	44009	mfloat8x8_t __ret;
	44010	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	44011	mfloat8x8x4_t __rev1;
	44012	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_64_8);
	44013	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_64_8);
	44014	__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], __lane_reverse_64_8);
	44015	__rev1.val[3] = __builtin_shufflevector(__p1.val[3], __p1.val[3], __lane_reverse_64_8);
	44016	mfloat8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	44017	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vtbx4_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), __builtin_bit_cast(int8x8_t, __rev2), 12));
	44018	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
	44019	return __ret;
	44020	}
	44021	#endif
	44022
	44023	#ifdef __LITTLE_ENDIAN__
	44024	__ai __attribute__((target("neon"))) mfloat8x16x2_t vtrnq_mf8(mfloat8x16_t __p0, mfloat8x16_t __p1) {
	44025	mfloat8x16x2_t __ret;
	44026	__builtin_neon_vtrnq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 44);
	44027	return __ret;
	44028	}
	44029	#else
	44030	__ai __attribute__((target("neon"))) mfloat8x16x2_t vtrnq_mf8(mfloat8x16_t __p0, mfloat8x16_t __p1) {
	44031	mfloat8x16x2_t __ret;
	44032	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	44033	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	44034	__builtin_neon_vtrnq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 44);
	44035
	44036	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8);
	44037	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8);
	44038	return __ret;
	44039	}
	44040	#endif
	44041
	44042	#ifdef __LITTLE_ENDIAN__
	44043	__ai __attribute__((target("neon"))) mfloat8x8x2_t vtrn_mf8(mfloat8x8_t __p0, mfloat8x8_t __p1) {
	44044	mfloat8x8x2_t __ret;
	44045	__builtin_neon_vtrn_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 12);
	44046	return __ret;
	44047	}
	44048	#else
	44049	__ai __attribute__((target("neon"))) mfloat8x8x2_t vtrn_mf8(mfloat8x8_t __p0, mfloat8x8_t __p1) {
	44050	mfloat8x8x2_t __ret;
	44051	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	44052	mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	44053	__builtin_neon_vtrn_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 12);
	44054
	44055	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8);
	44056	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8);
	44057	return __ret;
	44058	}
	44059	#endif
	44060
	44061	#ifdef __LITTLE_ENDIAN__
	44062	__ai __attribute__((target("neon"))) mfloat8x16x2_t vuzpq_mf8(mfloat8x16_t __p0, mfloat8x16_t __p1) {
	44063	mfloat8x16x2_t __ret;
	44064	__builtin_neon_vuzpq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 44);
	44065	return __ret;
	44066	}
	44067	#else
	44068	__ai __attribute__((target("neon"))) mfloat8x16x2_t vuzpq_mf8(mfloat8x16_t __p0, mfloat8x16_t __p1) {
	44069	mfloat8x16x2_t __ret;
	44070	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	44071	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	44072	__builtin_neon_vuzpq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 44);
	44073
	44074	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8);
	44075	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8);
	44076	return __ret;
	44077	}
	44078	#endif
	44079
	44080	#ifdef __LITTLE_ENDIAN__
	44081	__ai __attribute__((target("neon"))) mfloat8x8x2_t vuzp_mf8(mfloat8x8_t __p0, mfloat8x8_t __p1) {
	44082	mfloat8x8x2_t __ret;
	44083	__builtin_neon_vuzp_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 12);
	44084	return __ret;
	44085	}
	44086	#else
	44087	__ai __attribute__((target("neon"))) mfloat8x8x2_t vuzp_mf8(mfloat8x8_t __p0, mfloat8x8_t __p1) {
	44088	mfloat8x8x2_t __ret;
	44089	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	44090	mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	44091	__builtin_neon_vuzp_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 12);
	44092
	44093	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8);
	44094	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8);
	44095	return __ret;
	44096	}
	44097	#endif
	44098
	44099	#ifdef __LITTLE_ENDIAN__
	44100	__ai __attribute__((target("neon"))) mfloat8x16x2_t vzipq_mf8(mfloat8x16_t __p0, mfloat8x16_t __p1) {
	44101	mfloat8x16x2_t __ret;
	44102	__builtin_neon_vzipq_v(&__ret, __builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 44);
	44103	return __ret;
	44104	}
	44105	#else
	44106	__ai __attribute__((target("neon"))) mfloat8x16x2_t vzipq_mf8(mfloat8x16_t __p0, mfloat8x16_t __p1) {
	44107	mfloat8x16x2_t __ret;
	44108	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	44109	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	44110	__builtin_neon_vzipq_v(&__ret, __builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 44);
	44111
	44112	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8);
	44113	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8);
	44114	return __ret;
	44115	}
	44116	#endif
	44117
	44118	#ifdef __LITTLE_ENDIAN__
	44119	__ai __attribute__((target("neon"))) mfloat8x8x2_t vzip_mf8(mfloat8x8_t __p0, mfloat8x8_t __p1) {
	44120	mfloat8x8x2_t __ret;
	44121	__builtin_neon_vzip_v(&__ret, __builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 12);
	44122	return __ret;
	44123	}
	44124	#else
	44125	__ai __attribute__((target("neon"))) mfloat8x8x2_t vzip_mf8(mfloat8x8_t __p0, mfloat8x8_t __p1) {
	44126	mfloat8x8x2_t __ret;
	44127	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	44128	mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	44129	__builtin_neon_vzip_v(&__ret, __builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 12);
	44130
	44131	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8);
	44132	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8);
	44133	return __ret;
	44134	}
43195	44135	#endif
43196	44136
43197	44137	#ifdef __LITTLE_ENDIAN__
43198	44138	__ai __attribute__((target("neon,faminmax"))) float64x2_t vamaxq_f64(float64x2_t __p0, float64x2_t __p1) {
43199	44139	float64x2_t __ret;
43200		__ret = (float64x2_t) __builtin_neon_vamaxq_f64((int8x16_t)__p0, (int8x16_t)__p1, 42);
	44140	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vamaxq_f64(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 42));
43201	44141	return __ret;
43202	44142	}
43203	44143	#else
43204	44144	__ai __attribute__((target("neon,faminmax"))) float64x2_t vamaxq_f64(float64x2_t __p0, float64x2_t __p1) {
43205	44145	float64x2_t __ret;
43206		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
43207		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
43208		__ret = (float64x2_t) __builtin_neon_vamaxq_f64((int8x16_t)__rev0, (int8x16_t)__rev1, 42);
43209		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	44146	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	44147	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	44148	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vamaxq_f64(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 42));
	44149	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
43210	44150	return __ret;
43211	44151	}
43212	44152	#endif
...	...	@@ -43214,16 +44154,16 @@ __ai __attribute__((target("neon,faminmax"))) float64x2_t vamaxq_f64(float64x2_t
43214	44154	#ifdef __LITTLE_ENDIAN__
43215	44155	__ai __attribute__((target("neon,faminmax"))) float32x4_t vamaxq_f32(float32x4_t __p0, float32x4_t __p1) {
43216	44156	float32x4_t __ret;
43217		__ret = (float32x4_t) __builtin_neon_vamaxq_f32((int8x16_t)__p0, (int8x16_t)__p1, 41);
	44157	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vamaxq_f32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 41));
43218	44158	return __ret;
43219	44159	}
43220	44160	#else
43221	44161	__ai __attribute__((target("neon,faminmax"))) float32x4_t vamaxq_f32(float32x4_t __p0, float32x4_t __p1) {
43222	44162	float32x4_t __ret;
43223		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
43224		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
43225		__ret = (float32x4_t) __builtin_neon_vamaxq_f32((int8x16_t)__rev0, (int8x16_t)__rev1, 41);
43226		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	44163	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	44164	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	44165	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vamaxq_f32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 41));
	44166	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
43227	44167	return __ret;
43228	44168	}
43229	44169	#endif
...	...	@@ -43231,16 +44171,16 @@ __ai __attribute__((target("neon,faminmax"))) float32x4_t vamaxq_f32(float32x4_t
43231	44171	#ifdef __LITTLE_ENDIAN__
43232	44172	__ai __attribute__((target("neon,faminmax"))) float16x8_t vamaxq_f16(float16x8_t __p0, float16x8_t __p1) {
43233	44173	float16x8_t __ret;
43234		__ret = (float16x8_t) __builtin_neon_vamaxq_f16((int8x16_t)__p0, (int8x16_t)__p1, 40);
	44174	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vamaxq_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 40));
43235	44175	return __ret;
43236	44176	}
43237	44177	#else
43238	44178	__ai __attribute__((target("neon,faminmax"))) float16x8_t vamaxq_f16(float16x8_t __p0, float16x8_t __p1) {
43239	44179	float16x8_t __ret;
43240		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
43241		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
43242		__ret = (float16x8_t) __builtin_neon_vamaxq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 40);
43243		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	44180	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	44181	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	44182	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vamaxq_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 40));
	44183	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
43244	44184	return __ret;
43245	44185	}
43246	44186	#endif
...	...	@@ -43248,16 +44188,16 @@ __ai __attribute__((target("neon,faminmax"))) float16x8_t vamaxq_f16(float16x8_t
43248	44188	#ifdef __LITTLE_ENDIAN__
43249	44189	__ai __attribute__((target("neon,faminmax"))) float32x2_t vamax_f32(float32x2_t __p0, float32x2_t __p1) {
43250	44190	float32x2_t __ret;
43251		__ret = (float32x2_t) __builtin_neon_vamax_f32((int8x8_t)__p0, (int8x8_t)__p1, 9);
	44191	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vamax_f32(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 9));
43252	44192	return __ret;
43253	44193	}
43254	44194	#else
43255	44195	__ai __attribute__((target("neon,faminmax"))) float32x2_t vamax_f32(float32x2_t __p0, float32x2_t __p1) {
43256	44196	float32x2_t __ret;
43257		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
43258		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
43259		__ret = (float32x2_t) __builtin_neon_vamax_f32((int8x8_t)__rev0, (int8x8_t)__rev1, 9);
43260		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	44197	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	44198	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	44199	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vamax_f32(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 9));
	44200	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
43261	44201	return __ret;
43262	44202	}
43263	44203	#endif
...	...	@@ -43265,16 +44205,16 @@ __ai __attribute__((target("neon,faminmax"))) float32x2_t vamax_f32(float32x2_t
43265	44205	#ifdef __LITTLE_ENDIAN__
43266	44206	__ai __attribute__((target("neon,faminmax"))) float16x4_t vamax_f16(float16x4_t __p0, float16x4_t __p1) {
43267	44207	float16x4_t __ret;
43268		__ret = (float16x4_t) __builtin_neon_vamax_f16((int8x8_t)__p0, (int8x8_t)__p1, 8);
	44208	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vamax_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 8));
43269	44209	return __ret;
43270	44210	}
43271	44211	#else
43272	44212	__ai __attribute__((target("neon,faminmax"))) float16x4_t vamax_f16(float16x4_t __p0, float16x4_t __p1) {
43273	44213	float16x4_t __ret;
43274		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
43275		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
43276		__ret = (float16x4_t) __builtin_neon_vamax_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 8);
43277		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	44214	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	44215	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	44216	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vamax_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 8));
	44217	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
43278	44218	return __ret;
43279	44219	}
43280	44220	#endif
...	...	@@ -43282,16 +44222,16 @@ __ai __attribute__((target("neon,faminmax"))) float16x4_t vamax_f16(float16x4_t
43282	44222	#ifdef __LITTLE_ENDIAN__
43283	44223	__ai __attribute__((target("neon,faminmax"))) float64x2_t vaminq_f64(float64x2_t __p0, float64x2_t __p1) {
43284	44224	float64x2_t __ret;
43285		__ret = (float64x2_t) __builtin_neon_vaminq_f64((int8x16_t)__p0, (int8x16_t)__p1, 42);
	44225	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vaminq_f64(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 42));
43286	44226	return __ret;
43287	44227	}
43288	44228	#else
43289	44229	__ai __attribute__((target("neon,faminmax"))) float64x2_t vaminq_f64(float64x2_t __p0, float64x2_t __p1) {
43290	44230	float64x2_t __ret;
43291		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
43292		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
43293		__ret = (float64x2_t) __builtin_neon_vaminq_f64((int8x16_t)__rev0, (int8x16_t)__rev1, 42);
43294		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	44231	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	44232	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	44233	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vaminq_f64(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 42));
	44234	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
43295	44235	return __ret;
43296	44236	}
43297	44237	#endif
...	...	@@ -43299,16 +44239,16 @@ __ai __attribute__((target("neon,faminmax"))) float64x2_t vaminq_f64(float64x2_t
43299	44239	#ifdef __LITTLE_ENDIAN__
43300	44240	__ai __attribute__((target("neon,faminmax"))) float32x4_t vaminq_f32(float32x4_t __p0, float32x4_t __p1) {
43301	44241	float32x4_t __ret;
43302		__ret = (float32x4_t) __builtin_neon_vaminq_f32((int8x16_t)__p0, (int8x16_t)__p1, 41);
	44242	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vaminq_f32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 41));
43303	44243	return __ret;
43304	44244	}
43305	44245	#else
43306	44246	__ai __attribute__((target("neon,faminmax"))) float32x4_t vaminq_f32(float32x4_t __p0, float32x4_t __p1) {
43307	44247	float32x4_t __ret;
43308		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
43309		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
43310		__ret = (float32x4_t) __builtin_neon_vaminq_f32((int8x16_t)__rev0, (int8x16_t)__rev1, 41);
43311		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	44248	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	44249	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	44250	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vaminq_f32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 41));
	44251	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
43312	44252	return __ret;
43313	44253	}
43314	44254	#endif
...	...	@@ -43316,16 +44256,16 @@ __ai __attribute__((target("neon,faminmax"))) float32x4_t vaminq_f32(float32x4_t
43316	44256	#ifdef __LITTLE_ENDIAN__
43317	44257	__ai __attribute__((target("neon,faminmax"))) float16x8_t vaminq_f16(float16x8_t __p0, float16x8_t __p1) {
43318	44258	float16x8_t __ret;
43319		__ret = (float16x8_t) __builtin_neon_vaminq_f16((int8x16_t)__p0, (int8x16_t)__p1, 40);
	44259	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vaminq_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 40));
43320	44260	return __ret;
43321	44261	}
43322	44262	#else
43323	44263	__ai __attribute__((target("neon,faminmax"))) float16x8_t vaminq_f16(float16x8_t __p0, float16x8_t __p1) {
43324	44264	float16x8_t __ret;
43325		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
43326		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
43327		__ret = (float16x8_t) __builtin_neon_vaminq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 40);
43328		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	44265	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	44266	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	44267	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vaminq_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 40));
	44268	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
43329	44269	return __ret;
43330	44270	}
43331	44271	#endif
...	...	@@ -43333,16 +44273,16 @@ __ai __attribute__((target("neon,faminmax"))) float16x8_t vaminq_f16(float16x8_t
43333	44273	#ifdef __LITTLE_ENDIAN__
43334	44274	__ai __attribute__((target("neon,faminmax"))) float32x2_t vamin_f32(float32x2_t __p0, float32x2_t __p1) {
43335	44275	float32x2_t __ret;
43336		__ret = (float32x2_t) __builtin_neon_vamin_f32((int8x8_t)__p0, (int8x8_t)__p1, 9);
	44276	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vamin_f32(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 9));
43337	44277	return __ret;
43338	44278	}
43339	44279	#else
43340	44280	__ai __attribute__((target("neon,faminmax"))) float32x2_t vamin_f32(float32x2_t __p0, float32x2_t __p1) {
43341	44281	float32x2_t __ret;
43342		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
43343		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
43344		__ret = (float32x2_t) __builtin_neon_vamin_f32((int8x8_t)__rev0, (int8x8_t)__rev1, 9);
43345		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	44282	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	44283	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	44284	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vamin_f32(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 9));
	44285	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
43346	44286	return __ret;
43347	44287	}
43348	44288	#endif
...	...	@@ -43350,16 +44290,16 @@ __ai __attribute__((target("neon,faminmax"))) float32x2_t vamin_f32(float32x2_t
43350	44290	#ifdef __LITTLE_ENDIAN__
43351	44291	__ai __attribute__((target("neon,faminmax"))) float16x4_t vamin_f16(float16x4_t __p0, float16x4_t __p1) {
43352	44292	float16x4_t __ret;
43353		__ret = (float16x4_t) __builtin_neon_vamin_f16((int8x8_t)__p0, (int8x8_t)__p1, 8);
	44293	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vamin_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 8));
43354	44294	return __ret;
43355	44295	}
43356	44296	#else
43357	44297	__ai __attribute__((target("neon,faminmax"))) float16x4_t vamin_f16(float16x4_t __p0, float16x4_t __p1) {
43358	44298	float16x4_t __ret;
43359		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
43360		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
43361		__ret = (float16x4_t) __builtin_neon_vamin_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 8);
43362		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	44299	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	44300	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	44301	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vamin_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 8));
	44302	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
43363	44303	return __ret;
43364	44304	}
43365	44305	#endif
...	...	@@ -43368,105 +44308,105 @@ __ai __attribute__((target("neon,faminmax"))) float16x4_t vamin_f16(float16x4_t
43368	44308	#if defined(__aarch64__) \|\| defined(__arm64ec__)
43369	44309	__ai __attribute__((target("aes,neon"))) poly128_t vmull_p64(poly64_t __p0, poly64_t __p1) {
43370	44310	poly128_t __ret;
43371		__ret = (poly128_t) __builtin_neon_vmull_p64(__p0, __p1);
	44311	__ret = __builtin_bit_cast(poly128_t, __builtin_neon_vmull_p64(__p0, __p1));
43372	44312	return __ret;
43373	44313	}
43374	44314	#ifdef __LITTLE_ENDIAN__
43375		#define vcopyq_lane_bf16(__p0_230, __p1_230, __p2_230, __p3_230) __extension__ ({ \
43376		bfloat16x8_t __ret_230; \
43377		bfloat16x8_t __s0_230 = __p0_230; \
43378		bfloat16x4_t __s2_230 = __p2_230; \
43379		__ret_230 = vsetq_lane_bf16(vget_lane_bf16(__s2_230, __p3_230), __s0_230, __p1_230); \
43380		__ret_230; \
43381		})
43382		#else
43383		#define vcopyq_lane_bf16(__p0_231, __p1_231, __p2_231, __p3_231) __extension__ ({ \
43384		bfloat16x8_t __ret_231; \
43385		bfloat16x8_t __s0_231 = __p0_231; \
43386		bfloat16x4_t __s2_231 = __p2_231; \
43387		bfloat16x8_t __rev0_231; __rev0_231 = __builtin_shufflevector(__s0_231, __s0_231, 7, 6, 5, 4, 3, 2, 1, 0); \
43388		bfloat16x4_t __rev2_231; __rev2_231 = __builtin_shufflevector(__s2_231, __s2_231, 3, 2, 1, 0); \
43389		__ret_231 = __noswap_vsetq_lane_bf16(__noswap_vget_lane_bf16(__rev2_231, __p3_231), __rev0_231, __p1_231); \
43390		__ret_231 = __builtin_shufflevector(__ret_231, __ret_231, 7, 6, 5, 4, 3, 2, 1, 0); \
43391		__ret_231; \
43392		})
43393		#endif
43394
43395		#ifdef __LITTLE_ENDIAN__
43396		#define vcopy_lane_bf16(__p0_232, __p1_232, __p2_232, __p3_232) __extension__ ({ \
43397		bfloat16x4_t __ret_232; \
43398		bfloat16x4_t __s0_232 = __p0_232; \
43399		bfloat16x4_t __s2_232 = __p2_232; \
43400		__ret_232 = vset_lane_bf16(vget_lane_bf16(__s2_232, __p3_232), __s0_232, __p1_232); \
43401		__ret_232; \
43402		})
43403		#else
43404		#define vcopy_lane_bf16(__p0_233, __p1_233, __p2_233, __p3_233) __extension__ ({ \
43405		bfloat16x4_t __ret_233; \
43406		bfloat16x4_t __s0_233 = __p0_233; \
43407		bfloat16x4_t __s2_233 = __p2_233; \
43408		bfloat16x4_t __rev0_233; __rev0_233 = __builtin_shufflevector(__s0_233, __s0_233, 3, 2, 1, 0); \
43409		bfloat16x4_t __rev2_233; __rev2_233 = __builtin_shufflevector(__s2_233, __s2_233, 3, 2, 1, 0); \
43410		__ret_233 = __noswap_vset_lane_bf16(__noswap_vget_lane_bf16(__rev2_233, __p3_233), __rev0_233, __p1_233); \
43411		__ret_233 = __builtin_shufflevector(__ret_233, __ret_233, 3, 2, 1, 0); \
43412		__ret_233; \
43413		})
43414		#endif
43415
43416		#ifdef __LITTLE_ENDIAN__
43417		#define vcopyq_laneq_bf16(__p0_234, __p1_234, __p2_234, __p3_234) __extension__ ({ \
	44315	#define vcopyq_lane_bf16(__p0_234, __p1_234, __p2_234, __p3_234) __extension__ ({ \
43418	44316	bfloat16x8_t __ret_234; \
43419	44317	bfloat16x8_t __s0_234 = __p0_234; \
43420		bfloat16x8_t __s2_234 = __p2_234; \
43421		__ret_234 = vsetq_lane_bf16(vgetq_lane_bf16(__s2_234, __p3_234), __s0_234, __p1_234); \
	44318	bfloat16x4_t __s2_234 = __p2_234; \
	44319	__ret_234 = vsetq_lane_bf16(vget_lane_bf16(__s2_234, __p3_234), __s0_234, __p1_234); \
43422	44320	__ret_234; \
43423	44321	})
43424	44322	#else
43425		#define vcopyq_laneq_bf16(__p0_235, __p1_235, __p2_235, __p3_235) __extension__ ({ \
	44323	#define vcopyq_lane_bf16(__p0_235, __p1_235, __p2_235, __p3_235) __extension__ ({ \
43426	44324	bfloat16x8_t __ret_235; \
43427	44325	bfloat16x8_t __s0_235 = __p0_235; \
43428		bfloat16x8_t __s2_235 = __p2_235; \
43429		bfloat16x8_t __rev0_235; __rev0_235 = __builtin_shufflevector(__s0_235, __s0_235, 7, 6, 5, 4, 3, 2, 1, 0); \
43430		bfloat16x8_t __rev2_235; __rev2_235 = __builtin_shufflevector(__s2_235, __s2_235, 7, 6, 5, 4, 3, 2, 1, 0); \
43431		__ret_235 = __noswap_vsetq_lane_bf16(__noswap_vgetq_lane_bf16(__rev2_235, __p3_235), __rev0_235, __p1_235); \
43432		__ret_235 = __builtin_shufflevector(__ret_235, __ret_235, 7, 6, 5, 4, 3, 2, 1, 0); \
	44326	bfloat16x4_t __s2_235 = __p2_235; \
	44327	bfloat16x8_t __rev0_235; __rev0_235 = __builtin_shufflevector(__s0_235, __s0_235, __lane_reverse_128_16); \
	44328	bfloat16x4_t __rev2_235; __rev2_235 = __builtin_shufflevector(__s2_235, __s2_235, __lane_reverse_64_16); \
	44329	__ret_235 = __noswap_vsetq_lane_bf16(__noswap_vget_lane_bf16(__rev2_235, __p3_235), __rev0_235, __p1_235); \
	44330	__ret_235 = __builtin_shufflevector(__ret_235, __ret_235, __lane_reverse_128_16); \
43433	44331	__ret_235; \
43434	44332	})
43435	44333	#endif
43436	44334
43437	44335	#ifdef __LITTLE_ENDIAN__
43438		#define vcopy_laneq_bf16(__p0_236, __p1_236, __p2_236, __p3_236) __extension__ ({ \
	44336	#define vcopy_lane_bf16(__p0_236, __p1_236, __p2_236, __p3_236) __extension__ ({ \
43439	44337	bfloat16x4_t __ret_236; \
43440	44338	bfloat16x4_t __s0_236 = __p0_236; \
43441		bfloat16x8_t __s2_236 = __p2_236; \
43442		__ret_236 = vset_lane_bf16(vgetq_lane_bf16(__s2_236, __p3_236), __s0_236, __p1_236); \
	44339	bfloat16x4_t __s2_236 = __p2_236; \
	44340	__ret_236 = vset_lane_bf16(vget_lane_bf16(__s2_236, __p3_236), __s0_236, __p1_236); \
43443	44341	__ret_236; \
43444	44342	})
43445	44343	#else
43446		#define vcopy_laneq_bf16(__p0_237, __p1_237, __p2_237, __p3_237) __extension__ ({ \
	44344	#define vcopy_lane_bf16(__p0_237, __p1_237, __p2_237, __p3_237) __extension__ ({ \
43447	44345	bfloat16x4_t __ret_237; \
43448	44346	bfloat16x4_t __s0_237 = __p0_237; \
43449		bfloat16x8_t __s2_237 = __p2_237; \
43450		bfloat16x4_t __rev0_237; __rev0_237 = __builtin_shufflevector(__s0_237, __s0_237, 3, 2, 1, 0); \
43451		bfloat16x8_t __rev2_237; __rev2_237 = __builtin_shufflevector(__s2_237, __s2_237, 7, 6, 5, 4, 3, 2, 1, 0); \
43452		__ret_237 = __noswap_vset_lane_bf16(__noswap_vgetq_lane_bf16(__rev2_237, __p3_237), __rev0_237, __p1_237); \
43453		__ret_237 = __builtin_shufflevector(__ret_237, __ret_237, 3, 2, 1, 0); \
	44347	bfloat16x4_t __s2_237 = __p2_237; \
	44348	bfloat16x4_t __rev0_237; __rev0_237 = __builtin_shufflevector(__s0_237, __s0_237, __lane_reverse_64_16); \
	44349	bfloat16x4_t __rev2_237; __rev2_237 = __builtin_shufflevector(__s2_237, __s2_237, __lane_reverse_64_16); \
	44350	__ret_237 = __noswap_vset_lane_bf16(__noswap_vget_lane_bf16(__rev2_237, __p3_237), __rev0_237, __p1_237); \
	44351	__ret_237 = __builtin_shufflevector(__ret_237, __ret_237, __lane_reverse_64_16); \
43454	44352	__ret_237; \
43455	44353	})
43456	44354	#endif
43457	44355
	44356	#ifdef __LITTLE_ENDIAN__
	44357	#define vcopyq_laneq_bf16(__p0_238, __p1_238, __p2_238, __p3_238) __extension__ ({ \
	44358	bfloat16x8_t __ret_238; \
	44359	bfloat16x8_t __s0_238 = __p0_238; \
	44360	bfloat16x8_t __s2_238 = __p2_238; \
	44361	__ret_238 = vsetq_lane_bf16(vgetq_lane_bf16(__s2_238, __p3_238), __s0_238, __p1_238); \
	44362	__ret_238; \
	44363	})
	44364	#else
	44365	#define vcopyq_laneq_bf16(__p0_239, __p1_239, __p2_239, __p3_239) __extension__ ({ \
	44366	bfloat16x8_t __ret_239; \
	44367	bfloat16x8_t __s0_239 = __p0_239; \
	44368	bfloat16x8_t __s2_239 = __p2_239; \
	44369	bfloat16x8_t __rev0_239; __rev0_239 = __builtin_shufflevector(__s0_239, __s0_239, __lane_reverse_128_16); \
	44370	bfloat16x8_t __rev2_239; __rev2_239 = __builtin_shufflevector(__s2_239, __s2_239, __lane_reverse_128_16); \
	44371	__ret_239 = __noswap_vsetq_lane_bf16(__noswap_vgetq_lane_bf16(__rev2_239, __p3_239), __rev0_239, __p1_239); \
	44372	__ret_239 = __builtin_shufflevector(__ret_239, __ret_239, __lane_reverse_128_16); \
	44373	__ret_239; \
	44374	})
	44375	#endif
	44376
	44377	#ifdef __LITTLE_ENDIAN__
	44378	#define vcopy_laneq_bf16(__p0_240, __p1_240, __p2_240, __p3_240) __extension__ ({ \
	44379	bfloat16x4_t __ret_240; \
	44380	bfloat16x4_t __s0_240 = __p0_240; \
	44381	bfloat16x8_t __s2_240 = __p2_240; \
	44382	__ret_240 = vset_lane_bf16(vgetq_lane_bf16(__s2_240, __p3_240), __s0_240, __p1_240); \
	44383	__ret_240; \
	44384	})
	44385	#else
	44386	#define vcopy_laneq_bf16(__p0_241, __p1_241, __p2_241, __p3_241) __extension__ ({ \
	44387	bfloat16x4_t __ret_241; \
	44388	bfloat16x4_t __s0_241 = __p0_241; \
	44389	bfloat16x8_t __s2_241 = __p2_241; \
	44390	bfloat16x4_t __rev0_241; __rev0_241 = __builtin_shufflevector(__s0_241, __s0_241, __lane_reverse_64_16); \
	44391	bfloat16x8_t __rev2_241; __rev2_241 = __builtin_shufflevector(__s2_241, __s2_241, __lane_reverse_128_16); \
	44392	__ret_241 = __noswap_vset_lane_bf16(__noswap_vgetq_lane_bf16(__rev2_241, __p3_241), __rev0_241, __p1_241); \
	44393	__ret_241 = __builtin_shufflevector(__ret_241, __ret_241, __lane_reverse_64_16); \
	44394	__ret_241; \
	44395	})
	44396	#endif
	44397
43458	44398	#ifdef __LITTLE_ENDIAN__
43459	44399	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vcvt_bf16_f32(float32x4_t __p0) {
43460	44400	bfloat16x4_t __ret;
43461		__ret = (bfloat16x4_t) __builtin_neon_vcvt_bf16_f32((int8x16_t)__p0, 11);
	44401	__ret = __builtin_bit_cast(bfloat16x4_t, __builtin_neon_vcvt_bf16_f32(__builtin_bit_cast(int8x16_t, __p0), 11));
43462	44402	return __ret;
43463	44403	}
43464	44404	#else
43465	44405	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vcvt_bf16_f32(float32x4_t __p0) {
43466	44406	bfloat16x4_t __ret;
43467		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
43468		__ret = (bfloat16x4_t) __builtin_neon_vcvt_bf16_f32((int8x16_t)__rev0, 11);
43469		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	44407	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	44408	__ret = __builtin_bit_cast(bfloat16x4_t, __builtin_neon_vcvt_bf16_f32(__builtin_bit_cast(int8x16_t, __rev0), 11));
	44409	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
43470	44410	return __ret;
43471	44411	}
43472	44412	#endif
...	...	@@ -43474,16 +44414,16 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x4_t vcvt_bf16_f32(float32x4_t
43474	44414	#ifdef __LITTLE_ENDIAN__
43475	44415	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vcvtq_high_bf16_f32(bfloat16x8_t __p0, float32x4_t __p1) {
43476	44416	bfloat16x8_t __ret;
43477		__ret = (bfloat16x8_t) __builtin_neon_vcvtq_high_bf16_f32((int8x16_t)__p0, (int8x16_t)__p1, 43);
	44417	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vcvtq_high_bf16_f32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 43));
43478	44418	return __ret;
43479	44419	}
43480	44420	#else
43481	44421	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vcvtq_high_bf16_f32(bfloat16x8_t __p0, float32x4_t __p1) {
43482	44422	bfloat16x8_t __ret;
43483		bfloat16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
43484		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
43485		__ret = (bfloat16x8_t) __builtin_neon_vcvtq_high_bf16_f32((int8x16_t)__rev0, (int8x16_t)__rev1, 43);
43486		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	44423	bfloat16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	44424	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	44425	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vcvtq_high_bf16_f32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 43));
	44426	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
43487	44427	return __ret;
43488	44428	}
43489	44429	#endif
...	...	@@ -43491,424 +44431,424 @@ __ai __attribute__((target("bf16,neon"))) bfloat16x8_t vcvtq_high_bf16_f32(bfloa
43491	44431	#ifdef __LITTLE_ENDIAN__
43492	44432	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vcvtq_low_bf16_f32(float32x4_t __p0) {
43493	44433	bfloat16x8_t __ret;
43494		__ret = (bfloat16x8_t) __builtin_neon_vcvtq_low_bf16_f32((int8x16_t)__p0, 43);
	44434	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vcvtq_low_bf16_f32(__builtin_bit_cast(int8x16_t, __p0), 43));
43495	44435	return __ret;
43496	44436	}
43497	44437	#else
43498	44438	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vcvtq_low_bf16_f32(float32x4_t __p0) {
43499	44439	bfloat16x8_t __ret;
43500		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
43501		__ret = (bfloat16x8_t) __builtin_neon_vcvtq_low_bf16_f32((int8x16_t)__rev0, 43);
43502		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	44440	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	44441	__ret = __builtin_bit_cast(bfloat16x8_t, __builtin_neon_vcvtq_low_bf16_f32(__builtin_bit_cast(int8x16_t, __rev0), 43));
	44442	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
43503	44443	return __ret;
43504	44444	}
43505	44445	#endif
43506	44446
43507	44447	__ai __attribute__((target("bf16,neon"))) poly8x8_t vreinterpret_p8_bf16(bfloat16x4_t __p0) {
43508	44448	poly8x8_t __ret;
43509		__ret = (poly8x8_t)(__p0);
	44449	__ret = __builtin_bit_cast(poly8x8_t, __p0);
43510	44450	return __ret;
43511	44451	}
43512	44452	__ai __attribute__((target("bf16,neon"))) poly64x1_t vreinterpret_p64_bf16(bfloat16x4_t __p0) {
43513	44453	poly64x1_t __ret;
43514		__ret = (poly64x1_t)(__p0);
	44454	__ret = __builtin_bit_cast(poly64x1_t, __p0);
43515	44455	return __ret;
43516	44456	}
43517	44457	__ai __attribute__((target("bf16,neon"))) poly16x4_t vreinterpret_p16_bf16(bfloat16x4_t __p0) {
43518	44458	poly16x4_t __ret;
43519		__ret = (poly16x4_t)(__p0);
	44459	__ret = __builtin_bit_cast(poly16x4_t, __p0);
43520	44460	return __ret;
43521	44461	}
43522	44462	__ai __attribute__((target("bf16,neon"))) poly8x16_t vreinterpretq_p8_bf16(bfloat16x8_t __p0) {
43523	44463	poly8x16_t __ret;
43524		__ret = (poly8x16_t)(__p0);
	44464	__ret = __builtin_bit_cast(poly8x16_t, __p0);
43525	44465	return __ret;
43526	44466	}
43527	44467	__ai __attribute__((target("bf16,neon"))) poly128_t vreinterpretq_p128_bf16(bfloat16x8_t __p0) {
43528	44468	poly128_t __ret;
43529		__ret = (poly128_t)(__p0);
	44469	__ret = __builtin_bit_cast(poly128_t, __p0);
43530	44470	return __ret;
43531	44471	}
43532	44472	__ai __attribute__((target("bf16,neon"))) poly64x2_t vreinterpretq_p64_bf16(bfloat16x8_t __p0) {
43533	44473	poly64x2_t __ret;
43534		__ret = (poly64x2_t)(__p0);
	44474	__ret = __builtin_bit_cast(poly64x2_t, __p0);
43535	44475	return __ret;
43536	44476	}
43537	44477	__ai __attribute__((target("bf16,neon"))) poly16x8_t vreinterpretq_p16_bf16(bfloat16x8_t __p0) {
43538	44478	poly16x8_t __ret;
43539		__ret = (poly16x8_t)(__p0);
	44479	__ret = __builtin_bit_cast(poly16x8_t, __p0);
43540	44480	return __ret;
43541	44481	}
43542	44482	__ai __attribute__((target("bf16,neon"))) uint8x16_t vreinterpretq_u8_bf16(bfloat16x8_t __p0) {
43543	44483	uint8x16_t __ret;
43544		__ret = (uint8x16_t)(__p0);
	44484	__ret = __builtin_bit_cast(uint8x16_t, __p0);
43545	44485	return __ret;
43546	44486	}
43547	44487	__ai __attribute__((target("bf16,neon"))) uint32x4_t vreinterpretq_u32_bf16(bfloat16x8_t __p0) {
43548	44488	uint32x4_t __ret;
43549		__ret = (uint32x4_t)(__p0);
	44489	__ret = __builtin_bit_cast(uint32x4_t, __p0);
43550	44490	return __ret;
43551	44491	}
43552	44492	__ai __attribute__((target("bf16,neon"))) uint64x2_t vreinterpretq_u64_bf16(bfloat16x8_t __p0) {
43553	44493	uint64x2_t __ret;
43554		__ret = (uint64x2_t)(__p0);
	44494	__ret = __builtin_bit_cast(uint64x2_t, __p0);
43555	44495	return __ret;
43556	44496	}
43557	44497	__ai __attribute__((target("bf16,neon"))) uint16x8_t vreinterpretq_u16_bf16(bfloat16x8_t __p0) {
43558	44498	uint16x8_t __ret;
43559		__ret = (uint16x8_t)(__p0);
	44499	__ret = __builtin_bit_cast(uint16x8_t, __p0);
43560	44500	return __ret;
43561	44501	}
43562	44502	__ai __attribute__((target("bf16,neon"))) int8x16_t vreinterpretq_s8_bf16(bfloat16x8_t __p0) {
43563	44503	int8x16_t __ret;
43564		__ret = (int8x16_t)(__p0);
	44504	__ret = __builtin_bit_cast(int8x16_t, __p0);
43565	44505	return __ret;
43566	44506	}
43567	44507	__ai __attribute__((target("bf16,neon"))) float64x2_t vreinterpretq_f64_bf16(bfloat16x8_t __p0) {
43568	44508	float64x2_t __ret;
43569		__ret = (float64x2_t)(__p0);
	44509	__ret = __builtin_bit_cast(float64x2_t, __p0);
43570	44510	return __ret;
43571	44511	}
43572	44512	__ai __attribute__((target("bf16,neon"))) float32x4_t vreinterpretq_f32_bf16(bfloat16x8_t __p0) {
43573	44513	float32x4_t __ret;
43574		__ret = (float32x4_t)(__p0);
	44514	__ret = __builtin_bit_cast(float32x4_t, __p0);
43575	44515	return __ret;
43576	44516	}
43577	44517	__ai __attribute__((target("bf16,neon"))) float16x8_t vreinterpretq_f16_bf16(bfloat16x8_t __p0) {
43578	44518	float16x8_t __ret;
43579		__ret = (float16x8_t)(__p0);
	44519	__ret = __builtin_bit_cast(float16x8_t, __p0);
43580	44520	return __ret;
43581	44521	}
43582	44522	__ai __attribute__((target("bf16,neon"))) int32x4_t vreinterpretq_s32_bf16(bfloat16x8_t __p0) {
43583	44523	int32x4_t __ret;
43584		__ret = (int32x4_t)(__p0);
	44524	__ret = __builtin_bit_cast(int32x4_t, __p0);
43585	44525	return __ret;
43586	44526	}
43587	44527	__ai __attribute__((target("bf16,neon"))) int64x2_t vreinterpretq_s64_bf16(bfloat16x8_t __p0) {
43588	44528	int64x2_t __ret;
43589		__ret = (int64x2_t)(__p0);
	44529	__ret = __builtin_bit_cast(int64x2_t, __p0);
43590	44530	return __ret;
43591	44531	}
43592	44532	__ai __attribute__((target("bf16,neon"))) int16x8_t vreinterpretq_s16_bf16(bfloat16x8_t __p0) {
43593	44533	int16x8_t __ret;
43594		__ret = (int16x8_t)(__p0);
	44534	__ret = __builtin_bit_cast(int16x8_t, __p0);
43595	44535	return __ret;
43596	44536	}
43597	44537	__ai __attribute__((target("bf16,neon"))) uint8x8_t vreinterpret_u8_bf16(bfloat16x4_t __p0) {
43598	44538	uint8x8_t __ret;
43599		__ret = (uint8x8_t)(__p0);
	44539	__ret = __builtin_bit_cast(uint8x8_t, __p0);
43600	44540	return __ret;
43601	44541	}
43602	44542	__ai __attribute__((target("bf16,neon"))) uint32x2_t vreinterpret_u32_bf16(bfloat16x4_t __p0) {
43603	44543	uint32x2_t __ret;
43604		__ret = (uint32x2_t)(__p0);
	44544	__ret = __builtin_bit_cast(uint32x2_t, __p0);
43605	44545	return __ret;
43606	44546	}
43607	44547	__ai __attribute__((target("bf16,neon"))) uint64x1_t vreinterpret_u64_bf16(bfloat16x4_t __p0) {
43608	44548	uint64x1_t __ret;
43609		__ret = (uint64x1_t)(__p0);
	44549	__ret = __builtin_bit_cast(uint64x1_t, __p0);
43610	44550	return __ret;
43611	44551	}
43612	44552	__ai __attribute__((target("bf16,neon"))) uint16x4_t vreinterpret_u16_bf16(bfloat16x4_t __p0) {
43613	44553	uint16x4_t __ret;
43614		__ret = (uint16x4_t)(__p0);
	44554	__ret = __builtin_bit_cast(uint16x4_t, __p0);
43615	44555	return __ret;
43616	44556	}
43617	44557	__ai __attribute__((target("bf16,neon"))) int8x8_t vreinterpret_s8_bf16(bfloat16x4_t __p0) {
43618	44558	int8x8_t __ret;
43619		__ret = (int8x8_t)(__p0);
	44559	__ret = __builtin_bit_cast(int8x8_t, __p0);
43620	44560	return __ret;
43621	44561	}
43622	44562	__ai __attribute__((target("bf16,neon"))) float64x1_t vreinterpret_f64_bf16(bfloat16x4_t __p0) {
43623	44563	float64x1_t __ret;
43624		__ret = (float64x1_t)(__p0);
	44564	__ret = __builtin_bit_cast(float64x1_t, __p0);
43625	44565	return __ret;
43626	44566	}
43627	44567	__ai __attribute__((target("bf16,neon"))) float32x2_t vreinterpret_f32_bf16(bfloat16x4_t __p0) {
43628	44568	float32x2_t __ret;
43629		__ret = (float32x2_t)(__p0);
	44569	__ret = __builtin_bit_cast(float32x2_t, __p0);
43630	44570	return __ret;
43631	44571	}
43632	44572	__ai __attribute__((target("bf16,neon"))) float16x4_t vreinterpret_f16_bf16(bfloat16x4_t __p0) {
43633	44573	float16x4_t __ret;
43634		__ret = (float16x4_t)(__p0);
	44574	__ret = __builtin_bit_cast(float16x4_t, __p0);
43635	44575	return __ret;
43636	44576	}
43637	44577	__ai __attribute__((target("bf16,neon"))) int32x2_t vreinterpret_s32_bf16(bfloat16x4_t __p0) {
43638	44578	int32x2_t __ret;
43639		__ret = (int32x2_t)(__p0);
	44579	__ret = __builtin_bit_cast(int32x2_t, __p0);
43640	44580	return __ret;
43641	44581	}
43642	44582	__ai __attribute__((target("bf16,neon"))) int64x1_t vreinterpret_s64_bf16(bfloat16x4_t __p0) {
43643	44583	int64x1_t __ret;
43644		__ret = (int64x1_t)(__p0);
	44584	__ret = __builtin_bit_cast(int64x1_t, __p0);
43645	44585	return __ret;
43646	44586	}
43647	44587	__ai __attribute__((target("bf16,neon"))) int16x4_t vreinterpret_s16_bf16(bfloat16x4_t __p0) {
43648	44588	int16x4_t __ret;
43649		__ret = (int16x4_t)(__p0);
	44589	__ret = __builtin_bit_cast(int16x4_t, __p0);
43650	44590	return __ret;
43651	44591	}
43652	44592	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vreinterpretq_bf16_p8(poly8x16_t __p0) {
43653	44593	bfloat16x8_t __ret;
43654		__ret = (bfloat16x8_t)(__p0);
	44594	__ret = __builtin_bit_cast(bfloat16x8_t, __p0);
43655	44595	return __ret;
43656	44596	}
43657	44597	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vreinterpretq_bf16_p128(poly128_t __p0) {
43658	44598	bfloat16x8_t __ret;
43659		__ret = (bfloat16x8_t)(__p0);
	44599	__ret = __builtin_bit_cast(bfloat16x8_t, __p0);
43660	44600	return __ret;
43661	44601	}
43662	44602	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vreinterpretq_bf16_p64(poly64x2_t __p0) {
43663	44603	bfloat16x8_t __ret;
43664		__ret = (bfloat16x8_t)(__p0);
	44604	__ret = __builtin_bit_cast(bfloat16x8_t, __p0);
43665	44605	return __ret;
43666	44606	}
43667	44607	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vreinterpretq_bf16_p16(poly16x8_t __p0) {
43668	44608	bfloat16x8_t __ret;
43669		__ret = (bfloat16x8_t)(__p0);
	44609	__ret = __builtin_bit_cast(bfloat16x8_t, __p0);
43670	44610	return __ret;
43671	44611	}
43672	44612	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vreinterpretq_bf16_u8(uint8x16_t __p0) {
43673	44613	bfloat16x8_t __ret;
43674		__ret = (bfloat16x8_t)(__p0);
	44614	__ret = __builtin_bit_cast(bfloat16x8_t, __p0);
43675	44615	return __ret;
43676	44616	}
43677	44617	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vreinterpretq_bf16_u32(uint32x4_t __p0) {
43678	44618	bfloat16x8_t __ret;
43679		__ret = (bfloat16x8_t)(__p0);
	44619	__ret = __builtin_bit_cast(bfloat16x8_t, __p0);
43680	44620	return __ret;
43681	44621	}
43682	44622	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vreinterpretq_bf16_u64(uint64x2_t __p0) {
43683	44623	bfloat16x8_t __ret;
43684		__ret = (bfloat16x8_t)(__p0);
	44624	__ret = __builtin_bit_cast(bfloat16x8_t, __p0);
43685	44625	return __ret;
43686	44626	}
43687	44627	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vreinterpretq_bf16_u16(uint16x8_t __p0) {
43688	44628	bfloat16x8_t __ret;
43689		__ret = (bfloat16x8_t)(__p0);
	44629	__ret = __builtin_bit_cast(bfloat16x8_t, __p0);
43690	44630	return __ret;
43691	44631	}
43692	44632	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vreinterpretq_bf16_s8(int8x16_t __p0) {
43693	44633	bfloat16x8_t __ret;
43694		__ret = (bfloat16x8_t)(__p0);
	44634	__ret = __builtin_bit_cast(bfloat16x8_t, __p0);
43695	44635	return __ret;
43696	44636	}
43697	44637	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vreinterpretq_bf16_f64(float64x2_t __p0) {
43698	44638	bfloat16x8_t __ret;
43699		__ret = (bfloat16x8_t)(__p0);
	44639	__ret = __builtin_bit_cast(bfloat16x8_t, __p0);
43700	44640	return __ret;
43701	44641	}
43702	44642	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vreinterpretq_bf16_f32(float32x4_t __p0) {
43703	44643	bfloat16x8_t __ret;
43704		__ret = (bfloat16x8_t)(__p0);
	44644	__ret = __builtin_bit_cast(bfloat16x8_t, __p0);
43705	44645	return __ret;
43706	44646	}
43707	44647	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vreinterpretq_bf16_f16(float16x8_t __p0) {
43708	44648	bfloat16x8_t __ret;
43709		__ret = (bfloat16x8_t)(__p0);
	44649	__ret = __builtin_bit_cast(bfloat16x8_t, __p0);
43710	44650	return __ret;
43711	44651	}
43712	44652	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vreinterpretq_bf16_s32(int32x4_t __p0) {
43713	44653	bfloat16x8_t __ret;
43714		__ret = (bfloat16x8_t)(__p0);
	44654	__ret = __builtin_bit_cast(bfloat16x8_t, __p0);
43715	44655	return __ret;
43716	44656	}
43717	44657	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vreinterpretq_bf16_s64(int64x2_t __p0) {
43718	44658	bfloat16x8_t __ret;
43719		__ret = (bfloat16x8_t)(__p0);
	44659	__ret = __builtin_bit_cast(bfloat16x8_t, __p0);
43720	44660	return __ret;
43721	44661	}
43722	44662	__ai __attribute__((target("bf16,neon"))) bfloat16x8_t vreinterpretq_bf16_s16(int16x8_t __p0) {
43723	44663	bfloat16x8_t __ret;
43724		__ret = (bfloat16x8_t)(__p0);
	44664	__ret = __builtin_bit_cast(bfloat16x8_t, __p0);
43725	44665	return __ret;
43726	44666	}
43727	44667	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_p8(poly8x8_t __p0) {
43728	44668	bfloat16x4_t __ret;
43729		__ret = (bfloat16x4_t)(__p0);
	44669	__ret = __builtin_bit_cast(bfloat16x4_t, __p0);
43730	44670	return __ret;
43731	44671	}
43732	44672	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_p64(poly64x1_t __p0) {
43733	44673	bfloat16x4_t __ret;
43734		__ret = (bfloat16x4_t)(__p0);
	44674	__ret = __builtin_bit_cast(bfloat16x4_t, __p0);
43735	44675	return __ret;
43736	44676	}
43737	44677	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_p16(poly16x4_t __p0) {
43738	44678	bfloat16x4_t __ret;
43739		__ret = (bfloat16x4_t)(__p0);
	44679	__ret = __builtin_bit_cast(bfloat16x4_t, __p0);
43740	44680	return __ret;
43741	44681	}
43742	44682	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_u8(uint8x8_t __p0) {
43743	44683	bfloat16x4_t __ret;
43744		__ret = (bfloat16x4_t)(__p0);
	44684	__ret = __builtin_bit_cast(bfloat16x4_t, __p0);
43745	44685	return __ret;
43746	44686	}
43747	44687	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_u32(uint32x2_t __p0) {
43748	44688	bfloat16x4_t __ret;
43749		__ret = (bfloat16x4_t)(__p0);
	44689	__ret = __builtin_bit_cast(bfloat16x4_t, __p0);
43750	44690	return __ret;
43751	44691	}
43752	44692	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_u64(uint64x1_t __p0) {
43753	44693	bfloat16x4_t __ret;
43754		__ret = (bfloat16x4_t)(__p0);
	44694	__ret = __builtin_bit_cast(bfloat16x4_t, __p0);
43755	44695	return __ret;
43756	44696	}
43757	44697	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_u16(uint16x4_t __p0) {
43758	44698	bfloat16x4_t __ret;
43759		__ret = (bfloat16x4_t)(__p0);
	44699	__ret = __builtin_bit_cast(bfloat16x4_t, __p0);
43760	44700	return __ret;
43761	44701	}
43762	44702	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_s8(int8x8_t __p0) {
43763	44703	bfloat16x4_t __ret;
43764		__ret = (bfloat16x4_t)(__p0);
	44704	__ret = __builtin_bit_cast(bfloat16x4_t, __p0);
43765	44705	return __ret;
43766	44706	}
43767	44707	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_f64(float64x1_t __p0) {
43768	44708	bfloat16x4_t __ret;
43769		__ret = (bfloat16x4_t)(__p0);
	44709	__ret = __builtin_bit_cast(bfloat16x4_t, __p0);
43770	44710	return __ret;
43771	44711	}
43772	44712	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_f32(float32x2_t __p0) {
43773	44713	bfloat16x4_t __ret;
43774		__ret = (bfloat16x4_t)(__p0);
	44714	__ret = __builtin_bit_cast(bfloat16x4_t, __p0);
43775	44715	return __ret;
43776	44716	}
43777	44717	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_f16(float16x4_t __p0) {
43778	44718	bfloat16x4_t __ret;
43779		__ret = (bfloat16x4_t)(__p0);
	44719	__ret = __builtin_bit_cast(bfloat16x4_t, __p0);
43780	44720	return __ret;
43781	44721	}
43782	44722	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_s32(int32x2_t __p0) {
43783	44723	bfloat16x4_t __ret;
43784		__ret = (bfloat16x4_t)(__p0);
	44724	__ret = __builtin_bit_cast(bfloat16x4_t, __p0);
43785	44725	return __ret;
43786	44726	}
43787	44727	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_s64(int64x1_t __p0) {
43788	44728	bfloat16x4_t __ret;
43789		__ret = (bfloat16x4_t)(__p0);
	44729	__ret = __builtin_bit_cast(bfloat16x4_t, __p0);
43790	44730	return __ret;
43791	44731	}
43792	44732	__ai __attribute__((target("bf16,neon"))) bfloat16x4_t vreinterpret_bf16_s16(int16x4_t __p0) {
43793	44733	bfloat16x4_t __ret;
43794		__ret = (bfloat16x4_t)(__p0);
	44734	__ret = __builtin_bit_cast(bfloat16x4_t, __p0);
43795	44735	return __ret;
43796	44736	}
43797	44737	#ifdef __LITTLE_ENDIAN__
43798		#define vdotq_laneq_u32(__p0_238, __p1_238, __p2_238, __p3_238) __extension__ ({ \
43799		uint32x4_t __ret_238; \
43800		uint32x4_t __s0_238 = __p0_238; \
43801		uint8x16_t __s1_238 = __p1_238; \
43802		uint8x16_t __s2_238 = __p2_238; \
43803		__ret_238 = vdotq_u32(__s0_238, __s1_238, __builtin_bit_cast(uint8x16_t, splatq_laneq_u32(__builtin_bit_cast(uint32x4_t, __s2_238), __p3_238))); \
43804		__ret_238; \
	44738	#define vdotq_laneq_u32(__p0_242, __p1_242, __p2_242, __p3_242) __extension__ ({ \
	44739	uint32x4_t __ret_242; \
	44740	uint32x4_t __s0_242 = __p0_242; \
	44741	uint8x16_t __s1_242 = __p1_242; \
	44742	uint8x16_t __s2_242 = __p2_242; \
	44743	__ret_242 = vdotq_u32(__s0_242, __s1_242, __builtin_bit_cast(uint8x16_t, splatq_laneq_u32(__builtin_bit_cast(uint32x4_t, __s2_242), __p3_242))); \
	44744	__ret_242; \
43805	44745	})
43806	44746	#else
43807		#define vdotq_laneq_u32(__p0_239, __p1_239, __p2_239, __p3_239) __extension__ ({ \
43808		uint32x4_t __ret_239; \
43809		uint32x4_t __s0_239 = __p0_239; \
43810		uint8x16_t __s1_239 = __p1_239; \
43811		uint8x16_t __s2_239 = __p2_239; \
43812		uint32x4_t __rev0_239; __rev0_239 = __builtin_shufflevector(__s0_239, __s0_239, 3, 2, 1, 0); \
43813		uint8x16_t __rev1_239; __rev1_239 = __builtin_shufflevector(__s1_239, __s1_239, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
43814		uint8x16_t __rev2_239; __rev2_239 = __builtin_shufflevector(__s2_239, __s2_239, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
43815		__ret_239 = __noswap_vdotq_u32(__rev0_239, __rev1_239, __builtin_bit_cast(uint8x16_t, __noswap_splatq_laneq_u32(__builtin_bit_cast(uint32x4_t, __rev2_239), __p3_239))); \
43816		__ret_239 = __builtin_shufflevector(__ret_239, __ret_239, 3, 2, 1, 0); \
43817		__ret_239; \
	44747	#define vdotq_laneq_u32(__p0_243, __p1_243, __p2_243, __p3_243) __extension__ ({ \
	44748	uint32x4_t __ret_243; \
	44749	uint32x4_t __s0_243 = __p0_243; \
	44750	uint8x16_t __s1_243 = __p1_243; \
	44751	uint8x16_t __s2_243 = __p2_243; \
	44752	uint32x4_t __rev0_243; __rev0_243 = __builtin_shufflevector(__s0_243, __s0_243, __lane_reverse_128_32); \
	44753	uint8x16_t __rev1_243; __rev1_243 = __builtin_shufflevector(__s1_243, __s1_243, __lane_reverse_128_8); \
	44754	uint8x16_t __rev2_243; __rev2_243 = __builtin_shufflevector(__s2_243, __s2_243, __lane_reverse_128_8); \
	44755	__ret_243 = __noswap_vdotq_u32(__rev0_243, __rev1_243, __builtin_bit_cast(uint8x16_t, __noswap_splatq_laneq_u32(__builtin_bit_cast(uint32x4_t, __rev2_243), __p3_243))); \
	44756	__ret_243 = __builtin_shufflevector(__ret_243, __ret_243, __lane_reverse_128_32); \
	44757	__ret_243; \
43818	44758	})
43819	44759	#endif
43820	44760
43821	44761	#ifdef __LITTLE_ENDIAN__
43822		#define vdotq_laneq_s32(__p0_240, __p1_240, __p2_240, __p3_240) __extension__ ({ \
43823		int32x4_t __ret_240; \
43824		int32x4_t __s0_240 = __p0_240; \
43825		int8x16_t __s1_240 = __p1_240; \
43826		int8x16_t __s2_240 = __p2_240; \
43827		__ret_240 = vdotq_s32(__s0_240, __s1_240, __builtin_bit_cast(int8x16_t, splatq_laneq_s32(__builtin_bit_cast(int32x4_t, __s2_240), __p3_240))); \
43828		__ret_240; \
	44762	#define vdotq_laneq_s32(__p0_244, __p1_244, __p2_244, __p3_244) __extension__ ({ \
	44763	int32x4_t __ret_244; \
	44764	int32x4_t __s0_244 = __p0_244; \
	44765	int8x16_t __s1_244 = __p1_244; \
	44766	int8x16_t __s2_244 = __p2_244; \
	44767	__ret_244 = vdotq_s32(__s0_244, __s1_244, __builtin_bit_cast(int8x16_t, splatq_laneq_s32(__builtin_bit_cast(int32x4_t, __s2_244), __p3_244))); \
	44768	__ret_244; \
43829	44769	})
43830	44770	#else
43831		#define vdotq_laneq_s32(__p0_241, __p1_241, __p2_241, __p3_241) __extension__ ({ \
43832		int32x4_t __ret_241; \
43833		int32x4_t __s0_241 = __p0_241; \
43834		int8x16_t __s1_241 = __p1_241; \
43835		int8x16_t __s2_241 = __p2_241; \
43836		int32x4_t __rev0_241; __rev0_241 = __builtin_shufflevector(__s0_241, __s0_241, 3, 2, 1, 0); \
43837		int8x16_t __rev1_241; __rev1_241 = __builtin_shufflevector(__s1_241, __s1_241, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
43838		int8x16_t __rev2_241; __rev2_241 = __builtin_shufflevector(__s2_241, __s2_241, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
43839		__ret_241 = __noswap_vdotq_s32(__rev0_241, __rev1_241, __builtin_bit_cast(int8x16_t, __noswap_splatq_laneq_s32(__builtin_bit_cast(int32x4_t, __rev2_241), __p3_241))); \
43840		__ret_241 = __builtin_shufflevector(__ret_241, __ret_241, 3, 2, 1, 0); \
43841		__ret_241; \
	44771	#define vdotq_laneq_s32(__p0_245, __p1_245, __p2_245, __p3_245) __extension__ ({ \
	44772	int32x4_t __ret_245; \
	44773	int32x4_t __s0_245 = __p0_245; \
	44774	int8x16_t __s1_245 = __p1_245; \
	44775	int8x16_t __s2_245 = __p2_245; \
	44776	int32x4_t __rev0_245; __rev0_245 = __builtin_shufflevector(__s0_245, __s0_245, __lane_reverse_128_32); \
	44777	int8x16_t __rev1_245; __rev1_245 = __builtin_shufflevector(__s1_245, __s1_245, __lane_reverse_128_8); \
	44778	int8x16_t __rev2_245; __rev2_245 = __builtin_shufflevector(__s2_245, __s2_245, __lane_reverse_128_8); \
	44779	__ret_245 = __noswap_vdotq_s32(__rev0_245, __rev1_245, __builtin_bit_cast(int8x16_t, __noswap_splatq_laneq_s32(__builtin_bit_cast(int32x4_t, __rev2_245), __p3_245))); \
	44780	__ret_245 = __builtin_shufflevector(__ret_245, __ret_245, __lane_reverse_128_32); \
	44781	__ret_245; \
43842	44782	})
43843	44783	#endif
43844	44784
43845	44785	#ifdef __LITTLE_ENDIAN__
43846		#define vdot_laneq_u32(__p0_242, __p1_242, __p2_242, __p3_242) __extension__ ({ \
43847		uint32x2_t __ret_242; \
43848		uint32x2_t __s0_242 = __p0_242; \
43849		uint8x8_t __s1_242 = __p1_242; \
43850		uint8x16_t __s2_242 = __p2_242; \
43851		__ret_242 = vdot_u32(__s0_242, __s1_242, __builtin_bit_cast(uint8x8_t, splat_laneq_u32(__builtin_bit_cast(uint32x4_t, __s2_242), __p3_242))); \
43852		__ret_242; \
	44786	#define vdot_laneq_u32(__p0_246, __p1_246, __p2_246, __p3_246) __extension__ ({ \
	44787	uint32x2_t __ret_246; \
	44788	uint32x2_t __s0_246 = __p0_246; \
	44789	uint8x8_t __s1_246 = __p1_246; \
	44790	uint8x16_t __s2_246 = __p2_246; \
	44791	__ret_246 = vdot_u32(__s0_246, __s1_246, __builtin_bit_cast(uint8x8_t, splat_laneq_u32(__builtin_bit_cast(uint32x4_t, __s2_246), __p3_246))); \
	44792	__ret_246; \
43853	44793	})
43854	44794	#else
43855		#define vdot_laneq_u32(__p0_243, __p1_243, __p2_243, __p3_243) __extension__ ({ \
43856		uint32x2_t __ret_243; \
43857		uint32x2_t __s0_243 = __p0_243; \
43858		uint8x8_t __s1_243 = __p1_243; \
43859		uint8x16_t __s2_243 = __p2_243; \
43860		uint32x2_t __rev0_243; __rev0_243 = __builtin_shufflevector(__s0_243, __s0_243, 1, 0); \
43861		uint8x8_t __rev1_243; __rev1_243 = __builtin_shufflevector(__s1_243, __s1_243, 7, 6, 5, 4, 3, 2, 1, 0); \
43862		uint8x16_t __rev2_243; __rev2_243 = __builtin_shufflevector(__s2_243, __s2_243, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
43863		__ret_243 = __noswap_vdot_u32(__rev0_243, __rev1_243, __builtin_bit_cast(uint8x8_t, __noswap_splat_laneq_u32(__builtin_bit_cast(uint32x4_t, __rev2_243), __p3_243))); \
43864		__ret_243 = __builtin_shufflevector(__ret_243, __ret_243, 1, 0); \
43865		__ret_243; \
	44795	#define vdot_laneq_u32(__p0_247, __p1_247, __p2_247, __p3_247) __extension__ ({ \
	44796	uint32x2_t __ret_247; \
	44797	uint32x2_t __s0_247 = __p0_247; \
	44798	uint8x8_t __s1_247 = __p1_247; \
	44799	uint8x16_t __s2_247 = __p2_247; \
	44800	uint32x2_t __rev0_247; __rev0_247 = __builtin_shufflevector(__s0_247, __s0_247, __lane_reverse_64_32); \
	44801	uint8x8_t __rev1_247; __rev1_247 = __builtin_shufflevector(__s1_247, __s1_247, __lane_reverse_64_8); \
	44802	uint8x16_t __rev2_247; __rev2_247 = __builtin_shufflevector(__s2_247, __s2_247, __lane_reverse_128_8); \
	44803	__ret_247 = __noswap_vdot_u32(__rev0_247, __rev1_247, __builtin_bit_cast(uint8x8_t, __noswap_splat_laneq_u32(__builtin_bit_cast(uint32x4_t, __rev2_247), __p3_247))); \
	44804	__ret_247 = __builtin_shufflevector(__ret_247, __ret_247, __lane_reverse_64_32); \
	44805	__ret_247; \
43866	44806	})
43867	44807	#endif
43868	44808
43869	44809	#ifdef __LITTLE_ENDIAN__
43870		#define vdot_laneq_s32(__p0_244, __p1_244, __p2_244, __p3_244) __extension__ ({ \
43871		int32x2_t __ret_244; \
43872		int32x2_t __s0_244 = __p0_244; \
43873		int8x8_t __s1_244 = __p1_244; \
43874		int8x16_t __s2_244 = __p2_244; \
43875		__ret_244 = vdot_s32(__s0_244, __s1_244, __builtin_bit_cast(int8x8_t, splat_laneq_s32(__builtin_bit_cast(int32x4_t, __s2_244), __p3_244))); \
43876		__ret_244; \
	44810	#define vdot_laneq_s32(__p0_248, __p1_248, __p2_248, __p3_248) __extension__ ({ \
	44811	int32x2_t __ret_248; \
	44812	int32x2_t __s0_248 = __p0_248; \
	44813	int8x8_t __s1_248 = __p1_248; \
	44814	int8x16_t __s2_248 = __p2_248; \
	44815	__ret_248 = vdot_s32(__s0_248, __s1_248, __builtin_bit_cast(int8x8_t, splat_laneq_s32(__builtin_bit_cast(int32x4_t, __s2_248), __p3_248))); \
	44816	__ret_248; \
43877	44817	})
43878	44818	#else
43879		#define vdot_laneq_s32(__p0_245, __p1_245, __p2_245, __p3_245) __extension__ ({ \
43880		int32x2_t __ret_245; \
43881		int32x2_t __s0_245 = __p0_245; \
43882		int8x8_t __s1_245 = __p1_245; \
43883		int8x16_t __s2_245 = __p2_245; \
43884		int32x2_t __rev0_245; __rev0_245 = __builtin_shufflevector(__s0_245, __s0_245, 1, 0); \
43885		int8x8_t __rev1_245; __rev1_245 = __builtin_shufflevector(__s1_245, __s1_245, 7, 6, 5, 4, 3, 2, 1, 0); \
43886		int8x16_t __rev2_245; __rev2_245 = __builtin_shufflevector(__s2_245, __s2_245, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
43887		__ret_245 = __noswap_vdot_s32(__rev0_245, __rev1_245, __builtin_bit_cast(int8x8_t, __noswap_splat_laneq_s32(__builtin_bit_cast(int32x4_t, __rev2_245), __p3_245))); \
43888		__ret_245 = __builtin_shufflevector(__ret_245, __ret_245, 1, 0); \
43889		__ret_245; \
	44819	#define vdot_laneq_s32(__p0_249, __p1_249, __p2_249, __p3_249) __extension__ ({ \
	44820	int32x2_t __ret_249; \
	44821	int32x2_t __s0_249 = __p0_249; \
	44822	int8x8_t __s1_249 = __p1_249; \
	44823	int8x16_t __s2_249 = __p2_249; \
	44824	int32x2_t __rev0_249; __rev0_249 = __builtin_shufflevector(__s0_249, __s0_249, __lane_reverse_64_32); \
	44825	int8x8_t __rev1_249; __rev1_249 = __builtin_shufflevector(__s1_249, __s1_249, __lane_reverse_64_8); \
	44826	int8x16_t __rev2_249; __rev2_249 = __builtin_shufflevector(__s2_249, __s2_249, __lane_reverse_128_8); \
	44827	__ret_249 = __noswap_vdot_s32(__rev0_249, __rev1_249, __builtin_bit_cast(int8x8_t, __noswap_splat_laneq_s32(__builtin_bit_cast(int32x4_t, __rev2_249), __p3_249))); \
	44828	__ret_249 = __builtin_shufflevector(__ret_249, __ret_249, __lane_reverse_64_32); \
	44829	__ret_249; \
43890	44830	})
43891	44831	#endif
43892	44832
43893	44833	#ifdef __LITTLE_ENDIAN__
43894	44834	__ai __attribute__((target("fp16fml,neon"))) float32x4_t vfmlalq_high_f16(float32x4_t __p0, float16x8_t __p1, float16x8_t __p2) {
43895	44835	float32x4_t __ret;
43896		__ret = (float32x4_t) __builtin_neon_vfmlalq_high_f16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
	44836	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vfmlalq_high_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 41));
43897	44837	return __ret;
43898	44838	}
43899	44839	#else
43900	44840	__ai __attribute__((target("fp16fml,neon"))) float32x4_t vfmlalq_high_f16(float32x4_t __p0, float16x8_t __p1, float16x8_t __p2) {
43901	44841	float32x4_t __ret;
43902		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
43903		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
43904		float16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
43905		__ret = (float32x4_t) __builtin_neon_vfmlalq_high_f16((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 41);
43906		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	44842	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	44843	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	44844	float16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
	44845	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vfmlalq_high_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 41));
	44846	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
43907	44847	return __ret;
43908	44848	}
43909	44849	__ai __attribute__((target("fp16fml,neon"))) float32x4_t __noswap_vfmlalq_high_f16(float32x4_t __p0, float16x8_t __p1, float16x8_t __p2) {
43910	44850	float32x4_t __ret;
43911		__ret = (float32x4_t) __builtin_neon_vfmlalq_high_f16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
	44851	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vfmlalq_high_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 41));
43912	44852	return __ret;
43913	44853	}
43914	44854	#endif
...	...	@@ -43916,22 +44856,22 @@ __ai __attribute__((target("fp16fml,neon"))) float32x4_t __noswap_vfmlalq_high_f
43916	44856	#ifdef __LITTLE_ENDIAN__
43917	44857	__ai __attribute__((target("fp16fml,neon"))) float32x2_t vfmlal_high_f16(float32x2_t __p0, float16x4_t __p1, float16x4_t __p2) {
43918	44858	float32x2_t __ret;
43919		__ret = (float32x2_t) __builtin_neon_vfmlal_high_f16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
	44859	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vfmlal_high_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 9));
43920	44860	return __ret;
43921	44861	}
43922	44862	#else
43923	44863	__ai __attribute__((target("fp16fml,neon"))) float32x2_t vfmlal_high_f16(float32x2_t __p0, float16x4_t __p1, float16x4_t __p2) {
43924	44864	float32x2_t __ret;
43925		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
43926		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
43927		float16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
43928		__ret = (float32x2_t) __builtin_neon_vfmlal_high_f16((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 9);
43929		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	44865	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	44866	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	44867	float16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
	44868	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vfmlal_high_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 9));
	44869	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
43930	44870	return __ret;
43931	44871	}
43932	44872	__ai __attribute__((target("fp16fml,neon"))) float32x2_t __noswap_vfmlal_high_f16(float32x2_t __p0, float16x4_t __p1, float16x4_t __p2) {
43933	44873	float32x2_t __ret;
43934		__ret = (float32x2_t) __builtin_neon_vfmlal_high_f16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
	44874	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vfmlal_high_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 9));
43935	44875	return __ret;
43936	44876	}
43937	44877	#endif
...	...	@@ -43939,22 +44879,22 @@ __ai __attribute__((target("fp16fml,neon"))) float32x2_t __noswap_vfmlal_high_f1
43939	44879	#ifdef __LITTLE_ENDIAN__
43940	44880	__ai __attribute__((target("fp16fml,neon"))) float32x4_t vfmlalq_low_f16(float32x4_t __p0, float16x8_t __p1, float16x8_t __p2) {
43941	44881	float32x4_t __ret;
43942		__ret = (float32x4_t) __builtin_neon_vfmlalq_low_f16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
	44882	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vfmlalq_low_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 41));
43943	44883	return __ret;
43944	44884	}
43945	44885	#else
43946	44886	__ai __attribute__((target("fp16fml,neon"))) float32x4_t vfmlalq_low_f16(float32x4_t __p0, float16x8_t __p1, float16x8_t __p2) {
43947	44887	float32x4_t __ret;
43948		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
43949		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
43950		float16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
43951		__ret = (float32x4_t) __builtin_neon_vfmlalq_low_f16((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 41);
43952		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	44888	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	44889	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	44890	float16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
	44891	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vfmlalq_low_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 41));
	44892	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
43953	44893	return __ret;
43954	44894	}
43955	44895	__ai __attribute__((target("fp16fml,neon"))) float32x4_t __noswap_vfmlalq_low_f16(float32x4_t __p0, float16x8_t __p1, float16x8_t __p2) {
43956	44896	float32x4_t __ret;
43957		__ret = (float32x4_t) __builtin_neon_vfmlalq_low_f16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
	44897	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vfmlalq_low_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 41));
43958	44898	return __ret;
43959	44899	}
43960	44900	#endif
...	...	@@ -43962,22 +44902,22 @@ __ai __attribute__((target("fp16fml,neon"))) float32x4_t __noswap_vfmlalq_low_f1
43962	44902	#ifdef __LITTLE_ENDIAN__
43963	44903	__ai __attribute__((target("fp16fml,neon"))) float32x2_t vfmlal_low_f16(float32x2_t __p0, float16x4_t __p1, float16x4_t __p2) {
43964	44904	float32x2_t __ret;
43965		__ret = (float32x2_t) __builtin_neon_vfmlal_low_f16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
	44905	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vfmlal_low_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 9));
43966	44906	return __ret;
43967	44907	}
43968	44908	#else
43969	44909	__ai __attribute__((target("fp16fml,neon"))) float32x2_t vfmlal_low_f16(float32x2_t __p0, float16x4_t __p1, float16x4_t __p2) {
43970	44910	float32x2_t __ret;
43971		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
43972		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
43973		float16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
43974		__ret = (float32x2_t) __builtin_neon_vfmlal_low_f16((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 9);
43975		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	44911	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	44912	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	44913	float16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
	44914	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vfmlal_low_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 9));
	44915	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
43976	44916	return __ret;
43977	44917	}
43978	44918	__ai __attribute__((target("fp16fml,neon"))) float32x2_t __noswap_vfmlal_low_f16(float32x2_t __p0, float16x4_t __p1, float16x4_t __p2) {
43979	44919	float32x2_t __ret;
43980		__ret = (float32x2_t) __builtin_neon_vfmlal_low_f16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
	44920	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vfmlal_low_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 9));
43981	44921	return __ret;
43982	44922	}
43983	44923	#endif
...	...	@@ -43985,22 +44925,22 @@ __ai __attribute__((target("fp16fml,neon"))) float32x2_t __noswap_vfmlal_low_f16
43985	44925	#ifdef __LITTLE_ENDIAN__
43986	44926	__ai __attribute__((target("fp16fml,neon"))) float32x4_t vfmlslq_high_f16(float32x4_t __p0, float16x8_t __p1, float16x8_t __p2) {
43987	44927	float32x4_t __ret;
43988		__ret = (float32x4_t) __builtin_neon_vfmlslq_high_f16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
	44928	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vfmlslq_high_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 41));
43989	44929	return __ret;
43990	44930	}
43991	44931	#else
43992	44932	__ai __attribute__((target("fp16fml,neon"))) float32x4_t vfmlslq_high_f16(float32x4_t __p0, float16x8_t __p1, float16x8_t __p2) {
43993	44933	float32x4_t __ret;
43994		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
43995		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
43996		float16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
43997		__ret = (float32x4_t) __builtin_neon_vfmlslq_high_f16((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 41);
43998		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	44934	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	44935	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	44936	float16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
	44937	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vfmlslq_high_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 41));
	44938	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
43999	44939	return __ret;
44000	44940	}
44001	44941	__ai __attribute__((target("fp16fml,neon"))) float32x4_t __noswap_vfmlslq_high_f16(float32x4_t __p0, float16x8_t __p1, float16x8_t __p2) {
44002	44942	float32x4_t __ret;
44003		__ret = (float32x4_t) __builtin_neon_vfmlslq_high_f16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
	44943	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vfmlslq_high_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 41));
44004	44944	return __ret;
44005	44945	}
44006	44946	#endif
...	...	@@ -44008,22 +44948,22 @@ __ai __attribute__((target("fp16fml,neon"))) float32x4_t __noswap_vfmlslq_high_f
44008	44948	#ifdef __LITTLE_ENDIAN__
44009	44949	__ai __attribute__((target("fp16fml,neon"))) float32x2_t vfmlsl_high_f16(float32x2_t __p0, float16x4_t __p1, float16x4_t __p2) {
44010	44950	float32x2_t __ret;
44011		__ret = (float32x2_t) __builtin_neon_vfmlsl_high_f16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
	44951	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vfmlsl_high_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 9));
44012	44952	return __ret;
44013	44953	}
44014	44954	#else
44015	44955	__ai __attribute__((target("fp16fml,neon"))) float32x2_t vfmlsl_high_f16(float32x2_t __p0, float16x4_t __p1, float16x4_t __p2) {
44016	44956	float32x2_t __ret;
44017		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
44018		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
44019		float16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
44020		__ret = (float32x2_t) __builtin_neon_vfmlsl_high_f16((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 9);
44021		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	44957	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	44958	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	44959	float16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
	44960	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vfmlsl_high_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 9));
	44961	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
44022	44962	return __ret;
44023	44963	}
44024	44964	__ai __attribute__((target("fp16fml,neon"))) float32x2_t __noswap_vfmlsl_high_f16(float32x2_t __p0, float16x4_t __p1, float16x4_t __p2) {
44025	44965	float32x2_t __ret;
44026		__ret = (float32x2_t) __builtin_neon_vfmlsl_high_f16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
	44966	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vfmlsl_high_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 9));
44027	44967	return __ret;
44028	44968	}
44029	44969	#endif
...	...	@@ -44031,22 +44971,22 @@ __ai __attribute__((target("fp16fml,neon"))) float32x2_t __noswap_vfmlsl_high_f1
44031	44971	#ifdef __LITTLE_ENDIAN__
44032	44972	__ai __attribute__((target("fp16fml,neon"))) float32x4_t vfmlslq_low_f16(float32x4_t __p0, float16x8_t __p1, float16x8_t __p2) {
44033	44973	float32x4_t __ret;
44034		__ret = (float32x4_t) __builtin_neon_vfmlslq_low_f16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
	44974	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vfmlslq_low_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 41));
44035	44975	return __ret;
44036	44976	}
44037	44977	#else
44038	44978	__ai __attribute__((target("fp16fml,neon"))) float32x4_t vfmlslq_low_f16(float32x4_t __p0, float16x8_t __p1, float16x8_t __p2) {
44039	44979	float32x4_t __ret;
44040		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
44041		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
44042		float16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
44043		__ret = (float32x4_t) __builtin_neon_vfmlslq_low_f16((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 41);
44044		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	44980	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	44981	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	44982	float16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
	44983	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vfmlslq_low_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 41));
	44984	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
44045	44985	return __ret;
44046	44986	}
44047	44987	__ai __attribute__((target("fp16fml,neon"))) float32x4_t __noswap_vfmlslq_low_f16(float32x4_t __p0, float16x8_t __p1, float16x8_t __p2) {
44048	44988	float32x4_t __ret;
44049		__ret = (float32x4_t) __builtin_neon_vfmlslq_low_f16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
	44989	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vfmlslq_low_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 41));
44050	44990	return __ret;
44051	44991	}
44052	44992	#endif
...	...	@@ -44054,22 +44994,22 @@ __ai __attribute__((target("fp16fml,neon"))) float32x4_t __noswap_vfmlslq_low_f1
44054	44994	#ifdef __LITTLE_ENDIAN__
44055	44995	__ai __attribute__((target("fp16fml,neon"))) float32x2_t vfmlsl_low_f16(float32x2_t __p0, float16x4_t __p1, float16x4_t __p2) {
44056	44996	float32x2_t __ret;
44057		__ret = (float32x2_t) __builtin_neon_vfmlsl_low_f16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
	44997	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vfmlsl_low_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 9));
44058	44998	return __ret;
44059	44999	}
44060	45000	#else
44061	45001	__ai __attribute__((target("fp16fml,neon"))) float32x2_t vfmlsl_low_f16(float32x2_t __p0, float16x4_t __p1, float16x4_t __p2) {
44062	45002	float32x2_t __ret;
44063		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
44064		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
44065		float16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
44066		__ret = (float32x2_t) __builtin_neon_vfmlsl_low_f16((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 9);
44067		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	45003	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	45004	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	45005	float16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
	45006	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vfmlsl_low_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 9));
	45007	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
44068	45008	return __ret;
44069	45009	}
44070	45010	__ai __attribute__((target("fp16fml,neon"))) float32x2_t __noswap_vfmlsl_low_f16(float32x2_t __p0, float16x4_t __p1, float16x4_t __p2) {
44071	45011	float32x2_t __ret;
44072		__ret = (float32x2_t) __builtin_neon_vfmlsl_low_f16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
	45012	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vfmlsl_low_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 9));
44073	45013	return __ret;
44074	45014	}
44075	45015	#endif
...	...	@@ -44083,10 +45023,10 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vdivq_f16(float16x8_t
44083	45023	#else
44084	45024	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vdivq_f16(float16x8_t __p0, float16x8_t __p1) {
44085	45025	float16x8_t __ret;
44086		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
44087		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	45026	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	45027	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
44088	45028	__ret = __rev0 / __rev1;
44089		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	45029	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
44090	45030	return __ret;
44091	45031	}
44092	45032	#endif
...	...	@@ -44100,10 +45040,10 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44100	45040	#else
44101	45041	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t __p0, float16x4_t __p1) {
44102	45042	float16x4_t __ret;
44103		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
44104		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	45043	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	45044	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
44105	45045	__ret = __rev0 / __rev1;
44106		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	45046	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
44107	45047	return __ret;
44108	45048	}
44109	45049	#endif
...	...	@@ -44114,7 +45054,7 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44114	45054	float16_t __s0 = __p0; \
44115	45055	float16_t __s1 = __p1; \
44116	45056	float16x4_t __s2 = __p2; \
44117		__ret = (float16_t) __builtin_neon_vfmah_lane_f16(__s0, __s1, (float16x4_t)__s2, __p3); \
	45057	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vfmah_lane_f16(__s0, __s1, __s2, __p3)); \
44118	45058	__ret; \
44119	45059	})
44120	45060	#else
...	...	@@ -44123,8 +45063,8 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44123	45063	float16_t __s0 = __p0; \
44124	45064	float16_t __s1 = __p1; \
44125	45065	float16x4_t __s2 = __p2; \
44126		float16x4_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 3, 2, 1, 0); \
44127		__ret = (float16_t) __builtin_neon_vfmah_lane_f16(__s0, __s1, (float16x4_t)__rev2, __p3); \
	45066	float16x4_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_64_16); \
	45067	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vfmah_lane_f16(__s0, __s1, __rev2, __p3)); \
44128	45068	__ret; \
44129	45069	})
44130	45070	#define __noswap_vfmah_lane_f16(__p0, __p1, __p2, __p3) __extension__ ({ \
...	...	@@ -44132,7 +45072,7 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44132	45072	float16_t __s0 = __p0; \
44133	45073	float16_t __s1 = __p1; \
44134	45074	float16x4_t __s2 = __p2; \
44135		__ret = (float16_t) __builtin_neon_vfmah_lane_f16(__s0, __s1, (float16x4_t)__s2, __p3); \
	45075	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vfmah_lane_f16(__s0, __s1, __s2, __p3)); \
44136	45076	__ret; \
44137	45077	})
44138	45078	#endif
...	...	@@ -44143,7 +45083,7 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44143	45083	float16x8_t __s0 = __p0; \
44144	45084	float16x8_t __s1 = __p1; \
44145	45085	float16x4_t __s2 = __p2; \
44146		__ret = (float16x8_t) __builtin_neon_vfmaq_lane_f16((int8x16_t)__s0, (int8x16_t)__s1, (int8x8_t)__s2, __p3, 40); \
	45086	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vfmaq_lane_f16(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __builtin_bit_cast(int8x8_t, __s2), __p3, 40)); \
44147	45087	__ret; \
44148	45088	})
44149	45089	#else
...	...	@@ -44152,11 +45092,11 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44152	45092	float16x8_t __s0 = __p0; \
44153	45093	float16x8_t __s1 = __p1; \
44154	45094	float16x4_t __s2 = __p2; \
44155		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
44156		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
44157		float16x4_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 3, 2, 1, 0); \
44158		__ret = (float16x8_t) __builtin_neon_vfmaq_lane_f16((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x8_t)__rev2, __p3, 40); \
44159		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	45095	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	45096	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	45097	float16x4_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_64_16); \
	45098	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vfmaq_lane_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), __p3, 40)); \
	45099	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
44160	45100	__ret; \
44161	45101	})
44162	45102	#define __noswap_vfmaq_lane_f16(__p0, __p1, __p2, __p3) __extension__ ({ \
...	...	@@ -44164,7 +45104,7 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44164	45104	float16x8_t __s0 = __p0; \
44165	45105	float16x8_t __s1 = __p1; \
44166	45106	float16x4_t __s2 = __p2; \
44167		__ret = (float16x8_t) __builtin_neon_vfmaq_lane_f16((int8x16_t)__s0, (int8x16_t)__s1, (int8x8_t)__s2, __p3, 40); \
	45107	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vfmaq_lane_f16(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __builtin_bit_cast(int8x8_t, __s2), __p3, 40)); \
44168	45108	__ret; \
44169	45109	})
44170	45110	#endif
...	...	@@ -44175,7 +45115,7 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44175	45115	float16x4_t __s0 = __p0; \
44176	45116	float16x4_t __s1 = __p1; \
44177	45117	float16x4_t __s2 = __p2; \
44178		__ret = (float16x4_t) __builtin_neon_vfma_lane_f16((int8x8_t)__s0, (int8x8_t)__s1, (int8x8_t)__s2, __p3, 8); \
	45118	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vfma_lane_f16(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __builtin_bit_cast(int8x8_t, __s2), __p3, 8)); \
44179	45119	__ret; \
44180	45120	})
44181	45121	#else
...	...	@@ -44184,11 +45124,11 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44184	45124	float16x4_t __s0 = __p0; \
44185	45125	float16x4_t __s1 = __p1; \
44186	45126	float16x4_t __s2 = __p2; \
44187		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
44188		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
44189		float16x4_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 3, 2, 1, 0); \
44190		__ret = (float16x4_t) __builtin_neon_vfma_lane_f16((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, __p3, 8); \
44191		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	45127	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	45128	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	45129	float16x4_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_64_16); \
	45130	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vfma_lane_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), __p3, 8)); \
	45131	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
44192	45132	__ret; \
44193	45133	})
44194	45134	#define __noswap_vfma_lane_f16(__p0, __p1, __p2, __p3) __extension__ ({ \
...	...	@@ -44196,7 +45136,7 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44196	45136	float16x4_t __s0 = __p0; \
44197	45137	float16x4_t __s1 = __p1; \
44198	45138	float16x4_t __s2 = __p2; \
44199		__ret = (float16x4_t) __builtin_neon_vfma_lane_f16((int8x8_t)__s0, (int8x8_t)__s1, (int8x8_t)__s2, __p3, 8); \
	45139	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vfma_lane_f16(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __builtin_bit_cast(int8x8_t, __s2), __p3, 8)); \
44200	45140	__ret; \
44201	45141	})
44202	45142	#endif
...	...	@@ -44207,7 +45147,7 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44207	45147	float16_t __s0 = __p0; \
44208	45148	float16_t __s1 = __p1; \
44209	45149	float16x8_t __s2 = __p2; \
44210		__ret = (float16_t) __builtin_neon_vfmah_laneq_f16(__s0, __s1, (float16x8_t)__s2, __p3); \
	45150	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vfmah_laneq_f16(__s0, __s1, __s2, __p3)); \
44211	45151	__ret; \
44212	45152	})
44213	45153	#else
...	...	@@ -44216,8 +45156,8 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44216	45156	float16_t __s0 = __p0; \
44217	45157	float16_t __s1 = __p1; \
44218	45158	float16x8_t __s2 = __p2; \
44219		float16x8_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 7, 6, 5, 4, 3, 2, 1, 0); \
44220		__ret = (float16_t) __builtin_neon_vfmah_laneq_f16(__s0, __s1, (float16x8_t)__rev2, __p3); \
	45159	float16x8_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_128_16); \
	45160	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vfmah_laneq_f16(__s0, __s1, __rev2, __p3)); \
44221	45161	__ret; \
44222	45162	})
44223	45163	#define __noswap_vfmah_laneq_f16(__p0, __p1, __p2, __p3) __extension__ ({ \
...	...	@@ -44225,7 +45165,7 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44225	45165	float16_t __s0 = __p0; \
44226	45166	float16_t __s1 = __p1; \
44227	45167	float16x8_t __s2 = __p2; \
44228		__ret = (float16_t) __builtin_neon_vfmah_laneq_f16(__s0, __s1, (float16x8_t)__s2, __p3); \
	45168	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vfmah_laneq_f16(__s0, __s1, __s2, __p3)); \
44229	45169	__ret; \
44230	45170	})
44231	45171	#endif
...	...	@@ -44236,7 +45176,7 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44236	45176	float16x8_t __s0 = __p0; \
44237	45177	float16x8_t __s1 = __p1; \
44238	45178	float16x8_t __s2 = __p2; \
44239		__ret = (float16x8_t) __builtin_neon_vfmaq_laneq_f16((int8x16_t)__s0, (int8x16_t)__s1, (int8x16_t)__s2, __p3, 40); \
	45179	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vfmaq_laneq_f16(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __builtin_bit_cast(int8x16_t, __s2), __p3, 40)); \
44240	45180	__ret; \
44241	45181	})
44242	45182	#else
...	...	@@ -44245,11 +45185,11 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44245	45185	float16x8_t __s0 = __p0; \
44246	45186	float16x8_t __s1 = __p1; \
44247	45187	float16x8_t __s2 = __p2; \
44248		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
44249		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
44250		float16x8_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 7, 6, 5, 4, 3, 2, 1, 0); \
44251		__ret = (float16x8_t) __builtin_neon_vfmaq_laneq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, __p3, 40); \
44252		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	45188	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	45189	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	45190	float16x8_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_128_16); \
	45191	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vfmaq_laneq_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), __p3, 40)); \
	45192	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
44253	45193	__ret; \
44254	45194	})
44255	45195	#define __noswap_vfmaq_laneq_f16(__p0, __p1, __p2, __p3) __extension__ ({ \
...	...	@@ -44257,7 +45197,7 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44257	45197	float16x8_t __s0 = __p0; \
44258	45198	float16x8_t __s1 = __p1; \
44259	45199	float16x8_t __s2 = __p2; \
44260		__ret = (float16x8_t) __builtin_neon_vfmaq_laneq_f16((int8x16_t)__s0, (int8x16_t)__s1, (int8x16_t)__s2, __p3, 40); \
	45200	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vfmaq_laneq_f16(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __builtin_bit_cast(int8x16_t, __s2), __p3, 40)); \
44261	45201	__ret; \
44262	45202	})
44263	45203	#endif
...	...	@@ -44268,7 +45208,7 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44268	45208	float16x4_t __s0 = __p0; \
44269	45209	float16x4_t __s1 = __p1; \
44270	45210	float16x8_t __s2 = __p2; \
44271		__ret = (float16x4_t) __builtin_neon_vfma_laneq_f16((int8x8_t)__s0, (int8x8_t)__s1, (int8x16_t)__s2, __p3, 8); \
	45211	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vfma_laneq_f16(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __builtin_bit_cast(int8x16_t, __s2), __p3, 8)); \
44272	45212	__ret; \
44273	45213	})
44274	45214	#else
...	...	@@ -44277,11 +45217,11 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44277	45217	float16x4_t __s0 = __p0; \
44278	45218	float16x4_t __s1 = __p1; \
44279	45219	float16x8_t __s2 = __p2; \
44280		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
44281		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
44282		float16x8_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 7, 6, 5, 4, 3, 2, 1, 0); \
44283		__ret = (float16x4_t) __builtin_neon_vfma_laneq_f16((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x16_t)__rev2, __p3, 8); \
44284		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	45220	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	45221	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	45222	float16x8_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_128_16); \
	45223	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vfma_laneq_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), __p3, 8)); \
	45224	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
44285	45225	__ret; \
44286	45226	})
44287	45227	#define __noswap_vfma_laneq_f16(__p0, __p1, __p2, __p3) __extension__ ({ \
...	...	@@ -44289,7 +45229,7 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44289	45229	float16x4_t __s0 = __p0; \
44290	45230	float16x4_t __s1 = __p1; \
44291	45231	float16x8_t __s2 = __p2; \
44292		__ret = (float16x4_t) __builtin_neon_vfma_laneq_f16((int8x8_t)__s0, (int8x8_t)__s1, (int8x16_t)__s2, __p3, 8); \
	45232	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vfma_laneq_f16(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __builtin_bit_cast(int8x16_t, __s2), __p3, 8)); \
44293	45233	__ret; \
44294	45234	})
44295	45235	#endif
...	...	@@ -44309,10 +45249,10 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44309	45249	float16x8_t __s0 = __p0; \
44310	45250	float16x8_t __s1 = __p1; \
44311	45251	float16_t __s2 = __p2; \
44312		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
44313		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
	45252	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	45253	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
44314	45254	__ret = __noswap_vfmaq_f16(__rev0, __rev1, (float16x8_t) {__s2, __s2, __s2, __s2, __s2, __s2, __s2, __s2}); \
44315		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	45255	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
44316	45256	__ret; \
44317	45257	})
44318	45258	#endif
...	...	@@ -44332,152 +45272,152 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44332	45272	float16x4_t __s0 = __p0; \
44333	45273	float16x4_t __s1 = __p1; \
44334	45274	float16_t __s2 = __p2; \
44335		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
44336		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
	45275	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	45276	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
44337	45277	__ret = __noswap_vfma_f16(__rev0, __rev1, (float16x4_t) {__s2, __s2, __s2, __s2}); \
44338		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	45278	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
44339	45279	__ret; \
44340	45280	})
44341	45281	#endif
44342	45282
44343	45283	#ifdef __LITTLE_ENDIAN__
44344		#define vfmsh_lane_f16(__p0_246, __p1_246, __p2_246, __p3_246) __extension__ ({ \
44345		float16_t __ret_246; \
44346		float16_t __s0_246 = __p0_246; \
44347		float16_t __s1_246 = __p1_246; \
44348		float16x4_t __s2_246 = __p2_246; \
44349		__ret_246 = vfmah_lane_f16(__s0_246, -__s1_246, __s2_246, __p3_246); \
44350		__ret_246; \
44351		})
44352		#else
44353		#define vfmsh_lane_f16(__p0_247, __p1_247, __p2_247, __p3_247) __extension__ ({ \
44354		float16_t __ret_247; \
44355		float16_t __s0_247 = __p0_247; \
44356		float16_t __s1_247 = __p1_247; \
44357		float16x4_t __s2_247 = __p2_247; \
44358		float16x4_t __rev2_247; __rev2_247 = __builtin_shufflevector(__s2_247, __s2_247, 3, 2, 1, 0); \
44359		__ret_247 = __noswap_vfmah_lane_f16(__s0_247, -__s1_247, __rev2_247, __p3_247); \
44360		__ret_247; \
44361		})
44362		#endif
44363
44364		#ifdef __LITTLE_ENDIAN__
44365		#define vfmsq_lane_f16(__p0_248, __p1_248, __p2_248, __p3_248) __extension__ ({ \
44366		float16x8_t __ret_248; \
44367		float16x8_t __s0_248 = __p0_248; \
44368		float16x8_t __s1_248 = __p1_248; \
44369		float16x4_t __s2_248 = __p2_248; \
44370		__ret_248 = vfmaq_lane_f16(__s0_248, -__s1_248, __s2_248, __p3_248); \
44371		__ret_248; \
44372		})
44373		#else
44374		#define vfmsq_lane_f16(__p0_249, __p1_249, __p2_249, __p3_249) __extension__ ({ \
44375		float16x8_t __ret_249; \
44376		float16x8_t __s0_249 = __p0_249; \
44377		float16x8_t __s1_249 = __p1_249; \
44378		float16x4_t __s2_249 = __p2_249; \
44379		float16x8_t __rev0_249; __rev0_249 = __builtin_shufflevector(__s0_249, __s0_249, 7, 6, 5, 4, 3, 2, 1, 0); \
44380		float16x8_t __rev1_249; __rev1_249 = __builtin_shufflevector(__s1_249, __s1_249, 7, 6, 5, 4, 3, 2, 1, 0); \
44381		float16x4_t __rev2_249; __rev2_249 = __builtin_shufflevector(__s2_249, __s2_249, 3, 2, 1, 0); \
44382		__ret_249 = __noswap_vfmaq_lane_f16(__rev0_249, -__rev1_249, __rev2_249, __p3_249); \
44383		__ret_249 = __builtin_shufflevector(__ret_249, __ret_249, 7, 6, 5, 4, 3, 2, 1, 0); \
44384		__ret_249; \
44385		})
44386		#endif
44387
44388		#ifdef __LITTLE_ENDIAN__
44389		#define vfms_lane_f16(__p0_250, __p1_250, __p2_250, __p3_250) __extension__ ({ \
44390		float16x4_t __ret_250; \
44391		float16x4_t __s0_250 = __p0_250; \
44392		float16x4_t __s1_250 = __p1_250; \
	45284	#define vfmsh_lane_f16(__p0_250, __p1_250, __p2_250, __p3_250) __extension__ ({ \
	45285	float16_t __ret_250; \
	45286	float16_t __s0_250 = __p0_250; \
	45287	float16_t __s1_250 = __p1_250; \
44393	45288	float16x4_t __s2_250 = __p2_250; \
44394		__ret_250 = vfma_lane_f16(__s0_250, -__s1_250, __s2_250, __p3_250); \
	45289	__ret_250 = vfmah_lane_f16(__s0_250, -__s1_250, __s2_250, __p3_250); \
44395	45290	__ret_250; \
44396	45291	})
44397	45292	#else
44398		#define vfms_lane_f16(__p0_251, __p1_251, __p2_251, __p3_251) __extension__ ({ \
44399		float16x4_t __ret_251; \
44400		float16x4_t __s0_251 = __p0_251; \
44401		float16x4_t __s1_251 = __p1_251; \
	45293	#define vfmsh_lane_f16(__p0_251, __p1_251, __p2_251, __p3_251) __extension__ ({ \
	45294	float16_t __ret_251; \
	45295	float16_t __s0_251 = __p0_251; \
	45296	float16_t __s1_251 = __p1_251; \
44402	45297	float16x4_t __s2_251 = __p2_251; \
44403		float16x4_t __rev0_251; __rev0_251 = __builtin_shufflevector(__s0_251, __s0_251, 3, 2, 1, 0); \
44404		float16x4_t __rev1_251; __rev1_251 = __builtin_shufflevector(__s1_251, __s1_251, 3, 2, 1, 0); \
44405		float16x4_t __rev2_251; __rev2_251 = __builtin_shufflevector(__s2_251, __s2_251, 3, 2, 1, 0); \
44406		__ret_251 = __noswap_vfma_lane_f16(__rev0_251, -__rev1_251, __rev2_251, __p3_251); \
44407		__ret_251 = __builtin_shufflevector(__ret_251, __ret_251, 3, 2, 1, 0); \
	45298	float16x4_t __rev2_251; __rev2_251 = __builtin_shufflevector(__s2_251, __s2_251, __lane_reverse_64_16); \
	45299	__ret_251 = __noswap_vfmah_lane_f16(__s0_251, -__s1_251, __rev2_251, __p3_251); \
44408	45300	__ret_251; \
44409	45301	})
44410	45302	#endif
44411	45303
44412	45304	#ifdef __LITTLE_ENDIAN__
44413		#define vfmsh_laneq_f16(__p0_252, __p1_252, __p2_252, __p3_252) __extension__ ({ \
44414		float16_t __ret_252; \
44415		float16_t __s0_252 = __p0_252; \
44416		float16_t __s1_252 = __p1_252; \
44417		float16x8_t __s2_252 = __p2_252; \
44418		__ret_252 = vfmah_laneq_f16(__s0_252, -__s1_252, __s2_252, __p3_252); \
	45305	#define vfmsq_lane_f16(__p0_252, __p1_252, __p2_252, __p3_252) __extension__ ({ \
	45306	float16x8_t __ret_252; \
	45307	float16x8_t __s0_252 = __p0_252; \
	45308	float16x8_t __s1_252 = __p1_252; \
	45309	float16x4_t __s2_252 = __p2_252; \
	45310	__ret_252 = vfmaq_lane_f16(__s0_252, -__s1_252, __s2_252, __p3_252); \
44419	45311	__ret_252; \
44420	45312	})
44421	45313	#else
44422		#define vfmsh_laneq_f16(__p0_253, __p1_253, __p2_253, __p3_253) __extension__ ({ \
44423		float16_t __ret_253; \
44424		float16_t __s0_253 = __p0_253; \
44425		float16_t __s1_253 = __p1_253; \
44426		float16x8_t __s2_253 = __p2_253; \
44427		float16x8_t __rev2_253; __rev2_253 = __builtin_shufflevector(__s2_253, __s2_253, 7, 6, 5, 4, 3, 2, 1, 0); \
44428		__ret_253 = __noswap_vfmah_laneq_f16(__s0_253, -__s1_253, __rev2_253, __p3_253); \
	45314	#define vfmsq_lane_f16(__p0_253, __p1_253, __p2_253, __p3_253) __extension__ ({ \
	45315	float16x8_t __ret_253; \
	45316	float16x8_t __s0_253 = __p0_253; \
	45317	float16x8_t __s1_253 = __p1_253; \
	45318	float16x4_t __s2_253 = __p2_253; \
	45319	float16x8_t __rev0_253; __rev0_253 = __builtin_shufflevector(__s0_253, __s0_253, __lane_reverse_128_16); \
	45320	float16x8_t __rev1_253; __rev1_253 = __builtin_shufflevector(__s1_253, __s1_253, __lane_reverse_128_16); \
	45321	float16x4_t __rev2_253; __rev2_253 = __builtin_shufflevector(__s2_253, __s2_253, __lane_reverse_64_16); \
	45322	__ret_253 = __noswap_vfmaq_lane_f16(__rev0_253, -__rev1_253, __rev2_253, __p3_253); \
	45323	__ret_253 = __builtin_shufflevector(__ret_253, __ret_253, __lane_reverse_128_16); \
44429	45324	__ret_253; \
44430	45325	})
44431	45326	#endif
44432	45327
44433	45328	#ifdef __LITTLE_ENDIAN__
44434		#define vfmsq_laneq_f16(__p0_254, __p1_254, __p2_254, __p3_254) __extension__ ({ \
44435		float16x8_t __ret_254; \
44436		float16x8_t __s0_254 = __p0_254; \
44437		float16x8_t __s1_254 = __p1_254; \
44438		float16x8_t __s2_254 = __p2_254; \
44439		__ret_254 = vfmaq_laneq_f16(__s0_254, -__s1_254, __s2_254, __p3_254); \
	45329	#define vfms_lane_f16(__p0_254, __p1_254, __p2_254, __p3_254) __extension__ ({ \
	45330	float16x4_t __ret_254; \
	45331	float16x4_t __s0_254 = __p0_254; \
	45332	float16x4_t __s1_254 = __p1_254; \
	45333	float16x4_t __s2_254 = __p2_254; \
	45334	__ret_254 = vfma_lane_f16(__s0_254, -__s1_254, __s2_254, __p3_254); \
44440	45335	__ret_254; \
44441	45336	})
44442	45337	#else
44443		#define vfmsq_laneq_f16(__p0_255, __p1_255, __p2_255, __p3_255) __extension__ ({ \
44444		float16x8_t __ret_255; \
44445		float16x8_t __s0_255 = __p0_255; \
44446		float16x8_t __s1_255 = __p1_255; \
44447		float16x8_t __s2_255 = __p2_255; \
44448		float16x8_t __rev0_255; __rev0_255 = __builtin_shufflevector(__s0_255, __s0_255, 7, 6, 5, 4, 3, 2, 1, 0); \
44449		float16x8_t __rev1_255; __rev1_255 = __builtin_shufflevector(__s1_255, __s1_255, 7, 6, 5, 4, 3, 2, 1, 0); \
44450		float16x8_t __rev2_255; __rev2_255 = __builtin_shufflevector(__s2_255, __s2_255, 7, 6, 5, 4, 3, 2, 1, 0); \
44451		__ret_255 = __noswap_vfmaq_laneq_f16(__rev0_255, -__rev1_255, __rev2_255, __p3_255); \
44452		__ret_255 = __builtin_shufflevector(__ret_255, __ret_255, 7, 6, 5, 4, 3, 2, 1, 0); \
	45338	#define vfms_lane_f16(__p0_255, __p1_255, __p2_255, __p3_255) __extension__ ({ \
	45339	float16x4_t __ret_255; \
	45340	float16x4_t __s0_255 = __p0_255; \
	45341	float16x4_t __s1_255 = __p1_255; \
	45342	float16x4_t __s2_255 = __p2_255; \
	45343	float16x4_t __rev0_255; __rev0_255 = __builtin_shufflevector(__s0_255, __s0_255, __lane_reverse_64_16); \
	45344	float16x4_t __rev1_255; __rev1_255 = __builtin_shufflevector(__s1_255, __s1_255, __lane_reverse_64_16); \
	45345	float16x4_t __rev2_255; __rev2_255 = __builtin_shufflevector(__s2_255, __s2_255, __lane_reverse_64_16); \
	45346	__ret_255 = __noswap_vfma_lane_f16(__rev0_255, -__rev1_255, __rev2_255, __p3_255); \
	45347	__ret_255 = __builtin_shufflevector(__ret_255, __ret_255, __lane_reverse_64_16); \
44453	45348	__ret_255; \
44454	45349	})
44455	45350	#endif
44456	45351
44457	45352	#ifdef __LITTLE_ENDIAN__
44458		#define vfms_laneq_f16(__p0_256, __p1_256, __p2_256, __p3_256) __extension__ ({ \
44459		float16x4_t __ret_256; \
44460		float16x4_t __s0_256 = __p0_256; \
44461		float16x4_t __s1_256 = __p1_256; \
	45353	#define vfmsh_laneq_f16(__p0_256, __p1_256, __p2_256, __p3_256) __extension__ ({ \
	45354	float16_t __ret_256; \
	45355	float16_t __s0_256 = __p0_256; \
	45356	float16_t __s1_256 = __p1_256; \
44462	45357	float16x8_t __s2_256 = __p2_256; \
44463		__ret_256 = vfma_laneq_f16(__s0_256, -__s1_256, __s2_256, __p3_256); \
	45358	__ret_256 = vfmah_laneq_f16(__s0_256, -__s1_256, __s2_256, __p3_256); \
44464	45359	__ret_256; \
44465	45360	})
44466	45361	#else
44467		#define vfms_laneq_f16(__p0_257, __p1_257, __p2_257, __p3_257) __extension__ ({ \
44468		float16x4_t __ret_257; \
44469		float16x4_t __s0_257 = __p0_257; \
44470		float16x4_t __s1_257 = __p1_257; \
	45362	#define vfmsh_laneq_f16(__p0_257, __p1_257, __p2_257, __p3_257) __extension__ ({ \
	45363	float16_t __ret_257; \
	45364	float16_t __s0_257 = __p0_257; \
	45365	float16_t __s1_257 = __p1_257; \
44471	45366	float16x8_t __s2_257 = __p2_257; \
44472		float16x4_t __rev0_257; __rev0_257 = __builtin_shufflevector(__s0_257, __s0_257, 3, 2, 1, 0); \
44473		float16x4_t __rev1_257; __rev1_257 = __builtin_shufflevector(__s1_257, __s1_257, 3, 2, 1, 0); \
44474		float16x8_t __rev2_257; __rev2_257 = __builtin_shufflevector(__s2_257, __s2_257, 7, 6, 5, 4, 3, 2, 1, 0); \
44475		__ret_257 = __noswap_vfma_laneq_f16(__rev0_257, -__rev1_257, __rev2_257, __p3_257); \
44476		__ret_257 = __builtin_shufflevector(__ret_257, __ret_257, 3, 2, 1, 0); \
	45367	float16x8_t __rev2_257; __rev2_257 = __builtin_shufflevector(__s2_257, __s2_257, __lane_reverse_128_16); \
	45368	__ret_257 = __noswap_vfmah_laneq_f16(__s0_257, -__s1_257, __rev2_257, __p3_257); \
44477	45369	__ret_257; \
44478	45370	})
44479	45371	#endif
44480	45372
	45373	#ifdef __LITTLE_ENDIAN__
	45374	#define vfmsq_laneq_f16(__p0_258, __p1_258, __p2_258, __p3_258) __extension__ ({ \
	45375	float16x8_t __ret_258; \
	45376	float16x8_t __s0_258 = __p0_258; \
	45377	float16x8_t __s1_258 = __p1_258; \
	45378	float16x8_t __s2_258 = __p2_258; \
	45379	__ret_258 = vfmaq_laneq_f16(__s0_258, -__s1_258, __s2_258, __p3_258); \
	45380	__ret_258; \
	45381	})
	45382	#else
	45383	#define vfmsq_laneq_f16(__p0_259, __p1_259, __p2_259, __p3_259) __extension__ ({ \
	45384	float16x8_t __ret_259; \
	45385	float16x8_t __s0_259 = __p0_259; \
	45386	float16x8_t __s1_259 = __p1_259; \
	45387	float16x8_t __s2_259 = __p2_259; \
	45388	float16x8_t __rev0_259; __rev0_259 = __builtin_shufflevector(__s0_259, __s0_259, __lane_reverse_128_16); \
	45389	float16x8_t __rev1_259; __rev1_259 = __builtin_shufflevector(__s1_259, __s1_259, __lane_reverse_128_16); \
	45390	float16x8_t __rev2_259; __rev2_259 = __builtin_shufflevector(__s2_259, __s2_259, __lane_reverse_128_16); \
	45391	__ret_259 = __noswap_vfmaq_laneq_f16(__rev0_259, -__rev1_259, __rev2_259, __p3_259); \
	45392	__ret_259 = __builtin_shufflevector(__ret_259, __ret_259, __lane_reverse_128_16); \
	45393	__ret_259; \
	45394	})
	45395	#endif
	45396
	45397	#ifdef __LITTLE_ENDIAN__
	45398	#define vfms_laneq_f16(__p0_260, __p1_260, __p2_260, __p3_260) __extension__ ({ \
	45399	float16x4_t __ret_260; \
	45400	float16x4_t __s0_260 = __p0_260; \
	45401	float16x4_t __s1_260 = __p1_260; \
	45402	float16x8_t __s2_260 = __p2_260; \
	45403	__ret_260 = vfma_laneq_f16(__s0_260, -__s1_260, __s2_260, __p3_260); \
	45404	__ret_260; \
	45405	})
	45406	#else
	45407	#define vfms_laneq_f16(__p0_261, __p1_261, __p2_261, __p3_261) __extension__ ({ \
	45408	float16x4_t __ret_261; \
	45409	float16x4_t __s0_261 = __p0_261; \
	45410	float16x4_t __s1_261 = __p1_261; \
	45411	float16x8_t __s2_261 = __p2_261; \
	45412	float16x4_t __rev0_261; __rev0_261 = __builtin_shufflevector(__s0_261, __s0_261, __lane_reverse_64_16); \
	45413	float16x4_t __rev1_261; __rev1_261 = __builtin_shufflevector(__s1_261, __s1_261, __lane_reverse_64_16); \
	45414	float16x8_t __rev2_261; __rev2_261 = __builtin_shufflevector(__s2_261, __s2_261, __lane_reverse_128_16); \
	45415	__ret_261 = __noswap_vfma_laneq_f16(__rev0_261, -__rev1_261, __rev2_261, __p3_261); \
	45416	__ret_261 = __builtin_shufflevector(__ret_261, __ret_261, __lane_reverse_64_16); \
	45417	__ret_261; \
	45418	})
	45419	#endif
	45420
44481	45421	#ifdef __LITTLE_ENDIAN__
44482	45422	#define vfmsq_n_f16(__p0, __p1, __p2) __extension__ ({ \
44483	45423	float16x8_t __ret; \
...	...	@@ -44493,10 +45433,10 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44493	45433	float16x8_t __s0 = __p0; \
44494	45434	float16x8_t __s1 = __p1; \
44495	45435	float16_t __s2 = __p2; \
44496		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
44497		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
	45436	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	45437	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
44498	45438	__ret = __noswap_vfmaq_f16(__rev0, -__rev1, (float16x8_t) {__s2, __s2, __s2, __s2, __s2, __s2, __s2, __s2}); \
44499		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	45439	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
44500	45440	__ret; \
44501	45441	})
44502	45442	#endif
...	...	@@ -44516,10 +45456,10 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44516	45456	float16x4_t __s0 = __p0; \
44517	45457	float16x4_t __s1 = __p1; \
44518	45458	float16_t __s2 = __p2; \
44519		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
44520		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
	45459	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	45460	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
44521	45461	__ret = __noswap_vfma_f16(__rev0, -__rev1, (float16x4_t) {__s2, __s2, __s2, __s2}); \
44522		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	45462	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
44523	45463	__ret; \
44524	45464	})
44525	45465	#endif
...	...	@@ -44528,15 +45468,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44528	45468	#define vmaxnmvq_f16(__p0) __extension__ ({ \
44529	45469	float16_t __ret; \
44530	45470	float16x8_t __s0 = __p0; \
44531		__ret = (float16_t) __builtin_neon_vmaxnmvq_f16((int8x16_t)__s0); \
	45471	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vmaxnmvq_f16(__builtin_bit_cast(int8x16_t, __s0))); \
44532	45472	__ret; \
44533	45473	})
44534	45474	#else
44535	45475	#define vmaxnmvq_f16(__p0) __extension__ ({ \
44536	45476	float16_t __ret; \
44537	45477	float16x8_t __s0 = __p0; \
44538		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
44539		__ret = (float16_t) __builtin_neon_vmaxnmvq_f16((int8x16_t)__rev0); \
	45478	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	45479	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vmaxnmvq_f16(__builtin_bit_cast(int8x16_t, __rev0))); \
44540	45480	__ret; \
44541	45481	})
44542	45482	#endif
...	...	@@ -44545,15 +45485,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44545	45485	#define vmaxnmv_f16(__p0) __extension__ ({ \
44546	45486	float16_t __ret; \
44547	45487	float16x4_t __s0 = __p0; \
44548		__ret = (float16_t) __builtin_neon_vmaxnmv_f16((int8x8_t)__s0); \
	45488	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vmaxnmv_f16(__builtin_bit_cast(int8x8_t, __s0))); \
44549	45489	__ret; \
44550	45490	})
44551	45491	#else
44552	45492	#define vmaxnmv_f16(__p0) __extension__ ({ \
44553	45493	float16_t __ret; \
44554	45494	float16x4_t __s0 = __p0; \
44555		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
44556		__ret = (float16_t) __builtin_neon_vmaxnmv_f16((int8x8_t)__rev0); \
	45495	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	45496	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vmaxnmv_f16(__builtin_bit_cast(int8x8_t, __rev0))); \
44557	45497	__ret; \
44558	45498	})
44559	45499	#endif
...	...	@@ -44562,15 +45502,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44562	45502	#define vmaxvq_f16(__p0) __extension__ ({ \
44563	45503	float16_t __ret; \
44564	45504	float16x8_t __s0 = __p0; \
44565		__ret = (float16_t) __builtin_neon_vmaxvq_f16((int8x16_t)__s0); \
	45505	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vmaxvq_f16(__builtin_bit_cast(int8x16_t, __s0))); \
44566	45506	__ret; \
44567	45507	})
44568	45508	#else
44569	45509	#define vmaxvq_f16(__p0) __extension__ ({ \
44570	45510	float16_t __ret; \
44571	45511	float16x8_t __s0 = __p0; \
44572		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
44573		__ret = (float16_t) __builtin_neon_vmaxvq_f16((int8x16_t)__rev0); \
	45512	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	45513	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vmaxvq_f16(__builtin_bit_cast(int8x16_t, __rev0))); \
44574	45514	__ret; \
44575	45515	})
44576	45516	#endif
...	...	@@ -44579,15 +45519,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44579	45519	#define vmaxv_f16(__p0) __extension__ ({ \
44580	45520	float16_t __ret; \
44581	45521	float16x4_t __s0 = __p0; \
44582		__ret = (float16_t) __builtin_neon_vmaxv_f16((int8x8_t)__s0); \
	45522	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vmaxv_f16(__builtin_bit_cast(int8x8_t, __s0))); \
44583	45523	__ret; \
44584	45524	})
44585	45525	#else
44586	45526	#define vmaxv_f16(__p0) __extension__ ({ \
44587	45527	float16_t __ret; \
44588	45528	float16x4_t __s0 = __p0; \
44589		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
44590		__ret = (float16_t) __builtin_neon_vmaxv_f16((int8x8_t)__rev0); \
	45529	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	45530	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vmaxv_f16(__builtin_bit_cast(int8x8_t, __rev0))); \
44591	45531	__ret; \
44592	45532	})
44593	45533	#endif
...	...	@@ -44596,15 +45536,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44596	45536	#define vminnmvq_f16(__p0) __extension__ ({ \
44597	45537	float16_t __ret; \
44598	45538	float16x8_t __s0 = __p0; \
44599		__ret = (float16_t) __builtin_neon_vminnmvq_f16((int8x16_t)__s0); \
	45539	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vminnmvq_f16(__builtin_bit_cast(int8x16_t, __s0))); \
44600	45540	__ret; \
44601	45541	})
44602	45542	#else
44603	45543	#define vminnmvq_f16(__p0) __extension__ ({ \
44604	45544	float16_t __ret; \
44605	45545	float16x8_t __s0 = __p0; \
44606		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
44607		__ret = (float16_t) __builtin_neon_vminnmvq_f16((int8x16_t)__rev0); \
	45546	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	45547	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vminnmvq_f16(__builtin_bit_cast(int8x16_t, __rev0))); \
44608	45548	__ret; \
44609	45549	})
44610	45550	#endif
...	...	@@ -44613,15 +45553,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44613	45553	#define vminnmv_f16(__p0) __extension__ ({ \
44614	45554	float16_t __ret; \
44615	45555	float16x4_t __s0 = __p0; \
44616		__ret = (float16_t) __builtin_neon_vminnmv_f16((int8x8_t)__s0); \
	45556	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vminnmv_f16(__builtin_bit_cast(int8x8_t, __s0))); \
44617	45557	__ret; \
44618	45558	})
44619	45559	#else
44620	45560	#define vminnmv_f16(__p0) __extension__ ({ \
44621	45561	float16_t __ret; \
44622	45562	float16x4_t __s0 = __p0; \
44623		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
44624		__ret = (float16_t) __builtin_neon_vminnmv_f16((int8x8_t)__rev0); \
	45563	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	45564	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vminnmv_f16(__builtin_bit_cast(int8x8_t, __rev0))); \
44625	45565	__ret; \
44626	45566	})
44627	45567	#endif
...	...	@@ -44630,15 +45570,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44630	45570	#define vminvq_f16(__p0) __extension__ ({ \
44631	45571	float16_t __ret; \
44632	45572	float16x8_t __s0 = __p0; \
44633		__ret = (float16_t) __builtin_neon_vminvq_f16((int8x16_t)__s0); \
	45573	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vminvq_f16(__builtin_bit_cast(int8x16_t, __s0))); \
44634	45574	__ret; \
44635	45575	})
44636	45576	#else
44637	45577	#define vminvq_f16(__p0) __extension__ ({ \
44638	45578	float16_t __ret; \
44639	45579	float16x8_t __s0 = __p0; \
44640		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
44641		__ret = (float16_t) __builtin_neon_vminvq_f16((int8x16_t)__rev0); \
	45580	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	45581	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vminvq_f16(__builtin_bit_cast(int8x16_t, __rev0))); \
44642	45582	__ret; \
44643	45583	})
44644	45584	#endif
...	...	@@ -44647,79 +45587,79 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vdiv_f16(float16x4_t _
44647	45587	#define vminv_f16(__p0) __extension__ ({ \
44648	45588	float16_t __ret; \
44649	45589	float16x4_t __s0 = __p0; \
44650		__ret = (float16_t) __builtin_neon_vminv_f16((int8x8_t)__s0); \
	45590	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vminv_f16(__builtin_bit_cast(int8x8_t, __s0))); \
44651	45591	__ret; \
44652	45592	})
44653	45593	#else
44654	45594	#define vminv_f16(__p0) __extension__ ({ \
44655	45595	float16_t __ret; \
44656	45596	float16x4_t __s0 = __p0; \
44657		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
44658		__ret = (float16_t) __builtin_neon_vminv_f16((int8x8_t)__rev0); \
	45597	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	45598	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vminv_f16(__builtin_bit_cast(int8x8_t, __rev0))); \
44659	45599	__ret; \
44660	45600	})
44661	45601	#endif
44662	45602
44663	45603	#ifdef __LITTLE_ENDIAN__
44664		#define vmulq_laneq_f16(__p0_258, __p1_258, __p2_258) __extension__ ({ \
44665		float16x8_t __ret_258; \
44666		float16x8_t __s0_258 = __p0_258; \
44667		float16x8_t __s1_258 = __p1_258; \
44668		__ret_258 = __s0_258 * splatq_laneq_f16(__s1_258, __p2_258); \
44669		__ret_258; \
	45604	#define vmulq_laneq_f16(__p0_262, __p1_262, __p2_262) __extension__ ({ \
	45605	float16x8_t __ret_262; \
	45606	float16x8_t __s0_262 = __p0_262; \
	45607	float16x8_t __s1_262 = __p1_262; \
	45608	__ret_262 = __s0_262 * splatq_laneq_f16(__s1_262, __p2_262); \
	45609	__ret_262; \
44670	45610	})
44671	45611	#else
44672		#define vmulq_laneq_f16(__p0_259, __p1_259, __p2_259) __extension__ ({ \
44673		float16x8_t __ret_259; \
44674		float16x8_t __s0_259 = __p0_259; \
44675		float16x8_t __s1_259 = __p1_259; \
44676		float16x8_t __rev0_259; __rev0_259 = __builtin_shufflevector(__s0_259, __s0_259, 7, 6, 5, 4, 3, 2, 1, 0); \
44677		float16x8_t __rev1_259; __rev1_259 = __builtin_shufflevector(__s1_259, __s1_259, 7, 6, 5, 4, 3, 2, 1, 0); \
44678		__ret_259 = __rev0_259 * __noswap_splatq_laneq_f16(__rev1_259, __p2_259); \
44679		__ret_259 = __builtin_shufflevector(__ret_259, __ret_259, 7, 6, 5, 4, 3, 2, 1, 0); \
44680		__ret_259; \
	45612	#define vmulq_laneq_f16(__p0_263, __p1_263, __p2_263) __extension__ ({ \
	45613	float16x8_t __ret_263; \
	45614	float16x8_t __s0_263 = __p0_263; \
	45615	float16x8_t __s1_263 = __p1_263; \
	45616	float16x8_t __rev0_263; __rev0_263 = __builtin_shufflevector(__s0_263, __s0_263, __lane_reverse_128_16); \
	45617	float16x8_t __rev1_263; __rev1_263 = __builtin_shufflevector(__s1_263, __s1_263, __lane_reverse_128_16); \
	45618	__ret_263 = __rev0_263 * __noswap_splatq_laneq_f16(__rev1_263, __p2_263); \
	45619	__ret_263 = __builtin_shufflevector(__ret_263, __ret_263, __lane_reverse_128_16); \
	45620	__ret_263; \
44681	45621	})
44682	45622	#endif
44683	45623
44684	45624	#ifdef __LITTLE_ENDIAN__
44685		#define vmul_laneq_f16(__p0_260, __p1_260, __p2_260) __extension__ ({ \
44686		float16x4_t __ret_260; \
44687		float16x4_t __s0_260 = __p0_260; \
44688		float16x8_t __s1_260 = __p1_260; \
44689		__ret_260 = __s0_260 * splat_laneq_f16(__s1_260, __p2_260); \
44690		__ret_260; \
	45625	#define vmul_laneq_f16(__p0_264, __p1_264, __p2_264) __extension__ ({ \
	45626	float16x4_t __ret_264; \
	45627	float16x4_t __s0_264 = __p0_264; \
	45628	float16x8_t __s1_264 = __p1_264; \
	45629	__ret_264 = __s0_264 * splat_laneq_f16(__s1_264, __p2_264); \
	45630	__ret_264; \
44691	45631	})
44692	45632	#else
44693		#define vmul_laneq_f16(__p0_261, __p1_261, __p2_261) __extension__ ({ \
44694		float16x4_t __ret_261; \
44695		float16x4_t __s0_261 = __p0_261; \
44696		float16x8_t __s1_261 = __p1_261; \
44697		float16x4_t __rev0_261; __rev0_261 = __builtin_shufflevector(__s0_261, __s0_261, 3, 2, 1, 0); \
44698		float16x8_t __rev1_261; __rev1_261 = __builtin_shufflevector(__s1_261, __s1_261, 7, 6, 5, 4, 3, 2, 1, 0); \
44699		__ret_261 = __rev0_261 * __noswap_splat_laneq_f16(__rev1_261, __p2_261); \
44700		__ret_261 = __builtin_shufflevector(__ret_261, __ret_261, 3, 2, 1, 0); \
44701		__ret_261; \
	45633	#define vmul_laneq_f16(__p0_265, __p1_265, __p2_265) __extension__ ({ \
	45634	float16x4_t __ret_265; \
	45635	float16x4_t __s0_265 = __p0_265; \
	45636	float16x8_t __s1_265 = __p1_265; \
	45637	float16x4_t __rev0_265; __rev0_265 = __builtin_shufflevector(__s0_265, __s0_265, __lane_reverse_64_16); \
	45638	float16x8_t __rev1_265; __rev1_265 = __builtin_shufflevector(__s1_265, __s1_265, __lane_reverse_128_16); \
	45639	__ret_265 = __rev0_265 * __noswap_splat_laneq_f16(__rev1_265, __p2_265); \
	45640	__ret_265 = __builtin_shufflevector(__ret_265, __ret_265, __lane_reverse_64_16); \
	45641	__ret_265; \
44702	45642	})
44703	45643	#endif
44704	45644
44705	45645	#ifdef __LITTLE_ENDIAN__
44706	45646	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vmulxq_f16(float16x8_t __p0, float16x8_t __p1) {
44707	45647	float16x8_t __ret;
44708		__ret = (float16x8_t) __builtin_neon_vmulxq_f16((int8x16_t)__p0, (int8x16_t)__p1, 40);
	45648	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vmulxq_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 40));
44709	45649	return __ret;
44710	45650	}
44711	45651	#else
44712	45652	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vmulxq_f16(float16x8_t __p0, float16x8_t __p1) {
44713	45653	float16x8_t __ret;
44714		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
44715		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
44716		__ret = (float16x8_t) __builtin_neon_vmulxq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 40);
44717		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	45654	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	45655	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	45656	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vmulxq_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 40));
	45657	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
44718	45658	return __ret;
44719	45659	}
44720	45660	__ai __attribute__((target("fullfp16,neon"))) float16x8_t __noswap_vmulxq_f16(float16x8_t __p0, float16x8_t __p1) {
44721	45661	float16x8_t __ret;
44722		__ret = (float16x8_t) __builtin_neon_vmulxq_f16((int8x16_t)__p0, (int8x16_t)__p1, 40);
	45662	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vmulxq_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 40));
44723	45663	return __ret;
44724	45664	}
44725	45665	#endif
...	...	@@ -44727,21 +45667,21 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t __noswap_vmulxq_f16(fl
44727	45667	#ifdef __LITTLE_ENDIAN__
44728	45668	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vmulx_f16(float16x4_t __p0, float16x4_t __p1) {
44729	45669	float16x4_t __ret;
44730		__ret = (float16x4_t) __builtin_neon_vmulx_f16((int8x8_t)__p0, (int8x8_t)__p1, 8);
	45670	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vmulx_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 8));
44731	45671	return __ret;
44732	45672	}
44733	45673	#else
44734	45674	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vmulx_f16(float16x4_t __p0, float16x4_t __p1) {
44735	45675	float16x4_t __ret;
44736		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
44737		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
44738		__ret = (float16x4_t) __builtin_neon_vmulx_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 8);
44739		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	45676	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	45677	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	45678	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vmulx_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 8));
	45679	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
44740	45680	return __ret;
44741	45681	}
44742	45682	__ai __attribute__((target("fullfp16,neon"))) float16x4_t __noswap_vmulx_f16(float16x4_t __p0, float16x4_t __p1) {
44743	45683	float16x4_t __ret;
44744		__ret = (float16x4_t) __builtin_neon_vmulx_f16((int8x8_t)__p0, (int8x8_t)__p1, 8);
	45684	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vmulx_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 8));
44745	45685	return __ret;
44746	45686	}
44747	45687	#endif
...	...	@@ -44751,7 +45691,7 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t __noswap_vmulx_f16(flo
44751	45691	float16_t __ret; \
44752	45692	float16_t __s0 = __p0; \
44753	45693	float16x4_t __s1 = __p1; \
44754		__ret = (float16_t) __builtin_neon_vmulxh_lane_f16(__s0, (float16x4_t)__s1, __p2); \
	45694	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vmulxh_lane_f16(__s0, __s1, __p2)); \
44755	45695	__ret; \
44756	45696	})
44757	45697	#else
...	...	@@ -44759,51 +45699,51 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t __noswap_vmulx_f16(flo
44759	45699	float16_t __ret; \
44760	45700	float16_t __s0 = __p0; \
44761	45701	float16x4_t __s1 = __p1; \
44762		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
44763		__ret = (float16_t) __builtin_neon_vmulxh_lane_f16(__s0, (float16x4_t)__rev1, __p2); \
	45702	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	45703	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vmulxh_lane_f16(__s0, __rev1, __p2)); \
44764	45704	__ret; \
44765	45705	})
44766	45706	#endif
44767	45707
44768	45708	#ifdef __LITTLE_ENDIAN__
44769		#define vmulxq_lane_f16(__p0_262, __p1_262, __p2_262) __extension__ ({ \
44770		float16x8_t __ret_262; \
44771		float16x8_t __s0_262 = __p0_262; \
44772		float16x4_t __s1_262 = __p1_262; \
44773		__ret_262 = vmulxq_f16(__s0_262, splatq_lane_f16(__s1_262, __p2_262)); \
44774		__ret_262; \
	45709	#define vmulxq_lane_f16(__p0_266, __p1_266, __p2_266) __extension__ ({ \
	45710	float16x8_t __ret_266; \
	45711	float16x8_t __s0_266 = __p0_266; \
	45712	float16x4_t __s1_266 = __p1_266; \
	45713	__ret_266 = vmulxq_f16(__s0_266, splatq_lane_f16(__s1_266, __p2_266)); \
	45714	__ret_266; \
44775	45715	})
44776	45716	#else
44777		#define vmulxq_lane_f16(__p0_263, __p1_263, __p2_263) __extension__ ({ \
44778		float16x8_t __ret_263; \
44779		float16x8_t __s0_263 = __p0_263; \
44780		float16x4_t __s1_263 = __p1_263; \
44781		float16x8_t __rev0_263; __rev0_263 = __builtin_shufflevector(__s0_263, __s0_263, 7, 6, 5, 4, 3, 2, 1, 0); \
44782		float16x4_t __rev1_263; __rev1_263 = __builtin_shufflevector(__s1_263, __s1_263, 3, 2, 1, 0); \
44783		__ret_263 = __noswap_vmulxq_f16(__rev0_263, __noswap_splatq_lane_f16(__rev1_263, __p2_263)); \
44784		__ret_263 = __builtin_shufflevector(__ret_263, __ret_263, 7, 6, 5, 4, 3, 2, 1, 0); \
44785		__ret_263; \
	45717	#define vmulxq_lane_f16(__p0_267, __p1_267, __p2_267) __extension__ ({ \
	45718	float16x8_t __ret_267; \
	45719	float16x8_t __s0_267 = __p0_267; \
	45720	float16x4_t __s1_267 = __p1_267; \
	45721	float16x8_t __rev0_267; __rev0_267 = __builtin_shufflevector(__s0_267, __s0_267, __lane_reverse_128_16); \
	45722	float16x4_t __rev1_267; __rev1_267 = __builtin_shufflevector(__s1_267, __s1_267, __lane_reverse_64_16); \
	45723	__ret_267 = __noswap_vmulxq_f16(__rev0_267, __noswap_splatq_lane_f16(__rev1_267, __p2_267)); \
	45724	__ret_267 = __builtin_shufflevector(__ret_267, __ret_267, __lane_reverse_128_16); \
	45725	__ret_267; \
44786	45726	})
44787	45727	#endif
44788	45728
44789	45729	#ifdef __LITTLE_ENDIAN__
44790		#define vmulx_lane_f16(__p0_264, __p1_264, __p2_264) __extension__ ({ \
44791		float16x4_t __ret_264; \
44792		float16x4_t __s0_264 = __p0_264; \
44793		float16x4_t __s1_264 = __p1_264; \
44794		__ret_264 = vmulx_f16(__s0_264, splat_lane_f16(__s1_264, __p2_264)); \
44795		__ret_264; \
	45730	#define vmulx_lane_f16(__p0_268, __p1_268, __p2_268) __extension__ ({ \
	45731	float16x4_t __ret_268; \
	45732	float16x4_t __s0_268 = __p0_268; \
	45733	float16x4_t __s1_268 = __p1_268; \
	45734	__ret_268 = vmulx_f16(__s0_268, splat_lane_f16(__s1_268, __p2_268)); \
	45735	__ret_268; \
44796	45736	})
44797	45737	#else
44798		#define vmulx_lane_f16(__p0_265, __p1_265, __p2_265) __extension__ ({ \
44799		float16x4_t __ret_265; \
44800		float16x4_t __s0_265 = __p0_265; \
44801		float16x4_t __s1_265 = __p1_265; \
44802		float16x4_t __rev0_265; __rev0_265 = __builtin_shufflevector(__s0_265, __s0_265, 3, 2, 1, 0); \
44803		float16x4_t __rev1_265; __rev1_265 = __builtin_shufflevector(__s1_265, __s1_265, 3, 2, 1, 0); \
44804		__ret_265 = __noswap_vmulx_f16(__rev0_265, __noswap_splat_lane_f16(__rev1_265, __p2_265)); \
44805		__ret_265 = __builtin_shufflevector(__ret_265, __ret_265, 3, 2, 1, 0); \
44806		__ret_265; \
	45738	#define vmulx_lane_f16(__p0_269, __p1_269, __p2_269) __extension__ ({ \
	45739	float16x4_t __ret_269; \
	45740	float16x4_t __s0_269 = __p0_269; \
	45741	float16x4_t __s1_269 = __p1_269; \
	45742	float16x4_t __rev0_269; __rev0_269 = __builtin_shufflevector(__s0_269, __s0_269, __lane_reverse_64_16); \
	45743	float16x4_t __rev1_269; __rev1_269 = __builtin_shufflevector(__s1_269, __s1_269, __lane_reverse_64_16); \
	45744	__ret_269 = __noswap_vmulx_f16(__rev0_269, __noswap_splat_lane_f16(__rev1_269, __p2_269)); \
	45745	__ret_269 = __builtin_shufflevector(__ret_269, __ret_269, __lane_reverse_64_16); \
	45746	__ret_269; \
44807	45747	})
44808	45748	#endif
44809	45749
...	...	@@ -44812,7 +45752,7 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t __noswap_vmulx_f16(flo
44812	45752	float16_t __ret; \
44813	45753	float16_t __s0 = __p0; \
44814	45754	float16x8_t __s1 = __p1; \
44815		__ret = (float16_t) __builtin_neon_vmulxh_laneq_f16(__s0, (float16x8_t)__s1, __p2); \
	45755	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vmulxh_laneq_f16(__s0, __s1, __p2)); \
44816	45756	__ret; \
44817	45757	})
44818	45758	#else
...	...	@@ -44820,51 +45760,51 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t __noswap_vmulx_f16(flo
44820	45760	float16_t __ret; \
44821	45761	float16_t __s0 = __p0; \
44822	45762	float16x8_t __s1 = __p1; \
44823		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
44824		__ret = (float16_t) __builtin_neon_vmulxh_laneq_f16(__s0, (float16x8_t)__rev1, __p2); \
	45763	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	45764	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vmulxh_laneq_f16(__s0, __rev1, __p2)); \
44825	45765	__ret; \
44826	45766	})
44827	45767	#endif
44828	45768
44829	45769	#ifdef __LITTLE_ENDIAN__
44830		#define vmulxq_laneq_f16(__p0_266, __p1_266, __p2_266) __extension__ ({ \
44831		float16x8_t __ret_266; \
44832		float16x8_t __s0_266 = __p0_266; \
44833		float16x8_t __s1_266 = __p1_266; \
44834		__ret_266 = vmulxq_f16(__s0_266, splatq_laneq_f16(__s1_266, __p2_266)); \
44835		__ret_266; \
	45770	#define vmulxq_laneq_f16(__p0_270, __p1_270, __p2_270) __extension__ ({ \
	45771	float16x8_t __ret_270; \
	45772	float16x8_t __s0_270 = __p0_270; \
	45773	float16x8_t __s1_270 = __p1_270; \
	45774	__ret_270 = vmulxq_f16(__s0_270, splatq_laneq_f16(__s1_270, __p2_270)); \
	45775	__ret_270; \
44836	45776	})
44837	45777	#else
44838		#define vmulxq_laneq_f16(__p0_267, __p1_267, __p2_267) __extension__ ({ \
44839		float16x8_t __ret_267; \
44840		float16x8_t __s0_267 = __p0_267; \
44841		float16x8_t __s1_267 = __p1_267; \
44842		float16x8_t __rev0_267; __rev0_267 = __builtin_shufflevector(__s0_267, __s0_267, 7, 6, 5, 4, 3, 2, 1, 0); \
44843		float16x8_t __rev1_267; __rev1_267 = __builtin_shufflevector(__s1_267, __s1_267, 7, 6, 5, 4, 3, 2, 1, 0); \
44844		__ret_267 = __noswap_vmulxq_f16(__rev0_267, __noswap_splatq_laneq_f16(__rev1_267, __p2_267)); \
44845		__ret_267 = __builtin_shufflevector(__ret_267, __ret_267, 7, 6, 5, 4, 3, 2, 1, 0); \
44846		__ret_267; \
	45778	#define vmulxq_laneq_f16(__p0_271, __p1_271, __p2_271) __extension__ ({ \
	45779	float16x8_t __ret_271; \
	45780	float16x8_t __s0_271 = __p0_271; \
	45781	float16x8_t __s1_271 = __p1_271; \
	45782	float16x8_t __rev0_271; __rev0_271 = __builtin_shufflevector(__s0_271, __s0_271, __lane_reverse_128_16); \
	45783	float16x8_t __rev1_271; __rev1_271 = __builtin_shufflevector(__s1_271, __s1_271, __lane_reverse_128_16); \
	45784	__ret_271 = __noswap_vmulxq_f16(__rev0_271, __noswap_splatq_laneq_f16(__rev1_271, __p2_271)); \
	45785	__ret_271 = __builtin_shufflevector(__ret_271, __ret_271, __lane_reverse_128_16); \
	45786	__ret_271; \
44847	45787	})
44848	45788	#endif
44849	45789
44850	45790	#ifdef __LITTLE_ENDIAN__
44851		#define vmulx_laneq_f16(__p0_268, __p1_268, __p2_268) __extension__ ({ \
44852		float16x4_t __ret_268; \
44853		float16x4_t __s0_268 = __p0_268; \
44854		float16x8_t __s1_268 = __p1_268; \
44855		__ret_268 = vmulx_f16(__s0_268, splat_laneq_f16(__s1_268, __p2_268)); \
44856		__ret_268; \
	45791	#define vmulx_laneq_f16(__p0_272, __p1_272, __p2_272) __extension__ ({ \
	45792	float16x4_t __ret_272; \
	45793	float16x4_t __s0_272 = __p0_272; \
	45794	float16x8_t __s1_272 = __p1_272; \
	45795	__ret_272 = vmulx_f16(__s0_272, splat_laneq_f16(__s1_272, __p2_272)); \
	45796	__ret_272; \
44857	45797	})
44858	45798	#else
44859		#define vmulx_laneq_f16(__p0_269, __p1_269, __p2_269) __extension__ ({ \
44860		float16x4_t __ret_269; \
44861		float16x4_t __s0_269 = __p0_269; \
44862		float16x8_t __s1_269 = __p1_269; \
44863		float16x4_t __rev0_269; __rev0_269 = __builtin_shufflevector(__s0_269, __s0_269, 3, 2, 1, 0); \
44864		float16x8_t __rev1_269; __rev1_269 = __builtin_shufflevector(__s1_269, __s1_269, 7, 6, 5, 4, 3, 2, 1, 0); \
44865		__ret_269 = __noswap_vmulx_f16(__rev0_269, __noswap_splat_laneq_f16(__rev1_269, __p2_269)); \
44866		__ret_269 = __builtin_shufflevector(__ret_269, __ret_269, 3, 2, 1, 0); \
44867		__ret_269; \
	45799	#define vmulx_laneq_f16(__p0_273, __p1_273, __p2_273) __extension__ ({ \
	45800	float16x4_t __ret_273; \
	45801	float16x4_t __s0_273 = __p0_273; \
	45802	float16x8_t __s1_273 = __p1_273; \
	45803	float16x4_t __rev0_273; __rev0_273 = __builtin_shufflevector(__s0_273, __s0_273, __lane_reverse_64_16); \
	45804	float16x8_t __rev1_273; __rev1_273 = __builtin_shufflevector(__s1_273, __s1_273, __lane_reverse_128_16); \
	45805	__ret_273 = __noswap_vmulx_f16(__rev0_273, __noswap_splat_laneq_f16(__rev1_273, __p2_273)); \
	45806	__ret_273 = __builtin_shufflevector(__ret_273, __ret_273, __lane_reverse_64_16); \
	45807	__ret_273; \
44868	45808	})
44869	45809	#endif
44870	45810
...	...	@@ -44881,9 +45821,9 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t __noswap_vmulx_f16(flo
44881	45821	float16x8_t __ret; \
44882	45822	float16x8_t __s0 = __p0; \
44883	45823	float16_t __s1 = __p1; \
44884		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
	45824	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
44885	45825	__ret = __noswap_vmulxq_f16(__rev0, (float16x8_t) {__s1, __s1, __s1, __s1, __s1, __s1, __s1, __s1}); \
44886		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	45826	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
44887	45827	__ret; \
44888	45828	})
44889	45829	#endif
...	...	@@ -44901,9 +45841,9 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t __noswap_vmulx_f16(flo
44901	45841	float16x4_t __ret; \
44902	45842	float16x4_t __s0 = __p0; \
44903	45843	float16_t __s1 = __p1; \
44904		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
	45844	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
44905	45845	__ret = __noswap_vmulx_f16(__rev0, (float16x4_t) {__s1, __s1, __s1, __s1}); \
44906		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	45846	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
44907	45847	__ret; \
44908	45848	})
44909	45849	#endif
...	...	@@ -44911,16 +45851,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t __noswap_vmulx_f16(flo
44911	45851	#ifdef __LITTLE_ENDIAN__
44912	45852	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vpaddq_f16(float16x8_t __p0, float16x8_t __p1) {
44913	45853	float16x8_t __ret;
44914		__ret = (float16x8_t) __builtin_neon_vpaddq_f16((int8x16_t)__p0, (int8x16_t)__p1, 40);
	45854	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vpaddq_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 40));
44915	45855	return __ret;
44916	45856	}
44917	45857	#else
44918	45858	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vpaddq_f16(float16x8_t __p0, float16x8_t __p1) {
44919	45859	float16x8_t __ret;
44920		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
44921		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
44922		__ret = (float16x8_t) __builtin_neon_vpaddq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 40);
44923		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	45860	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	45861	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	45862	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vpaddq_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 40));
	45863	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
44924	45864	return __ret;
44925	45865	}
44926	45866	#endif
...	...	@@ -44928,16 +45868,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vpaddq_f16(float16x8_t
44928	45868	#ifdef __LITTLE_ENDIAN__
44929	45869	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vpmaxq_f16(float16x8_t __p0, float16x8_t __p1) {
44930	45870	float16x8_t __ret;
44931		__ret = (float16x8_t) __builtin_neon_vpmaxq_f16((int8x16_t)__p0, (int8x16_t)__p1, 40);
	45871	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vpmaxq_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 40));
44932	45872	return __ret;
44933	45873	}
44934	45874	#else
44935	45875	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vpmaxq_f16(float16x8_t __p0, float16x8_t __p1) {
44936	45876	float16x8_t __ret;
44937		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
44938		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
44939		__ret = (float16x8_t) __builtin_neon_vpmaxq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 40);
44940		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	45877	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	45878	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	45879	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vpmaxq_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 40));
	45880	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
44941	45881	return __ret;
44942	45882	}
44943	45883	#endif
...	...	@@ -44945,16 +45885,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vpmaxq_f16(float16x8_t
44945	45885	#ifdef __LITTLE_ENDIAN__
44946	45886	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vpmaxnmq_f16(float16x8_t __p0, float16x8_t __p1) {
44947	45887	float16x8_t __ret;
44948		__ret = (float16x8_t) __builtin_neon_vpmaxnmq_f16((int8x16_t)__p0, (int8x16_t)__p1, 40);
	45888	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vpmaxnmq_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 40));
44949	45889	return __ret;
44950	45890	}
44951	45891	#else
44952	45892	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vpmaxnmq_f16(float16x8_t __p0, float16x8_t __p1) {
44953	45893	float16x8_t __ret;
44954		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
44955		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
44956		__ret = (float16x8_t) __builtin_neon_vpmaxnmq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 40);
44957		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	45894	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	45895	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	45896	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vpmaxnmq_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 40));
	45897	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
44958	45898	return __ret;
44959	45899	}
44960	45900	#endif
...	...	@@ -44962,16 +45902,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vpmaxnmq_f16(float16x8
44962	45902	#ifdef __LITTLE_ENDIAN__
44963	45903	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vpmaxnm_f16(float16x4_t __p0, float16x4_t __p1) {
44964	45904	float16x4_t __ret;
44965		__ret = (float16x4_t) __builtin_neon_vpmaxnm_f16((int8x8_t)__p0, (int8x8_t)__p1, 8);
	45905	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vpmaxnm_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 8));
44966	45906	return __ret;
44967	45907	}
44968	45908	#else
44969	45909	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vpmaxnm_f16(float16x4_t __p0, float16x4_t __p1) {
44970	45910	float16x4_t __ret;
44971		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
44972		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
44973		__ret = (float16x4_t) __builtin_neon_vpmaxnm_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 8);
44974		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	45911	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	45912	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	45913	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vpmaxnm_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 8));
	45914	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
44975	45915	return __ret;
44976	45916	}
44977	45917	#endif
...	...	@@ -44979,16 +45919,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vpmaxnm_f16(float16x4_
44979	45919	#ifdef __LITTLE_ENDIAN__
44980	45920	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vpminq_f16(float16x8_t __p0, float16x8_t __p1) {
44981	45921	float16x8_t __ret;
44982		__ret = (float16x8_t) __builtin_neon_vpminq_f16((int8x16_t)__p0, (int8x16_t)__p1, 40);
	45922	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vpminq_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 40));
44983	45923	return __ret;
44984	45924	}
44985	45925	#else
44986	45926	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vpminq_f16(float16x8_t __p0, float16x8_t __p1) {
44987	45927	float16x8_t __ret;
44988		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
44989		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
44990		__ret = (float16x8_t) __builtin_neon_vpminq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 40);
44991		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	45928	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	45929	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	45930	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vpminq_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 40));
	45931	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
44992	45932	return __ret;
44993	45933	}
44994	45934	#endif
...	...	@@ -44996,16 +45936,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vpminq_f16(float16x8_t
44996	45936	#ifdef __LITTLE_ENDIAN__
44997	45937	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vpminnmq_f16(float16x8_t __p0, float16x8_t __p1) {
44998	45938	float16x8_t __ret;
44999		__ret = (float16x8_t) __builtin_neon_vpminnmq_f16((int8x16_t)__p0, (int8x16_t)__p1, 40);
	45939	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vpminnmq_f16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 40));
45000	45940	return __ret;
45001	45941	}
45002	45942	#else
45003	45943	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vpminnmq_f16(float16x8_t __p0, float16x8_t __p1) {
45004	45944	float16x8_t __ret;
45005		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
45006		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
45007		__ret = (float16x8_t) __builtin_neon_vpminnmq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 40);
45008		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	45945	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	45946	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	45947	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vpminnmq_f16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 40));
	45948	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
45009	45949	return __ret;
45010	45950	}
45011	45951	#endif
...	...	@@ -45013,16 +45953,16 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vpminnmq_f16(float16x8
45013	45953	#ifdef __LITTLE_ENDIAN__
45014	45954	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vpminnm_f16(float16x4_t __p0, float16x4_t __p1) {
45015	45955	float16x4_t __ret;
45016		__ret = (float16x4_t) __builtin_neon_vpminnm_f16((int8x8_t)__p0, (int8x8_t)__p1, 8);
	45956	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vpminnm_f16(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 8));
45017	45957	return __ret;
45018	45958	}
45019	45959	#else
45020	45960	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vpminnm_f16(float16x4_t __p0, float16x4_t __p1) {
45021	45961	float16x4_t __ret;
45022		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
45023		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
45024		__ret = (float16x4_t) __builtin_neon_vpminnm_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 8);
45025		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	45962	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	45963	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	45964	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vpminnm_f16(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 8));
	45965	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
45026	45966	return __ret;
45027	45967	}
45028	45968	#endif
...	...	@@ -45030,15 +45970,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vpminnm_f16(float16x4_
45030	45970	#ifdef __LITTLE_ENDIAN__
45031	45971	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vrndiq_f16(float16x8_t __p0) {
45032	45972	float16x8_t __ret;
45033		__ret = (float16x8_t) __builtin_neon_vrndiq_f16((int8x16_t)__p0, 40);
	45973	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vrndiq_f16(__builtin_bit_cast(int8x16_t, __p0), 40));
45034	45974	return __ret;
45035	45975	}
45036	45976	#else
45037	45977	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vrndiq_f16(float16x8_t __p0) {
45038	45978	float16x8_t __ret;
45039		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
45040		__ret = (float16x8_t) __builtin_neon_vrndiq_f16((int8x16_t)__rev0, 40);
45041		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	45979	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	45980	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vrndiq_f16(__builtin_bit_cast(int8x16_t, __rev0), 40));
	45981	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
45042	45982	return __ret;
45043	45983	}
45044	45984	#endif
...	...	@@ -45046,15 +45986,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vrndiq_f16(float16x8_t
45046	45986	#ifdef __LITTLE_ENDIAN__
45047	45987	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vrndi_f16(float16x4_t __p0) {
45048	45988	float16x4_t __ret;
45049		__ret = (float16x4_t) __builtin_neon_vrndi_f16((int8x8_t)__p0, 8);
	45989	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vrndi_f16(__builtin_bit_cast(int8x8_t, __p0), 8));
45050	45990	return __ret;
45051	45991	}
45052	45992	#else
45053	45993	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vrndi_f16(float16x4_t __p0) {
45054	45994	float16x4_t __ret;
45055		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
45056		__ret = (float16x4_t) __builtin_neon_vrndi_f16((int8x8_t)__rev0, 8);
45057		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	45995	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	45996	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vrndi_f16(__builtin_bit_cast(int8x8_t, __rev0), 8));
	45997	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
45058	45998	return __ret;
45059	45999	}
45060	46000	#endif
...	...	@@ -45062,15 +46002,15 @@ __ai __attribute__((target("fullfp16,neon"))) float16x4_t vrndi_f16(float16x4_t
45062	46002	#ifdef __LITTLE_ENDIAN__
45063	46003	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vsqrtq_f16(float16x8_t __p0) {
45064	46004	float16x8_t __ret;
45065		__ret = (float16x8_t) __builtin_neon_vsqrtq_f16((int8x16_t)__p0, 40);
	46005	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vsqrtq_f16(__builtin_bit_cast(int8x16_t, __p0), 40));
45066	46006	return __ret;
45067	46007	}
45068	46008	#else
45069	46009	__ai __attribute__((target("fullfp16,neon"))) float16x8_t vsqrtq_f16(float16x8_t __p0) {
45070	46010	float16x8_t __ret;
45071		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
45072		__ret = (float16x8_t) __builtin_neon_vsqrtq_f16((int8x16_t)__rev0, 40);
45073		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	46011	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	46012	__ret = __builtin_bit_cast(float16x8_t, __builtin_neon_vsqrtq_f16(__builtin_bit_cast(int8x16_t, __rev0), 40));
	46013	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
45074	46014	return __ret;
45075	46015	}
45076	46016	#endif
...	...	@@ -45078,159 +46018,159 @@ __ai __attribute__((target("fullfp16,neon"))) float16x8_t vsqrtq_f16(float16x8_t
45078	46018	#ifdef __LITTLE_ENDIAN__
45079	46019	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vsqrt_f16(float16x4_t __p0) {
45080	46020	float16x4_t __ret;
45081		__ret = (float16x4_t) __builtin_neon_vsqrt_f16((int8x8_t)__p0, 8);
	46021	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vsqrt_f16(__builtin_bit_cast(int8x8_t, __p0), 8));
45082	46022	return __ret;
45083	46023	}
45084	46024	#else
45085	46025	__ai __attribute__((target("fullfp16,neon"))) float16x4_t vsqrt_f16(float16x4_t __p0) {
45086	46026	float16x4_t __ret;
45087		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
45088		__ret = (float16x4_t) __builtin_neon_vsqrt_f16((int8x8_t)__rev0, 8);
45089		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	46027	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	46028	__ret = __builtin_bit_cast(float16x4_t, __builtin_neon_vsqrt_f16(__builtin_bit_cast(int8x8_t, __rev0), 8));
	46029	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
45090	46030	return __ret;
45091	46031	}
45092	46032	#endif
45093	46033
45094	46034	#ifdef __LITTLE_ENDIAN__
45095		#define vsudotq_laneq_s32(__p0_270, __p1_270, __p2_270, __p3_270) __extension__ ({ \
45096		int32x4_t __ret_270; \
45097		int32x4_t __s0_270 = __p0_270; \
45098		int8x16_t __s1_270 = __p1_270; \
45099		uint8x16_t __s2_270 = __p2_270; \
45100		__ret_270 = vusdotq_s32(__s0_270, (uint8x16_t)(splatq_laneq_s32(__builtin_bit_cast(int32x4_t, __s2_270), __p3_270)), __s1_270); \
45101		__ret_270; \
45102		})
45103		#else
45104		#define vsudotq_laneq_s32(__p0_271, __p1_271, __p2_271, __p3_271) __extension__ ({ \
45105		int32x4_t __ret_271; \
45106		int32x4_t __s0_271 = __p0_271; \
45107		int8x16_t __s1_271 = __p1_271; \
45108		uint8x16_t __s2_271 = __p2_271; \
45109		int32x4_t __rev0_271; __rev0_271 = __builtin_shufflevector(__s0_271, __s0_271, 3, 2, 1, 0); \
45110		int8x16_t __rev1_271; __rev1_271 = __builtin_shufflevector(__s1_271, __s1_271, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
45111		uint8x16_t __rev2_271; __rev2_271 = __builtin_shufflevector(__s2_271, __s2_271, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
45112		__ret_271 = __noswap_vusdotq_s32(__rev0_271, (uint8x16_t)(__noswap_splatq_laneq_s32(__builtin_bit_cast(int32x4_t, __rev2_271), __p3_271)), __rev1_271); \
45113		__ret_271 = __builtin_shufflevector(__ret_271, __ret_271, 3, 2, 1, 0); \
45114		__ret_271; \
45115		})
45116		#endif
45117
45118		#ifdef __LITTLE_ENDIAN__
45119		#define vsudot_laneq_s32(__p0_272, __p1_272, __p2_272, __p3_272) __extension__ ({ \
45120		int32x2_t __ret_272; \
45121		int32x2_t __s0_272 = __p0_272; \
45122		int8x8_t __s1_272 = __p1_272; \
45123		uint8x16_t __s2_272 = __p2_272; \
45124		__ret_272 = vusdot_s32(__s0_272, (uint8x8_t)(splat_laneq_s32(__builtin_bit_cast(int32x4_t, __s2_272), __p3_272)), __s1_272); \
45125		__ret_272; \
45126		})
45127		#else
45128		#define vsudot_laneq_s32(__p0_273, __p1_273, __p2_273, __p3_273) __extension__ ({ \
45129		int32x2_t __ret_273; \
45130		int32x2_t __s0_273 = __p0_273; \
45131		int8x8_t __s1_273 = __p1_273; \
45132		uint8x16_t __s2_273 = __p2_273; \
45133		int32x2_t __rev0_273; __rev0_273 = __builtin_shufflevector(__s0_273, __s0_273, 1, 0); \
45134		int8x8_t __rev1_273; __rev1_273 = __builtin_shufflevector(__s1_273, __s1_273, 7, 6, 5, 4, 3, 2, 1, 0); \
45135		uint8x16_t __rev2_273; __rev2_273 = __builtin_shufflevector(__s2_273, __s2_273, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
45136		__ret_273 = __noswap_vusdot_s32(__rev0_273, (uint8x8_t)(__noswap_splat_laneq_s32(__builtin_bit_cast(int32x4_t, __rev2_273), __p3_273)), __rev1_273); \
45137		__ret_273 = __builtin_shufflevector(__ret_273, __ret_273, 1, 0); \
45138		__ret_273; \
45139		})
45140		#endif
45141
45142		#ifdef __LITTLE_ENDIAN__
45143		#define vusdotq_laneq_s32(__p0_274, __p1_274, __p2_274, __p3_274) __extension__ ({ \
	46035	#define vsudotq_laneq_s32(__p0_274, __p1_274, __p2_274, __p3_274) __extension__ ({ \
45144	46036	int32x4_t __ret_274; \
45145	46037	int32x4_t __s0_274 = __p0_274; \
45146		uint8x16_t __s1_274 = __p1_274; \
45147		int8x16_t __s2_274 = __p2_274; \
45148		__ret_274 = vusdotq_s32(__s0_274, __s1_274, (int8x16_t)(splatq_laneq_s32(__builtin_bit_cast(int32x4_t, __s2_274), __p3_274))); \
	46038	int8x16_t __s1_274 = __p1_274; \
	46039	uint8x16_t __s2_274 = __p2_274; \
	46040	__ret_274 = vusdotq_s32(__s0_274, __builtin_bit_cast(uint8x16_t, splatq_laneq_s32(__builtin_bit_cast(int32x4_t, __s2_274), __p3_274)), __s1_274); \
45149	46041	__ret_274; \
45150	46042	})
45151	46043	#else
45152		#define vusdotq_laneq_s32(__p0_275, __p1_275, __p2_275, __p3_275) __extension__ ({ \
	46044	#define vsudotq_laneq_s32(__p0_275, __p1_275, __p2_275, __p3_275) __extension__ ({ \
45153	46045	int32x4_t __ret_275; \
45154	46046	int32x4_t __s0_275 = __p0_275; \
45155		uint8x16_t __s1_275 = __p1_275; \
45156		int8x16_t __s2_275 = __p2_275; \
45157		int32x4_t __rev0_275; __rev0_275 = __builtin_shufflevector(__s0_275, __s0_275, 3, 2, 1, 0); \
45158		uint8x16_t __rev1_275; __rev1_275 = __builtin_shufflevector(__s1_275, __s1_275, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
45159		int8x16_t __rev2_275; __rev2_275 = __builtin_shufflevector(__s2_275, __s2_275, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
45160		__ret_275 = __noswap_vusdotq_s32(__rev0_275, __rev1_275, (int8x16_t)(__noswap_splatq_laneq_s32(__builtin_bit_cast(int32x4_t, __rev2_275), __p3_275))); \
45161		__ret_275 = __builtin_shufflevector(__ret_275, __ret_275, 3, 2, 1, 0); \
	46047	int8x16_t __s1_275 = __p1_275; \
	46048	uint8x16_t __s2_275 = __p2_275; \
	46049	int32x4_t __rev0_275; __rev0_275 = __builtin_shufflevector(__s0_275, __s0_275, __lane_reverse_128_32); \
	46050	int8x16_t __rev1_275; __rev1_275 = __builtin_shufflevector(__s1_275, __s1_275, __lane_reverse_128_8); \
	46051	uint8x16_t __rev2_275; __rev2_275 = __builtin_shufflevector(__s2_275, __s2_275, __lane_reverse_128_8); \
	46052	__ret_275 = __noswap_vusdotq_s32(__rev0_275, __builtin_bit_cast(uint8x16_t, __noswap_splatq_laneq_s32(__builtin_bit_cast(int32x4_t, __rev2_275), __p3_275)), __rev1_275); \
	46053	__ret_275 = __builtin_shufflevector(__ret_275, __ret_275, __lane_reverse_128_32); \
45162	46054	__ret_275; \
45163	46055	})
45164	46056	#endif
45165	46057
45166	46058	#ifdef __LITTLE_ENDIAN__
45167		#define vusdot_laneq_s32(__p0_276, __p1_276, __p2_276, __p3_276) __extension__ ({ \
	46059	#define vsudot_laneq_s32(__p0_276, __p1_276, __p2_276, __p3_276) __extension__ ({ \
45168	46060	int32x2_t __ret_276; \
45169	46061	int32x2_t __s0_276 = __p0_276; \
45170		uint8x8_t __s1_276 = __p1_276; \
45171		int8x16_t __s2_276 = __p2_276; \
45172		__ret_276 = vusdot_s32(__s0_276, __s1_276, (int8x8_t)(splat_laneq_s32(__builtin_bit_cast(int32x4_t, __s2_276), __p3_276))); \
	46062	int8x8_t __s1_276 = __p1_276; \
	46063	uint8x16_t __s2_276 = __p2_276; \
	46064	__ret_276 = vusdot_s32(__s0_276, __builtin_bit_cast(uint8x8_t, splat_laneq_s32(__builtin_bit_cast(int32x4_t, __s2_276), __p3_276)), __s1_276); \
45173	46065	__ret_276; \
45174	46066	})
45175	46067	#else
45176		#define vusdot_laneq_s32(__p0_277, __p1_277, __p2_277, __p3_277) __extension__ ({ \
	46068	#define vsudot_laneq_s32(__p0_277, __p1_277, __p2_277, __p3_277) __extension__ ({ \
45177	46069	int32x2_t __ret_277; \
45178	46070	int32x2_t __s0_277 = __p0_277; \
45179		uint8x8_t __s1_277 = __p1_277; \
45180		int8x16_t __s2_277 = __p2_277; \
45181		int32x2_t __rev0_277; __rev0_277 = __builtin_shufflevector(__s0_277, __s0_277, 1, 0); \
45182		uint8x8_t __rev1_277; __rev1_277 = __builtin_shufflevector(__s1_277, __s1_277, 7, 6, 5, 4, 3, 2, 1, 0); \
45183		int8x16_t __rev2_277; __rev2_277 = __builtin_shufflevector(__s2_277, __s2_277, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
45184		__ret_277 = __noswap_vusdot_s32(__rev0_277, __rev1_277, (int8x8_t)(__noswap_splat_laneq_s32(__builtin_bit_cast(int32x4_t, __rev2_277), __p3_277))); \
45185		__ret_277 = __builtin_shufflevector(__ret_277, __ret_277, 1, 0); \
	46071	int8x8_t __s1_277 = __p1_277; \
	46072	uint8x16_t __s2_277 = __p2_277; \
	46073	int32x2_t __rev0_277; __rev0_277 = __builtin_shufflevector(__s0_277, __s0_277, __lane_reverse_64_32); \
	46074	int8x8_t __rev1_277; __rev1_277 = __builtin_shufflevector(__s1_277, __s1_277, __lane_reverse_64_8); \
	46075	uint8x16_t __rev2_277; __rev2_277 = __builtin_shufflevector(__s2_277, __s2_277, __lane_reverse_128_8); \
	46076	__ret_277 = __noswap_vusdot_s32(__rev0_277, __builtin_bit_cast(uint8x8_t, __noswap_splat_laneq_s32(__builtin_bit_cast(int32x4_t, __rev2_277), __p3_277)), __rev1_277); \
	46077	__ret_277 = __builtin_shufflevector(__ret_277, __ret_277, __lane_reverse_64_32); \
45186	46078	__ret_277; \
45187	46079	})
45188	46080	#endif
45189	46081
	46082	#ifdef __LITTLE_ENDIAN__
	46083	#define vusdotq_laneq_s32(__p0_278, __p1_278, __p2_278, __p3_278) __extension__ ({ \
	46084	int32x4_t __ret_278; \
	46085	int32x4_t __s0_278 = __p0_278; \
	46086	uint8x16_t __s1_278 = __p1_278; \
	46087	int8x16_t __s2_278 = __p2_278; \
	46088	__ret_278 = vusdotq_s32(__s0_278, __s1_278, __builtin_bit_cast(int8x16_t, splatq_laneq_s32(__builtin_bit_cast(int32x4_t, __s2_278), __p3_278))); \
	46089	__ret_278; \
	46090	})
	46091	#else
	46092	#define vusdotq_laneq_s32(__p0_279, __p1_279, __p2_279, __p3_279) __extension__ ({ \
	46093	int32x4_t __ret_279; \
	46094	int32x4_t __s0_279 = __p0_279; \
	46095	uint8x16_t __s1_279 = __p1_279; \
	46096	int8x16_t __s2_279 = __p2_279; \
	46097	int32x4_t __rev0_279; __rev0_279 = __builtin_shufflevector(__s0_279, __s0_279, __lane_reverse_128_32); \
	46098	uint8x16_t __rev1_279; __rev1_279 = __builtin_shufflevector(__s1_279, __s1_279, __lane_reverse_128_8); \
	46099	int8x16_t __rev2_279; __rev2_279 = __builtin_shufflevector(__s2_279, __s2_279, __lane_reverse_128_8); \
	46100	__ret_279 = __noswap_vusdotq_s32(__rev0_279, __rev1_279, __builtin_bit_cast(int8x16_t, __noswap_splatq_laneq_s32(__builtin_bit_cast(int32x4_t, __rev2_279), __p3_279))); \
	46101	__ret_279 = __builtin_shufflevector(__ret_279, __ret_279, __lane_reverse_128_32); \
	46102	__ret_279; \
	46103	})
	46104	#endif
	46105
	46106	#ifdef __LITTLE_ENDIAN__
	46107	#define vusdot_laneq_s32(__p0_280, __p1_280, __p2_280, __p3_280) __extension__ ({ \
	46108	int32x2_t __ret_280; \
	46109	int32x2_t __s0_280 = __p0_280; \
	46110	uint8x8_t __s1_280 = __p1_280; \
	46111	int8x16_t __s2_280 = __p2_280; \
	46112	__ret_280 = vusdot_s32(__s0_280, __s1_280, __builtin_bit_cast(int8x8_t, splat_laneq_s32(__builtin_bit_cast(int32x4_t, __s2_280), __p3_280))); \
	46113	__ret_280; \
	46114	})
	46115	#else
	46116	#define vusdot_laneq_s32(__p0_281, __p1_281, __p2_281, __p3_281) __extension__ ({ \
	46117	int32x2_t __ret_281; \
	46118	int32x2_t __s0_281 = __p0_281; \
	46119	uint8x8_t __s1_281 = __p1_281; \
	46120	int8x16_t __s2_281 = __p2_281; \
	46121	int32x2_t __rev0_281; __rev0_281 = __builtin_shufflevector(__s0_281, __s0_281, __lane_reverse_64_32); \
	46122	uint8x8_t __rev1_281; __rev1_281 = __builtin_shufflevector(__s1_281, __s1_281, __lane_reverse_64_8); \
	46123	int8x16_t __rev2_281; __rev2_281 = __builtin_shufflevector(__s2_281, __s2_281, __lane_reverse_128_8); \
	46124	__ret_281 = __noswap_vusdot_s32(__rev0_281, __rev1_281, __builtin_bit_cast(int8x8_t, __noswap_splat_laneq_s32(__builtin_bit_cast(int32x4_t, __rev2_281), __p3_281))); \
	46125	__ret_281 = __builtin_shufflevector(__ret_281, __ret_281, __lane_reverse_64_32); \
	46126	__ret_281; \
	46127	})
	46128	#endif
	46129
45190	46130	#ifdef __LITTLE_ENDIAN__
45191	46131	__ai __attribute__((target("neon"))) float64x2_t vabdq_f64(float64x2_t __p0, float64x2_t __p1) {
45192	46132	float64x2_t __ret;
45193		__ret = (float64x2_t) __builtin_neon_vabdq_v((int8x16_t)__p0, (int8x16_t)__p1, 42);
	46133	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vabdq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 42));
45194	46134	return __ret;
45195	46135	}
45196	46136	#else
45197	46137	__ai __attribute__((target("neon"))) float64x2_t vabdq_f64(float64x2_t __p0, float64x2_t __p1) {
45198	46138	float64x2_t __ret;
45199		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
45200		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
45201		__ret = (float64x2_t) __builtin_neon_vabdq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 42);
45202		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	46139	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	46140	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	46141	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vabdq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 42));
	46142	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
45203	46143	return __ret;
45204	46144	}
45205	46145	#endif
45206	46146
45207	46147	__ai __attribute__((target("neon"))) float64x1_t vabd_f64(float64x1_t __p0, float64x1_t __p1) {
45208	46148	float64x1_t __ret;
45209		__ret = (float64x1_t) __builtin_neon_vabd_v((int8x8_t)__p0, (int8x8_t)__p1, 10);
	46149	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vabd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 10));
45210	46150	return __ret;
45211	46151	}
45212	46152	__ai __attribute__((target("neon"))) float64_t vabdd_f64(float64_t __p0, float64_t __p1) {
45213	46153	float64_t __ret;
45214		__ret = (float64_t) __builtin_neon_vabdd_f64(__p0, __p1);
	46154	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vabdd_f64(__p0, __p1));
45215	46155	return __ret;
45216	46156	}
45217	46157	__ai __attribute__((target("neon"))) float32_t vabds_f32(float32_t __p0, float32_t __p1) {
45218	46158	float32_t __ret;
45219		__ret = (float32_t) __builtin_neon_vabds_f32(__p0, __p1);
	46159	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vabds_f32(__p0, __p1));
45220	46160	return __ret;
45221	46161	}
45222	46162	#ifdef __LITTLE_ENDIAN__
45223	46163	__ai __attribute__((target("neon"))) float64x2_t vabsq_f64(float64x2_t __p0) {
45224	46164	float64x2_t __ret;
45225		__ret = (float64x2_t) __builtin_neon_vabsq_v((int8x16_t)__p0, 42);
	46165	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vabsq_v(__builtin_bit_cast(int8x16_t, __p0), 42));
45226	46166	return __ret;
45227	46167	}
45228	46168	#else
45229	46169	__ai __attribute__((target("neon"))) float64x2_t vabsq_f64(float64x2_t __p0) {
45230	46170	float64x2_t __ret;
45231		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
45232		__ret = (float64x2_t) __builtin_neon_vabsq_v((int8x16_t)__rev0, 42);
45233		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	46171	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	46172	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vabsq_v(__builtin_bit_cast(int8x16_t, __rev0), 42));
	46173	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
45234	46174	return __ret;
45235	46175	}
45236	46176	#endif
...	...	@@ -45238,32 +46178,32 @@ __ai __attribute__((target("neon"))) float64x2_t vabsq_f64(float64x2_t __p0) {
45238	46178	#ifdef __LITTLE_ENDIAN__
45239	46179	__ai __attribute__((target("neon"))) int64x2_t vabsq_s64(int64x2_t __p0) {
45240	46180	int64x2_t __ret;
45241		__ret = (int64x2_t) __builtin_neon_vabsq_v((int8x16_t)__p0, 35);
	46181	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vabsq_v(__builtin_bit_cast(int8x16_t, __p0), 35));
45242	46182	return __ret;
45243	46183	}
45244	46184	#else
45245	46185	__ai __attribute__((target("neon"))) int64x2_t vabsq_s64(int64x2_t __p0) {
45246	46186	int64x2_t __ret;
45247		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
45248		__ret = (int64x2_t) __builtin_neon_vabsq_v((int8x16_t)__rev0, 35);
45249		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	46187	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	46188	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vabsq_v(__builtin_bit_cast(int8x16_t, __rev0), 35));
	46189	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
45250	46190	return __ret;
45251	46191	}
45252	46192	#endif
45253	46193
45254	46194	__ai __attribute__((target("neon"))) float64x1_t vabs_f64(float64x1_t __p0) {
45255	46195	float64x1_t __ret;
45256		__ret = (float64x1_t) __builtin_neon_vabs_v((int8x8_t)__p0, 10);
	46196	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vabs_v(__builtin_bit_cast(int8x8_t, __p0), 10));
45257	46197	return __ret;
45258	46198	}
45259	46199	__ai __attribute__((target("neon"))) int64x1_t vabs_s64(int64x1_t __p0) {
45260	46200	int64x1_t __ret;
45261		__ret = (int64x1_t) __builtin_neon_vabs_v((int8x8_t)__p0, 3);
	46201	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vabs_v(__builtin_bit_cast(int8x8_t, __p0), 3));
45262	46202	return __ret;
45263	46203	}
45264	46204	__ai __attribute__((target("neon"))) int64_t vabsd_s64(int64_t __p0) {
45265	46205	int64_t __ret;
45266		__ret = (int64_t) __builtin_neon_vabsd_s64(__p0);
	46206	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vabsd_s64(__p0));
45267	46207	return __ret;
45268	46208	}
45269	46209	#ifdef __LITTLE_ENDIAN__
...	...	@@ -45275,10 +46215,10 @@ __ai __attribute__((target("neon"))) float64x2_t vaddq_f64(float64x2_t __p0, flo
45275	46215	#else
45276	46216	__ai __attribute__((target("neon"))) float64x2_t vaddq_f64(float64x2_t __p0, float64x2_t __p1) {
45277	46217	float64x2_t __ret;
45278		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
45279		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	46218	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	46219	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
45280	46220	__ret = __rev0 + __rev1;
45281		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	46221	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
45282	46222	return __ret;
45283	46223	}
45284	46224	#endif
...	...	@@ -45290,17 +46230,17 @@ __ai __attribute__((target("neon"))) float64x1_t vadd_f64(float64x1_t __p0, floa
45290	46230	}
45291	46231	__ai __attribute__((target("neon"))) uint64_t vaddd_u64(uint64_t __p0, uint64_t __p1) {
45292	46232	uint64_t __ret;
45293		__ret = (uint64_t) __builtin_neon_vaddd_u64(__p0, __p1);
	46233	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vaddd_u64(__p0, __p1));
45294	46234	return __ret;
45295	46235	}
45296	46236	__ai __attribute__((target("neon"))) int64_t vaddd_s64(int64_t __p0, int64_t __p1) {
45297	46237	int64_t __ret;
45298		__ret = (int64_t) __builtin_neon_vaddd_s64(__p0, __p1);
	46238	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vaddd_s64(__p0, __p1));
45299	46239	return __ret;
45300	46240	}
45301	46241	__ai __attribute__((target("neon"))) poly128_t vaddq_p128(poly128_t __p0, poly128_t __p1) {
45302	46242	poly128_t __ret;
45303		__ret = (poly128_t) __builtin_neon_vaddq_p128(__p0, __p1);
	46243	__ret = __builtin_bit_cast(poly128_t, __builtin_neon_vaddq_p128(__p0, __p1));
45304	46244	return __ret;
45305	46245	}
45306	46246	#ifdef __LITTLE_ENDIAN__
...	...	@@ -45312,11 +46252,11 @@ __ai __attribute__((target("neon"))) uint16x8_t vaddhn_high_u32(uint16x4_t __p0,
45312	46252	#else
45313	46253	__ai __attribute__((target("neon"))) uint16x8_t vaddhn_high_u32(uint16x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
45314	46254	uint16x8_t __ret;
45315		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
45316		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
45317		uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	46255	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	46256	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	46257	uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
45318	46258	__ret = __noswap_vcombine_u16(__rev0, __noswap_vaddhn_u32(__rev1, __rev2));
45319		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	46259	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
45320	46260	return __ret;
45321	46261	}
45322	46262	#endif
...	...	@@ -45330,11 +46270,11 @@ __ai __attribute__((target("neon"))) uint32x4_t vaddhn_high_u64(uint32x2_t __p0,
45330	46270	#else
45331	46271	__ai __attribute__((target("neon"))) uint32x4_t vaddhn_high_u64(uint32x2_t __p0, uint64x2_t __p1, uint64x2_t __p2) {
45332	46272	uint32x4_t __ret;
45333		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
45334		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
45335		uint64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
	46273	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	46274	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	46275	uint64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_64);
45336	46276	__ret = __noswap_vcombine_u32(__rev0, __noswap_vaddhn_u64(__rev1, __rev2));
45337		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	46277	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
45338	46278	return __ret;
45339	46279	}
45340	46280	#endif
...	...	@@ -45348,11 +46288,11 @@ __ai __attribute__((target("neon"))) uint8x16_t vaddhn_high_u16(uint8x8_t __p0,
45348	46288	#else
45349	46289	__ai __attribute__((target("neon"))) uint8x16_t vaddhn_high_u16(uint8x8_t __p0, uint16x8_t __p1, uint16x8_t __p2) {
45350	46290	uint8x16_t __ret;
45351		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
45352		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
45353		uint16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	46291	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	46292	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	46293	uint16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
45354	46294	__ret = __noswap_vcombine_u8(__rev0, __noswap_vaddhn_u16(__rev1, __rev2));
45355		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	46295	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
45356	46296	return __ret;
45357	46297	}
45358	46298	#endif
...	...	@@ -45366,11 +46306,11 @@ __ai __attribute__((target("neon"))) int16x8_t vaddhn_high_s32(int16x4_t __p0, i
45366	46306	#else
45367	46307	__ai __attribute__((target("neon"))) int16x8_t vaddhn_high_s32(int16x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
45368	46308	int16x8_t __ret;
45369		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
45370		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
45371		int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	46309	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	46310	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	46311	int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
45372	46312	__ret = __noswap_vcombine_s16(__rev0, __noswap_vaddhn_s32(__rev1, __rev2));
45373		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	46313	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
45374	46314	return __ret;
45375	46315	}
45376	46316	#endif
...	...	@@ -45384,11 +46324,11 @@ __ai __attribute__((target("neon"))) int32x4_t vaddhn_high_s64(int32x2_t __p0, i
45384	46324	#else
45385	46325	__ai __attribute__((target("neon"))) int32x4_t vaddhn_high_s64(int32x2_t __p0, int64x2_t __p1, int64x2_t __p2) {
45386	46326	int32x4_t __ret;
45387		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
45388		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
45389		int64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
	46327	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	46328	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	46329	int64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_64);
45390	46330	__ret = __noswap_vcombine_s32(__rev0, __noswap_vaddhn_s64(__rev1, __rev2));
45391		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	46331	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
45392	46332	return __ret;
45393	46333	}
45394	46334	#endif
...	...	@@ -45402,11 +46342,11 @@ __ai __attribute__((target("neon"))) int8x16_t vaddhn_high_s16(int8x8_t __p0, in
45402	46342	#else
45403	46343	__ai __attribute__((target("neon"))) int8x16_t vaddhn_high_s16(int8x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
45404	46344	int8x16_t __ret;
45405		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
45406		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
45407		int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	46345	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	46346	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	46347	int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
45408	46348	__ret = __noswap_vcombine_s8(__rev0, __noswap_vaddhn_s16(__rev1, __rev2));
45409		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	46349	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
45410	46350	return __ret;
45411	46351	}
45412	46352	#endif
...	...	@@ -45414,14 +46354,14 @@ __ai __attribute__((target("neon"))) int8x16_t vaddhn_high_s16(int8x8_t __p0, in
45414	46354	#ifdef __LITTLE_ENDIAN__
45415	46355	__ai __attribute__((target("neon"))) uint16_t vaddlvq_u8(uint8x16_t __p0) {
45416	46356	uint16_t __ret;
45417		__ret = (uint16_t) __builtin_neon_vaddlvq_u8(__p0);
	46357	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vaddlvq_u8(__p0));
45418	46358	return __ret;
45419	46359	}
45420	46360	#else
45421	46361	__ai __attribute__((target("neon"))) uint16_t vaddlvq_u8(uint8x16_t __p0) {
45422	46362	uint16_t __ret;
45423		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
45424		__ret = (uint16_t) __builtin_neon_vaddlvq_u8(__rev0);
	46363	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	46364	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vaddlvq_u8(__rev0));
45425	46365	return __ret;
45426	46366	}
45427	46367	#endif
...	...	@@ -45429,14 +46369,14 @@ __ai __attribute__((target("neon"))) uint16_t vaddlvq_u8(uint8x16_t __p0) {
45429	46369	#ifdef __LITTLE_ENDIAN__
45430	46370	__ai __attribute__((target("neon"))) uint64_t vaddlvq_u32(uint32x4_t __p0) {
45431	46371	uint64_t __ret;
45432		__ret = (uint64_t) __builtin_neon_vaddlvq_u32(__p0);
	46372	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vaddlvq_u32(__p0));
45433	46373	return __ret;
45434	46374	}
45435	46375	#else
45436	46376	__ai __attribute__((target("neon"))) uint64_t vaddlvq_u32(uint32x4_t __p0) {
45437	46377	uint64_t __ret;
45438		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
45439		__ret = (uint64_t) __builtin_neon_vaddlvq_u32(__rev0);
	46378	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	46379	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vaddlvq_u32(__rev0));
45440	46380	return __ret;
45441	46381	}
45442	46382	#endif
...	...	@@ -45444,14 +46384,14 @@ __ai __attribute__((target("neon"))) uint64_t vaddlvq_u32(uint32x4_t __p0) {
45444	46384	#ifdef __LITTLE_ENDIAN__
45445	46385	__ai __attribute__((target("neon"))) uint32_t vaddlvq_u16(uint16x8_t __p0) {
45446	46386	uint32_t __ret;
45447		__ret = (uint32_t) __builtin_neon_vaddlvq_u16(__p0);
	46387	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vaddlvq_u16(__p0));
45448	46388	return __ret;
45449	46389	}
45450	46390	#else
45451	46391	__ai __attribute__((target("neon"))) uint32_t vaddlvq_u16(uint16x8_t __p0) {
45452	46392	uint32_t __ret;
45453		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
45454		__ret = (uint32_t) __builtin_neon_vaddlvq_u16(__rev0);
	46393	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	46394	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vaddlvq_u16(__rev0));
45455	46395	return __ret;
45456	46396	}
45457	46397	#endif
...	...	@@ -45459,14 +46399,14 @@ __ai __attribute__((target("neon"))) uint32_t vaddlvq_u16(uint16x8_t __p0) {
45459	46399	#ifdef __LITTLE_ENDIAN__
45460	46400	__ai __attribute__((target("neon"))) int16_t vaddlvq_s8(int8x16_t __p0) {
45461	46401	int16_t __ret;
45462		__ret = (int16_t) __builtin_neon_vaddlvq_s8(__p0);
	46402	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vaddlvq_s8(__p0));
45463	46403	return __ret;
45464	46404	}
45465	46405	#else
45466	46406	__ai __attribute__((target("neon"))) int16_t vaddlvq_s8(int8x16_t __p0) {
45467	46407	int16_t __ret;
45468		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
45469		__ret = (int16_t) __builtin_neon_vaddlvq_s8(__rev0);
	46408	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	46409	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vaddlvq_s8(__rev0));
45470	46410	return __ret;
45471	46411	}
45472	46412	#endif
...	...	@@ -45474,14 +46414,14 @@ __ai __attribute__((target("neon"))) int16_t vaddlvq_s8(int8x16_t __p0) {
45474	46414	#ifdef __LITTLE_ENDIAN__
45475	46415	__ai __attribute__((target("neon"))) int64_t vaddlvq_s32(int32x4_t __p0) {
45476	46416	int64_t __ret;
45477		__ret = (int64_t) __builtin_neon_vaddlvq_s32(__p0);
	46417	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vaddlvq_s32(__p0));
45478	46418	return __ret;
45479	46419	}
45480	46420	#else
45481	46421	__ai __attribute__((target("neon"))) int64_t vaddlvq_s32(int32x4_t __p0) {
45482	46422	int64_t __ret;
45483		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
45484		__ret = (int64_t) __builtin_neon_vaddlvq_s32(__rev0);
	46423	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	46424	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vaddlvq_s32(__rev0));
45485	46425	return __ret;
45486	46426	}
45487	46427	#endif
...	...	@@ -45489,14 +46429,14 @@ __ai __attribute__((target("neon"))) int64_t vaddlvq_s32(int32x4_t __p0) {
45489	46429	#ifdef __LITTLE_ENDIAN__
45490	46430	__ai __attribute__((target("neon"))) int32_t vaddlvq_s16(int16x8_t __p0) {
45491	46431	int32_t __ret;
45492		__ret = (int32_t) __builtin_neon_vaddlvq_s16(__p0);
	46432	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vaddlvq_s16(__p0));
45493	46433	return __ret;
45494	46434	}
45495	46435	#else
45496	46436	__ai __attribute__((target("neon"))) int32_t vaddlvq_s16(int16x8_t __p0) {
45497	46437	int32_t __ret;
45498		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
45499		__ret = (int32_t) __builtin_neon_vaddlvq_s16(__rev0);
	46438	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	46439	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vaddlvq_s16(__rev0));
45500	46440	return __ret;
45501	46441	}
45502	46442	#endif
...	...	@@ -45504,14 +46444,14 @@ __ai __attribute__((target("neon"))) int32_t vaddlvq_s16(int16x8_t __p0) {
45504	46444	#ifdef __LITTLE_ENDIAN__
45505	46445	__ai __attribute__((target("neon"))) uint16_t vaddlv_u8(uint8x8_t __p0) {
45506	46446	uint16_t __ret;
45507		__ret = (uint16_t) __builtin_neon_vaddlv_u8(__p0);
	46447	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vaddlv_u8(__p0));
45508	46448	return __ret;
45509	46449	}
45510	46450	#else
45511	46451	__ai __attribute__((target("neon"))) uint16_t vaddlv_u8(uint8x8_t __p0) {
45512	46452	uint16_t __ret;
45513		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
45514		__ret = (uint16_t) __builtin_neon_vaddlv_u8(__rev0);
	46453	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	46454	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vaddlv_u8(__rev0));
45515	46455	return __ret;
45516	46456	}
45517	46457	#endif
...	...	@@ -45519,14 +46459,14 @@ __ai __attribute__((target("neon"))) uint16_t vaddlv_u8(uint8x8_t __p0) {
45519	46459	#ifdef __LITTLE_ENDIAN__
45520	46460	__ai __attribute__((target("neon"))) uint64_t vaddlv_u32(uint32x2_t __p0) {
45521	46461	uint64_t __ret;
45522		__ret = (uint64_t) __builtin_neon_vaddlv_u32(__p0);
	46462	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vaddlv_u32(__p0));
45523	46463	return __ret;
45524	46464	}
45525	46465	#else
45526	46466	__ai __attribute__((target("neon"))) uint64_t vaddlv_u32(uint32x2_t __p0) {
45527	46467	uint64_t __ret;
45528		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
45529		__ret = (uint64_t) __builtin_neon_vaddlv_u32(__rev0);
	46468	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	46469	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vaddlv_u32(__rev0));
45530	46470	return __ret;
45531	46471	}
45532	46472	#endif
...	...	@@ -45534,14 +46474,14 @@ __ai __attribute__((target("neon"))) uint64_t vaddlv_u32(uint32x2_t __p0) {
45534	46474	#ifdef __LITTLE_ENDIAN__
45535	46475	__ai __attribute__((target("neon"))) uint32_t vaddlv_u16(uint16x4_t __p0) {
45536	46476	uint32_t __ret;
45537		__ret = (uint32_t) __builtin_neon_vaddlv_u16(__p0);
	46477	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vaddlv_u16(__p0));
45538	46478	return __ret;
45539	46479	}
45540	46480	#else
45541	46481	__ai __attribute__((target("neon"))) uint32_t vaddlv_u16(uint16x4_t __p0) {
45542	46482	uint32_t __ret;
45543		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
45544		__ret = (uint32_t) __builtin_neon_vaddlv_u16(__rev0);
	46483	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	46484	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vaddlv_u16(__rev0));
45545	46485	return __ret;
45546	46486	}
45547	46487	#endif
...	...	@@ -45549,14 +46489,14 @@ __ai __attribute__((target("neon"))) uint32_t vaddlv_u16(uint16x4_t __p0) {
45549	46489	#ifdef __LITTLE_ENDIAN__
45550	46490	__ai __attribute__((target("neon"))) int16_t vaddlv_s8(int8x8_t __p0) {
45551	46491	int16_t __ret;
45552		__ret = (int16_t) __builtin_neon_vaddlv_s8(__p0);
	46492	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vaddlv_s8(__p0));
45553	46493	return __ret;
45554	46494	}
45555	46495	#else
45556	46496	__ai __attribute__((target("neon"))) int16_t vaddlv_s8(int8x8_t __p0) {
45557	46497	int16_t __ret;
45558		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
45559		__ret = (int16_t) __builtin_neon_vaddlv_s8(__rev0);
	46498	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	46499	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vaddlv_s8(__rev0));
45560	46500	return __ret;
45561	46501	}
45562	46502	#endif
...	...	@@ -45564,14 +46504,14 @@ __ai __attribute__((target("neon"))) int16_t vaddlv_s8(int8x8_t __p0) {
45564	46504	#ifdef __LITTLE_ENDIAN__
45565	46505	__ai __attribute__((target("neon"))) int64_t vaddlv_s32(int32x2_t __p0) {
45566	46506	int64_t __ret;
45567		__ret = (int64_t) __builtin_neon_vaddlv_s32(__p0);
	46507	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vaddlv_s32(__p0));
45568	46508	return __ret;
45569	46509	}
45570	46510	#else
45571	46511	__ai __attribute__((target("neon"))) int64_t vaddlv_s32(int32x2_t __p0) {
45572	46512	int64_t __ret;
45573		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
45574		__ret = (int64_t) __builtin_neon_vaddlv_s32(__rev0);
	46513	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	46514	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vaddlv_s32(__rev0));
45575	46515	return __ret;
45576	46516	}
45577	46517	#endif
...	...	@@ -45579,14 +46519,14 @@ __ai __attribute__((target("neon"))) int64_t vaddlv_s32(int32x2_t __p0) {
45579	46519	#ifdef __LITTLE_ENDIAN__
45580	46520	__ai __attribute__((target("neon"))) int32_t vaddlv_s16(int16x4_t __p0) {
45581	46521	int32_t __ret;
45582		__ret = (int32_t) __builtin_neon_vaddlv_s16(__p0);
	46522	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vaddlv_s16(__p0));
45583	46523	return __ret;
45584	46524	}
45585	46525	#else
45586	46526	__ai __attribute__((target("neon"))) int32_t vaddlv_s16(int16x4_t __p0) {
45587	46527	int32_t __ret;
45588		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
45589		__ret = (int32_t) __builtin_neon_vaddlv_s16(__rev0);
	46528	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	46529	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vaddlv_s16(__rev0));
45590	46530	return __ret;
45591	46531	}
45592	46532	#endif
...	...	@@ -45594,14 +46534,14 @@ __ai __attribute__((target("neon"))) int32_t vaddlv_s16(int16x4_t __p0) {
45594	46534	#ifdef __LITTLE_ENDIAN__
45595	46535	__ai __attribute__((target("neon"))) uint8_t vaddvq_u8(uint8x16_t __p0) {
45596	46536	uint8_t __ret;
45597		__ret = (uint8_t) __builtin_neon_vaddvq_u8(__p0);
	46537	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vaddvq_u8(__p0));
45598	46538	return __ret;
45599	46539	}
45600	46540	#else
45601	46541	__ai __attribute__((target("neon"))) uint8_t vaddvq_u8(uint8x16_t __p0) {
45602	46542	uint8_t __ret;
45603		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
45604		__ret = (uint8_t) __builtin_neon_vaddvq_u8(__rev0);
	46543	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	46544	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vaddvq_u8(__rev0));
45605	46545	return __ret;
45606	46546	}
45607	46547	#endif
...	...	@@ -45609,14 +46549,14 @@ __ai __attribute__((target("neon"))) uint8_t vaddvq_u8(uint8x16_t __p0) {
45609	46549	#ifdef __LITTLE_ENDIAN__
45610	46550	__ai __attribute__((target("neon"))) uint32_t vaddvq_u32(uint32x4_t __p0) {
45611	46551	uint32_t __ret;
45612		__ret = (uint32_t) __builtin_neon_vaddvq_u32(__p0);
	46552	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vaddvq_u32(__p0));
45613	46553	return __ret;
45614	46554	}
45615	46555	#else
45616	46556	__ai __attribute__((target("neon"))) uint32_t vaddvq_u32(uint32x4_t __p0) {
45617	46557	uint32_t __ret;
45618		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
45619		__ret = (uint32_t) __builtin_neon_vaddvq_u32(__rev0);
	46558	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	46559	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vaddvq_u32(__rev0));
45620	46560	return __ret;
45621	46561	}
45622	46562	#endif
...	...	@@ -45624,14 +46564,14 @@ __ai __attribute__((target("neon"))) uint32_t vaddvq_u32(uint32x4_t __p0) {
45624	46564	#ifdef __LITTLE_ENDIAN__
45625	46565	__ai __attribute__((target("neon"))) uint64_t vaddvq_u64(uint64x2_t __p0) {
45626	46566	uint64_t __ret;
45627		__ret = (uint64_t) __builtin_neon_vaddvq_u64(__p0);
	46567	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vaddvq_u64(__p0));
45628	46568	return __ret;
45629	46569	}
45630	46570	#else
45631	46571	__ai __attribute__((target("neon"))) uint64_t vaddvq_u64(uint64x2_t __p0) {
45632	46572	uint64_t __ret;
45633		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
45634		__ret = (uint64_t) __builtin_neon_vaddvq_u64(__rev0);
	46573	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	46574	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vaddvq_u64(__rev0));
45635	46575	return __ret;
45636	46576	}
45637	46577	#endif
...	...	@@ -45639,14 +46579,14 @@ __ai __attribute__((target("neon"))) uint64_t vaddvq_u64(uint64x2_t __p0) {
45639	46579	#ifdef __LITTLE_ENDIAN__
45640	46580	__ai __attribute__((target("neon"))) uint16_t vaddvq_u16(uint16x8_t __p0) {
45641	46581	uint16_t __ret;
45642		__ret = (uint16_t) __builtin_neon_vaddvq_u16(__p0);
	46582	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vaddvq_u16(__p0));
45643	46583	return __ret;
45644	46584	}
45645	46585	#else
45646	46586	__ai __attribute__((target("neon"))) uint16_t vaddvq_u16(uint16x8_t __p0) {
45647	46587	uint16_t __ret;
45648		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
45649		__ret = (uint16_t) __builtin_neon_vaddvq_u16(__rev0);
	46588	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	46589	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vaddvq_u16(__rev0));
45650	46590	return __ret;
45651	46591	}
45652	46592	#endif
...	...	@@ -45654,14 +46594,14 @@ __ai __attribute__((target("neon"))) uint16_t vaddvq_u16(uint16x8_t __p0) {
45654	46594	#ifdef __LITTLE_ENDIAN__
45655	46595	__ai __attribute__((target("neon"))) int8_t vaddvq_s8(int8x16_t __p0) {
45656	46596	int8_t __ret;
45657		__ret = (int8_t) __builtin_neon_vaddvq_s8(__p0);
	46597	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vaddvq_s8(__p0));
45658	46598	return __ret;
45659	46599	}
45660	46600	#else
45661	46601	__ai __attribute__((target("neon"))) int8_t vaddvq_s8(int8x16_t __p0) {
45662	46602	int8_t __ret;
45663		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
45664		__ret = (int8_t) __builtin_neon_vaddvq_s8(__rev0);
	46603	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	46604	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vaddvq_s8(__rev0));
45665	46605	return __ret;
45666	46606	}
45667	46607	#endif
...	...	@@ -45669,14 +46609,14 @@ __ai __attribute__((target("neon"))) int8_t vaddvq_s8(int8x16_t __p0) {
45669	46609	#ifdef __LITTLE_ENDIAN__
45670	46610	__ai __attribute__((target("neon"))) float64_t vaddvq_f64(float64x2_t __p0) {
45671	46611	float64_t __ret;
45672		__ret = (float64_t) __builtin_neon_vaddvq_f64(__p0);
	46612	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vaddvq_f64(__p0));
45673	46613	return __ret;
45674	46614	}
45675	46615	#else
45676	46616	__ai __attribute__((target("neon"))) float64_t vaddvq_f64(float64x2_t __p0) {
45677	46617	float64_t __ret;
45678		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
45679		__ret = (float64_t) __builtin_neon_vaddvq_f64(__rev0);
	46618	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	46619	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vaddvq_f64(__rev0));
45680	46620	return __ret;
45681	46621	}
45682	46622	#endif
...	...	@@ -45684,14 +46624,14 @@ __ai __attribute__((target("neon"))) float64_t vaddvq_f64(float64x2_t __p0) {
45684	46624	#ifdef __LITTLE_ENDIAN__
45685	46625	__ai __attribute__((target("neon"))) float32_t vaddvq_f32(float32x4_t __p0) {
45686	46626	float32_t __ret;
45687		__ret = (float32_t) __builtin_neon_vaddvq_f32(__p0);
	46627	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vaddvq_f32(__p0));
45688	46628	return __ret;
45689	46629	}
45690	46630	#else
45691	46631	__ai __attribute__((target("neon"))) float32_t vaddvq_f32(float32x4_t __p0) {
45692	46632	float32_t __ret;
45693		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
45694		__ret = (float32_t) __builtin_neon_vaddvq_f32(__rev0);
	46633	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	46634	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vaddvq_f32(__rev0));
45695	46635	return __ret;
45696	46636	}
45697	46637	#endif
...	...	@@ -45699,14 +46639,14 @@ __ai __attribute__((target("neon"))) float32_t vaddvq_f32(float32x4_t __p0) {
45699	46639	#ifdef __LITTLE_ENDIAN__
45700	46640	__ai __attribute__((target("neon"))) int32_t vaddvq_s32(int32x4_t __p0) {
45701	46641	int32_t __ret;
45702		__ret = (int32_t) __builtin_neon_vaddvq_s32(__p0);
	46642	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vaddvq_s32(__p0));
45703	46643	return __ret;
45704	46644	}
45705	46645	#else
45706	46646	__ai __attribute__((target("neon"))) int32_t vaddvq_s32(int32x4_t __p0) {
45707	46647	int32_t __ret;
45708		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
45709		__ret = (int32_t) __builtin_neon_vaddvq_s32(__rev0);
	46648	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	46649	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vaddvq_s32(__rev0));
45710	46650	return __ret;
45711	46651	}
45712	46652	#endif
...	...	@@ -45714,14 +46654,14 @@ __ai __attribute__((target("neon"))) int32_t vaddvq_s32(int32x4_t __p0) {
45714	46654	#ifdef __LITTLE_ENDIAN__
45715	46655	__ai __attribute__((target("neon"))) int64_t vaddvq_s64(int64x2_t __p0) {
45716	46656	int64_t __ret;
45717		__ret = (int64_t) __builtin_neon_vaddvq_s64(__p0);
	46657	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vaddvq_s64(__p0));
45718	46658	return __ret;
45719	46659	}
45720	46660	#else
45721	46661	__ai __attribute__((target("neon"))) int64_t vaddvq_s64(int64x2_t __p0) {
45722	46662	int64_t __ret;
45723		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
45724		__ret = (int64_t) __builtin_neon_vaddvq_s64(__rev0);
	46663	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	46664	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vaddvq_s64(__rev0));
45725	46665	return __ret;
45726	46666	}
45727	46667	#endif
...	...	@@ -45729,14 +46669,14 @@ __ai __attribute__((target("neon"))) int64_t vaddvq_s64(int64x2_t __p0) {
45729	46669	#ifdef __LITTLE_ENDIAN__
45730	46670	__ai __attribute__((target("neon"))) int16_t vaddvq_s16(int16x8_t __p0) {
45731	46671	int16_t __ret;
45732		__ret = (int16_t) __builtin_neon_vaddvq_s16(__p0);
	46672	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vaddvq_s16(__p0));
45733	46673	return __ret;
45734	46674	}
45735	46675	#else
45736	46676	__ai __attribute__((target("neon"))) int16_t vaddvq_s16(int16x8_t __p0) {
45737	46677	int16_t __ret;
45738		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
45739		__ret = (int16_t) __builtin_neon_vaddvq_s16(__rev0);
	46678	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	46679	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vaddvq_s16(__rev0));
45740	46680	return __ret;
45741	46681	}
45742	46682	#endif
...	...	@@ -45744,14 +46684,14 @@ __ai __attribute__((target("neon"))) int16_t vaddvq_s16(int16x8_t __p0) {
45744	46684	#ifdef __LITTLE_ENDIAN__
45745	46685	__ai __attribute__((target("neon"))) uint8_t vaddv_u8(uint8x8_t __p0) {
45746	46686	uint8_t __ret;
45747		__ret = (uint8_t) __builtin_neon_vaddv_u8(__p0);
	46687	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vaddv_u8(__p0));
45748	46688	return __ret;
45749	46689	}
45750	46690	#else
45751	46691	__ai __attribute__((target("neon"))) uint8_t vaddv_u8(uint8x8_t __p0) {
45752	46692	uint8_t __ret;
45753		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
45754		__ret = (uint8_t) __builtin_neon_vaddv_u8(__rev0);
	46693	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	46694	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vaddv_u8(__rev0));
45755	46695	return __ret;
45756	46696	}
45757	46697	#endif
...	...	@@ -45759,14 +46699,14 @@ __ai __attribute__((target("neon"))) uint8_t vaddv_u8(uint8x8_t __p0) {
45759	46699	#ifdef __LITTLE_ENDIAN__
45760	46700	__ai __attribute__((target("neon"))) uint32_t vaddv_u32(uint32x2_t __p0) {
45761	46701	uint32_t __ret;
45762		__ret = (uint32_t) __builtin_neon_vaddv_u32(__p0);
	46702	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vaddv_u32(__p0));
45763	46703	return __ret;
45764	46704	}
45765	46705	#else
45766	46706	__ai __attribute__((target("neon"))) uint32_t vaddv_u32(uint32x2_t __p0) {
45767	46707	uint32_t __ret;
45768		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
45769		__ret = (uint32_t) __builtin_neon_vaddv_u32(__rev0);
	46708	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	46709	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vaddv_u32(__rev0));
45770	46710	return __ret;
45771	46711	}
45772	46712	#endif
...	...	@@ -45774,14 +46714,14 @@ __ai __attribute__((target("neon"))) uint32_t vaddv_u32(uint32x2_t __p0) {
45774	46714	#ifdef __LITTLE_ENDIAN__
45775	46715	__ai __attribute__((target("neon"))) uint16_t vaddv_u16(uint16x4_t __p0) {
45776	46716	uint16_t __ret;
45777		__ret = (uint16_t) __builtin_neon_vaddv_u16(__p0);
	46717	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vaddv_u16(__p0));
45778	46718	return __ret;
45779	46719	}
45780	46720	#else
45781	46721	__ai __attribute__((target("neon"))) uint16_t vaddv_u16(uint16x4_t __p0) {
45782	46722	uint16_t __ret;
45783		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
45784		__ret = (uint16_t) __builtin_neon_vaddv_u16(__rev0);
	46723	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	46724	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vaddv_u16(__rev0));
45785	46725	return __ret;
45786	46726	}
45787	46727	#endif
...	...	@@ -45789,14 +46729,14 @@ __ai __attribute__((target("neon"))) uint16_t vaddv_u16(uint16x4_t __p0) {
45789	46729	#ifdef __LITTLE_ENDIAN__
45790	46730	__ai __attribute__((target("neon"))) int8_t vaddv_s8(int8x8_t __p0) {
45791	46731	int8_t __ret;
45792		__ret = (int8_t) __builtin_neon_vaddv_s8(__p0);
	46732	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vaddv_s8(__p0));
45793	46733	return __ret;
45794	46734	}
45795	46735	#else
45796	46736	__ai __attribute__((target("neon"))) int8_t vaddv_s8(int8x8_t __p0) {
45797	46737	int8_t __ret;
45798		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
45799		__ret = (int8_t) __builtin_neon_vaddv_s8(__rev0);
	46738	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	46739	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vaddv_s8(__rev0));
45800	46740	return __ret;
45801	46741	}
45802	46742	#endif
...	...	@@ -45804,14 +46744,14 @@ __ai __attribute__((target("neon"))) int8_t vaddv_s8(int8x8_t __p0) {
45804	46744	#ifdef __LITTLE_ENDIAN__
45805	46745	__ai __attribute__((target("neon"))) float32_t vaddv_f32(float32x2_t __p0) {
45806	46746	float32_t __ret;
45807		__ret = (float32_t) __builtin_neon_vaddv_f32(__p0);
	46747	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vaddv_f32(__p0));
45808	46748	return __ret;
45809	46749	}
45810	46750	#else
45811	46751	__ai __attribute__((target("neon"))) float32_t vaddv_f32(float32x2_t __p0) {
45812	46752	float32_t __ret;
45813		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
45814		__ret = (float32_t) __builtin_neon_vaddv_f32(__rev0);
	46753	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	46754	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vaddv_f32(__rev0));
45815	46755	return __ret;
45816	46756	}
45817	46757	#endif
...	...	@@ -45819,14 +46759,14 @@ __ai __attribute__((target("neon"))) float32_t vaddv_f32(float32x2_t __p0) {
45819	46759	#ifdef __LITTLE_ENDIAN__
45820	46760	__ai __attribute__((target("neon"))) int32_t vaddv_s32(int32x2_t __p0) {
45821	46761	int32_t __ret;
45822		__ret = (int32_t) __builtin_neon_vaddv_s32(__p0);
	46762	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vaddv_s32(__p0));
45823	46763	return __ret;
45824	46764	}
45825	46765	#else
45826	46766	__ai __attribute__((target("neon"))) int32_t vaddv_s32(int32x2_t __p0) {
45827	46767	int32_t __ret;
45828		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
45829		__ret = (int32_t) __builtin_neon_vaddv_s32(__rev0);
	46768	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	46769	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vaddv_s32(__rev0));
45830	46770	return __ret;
45831	46771	}
45832	46772	#endif
...	...	@@ -45834,37 +46774,37 @@ __ai __attribute__((target("neon"))) int32_t vaddv_s32(int32x2_t __p0) {
45834	46774	#ifdef __LITTLE_ENDIAN__
45835	46775	__ai __attribute__((target("neon"))) int16_t vaddv_s16(int16x4_t __p0) {
45836	46776	int16_t __ret;
45837		__ret = (int16_t) __builtin_neon_vaddv_s16(__p0);
	46777	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vaddv_s16(__p0));
45838	46778	return __ret;
45839	46779	}
45840	46780	#else
45841	46781	__ai __attribute__((target("neon"))) int16_t vaddv_s16(int16x4_t __p0) {
45842	46782	int16_t __ret;
45843		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
45844		__ret = (int16_t) __builtin_neon_vaddv_s16(__rev0);
	46783	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	46784	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vaddv_s16(__rev0));
45845	46785	return __ret;
45846	46786	}
45847	46787	#endif
45848	46788
45849	46789	__ai __attribute__((target("neon"))) poly64x1_t vbsl_p64(uint64x1_t __p0, poly64x1_t __p1, poly64x1_t __p2) {
45850	46790	poly64x1_t __ret;
45851		__ret = (poly64x1_t) __builtin_neon_vbsl_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 6);
	46791	__ret = __builtin_bit_cast(poly64x1_t, __builtin_neon_vbsl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 6));
45852	46792	return __ret;
45853	46793	}
45854	46794	#ifdef __LITTLE_ENDIAN__
45855	46795	__ai __attribute__((target("neon"))) poly64x2_t vbslq_p64(uint64x2_t __p0, poly64x2_t __p1, poly64x2_t __p2) {
45856	46796	poly64x2_t __ret;
45857		__ret = (poly64x2_t) __builtin_neon_vbslq_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 38);
	46797	__ret = __builtin_bit_cast(poly64x2_t, __builtin_neon_vbslq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 38));
45858	46798	return __ret;
45859	46799	}
45860	46800	#else
45861	46801	__ai __attribute__((target("neon"))) poly64x2_t vbslq_p64(uint64x2_t __p0, poly64x2_t __p1, poly64x2_t __p2) {
45862	46802	poly64x2_t __ret;
45863		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
45864		poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
45865		poly64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
45866		__ret = (poly64x2_t) __builtin_neon_vbslq_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 38);
45867		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	46803	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	46804	poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	46805	poly64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_64);
	46806	__ret = __builtin_bit_cast(poly64x2_t, __builtin_neon_vbslq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 38));
	46807	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
45868	46808	return __ret;
45869	46809	}
45870	46810	#endif
...	...	@@ -45872,172 +46812,172 @@ __ai __attribute__((target("neon"))) poly64x2_t vbslq_p64(uint64x2_t __p0, poly6
45872	46812	#ifdef __LITTLE_ENDIAN__
45873	46813	__ai __attribute__((target("neon"))) float64x2_t vbslq_f64(uint64x2_t __p0, float64x2_t __p1, float64x2_t __p2) {
45874	46814	float64x2_t __ret;
45875		__ret = (float64x2_t) __builtin_neon_vbslq_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 42);
	46815	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vbslq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 42));
45876	46816	return __ret;
45877	46817	}
45878	46818	#else
45879	46819	__ai __attribute__((target("neon"))) float64x2_t vbslq_f64(uint64x2_t __p0, float64x2_t __p1, float64x2_t __p2) {
45880	46820	float64x2_t __ret;
45881		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
45882		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
45883		float64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
45884		__ret = (float64x2_t) __builtin_neon_vbslq_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 42);
45885		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	46821	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	46822	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	46823	float64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_64);
	46824	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vbslq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 42));
	46825	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
45886	46826	return __ret;
45887	46827	}
45888	46828	#endif
45889	46829
45890	46830	__ai __attribute__((target("neon"))) float64x1_t vbsl_f64(uint64x1_t __p0, float64x1_t __p1, float64x1_t __p2) {
45891	46831	float64x1_t __ret;
45892		__ret = (float64x1_t) __builtin_neon_vbsl_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 10);
	46832	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vbsl_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 10));
45893	46833	return __ret;
45894	46834	}
45895	46835	#ifdef __LITTLE_ENDIAN__
45896	46836	__ai __attribute__((target("neon"))) uint64x2_t vcageq_f64(float64x2_t __p0, float64x2_t __p1) {
45897	46837	uint64x2_t __ret;
45898		__ret = (uint64x2_t) __builtin_neon_vcageq_v((int8x16_t)__p0, (int8x16_t)__p1, 51);
	46838	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcageq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 51));
45899	46839	return __ret;
45900	46840	}
45901	46841	#else
45902	46842	__ai __attribute__((target("neon"))) uint64x2_t vcageq_f64(float64x2_t __p0, float64x2_t __p1) {
45903	46843	uint64x2_t __ret;
45904		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
45905		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
45906		__ret = (uint64x2_t) __builtin_neon_vcageq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 51);
45907		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	46844	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	46845	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	46846	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcageq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 51));
	46847	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
45908	46848	return __ret;
45909	46849	}
45910	46850	#endif
45911	46851
45912	46852	__ai __attribute__((target("neon"))) uint64x1_t vcage_f64(float64x1_t __p0, float64x1_t __p1) {
45913	46853	uint64x1_t __ret;
45914		__ret = (uint64x1_t) __builtin_neon_vcage_v((int8x8_t)__p0, (int8x8_t)__p1, 19);
	46854	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vcage_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 19));
45915	46855	return __ret;
45916	46856	}
45917	46857	__ai __attribute__((target("neon"))) uint64_t vcaged_f64(float64_t __p0, float64_t __p1) {
45918	46858	uint64_t __ret;
45919		__ret = (uint64_t) __builtin_neon_vcaged_f64(__p0, __p1);
	46859	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcaged_f64(__p0, __p1));
45920	46860	return __ret;
45921	46861	}
45922	46862	__ai __attribute__((target("neon"))) uint32_t vcages_f32(float32_t __p0, float32_t __p1) {
45923	46863	uint32_t __ret;
45924		__ret = (uint32_t) __builtin_neon_vcages_f32(__p0, __p1);
	46864	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vcages_f32(__p0, __p1));
45925	46865	return __ret;
45926	46866	}
45927	46867	#ifdef __LITTLE_ENDIAN__
45928	46868	__ai __attribute__((target("neon"))) uint64x2_t vcagtq_f64(float64x2_t __p0, float64x2_t __p1) {
45929	46869	uint64x2_t __ret;
45930		__ret = (uint64x2_t) __builtin_neon_vcagtq_v((int8x16_t)__p0, (int8x16_t)__p1, 51);
	46870	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcagtq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 51));
45931	46871	return __ret;
45932	46872	}
45933	46873	#else
45934	46874	__ai __attribute__((target("neon"))) uint64x2_t vcagtq_f64(float64x2_t __p0, float64x2_t __p1) {
45935	46875	uint64x2_t __ret;
45936		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
45937		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
45938		__ret = (uint64x2_t) __builtin_neon_vcagtq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 51);
45939		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	46876	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	46877	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	46878	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcagtq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 51));
	46879	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
45940	46880	return __ret;
45941	46881	}
45942	46882	#endif
45943	46883
45944	46884	__ai __attribute__((target("neon"))) uint64x1_t vcagt_f64(float64x1_t __p0, float64x1_t __p1) {
45945	46885	uint64x1_t __ret;
45946		__ret = (uint64x1_t) __builtin_neon_vcagt_v((int8x8_t)__p0, (int8x8_t)__p1, 19);
	46886	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vcagt_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 19));
45947	46887	return __ret;
45948	46888	}
45949	46889	__ai __attribute__((target("neon"))) uint64_t vcagtd_f64(float64_t __p0, float64_t __p1) {
45950	46890	uint64_t __ret;
45951		__ret = (uint64_t) __builtin_neon_vcagtd_f64(__p0, __p1);
	46891	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcagtd_f64(__p0, __p1));
45952	46892	return __ret;
45953	46893	}
45954	46894	__ai __attribute__((target("neon"))) uint32_t vcagts_f32(float32_t __p0, float32_t __p1) {
45955	46895	uint32_t __ret;
45956		__ret = (uint32_t) __builtin_neon_vcagts_f32(__p0, __p1);
	46896	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vcagts_f32(__p0, __p1));
45957	46897	return __ret;
45958	46898	}
45959	46899	#ifdef __LITTLE_ENDIAN__
45960	46900	__ai __attribute__((target("neon"))) uint64x2_t vcaleq_f64(float64x2_t __p0, float64x2_t __p1) {
45961	46901	uint64x2_t __ret;
45962		__ret = (uint64x2_t) __builtin_neon_vcaleq_v((int8x16_t)__p0, (int8x16_t)__p1, 51);
	46902	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcaleq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 51));
45963	46903	return __ret;
45964	46904	}
45965	46905	#else
45966	46906	__ai __attribute__((target("neon"))) uint64x2_t vcaleq_f64(float64x2_t __p0, float64x2_t __p1) {
45967	46907	uint64x2_t __ret;
45968		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
45969		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
45970		__ret = (uint64x2_t) __builtin_neon_vcaleq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 51);
45971		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	46908	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	46909	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	46910	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcaleq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 51));
	46911	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
45972	46912	return __ret;
45973	46913	}
45974	46914	#endif
45975	46915
45976	46916	__ai __attribute__((target("neon"))) uint64x1_t vcale_f64(float64x1_t __p0, float64x1_t __p1) {
45977	46917	uint64x1_t __ret;
45978		__ret = (uint64x1_t) __builtin_neon_vcale_v((int8x8_t)__p0, (int8x8_t)__p1, 19);
	46918	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vcale_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 19));
45979	46919	return __ret;
45980	46920	}
45981	46921	__ai __attribute__((target("neon"))) uint64_t vcaled_f64(float64_t __p0, float64_t __p1) {
45982	46922	uint64_t __ret;
45983		__ret = (uint64_t) __builtin_neon_vcaled_f64(__p0, __p1);
	46923	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcaled_f64(__p0, __p1));
45984	46924	return __ret;
45985	46925	}
45986	46926	__ai __attribute__((target("neon"))) uint32_t vcales_f32(float32_t __p0, float32_t __p1) {
45987	46927	uint32_t __ret;
45988		__ret = (uint32_t) __builtin_neon_vcales_f32(__p0, __p1);
	46928	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vcales_f32(__p0, __p1));
45989	46929	return __ret;
45990	46930	}
45991	46931	#ifdef __LITTLE_ENDIAN__
45992	46932	__ai __attribute__((target("neon"))) uint64x2_t vcaltq_f64(float64x2_t __p0, float64x2_t __p1) {
45993	46933	uint64x2_t __ret;
45994		__ret = (uint64x2_t) __builtin_neon_vcaltq_v((int8x16_t)__p0, (int8x16_t)__p1, 51);
	46934	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcaltq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 51));
45995	46935	return __ret;
45996	46936	}
45997	46937	#else
45998	46938	__ai __attribute__((target("neon"))) uint64x2_t vcaltq_f64(float64x2_t __p0, float64x2_t __p1) {
45999	46939	uint64x2_t __ret;
46000		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
46001		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
46002		__ret = (uint64x2_t) __builtin_neon_vcaltq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 51);
46003		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	46940	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	46941	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	46942	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcaltq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 51));
	46943	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
46004	46944	return __ret;
46005	46945	}
46006	46946	#endif
46007	46947
46008	46948	__ai __attribute__((target("neon"))) uint64x1_t vcalt_f64(float64x1_t __p0, float64x1_t __p1) {
46009	46949	uint64x1_t __ret;
46010		__ret = (uint64x1_t) __builtin_neon_vcalt_v((int8x8_t)__p0, (int8x8_t)__p1, 19);
	46950	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vcalt_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 19));
46011	46951	return __ret;
46012	46952	}
46013	46953	__ai __attribute__((target("neon"))) uint64_t vcaltd_f64(float64_t __p0, float64_t __p1) {
46014	46954	uint64_t __ret;
46015		__ret = (uint64_t) __builtin_neon_vcaltd_f64(__p0, __p1);
	46955	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcaltd_f64(__p0, __p1));
46016	46956	return __ret;
46017	46957	}
46018	46958	__ai __attribute__((target("neon"))) uint32_t vcalts_f32(float32_t __p0, float32_t __p1) {
46019	46959	uint32_t __ret;
46020		__ret = (uint32_t) __builtin_neon_vcalts_f32(__p0, __p1);
	46960	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vcalts_f32(__p0, __p1));
46021	46961	return __ret;
46022	46962	}
46023	46963	__ai __attribute__((target("neon"))) uint64x1_t vceq_p64(poly64x1_t __p0, poly64x1_t __p1) {
46024	46964	uint64x1_t __ret;
46025		__ret = (uint64x1_t)(__p0 == __p1);
	46965	__ret = __builtin_bit_cast(uint64x1_t, __p0 == __p1);
46026	46966	return __ret;
46027	46967	}
46028	46968	#ifdef __LITTLE_ENDIAN__
46029	46969	__ai __attribute__((target("neon"))) uint64x2_t vceqq_p64(poly64x2_t __p0, poly64x2_t __p1) {
46030	46970	uint64x2_t __ret;
46031		__ret = (uint64x2_t)(__p0 == __p1);
	46971	__ret = __builtin_bit_cast(uint64x2_t, __p0 == __p1);
46032	46972	return __ret;
46033	46973	}
46034	46974	#else
46035	46975	__ai __attribute__((target("neon"))) uint64x2_t vceqq_p64(poly64x2_t __p0, poly64x2_t __p1) {
46036	46976	uint64x2_t __ret;
46037		poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
46038		poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
46039		__ret = (uint64x2_t)(__rev0 == __rev1);
46040		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	46977	poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	46978	poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	46979	__ret = __builtin_bit_cast(uint64x2_t, __rev0 == __rev1);
	46980	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
46041	46981	return __ret;
46042	46982	}
46043	46983	#endif
...	...	@@ -46045,16 +46985,16 @@ __ai __attribute__((target("neon"))) uint64x2_t vceqq_p64(poly64x2_t __p0, poly6
46045	46985	#ifdef __LITTLE_ENDIAN__
46046	46986	__ai __attribute__((target("neon"))) uint64x2_t vceqq_u64(uint64x2_t __p0, uint64x2_t __p1) {
46047	46987	uint64x2_t __ret;
46048		__ret = (uint64x2_t)(__p0 == __p1);
	46988	__ret = __builtin_bit_cast(uint64x2_t, __p0 == __p1);
46049	46989	return __ret;
46050	46990	}
46051	46991	#else
46052	46992	__ai __attribute__((target("neon"))) uint64x2_t vceqq_u64(uint64x2_t __p0, uint64x2_t __p1) {
46053	46993	uint64x2_t __ret;
46054		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
46055		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
46056		__ret = (uint64x2_t)(__rev0 == __rev1);
46057		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	46994	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	46995	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	46996	__ret = __builtin_bit_cast(uint64x2_t, __rev0 == __rev1);
	46997	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
46058	46998	return __ret;
46059	46999	}
46060	47000	#endif
...	...	@@ -46062,16 +47002,16 @@ __ai __attribute__((target("neon"))) uint64x2_t vceqq_u64(uint64x2_t __p0, uint6
46062	47002	#ifdef __LITTLE_ENDIAN__
46063	47003	__ai __attribute__((target("neon"))) uint64x2_t vceqq_f64(float64x2_t __p0, float64x2_t __p1) {
46064	47004	uint64x2_t __ret;
46065		__ret = (uint64x2_t)(__p0 == __p1);
	47005	__ret = __builtin_bit_cast(uint64x2_t, __p0 == __p1);
46066	47006	return __ret;
46067	47007	}
46068	47008	#else
46069	47009	__ai __attribute__((target("neon"))) uint64x2_t vceqq_f64(float64x2_t __p0, float64x2_t __p1) {
46070	47010	uint64x2_t __ret;
46071		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
46072		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
46073		__ret = (uint64x2_t)(__rev0 == __rev1);
46074		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	47011	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	47012	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	47013	__ret = __builtin_bit_cast(uint64x2_t, __rev0 == __rev1);
	47014	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
46075	47015	return __ret;
46076	47016	}
46077	47017	#endif
...	...	@@ -46079,88 +47019,88 @@ __ai __attribute__((target("neon"))) uint64x2_t vceqq_f64(float64x2_t __p0, floa
46079	47019	#ifdef __LITTLE_ENDIAN__
46080	47020	__ai __attribute__((target("neon"))) uint64x2_t vceqq_s64(int64x2_t __p0, int64x2_t __p1) {
46081	47021	uint64x2_t __ret;
46082		__ret = (uint64x2_t)(__p0 == __p1);
	47022	__ret = __builtin_bit_cast(uint64x2_t, __p0 == __p1);
46083	47023	return __ret;
46084	47024	}
46085	47025	#else
46086	47026	__ai __attribute__((target("neon"))) uint64x2_t vceqq_s64(int64x2_t __p0, int64x2_t __p1) {
46087	47027	uint64x2_t __ret;
46088		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
46089		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
46090		__ret = (uint64x2_t)(__rev0 == __rev1);
46091		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	47028	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	47029	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	47030	__ret = __builtin_bit_cast(uint64x2_t, __rev0 == __rev1);
	47031	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
46092	47032	return __ret;
46093	47033	}
46094	47034	#endif
46095	47035
46096	47036	__ai __attribute__((target("neon"))) uint64x1_t vceq_u64(uint64x1_t __p0, uint64x1_t __p1) {
46097	47037	uint64x1_t __ret;
46098		__ret = (uint64x1_t)(__p0 == __p1);
	47038	__ret = __builtin_bit_cast(uint64x1_t, __p0 == __p1);
46099	47039	return __ret;
46100	47040	}
46101	47041	__ai __attribute__((target("neon"))) uint64x1_t vceq_f64(float64x1_t __p0, float64x1_t __p1) {
46102	47042	uint64x1_t __ret;
46103		__ret = (uint64x1_t)(__p0 == __p1);
	47043	__ret = __builtin_bit_cast(uint64x1_t, __p0 == __p1);
46104	47044	return __ret;
46105	47045	}
46106	47046	__ai __attribute__((target("neon"))) uint64x1_t vceq_s64(int64x1_t __p0, int64x1_t __p1) {
46107	47047	uint64x1_t __ret;
46108		__ret = (uint64x1_t)(__p0 == __p1);
	47048	__ret = __builtin_bit_cast(uint64x1_t, __p0 == __p1);
46109	47049	return __ret;
46110	47050	}
46111	47051	__ai __attribute__((target("neon"))) uint64_t vceqd_u64(uint64_t __p0, uint64_t __p1) {
46112	47052	uint64_t __ret;
46113		__ret = (uint64_t) __builtin_neon_vceqd_u64(__p0, __p1);
	47053	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vceqd_u64(__p0, __p1));
46114	47054	return __ret;
46115	47055	}
46116	47056	__ai __attribute__((target("neon"))) uint64_t vceqd_s64(int64_t __p0, int64_t __p1) {
46117	47057	uint64_t __ret;
46118		__ret = (uint64_t) __builtin_neon_vceqd_s64(__p0, __p1);
	47058	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vceqd_s64(__p0, __p1));
46119	47059	return __ret;
46120	47060	}
46121	47061	__ai __attribute__((target("neon"))) uint64_t vceqd_f64(float64_t __p0, float64_t __p1) {
46122	47062	uint64_t __ret;
46123		__ret = (uint64_t) __builtin_neon_vceqd_f64(__p0, __p1);
	47063	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vceqd_f64(__p0, __p1));
46124	47064	return __ret;
46125	47065	}
46126	47066	__ai __attribute__((target("neon"))) uint32_t vceqs_f32(float32_t __p0, float32_t __p1) {
46127	47067	uint32_t __ret;
46128		__ret = (uint32_t) __builtin_neon_vceqs_f32(__p0, __p1);
	47068	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vceqs_f32(__p0, __p1));
46129	47069	return __ret;
46130	47070	}
46131	47071	#ifdef __LITTLE_ENDIAN__
46132	47072	__ai __attribute__((target("neon"))) uint8x8_t vceqz_p8(poly8x8_t __p0) {
46133	47073	uint8x8_t __ret;
46134		__ret = (uint8x8_t) __builtin_neon_vceqz_v((int8x8_t)__p0, 16);
	47074	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vceqz_v(__builtin_bit_cast(int8x8_t, __p0), 4));
46135	47075	return __ret;
46136	47076	}
46137	47077	#else
46138	47078	__ai __attribute__((target("neon"))) uint8x8_t vceqz_p8(poly8x8_t __p0) {
46139	47079	uint8x8_t __ret;
46140		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
46141		__ret = (uint8x8_t) __builtin_neon_vceqz_v((int8x8_t)__rev0, 16);
46142		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	47080	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	47081	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vceqz_v(__builtin_bit_cast(int8x8_t, __rev0), 4));
	47082	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
46143	47083	return __ret;
46144	47084	}
46145	47085	#endif
46146	47086
46147	47087	__ai __attribute__((target("neon"))) uint64x1_t vceqz_p64(poly64x1_t __p0) {
46148	47088	uint64x1_t __ret;
46149		__ret = (uint64x1_t) __builtin_neon_vceqz_v((int8x8_t)__p0, 19);
	47089	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vceqz_v(__builtin_bit_cast(int8x8_t, __p0), 6));
46150	47090	return __ret;
46151	47091	}
46152	47092	#ifdef __LITTLE_ENDIAN__
46153	47093	__ai __attribute__((target("neon"))) uint8x16_t vceqzq_p8(poly8x16_t __p0) {
46154	47094	uint8x16_t __ret;
46155		__ret = (uint8x16_t) __builtin_neon_vceqzq_v((int8x16_t)__p0, 48);
	47095	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vceqzq_v(__builtin_bit_cast(int8x16_t, __p0), 36));
46156	47096	return __ret;
46157	47097	}
46158	47098	#else
46159	47099	__ai __attribute__((target("neon"))) uint8x16_t vceqzq_p8(poly8x16_t __p0) {
46160	47100	uint8x16_t __ret;
46161		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
46162		__ret = (uint8x16_t) __builtin_neon_vceqzq_v((int8x16_t)__rev0, 48);
46163		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	47101	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	47102	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vceqzq_v(__builtin_bit_cast(int8x16_t, __rev0), 36));
	47103	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
46164	47104	return __ret;
46165	47105	}
46166	47106	#endif
...	...	@@ -46168,15 +47108,15 @@ __ai __attribute__((target("neon"))) uint8x16_t vceqzq_p8(poly8x16_t __p0) {
46168	47108	#ifdef __LITTLE_ENDIAN__
46169	47109	__ai __attribute__((target("neon"))) uint64x2_t vceqzq_p64(poly64x2_t __p0) {
46170	47110	uint64x2_t __ret;
46171		__ret = (uint64x2_t) __builtin_neon_vceqzq_v((int8x16_t)__p0, 51);
	47111	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vceqzq_v(__builtin_bit_cast(int8x16_t, __p0), 38));
46172	47112	return __ret;
46173	47113	}
46174	47114	#else
46175	47115	__ai __attribute__((target("neon"))) uint64x2_t vceqzq_p64(poly64x2_t __p0) {
46176	47116	uint64x2_t __ret;
46177		poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
46178		__ret = (uint64x2_t) __builtin_neon_vceqzq_v((int8x16_t)__rev0, 51);
46179		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	47117	poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	47118	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vceqzq_v(__builtin_bit_cast(int8x16_t, __rev0), 38));
	47119	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
46180	47120	return __ret;
46181	47121	}
46182	47122	#endif
...	...	@@ -46184,15 +47124,15 @@ __ai __attribute__((target("neon"))) uint64x2_t vceqzq_p64(poly64x2_t __p0) {
46184	47124	#ifdef __LITTLE_ENDIAN__
46185	47125	__ai __attribute__((target("neon"))) uint8x16_t vceqzq_u8(uint8x16_t __p0) {
46186	47126	uint8x16_t __ret;
46187		__ret = (uint8x16_t) __builtin_neon_vceqzq_v((int8x16_t)__p0, 48);
	47127	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vceqzq_v(__builtin_bit_cast(int8x16_t, __p0), 48));
46188	47128	return __ret;
46189	47129	}
46190	47130	#else
46191	47131	__ai __attribute__((target("neon"))) uint8x16_t vceqzq_u8(uint8x16_t __p0) {
46192	47132	uint8x16_t __ret;
46193		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
46194		__ret = (uint8x16_t) __builtin_neon_vceqzq_v((int8x16_t)__rev0, 48);
46195		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	47133	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	47134	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vceqzq_v(__builtin_bit_cast(int8x16_t, __rev0), 48));
	47135	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
46196	47136	return __ret;
46197	47137	}
46198	47138	#endif
...	...	@@ -46200,15 +47140,15 @@ __ai __attribute__((target("neon"))) uint8x16_t vceqzq_u8(uint8x16_t __p0) {
46200	47140	#ifdef __LITTLE_ENDIAN__
46201	47141	__ai __attribute__((target("neon"))) uint32x4_t vceqzq_u32(uint32x4_t __p0) {
46202	47142	uint32x4_t __ret;
46203		__ret = (uint32x4_t) __builtin_neon_vceqzq_v((int8x16_t)__p0, 50);
	47143	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vceqzq_v(__builtin_bit_cast(int8x16_t, __p0), 50));
46204	47144	return __ret;
46205	47145	}
46206	47146	#else
46207	47147	__ai __attribute__((target("neon"))) uint32x4_t vceqzq_u32(uint32x4_t __p0) {
46208	47148	uint32x4_t __ret;
46209		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
46210		__ret = (uint32x4_t) __builtin_neon_vceqzq_v((int8x16_t)__rev0, 50);
46211		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	47149	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	47150	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vceqzq_v(__builtin_bit_cast(int8x16_t, __rev0), 50));
	47151	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
46212	47152	return __ret;
46213	47153	}
46214	47154	#endif
...	...	@@ -46216,15 +47156,15 @@ __ai __attribute__((target("neon"))) uint32x4_t vceqzq_u32(uint32x4_t __p0) {
46216	47156	#ifdef __LITTLE_ENDIAN__
46217	47157	__ai __attribute__((target("neon"))) uint64x2_t vceqzq_u64(uint64x2_t __p0) {
46218	47158	uint64x2_t __ret;
46219		__ret = (uint64x2_t) __builtin_neon_vceqzq_v((int8x16_t)__p0, 51);
	47159	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vceqzq_v(__builtin_bit_cast(int8x16_t, __p0), 51));
46220	47160	return __ret;
46221	47161	}
46222	47162	#else
46223	47163	__ai __attribute__((target("neon"))) uint64x2_t vceqzq_u64(uint64x2_t __p0) {
46224	47164	uint64x2_t __ret;
46225		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
46226		__ret = (uint64x2_t) __builtin_neon_vceqzq_v((int8x16_t)__rev0, 51);
46227		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	47165	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	47166	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vceqzq_v(__builtin_bit_cast(int8x16_t, __rev0), 51));
	47167	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
46228	47168	return __ret;
46229	47169	}
46230	47170	#endif
...	...	@@ -46232,15 +47172,15 @@ __ai __attribute__((target("neon"))) uint64x2_t vceqzq_u64(uint64x2_t __p0) {
46232	47172	#ifdef __LITTLE_ENDIAN__
46233	47173	__ai __attribute__((target("neon"))) uint16x8_t vceqzq_u16(uint16x8_t __p0) {
46234	47174	uint16x8_t __ret;
46235		__ret = (uint16x8_t) __builtin_neon_vceqzq_v((int8x16_t)__p0, 49);
	47175	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vceqzq_v(__builtin_bit_cast(int8x16_t, __p0), 49));
46236	47176	return __ret;
46237	47177	}
46238	47178	#else
46239	47179	__ai __attribute__((target("neon"))) uint16x8_t vceqzq_u16(uint16x8_t __p0) {
46240	47180	uint16x8_t __ret;
46241		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
46242		__ret = (uint16x8_t) __builtin_neon_vceqzq_v((int8x16_t)__rev0, 49);
46243		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	47181	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	47182	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vceqzq_v(__builtin_bit_cast(int8x16_t, __rev0), 49));
	47183	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
46244	47184	return __ret;
46245	47185	}
46246	47186	#endif
...	...	@@ -46248,15 +47188,15 @@ __ai __attribute__((target("neon"))) uint16x8_t vceqzq_u16(uint16x8_t __p0) {
46248	47188	#ifdef __LITTLE_ENDIAN__
46249	47189	__ai __attribute__((target("neon"))) uint8x16_t vceqzq_s8(int8x16_t __p0) {
46250	47190	uint8x16_t __ret;
46251		__ret = (uint8x16_t) __builtin_neon_vceqzq_v((int8x16_t)__p0, 48);
	47191	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vceqzq_v(__builtin_bit_cast(int8x16_t, __p0), 32));
46252	47192	return __ret;
46253	47193	}
46254	47194	#else
46255	47195	__ai __attribute__((target("neon"))) uint8x16_t vceqzq_s8(int8x16_t __p0) {
46256	47196	uint8x16_t __ret;
46257		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
46258		__ret = (uint8x16_t) __builtin_neon_vceqzq_v((int8x16_t)__rev0, 48);
46259		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	47197	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	47198	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vceqzq_v(__builtin_bit_cast(int8x16_t, __rev0), 32));
	47199	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
46260	47200	return __ret;
46261	47201	}
46262	47202	#endif
...	...	@@ -46264,15 +47204,15 @@ __ai __attribute__((target("neon"))) uint8x16_t vceqzq_s8(int8x16_t __p0) {
46264	47204	#ifdef __LITTLE_ENDIAN__
46265	47205	__ai __attribute__((target("neon"))) uint64x2_t vceqzq_f64(float64x2_t __p0) {
46266	47206	uint64x2_t __ret;
46267		__ret = (uint64x2_t) __builtin_neon_vceqzq_v((int8x16_t)__p0, 51);
	47207	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vceqzq_v(__builtin_bit_cast(int8x16_t, __p0), 42));
46268	47208	return __ret;
46269	47209	}
46270	47210	#else
46271	47211	__ai __attribute__((target("neon"))) uint64x2_t vceqzq_f64(float64x2_t __p0) {
46272	47212	uint64x2_t __ret;
46273		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
46274		__ret = (uint64x2_t) __builtin_neon_vceqzq_v((int8x16_t)__rev0, 51);
46275		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	47213	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	47214	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vceqzq_v(__builtin_bit_cast(int8x16_t, __rev0), 42));
	47215	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
46276	47216	return __ret;
46277	47217	}
46278	47218	#endif
...	...	@@ -46280,15 +47220,15 @@ __ai __attribute__((target("neon"))) uint64x2_t vceqzq_f64(float64x2_t __p0) {
46280	47220	#ifdef __LITTLE_ENDIAN__
46281	47221	__ai __attribute__((target("neon"))) uint32x4_t vceqzq_f32(float32x4_t __p0) {
46282	47222	uint32x4_t __ret;
46283		__ret = (uint32x4_t) __builtin_neon_vceqzq_v((int8x16_t)__p0, 50);
	47223	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vceqzq_v(__builtin_bit_cast(int8x16_t, __p0), 41));
46284	47224	return __ret;
46285	47225	}
46286	47226	#else
46287	47227	__ai __attribute__((target("neon"))) uint32x4_t vceqzq_f32(float32x4_t __p0) {
46288	47228	uint32x4_t __ret;
46289		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
46290		__ret = (uint32x4_t) __builtin_neon_vceqzq_v((int8x16_t)__rev0, 50);
46291		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	47229	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	47230	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vceqzq_v(__builtin_bit_cast(int8x16_t, __rev0), 41));
	47231	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
46292	47232	return __ret;
46293	47233	}
46294	47234	#endif
...	...	@@ -46296,15 +47236,15 @@ __ai __attribute__((target("neon"))) uint32x4_t vceqzq_f32(float32x4_t __p0) {
46296	47236	#ifdef __LITTLE_ENDIAN__
46297	47237	__ai __attribute__((target("neon"))) uint32x4_t vceqzq_s32(int32x4_t __p0) {
46298	47238	uint32x4_t __ret;
46299		__ret = (uint32x4_t) __builtin_neon_vceqzq_v((int8x16_t)__p0, 50);
	47239	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vceqzq_v(__builtin_bit_cast(int8x16_t, __p0), 34));
46300	47240	return __ret;
46301	47241	}
46302	47242	#else
46303	47243	__ai __attribute__((target("neon"))) uint32x4_t vceqzq_s32(int32x4_t __p0) {
46304	47244	uint32x4_t __ret;
46305		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
46306		__ret = (uint32x4_t) __builtin_neon_vceqzq_v((int8x16_t)__rev0, 50);
46307		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	47245	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	47246	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vceqzq_v(__builtin_bit_cast(int8x16_t, __rev0), 34));
	47247	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
46308	47248	return __ret;
46309	47249	}
46310	47250	#endif
...	...	@@ -46312,15 +47252,15 @@ __ai __attribute__((target("neon"))) uint32x4_t vceqzq_s32(int32x4_t __p0) {
46312	47252	#ifdef __LITTLE_ENDIAN__
46313	47253	__ai __attribute__((target("neon"))) uint64x2_t vceqzq_s64(int64x2_t __p0) {
46314	47254	uint64x2_t __ret;
46315		__ret = (uint64x2_t) __builtin_neon_vceqzq_v((int8x16_t)__p0, 51);
	47255	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vceqzq_v(__builtin_bit_cast(int8x16_t, __p0), 35));
46316	47256	return __ret;
46317	47257	}
46318	47258	#else
46319	47259	__ai __attribute__((target("neon"))) uint64x2_t vceqzq_s64(int64x2_t __p0) {
46320	47260	uint64x2_t __ret;
46321		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
46322		__ret = (uint64x2_t) __builtin_neon_vceqzq_v((int8x16_t)__rev0, 51);
46323		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	47261	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	47262	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vceqzq_v(__builtin_bit_cast(int8x16_t, __rev0), 35));
	47263	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
46324	47264	return __ret;
46325	47265	}
46326	47266	#endif
...	...	@@ -46328,15 +47268,15 @@ __ai __attribute__((target("neon"))) uint64x2_t vceqzq_s64(int64x2_t __p0) {
46328	47268	#ifdef __LITTLE_ENDIAN__
46329	47269	__ai __attribute__((target("neon"))) uint16x8_t vceqzq_s16(int16x8_t __p0) {
46330	47270	uint16x8_t __ret;
46331		__ret = (uint16x8_t) __builtin_neon_vceqzq_v((int8x16_t)__p0, 49);
	47271	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vceqzq_v(__builtin_bit_cast(int8x16_t, __p0), 33));
46332	47272	return __ret;
46333	47273	}
46334	47274	#else
46335	47275	__ai __attribute__((target("neon"))) uint16x8_t vceqzq_s16(int16x8_t __p0) {
46336	47276	uint16x8_t __ret;
46337		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
46338		__ret = (uint16x8_t) __builtin_neon_vceqzq_v((int8x16_t)__rev0, 49);
46339		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	47277	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	47278	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vceqzq_v(__builtin_bit_cast(int8x16_t, __rev0), 33));
	47279	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
46340	47280	return __ret;
46341	47281	}
46342	47282	#endif
...	...	@@ -46344,15 +47284,15 @@ __ai __attribute__((target("neon"))) uint16x8_t vceqzq_s16(int16x8_t __p0) {
46344	47284	#ifdef __LITTLE_ENDIAN__
46345	47285	__ai __attribute__((target("neon"))) uint8x8_t vceqz_u8(uint8x8_t __p0) {
46346	47286	uint8x8_t __ret;
46347		__ret = (uint8x8_t) __builtin_neon_vceqz_v((int8x8_t)__p0, 16);
	47287	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vceqz_v(__builtin_bit_cast(int8x8_t, __p0), 16));
46348	47288	return __ret;
46349	47289	}
46350	47290	#else
46351	47291	__ai __attribute__((target("neon"))) uint8x8_t vceqz_u8(uint8x8_t __p0) {
46352	47292	uint8x8_t __ret;
46353		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
46354		__ret = (uint8x8_t) __builtin_neon_vceqz_v((int8x8_t)__rev0, 16);
46355		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	47293	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	47294	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vceqz_v(__builtin_bit_cast(int8x8_t, __rev0), 16));
	47295	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
46356	47296	return __ret;
46357	47297	}
46358	47298	#endif
...	...	@@ -46360,36 +47300,36 @@ __ai __attribute__((target("neon"))) uint8x8_t vceqz_u8(uint8x8_t __p0) {
46360	47300	#ifdef __LITTLE_ENDIAN__
46361	47301	__ai __attribute__((target("neon"))) uint32x2_t vceqz_u32(uint32x2_t __p0) {
46362	47302	uint32x2_t __ret;
46363		__ret = (uint32x2_t) __builtin_neon_vceqz_v((int8x8_t)__p0, 18);
	47303	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vceqz_v(__builtin_bit_cast(int8x8_t, __p0), 18));
46364	47304	return __ret;
46365	47305	}
46366	47306	#else
46367	47307	__ai __attribute__((target("neon"))) uint32x2_t vceqz_u32(uint32x2_t __p0) {
46368	47308	uint32x2_t __ret;
46369		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
46370		__ret = (uint32x2_t) __builtin_neon_vceqz_v((int8x8_t)__rev0, 18);
46371		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	47309	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	47310	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vceqz_v(__builtin_bit_cast(int8x8_t, __rev0), 18));
	47311	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
46372	47312	return __ret;
46373	47313	}
46374	47314	#endif
46375	47315
46376	47316	__ai __attribute__((target("neon"))) uint64x1_t vceqz_u64(uint64x1_t __p0) {
46377	47317	uint64x1_t __ret;
46378		__ret = (uint64x1_t) __builtin_neon_vceqz_v((int8x8_t)__p0, 19);
	47318	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vceqz_v(__builtin_bit_cast(int8x8_t, __p0), 19));
46379	47319	return __ret;
46380	47320	}
46381	47321	#ifdef __LITTLE_ENDIAN__
46382	47322	__ai __attribute__((target("neon"))) uint16x4_t vceqz_u16(uint16x4_t __p0) {
46383	47323	uint16x4_t __ret;
46384		__ret = (uint16x4_t) __builtin_neon_vceqz_v((int8x8_t)__p0, 17);
	47324	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vceqz_v(__builtin_bit_cast(int8x8_t, __p0), 17));
46385	47325	return __ret;
46386	47326	}
46387	47327	#else
46388	47328	__ai __attribute__((target("neon"))) uint16x4_t vceqz_u16(uint16x4_t __p0) {
46389	47329	uint16x4_t __ret;
46390		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
46391		__ret = (uint16x4_t) __builtin_neon_vceqz_v((int8x8_t)__rev0, 17);
46392		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	47330	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	47331	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vceqz_v(__builtin_bit_cast(int8x8_t, __rev0), 17));
	47332	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
46393	47333	return __ret;
46394	47334	}
46395	47335	#endif
...	...	@@ -46397,36 +47337,36 @@ __ai __attribute__((target("neon"))) uint16x4_t vceqz_u16(uint16x4_t __p0) {
46397	47337	#ifdef __LITTLE_ENDIAN__
46398	47338	__ai __attribute__((target("neon"))) uint8x8_t vceqz_s8(int8x8_t __p0) {
46399	47339	uint8x8_t __ret;
46400		__ret = (uint8x8_t) __builtin_neon_vceqz_v((int8x8_t)__p0, 16);
	47340	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vceqz_v(__builtin_bit_cast(int8x8_t, __p0), 0));
46401	47341	return __ret;
46402	47342	}
46403	47343	#else
46404	47344	__ai __attribute__((target("neon"))) uint8x8_t vceqz_s8(int8x8_t __p0) {
46405	47345	uint8x8_t __ret;
46406		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
46407		__ret = (uint8x8_t) __builtin_neon_vceqz_v((int8x8_t)__rev0, 16);
46408		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	47346	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	47347	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vceqz_v(__builtin_bit_cast(int8x8_t, __rev0), 0));
	47348	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
46409	47349	return __ret;
46410	47350	}
46411	47351	#endif
46412	47352
46413	47353	__ai __attribute__((target("neon"))) uint64x1_t vceqz_f64(float64x1_t __p0) {
46414	47354	uint64x1_t __ret;
46415		__ret = (uint64x1_t) __builtin_neon_vceqz_v((int8x8_t)__p0, 19);
	47355	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vceqz_v(__builtin_bit_cast(int8x8_t, __p0), 10));
46416	47356	return __ret;
46417	47357	}
46418	47358	#ifdef __LITTLE_ENDIAN__
46419	47359	__ai __attribute__((target("neon"))) uint32x2_t vceqz_f32(float32x2_t __p0) {
46420	47360	uint32x2_t __ret;
46421		__ret = (uint32x2_t) __builtin_neon_vceqz_v((int8x8_t)__p0, 18);
	47361	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vceqz_v(__builtin_bit_cast(int8x8_t, __p0), 9));
46422	47362	return __ret;
46423	47363	}
46424	47364	#else
46425	47365	__ai __attribute__((target("neon"))) uint32x2_t vceqz_f32(float32x2_t __p0) {
46426	47366	uint32x2_t __ret;
46427		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
46428		__ret = (uint32x2_t) __builtin_neon_vceqz_v((int8x8_t)__rev0, 18);
46429		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	47367	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	47368	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vceqz_v(__builtin_bit_cast(int8x8_t, __rev0), 9));
	47369	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
46430	47370	return __ret;
46431	47371	}
46432	47372	#endif
...	...	@@ -46434,73 +47374,73 @@ __ai __attribute__((target("neon"))) uint32x2_t vceqz_f32(float32x2_t __p0) {
46434	47374	#ifdef __LITTLE_ENDIAN__
46435	47375	__ai __attribute__((target("neon"))) uint32x2_t vceqz_s32(int32x2_t __p0) {
46436	47376	uint32x2_t __ret;
46437		__ret = (uint32x2_t) __builtin_neon_vceqz_v((int8x8_t)__p0, 18);
	47377	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vceqz_v(__builtin_bit_cast(int8x8_t, __p0), 2));
46438	47378	return __ret;
46439	47379	}
46440	47380	#else
46441	47381	__ai __attribute__((target("neon"))) uint32x2_t vceqz_s32(int32x2_t __p0) {
46442	47382	uint32x2_t __ret;
46443		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
46444		__ret = (uint32x2_t) __builtin_neon_vceqz_v((int8x8_t)__rev0, 18);
46445		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	47383	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	47384	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vceqz_v(__builtin_bit_cast(int8x8_t, __rev0), 2));
	47385	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
46446	47386	return __ret;
46447	47387	}
46448	47388	#endif
46449	47389
46450	47390	__ai __attribute__((target("neon"))) uint64x1_t vceqz_s64(int64x1_t __p0) {
46451	47391	uint64x1_t __ret;
46452		__ret = (uint64x1_t) __builtin_neon_vceqz_v((int8x8_t)__p0, 19);
	47392	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vceqz_v(__builtin_bit_cast(int8x8_t, __p0), 3));
46453	47393	return __ret;
46454	47394	}
46455	47395	#ifdef __LITTLE_ENDIAN__
46456	47396	__ai __attribute__((target("neon"))) uint16x4_t vceqz_s16(int16x4_t __p0) {
46457	47397	uint16x4_t __ret;
46458		__ret = (uint16x4_t) __builtin_neon_vceqz_v((int8x8_t)__p0, 17);
	47398	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vceqz_v(__builtin_bit_cast(int8x8_t, __p0), 1));
46459	47399	return __ret;
46460	47400	}
46461	47401	#else
46462	47402	__ai __attribute__((target("neon"))) uint16x4_t vceqz_s16(int16x4_t __p0) {
46463	47403	uint16x4_t __ret;
46464		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
46465		__ret = (uint16x4_t) __builtin_neon_vceqz_v((int8x8_t)__rev0, 17);
46466		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	47404	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	47405	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vceqz_v(__builtin_bit_cast(int8x8_t, __rev0), 1));
	47406	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
46467	47407	return __ret;
46468	47408	}
46469	47409	#endif
46470	47410
46471	47411	__ai __attribute__((target("neon"))) uint64_t vceqzd_u64(uint64_t __p0) {
46472	47412	uint64_t __ret;
46473		__ret = (uint64_t) __builtin_neon_vceqzd_u64(__p0);
	47413	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vceqzd_u64(__p0));
46474	47414	return __ret;
46475	47415	}
46476	47416	__ai __attribute__((target("neon"))) uint64_t vceqzd_s64(int64_t __p0) {
46477	47417	uint64_t __ret;
46478		__ret = (uint64_t) __builtin_neon_vceqzd_s64(__p0);
	47418	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vceqzd_s64(__p0));
46479	47419	return __ret;
46480	47420	}
46481	47421	__ai __attribute__((target("neon"))) uint64_t vceqzd_f64(float64_t __p0) {
46482	47422	uint64_t __ret;
46483		__ret = (uint64_t) __builtin_neon_vceqzd_f64(__p0);
	47423	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vceqzd_f64(__p0));
46484	47424	return __ret;
46485	47425	}
46486	47426	__ai __attribute__((target("neon"))) uint32_t vceqzs_f32(float32_t __p0) {
46487	47427	uint32_t __ret;
46488		__ret = (uint32_t) __builtin_neon_vceqzs_f32(__p0);
	47428	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vceqzs_f32(__p0));
46489	47429	return __ret;
46490	47430	}
46491	47431	#ifdef __LITTLE_ENDIAN__
46492	47432	__ai __attribute__((target("neon"))) uint64x2_t vcgeq_u64(uint64x2_t __p0, uint64x2_t __p1) {
46493	47433	uint64x2_t __ret;
46494		__ret = (uint64x2_t)(__p0 >= __p1);
	47434	__ret = __builtin_bit_cast(uint64x2_t, __p0 >= __p1);
46495	47435	return __ret;
46496	47436	}
46497	47437	#else
46498	47438	__ai __attribute__((target("neon"))) uint64x2_t vcgeq_u64(uint64x2_t __p0, uint64x2_t __p1) {
46499	47439	uint64x2_t __ret;
46500		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
46501		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
46502		__ret = (uint64x2_t)(__rev0 >= __rev1);
46503		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	47440	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	47441	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	47442	__ret = __builtin_bit_cast(uint64x2_t, __rev0 >= __rev1);
	47443	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
46504	47444	return __ret;
46505	47445	}
46506	47446	#endif
...	...	@@ -46508,16 +47448,16 @@ __ai __attribute__((target("neon"))) uint64x2_t vcgeq_u64(uint64x2_t __p0, uint6
46508	47448	#ifdef __LITTLE_ENDIAN__
46509	47449	__ai __attribute__((target("neon"))) uint64x2_t vcgeq_f64(float64x2_t __p0, float64x2_t __p1) {
46510	47450	uint64x2_t __ret;
46511		__ret = (uint64x2_t)(__p0 >= __p1);
	47451	__ret = __builtin_bit_cast(uint64x2_t, __p0 >= __p1);
46512	47452	return __ret;
46513	47453	}
46514	47454	#else
46515	47455	__ai __attribute__((target("neon"))) uint64x2_t vcgeq_f64(float64x2_t __p0, float64x2_t __p1) {
46516	47456	uint64x2_t __ret;
46517		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
46518		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
46519		__ret = (uint64x2_t)(__rev0 >= __rev1);
46520		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	47457	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	47458	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	47459	__ret = __builtin_bit_cast(uint64x2_t, __rev0 >= __rev1);
	47460	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
46521	47461	return __ret;
46522	47462	}
46523	47463	#endif
...	...	@@ -46525,67 +47465,67 @@ __ai __attribute__((target("neon"))) uint64x2_t vcgeq_f64(float64x2_t __p0, floa
46525	47465	#ifdef __LITTLE_ENDIAN__
46526	47466	__ai __attribute__((target("neon"))) uint64x2_t vcgeq_s64(int64x2_t __p0, int64x2_t __p1) {
46527	47467	uint64x2_t __ret;
46528		__ret = (uint64x2_t)(__p0 >= __p1);
	47468	__ret = __builtin_bit_cast(uint64x2_t, __p0 >= __p1);
46529	47469	return __ret;
46530	47470	}
46531	47471	#else
46532	47472	__ai __attribute__((target("neon"))) uint64x2_t vcgeq_s64(int64x2_t __p0, int64x2_t __p1) {
46533	47473	uint64x2_t __ret;
46534		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
46535		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
46536		__ret = (uint64x2_t)(__rev0 >= __rev1);
46537		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	47474	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	47475	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	47476	__ret = __builtin_bit_cast(uint64x2_t, __rev0 >= __rev1);
	47477	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
46538	47478	return __ret;
46539	47479	}
46540	47480	#endif
46541	47481
46542	47482	__ai __attribute__((target("neon"))) uint64x1_t vcge_u64(uint64x1_t __p0, uint64x1_t __p1) {
46543	47483	uint64x1_t __ret;
46544		__ret = (uint64x1_t)(__p0 >= __p1);
	47484	__ret = __builtin_bit_cast(uint64x1_t, __p0 >= __p1);
46545	47485	return __ret;
46546	47486	}
46547	47487	__ai __attribute__((target("neon"))) uint64x1_t vcge_f64(float64x1_t __p0, float64x1_t __p1) {
46548	47488	uint64x1_t __ret;
46549		__ret = (uint64x1_t)(__p0 >= __p1);
	47489	__ret = __builtin_bit_cast(uint64x1_t, __p0 >= __p1);
46550	47490	return __ret;
46551	47491	}
46552	47492	__ai __attribute__((target("neon"))) uint64x1_t vcge_s64(int64x1_t __p0, int64x1_t __p1) {
46553	47493	uint64x1_t __ret;
46554		__ret = (uint64x1_t)(__p0 >= __p1);
	47494	__ret = __builtin_bit_cast(uint64x1_t, __p0 >= __p1);
46555	47495	return __ret;
46556	47496	}
46557	47497	__ai __attribute__((target("neon"))) uint64_t vcged_s64(int64_t __p0, int64_t __p1) {
46558	47498	uint64_t __ret;
46559		__ret = (uint64_t) __builtin_neon_vcged_s64(__p0, __p1);
	47499	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcged_s64(__p0, __p1));
46560	47500	return __ret;
46561	47501	}
46562	47502	__ai __attribute__((target("neon"))) uint64_t vcged_u64(uint64_t __p0, uint64_t __p1) {
46563	47503	uint64_t __ret;
46564		__ret = (uint64_t) __builtin_neon_vcged_u64(__p0, __p1);
	47504	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcged_u64(__p0, __p1));
46565	47505	return __ret;
46566	47506	}
46567	47507	__ai __attribute__((target("neon"))) uint64_t vcged_f64(float64_t __p0, float64_t __p1) {
46568	47508	uint64_t __ret;
46569		__ret = (uint64_t) __builtin_neon_vcged_f64(__p0, __p1);
	47509	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcged_f64(__p0, __p1));
46570	47510	return __ret;
46571	47511	}
46572	47512	__ai __attribute__((target("neon"))) uint32_t vcges_f32(float32_t __p0, float32_t __p1) {
46573	47513	uint32_t __ret;
46574		__ret = (uint32_t) __builtin_neon_vcges_f32(__p0, __p1);
	47514	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vcges_f32(__p0, __p1));
46575	47515	return __ret;
46576	47516	}
46577	47517	#ifdef __LITTLE_ENDIAN__
46578	47518	__ai __attribute__((target("neon"))) uint8x16_t vcgezq_s8(int8x16_t __p0) {
46579	47519	uint8x16_t __ret;
46580		__ret = (uint8x16_t) __builtin_neon_vcgezq_v((int8x16_t)__p0, 48);
	47520	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vcgezq_v(__builtin_bit_cast(int8x16_t, __p0), 32));
46581	47521	return __ret;
46582	47522	}
46583	47523	#else
46584	47524	__ai __attribute__((target("neon"))) uint8x16_t vcgezq_s8(int8x16_t __p0) {
46585	47525	uint8x16_t __ret;
46586		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
46587		__ret = (uint8x16_t) __builtin_neon_vcgezq_v((int8x16_t)__rev0, 48);
46588		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	47526	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	47527	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vcgezq_v(__builtin_bit_cast(int8x16_t, __rev0), 32));
	47528	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
46589	47529	return __ret;
46590	47530	}
46591	47531	#endif
...	...	@@ -46593,15 +47533,15 @@ __ai __attribute__((target("neon"))) uint8x16_t vcgezq_s8(int8x16_t __p0) {
46593	47533	#ifdef __LITTLE_ENDIAN__
46594	47534	__ai __attribute__((target("neon"))) uint64x2_t vcgezq_f64(float64x2_t __p0) {
46595	47535	uint64x2_t __ret;
46596		__ret = (uint64x2_t) __builtin_neon_vcgezq_v((int8x16_t)__p0, 51);
	47536	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcgezq_v(__builtin_bit_cast(int8x16_t, __p0), 42));
46597	47537	return __ret;
46598	47538	}
46599	47539	#else
46600	47540	__ai __attribute__((target("neon"))) uint64x2_t vcgezq_f64(float64x2_t __p0) {
46601	47541	uint64x2_t __ret;
46602		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
46603		__ret = (uint64x2_t) __builtin_neon_vcgezq_v((int8x16_t)__rev0, 51);
46604		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	47542	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	47543	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcgezq_v(__builtin_bit_cast(int8x16_t, __rev0), 42));
	47544	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
46605	47545	return __ret;
46606	47546	}
46607	47547	#endif
...	...	@@ -46609,15 +47549,15 @@ __ai __attribute__((target("neon"))) uint64x2_t vcgezq_f64(float64x2_t __p0) {
46609	47549	#ifdef __LITTLE_ENDIAN__
46610	47550	__ai __attribute__((target("neon"))) uint32x4_t vcgezq_f32(float32x4_t __p0) {
46611	47551	uint32x4_t __ret;
46612		__ret = (uint32x4_t) __builtin_neon_vcgezq_v((int8x16_t)__p0, 50);
	47552	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcgezq_v(__builtin_bit_cast(int8x16_t, __p0), 41));
46613	47553	return __ret;
46614	47554	}
46615	47555	#else
46616	47556	__ai __attribute__((target("neon"))) uint32x4_t vcgezq_f32(float32x4_t __p0) {
46617	47557	uint32x4_t __ret;
46618		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
46619		__ret = (uint32x4_t) __builtin_neon_vcgezq_v((int8x16_t)__rev0, 50);
46620		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	47558	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	47559	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcgezq_v(__builtin_bit_cast(int8x16_t, __rev0), 41));
	47560	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
46621	47561	return __ret;
46622	47562	}
46623	47563	#endif
...	...	@@ -46625,15 +47565,15 @@ __ai __attribute__((target("neon"))) uint32x4_t vcgezq_f32(float32x4_t __p0) {
46625	47565	#ifdef __LITTLE_ENDIAN__
46626	47566	__ai __attribute__((target("neon"))) uint32x4_t vcgezq_s32(int32x4_t __p0) {
46627	47567	uint32x4_t __ret;
46628		__ret = (uint32x4_t) __builtin_neon_vcgezq_v((int8x16_t)__p0, 50);
	47568	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcgezq_v(__builtin_bit_cast(int8x16_t, __p0), 34));
46629	47569	return __ret;
46630	47570	}
46631	47571	#else
46632	47572	__ai __attribute__((target("neon"))) uint32x4_t vcgezq_s32(int32x4_t __p0) {
46633	47573	uint32x4_t __ret;
46634		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
46635		__ret = (uint32x4_t) __builtin_neon_vcgezq_v((int8x16_t)__rev0, 50);
46636		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	47574	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	47575	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcgezq_v(__builtin_bit_cast(int8x16_t, __rev0), 34));
	47576	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
46637	47577	return __ret;
46638	47578	}
46639	47579	#endif
...	...	@@ -46641,15 +47581,15 @@ __ai __attribute__((target("neon"))) uint32x4_t vcgezq_s32(int32x4_t __p0) {
46641	47581	#ifdef __LITTLE_ENDIAN__
46642	47582	__ai __attribute__((target("neon"))) uint64x2_t vcgezq_s64(int64x2_t __p0) {
46643	47583	uint64x2_t __ret;
46644		__ret = (uint64x2_t) __builtin_neon_vcgezq_v((int8x16_t)__p0, 51);
	47584	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcgezq_v(__builtin_bit_cast(int8x16_t, __p0), 35));
46645	47585	return __ret;
46646	47586	}
46647	47587	#else
46648	47588	__ai __attribute__((target("neon"))) uint64x2_t vcgezq_s64(int64x2_t __p0) {
46649	47589	uint64x2_t __ret;
46650		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
46651		__ret = (uint64x2_t) __builtin_neon_vcgezq_v((int8x16_t)__rev0, 51);
46652		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	47590	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	47591	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcgezq_v(__builtin_bit_cast(int8x16_t, __rev0), 35));
	47592	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
46653	47593	return __ret;
46654	47594	}
46655	47595	#endif
...	...	@@ -46657,15 +47597,15 @@ __ai __attribute__((target("neon"))) uint64x2_t vcgezq_s64(int64x2_t __p0) {
46657	47597	#ifdef __LITTLE_ENDIAN__
46658	47598	__ai __attribute__((target("neon"))) uint16x8_t vcgezq_s16(int16x8_t __p0) {
46659	47599	uint16x8_t __ret;
46660		__ret = (uint16x8_t) __builtin_neon_vcgezq_v((int8x16_t)__p0, 49);
	47600	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcgezq_v(__builtin_bit_cast(int8x16_t, __p0), 33));
46661	47601	return __ret;
46662	47602	}
46663	47603	#else
46664	47604	__ai __attribute__((target("neon"))) uint16x8_t vcgezq_s16(int16x8_t __p0) {
46665	47605	uint16x8_t __ret;
46666		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
46667		__ret = (uint16x8_t) __builtin_neon_vcgezq_v((int8x16_t)__rev0, 49);
46668		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	47606	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	47607	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcgezq_v(__builtin_bit_cast(int8x16_t, __rev0), 33));
	47608	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
46669	47609	return __ret;
46670	47610	}
46671	47611	#endif
...	...	@@ -46673,36 +47613,36 @@ __ai __attribute__((target("neon"))) uint16x8_t vcgezq_s16(int16x8_t __p0) {
46673	47613	#ifdef __LITTLE_ENDIAN__
46674	47614	__ai __attribute__((target("neon"))) uint8x8_t vcgez_s8(int8x8_t __p0) {
46675	47615	uint8x8_t __ret;
46676		__ret = (uint8x8_t) __builtin_neon_vcgez_v((int8x8_t)__p0, 16);
	47616	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vcgez_v(__builtin_bit_cast(int8x8_t, __p0), 0));
46677	47617	return __ret;
46678	47618	}
46679	47619	#else
46680	47620	__ai __attribute__((target("neon"))) uint8x8_t vcgez_s8(int8x8_t __p0) {
46681	47621	uint8x8_t __ret;
46682		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
46683		__ret = (uint8x8_t) __builtin_neon_vcgez_v((int8x8_t)__rev0, 16);
46684		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	47622	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	47623	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vcgez_v(__builtin_bit_cast(int8x8_t, __rev0), 0));
	47624	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
46685	47625	return __ret;
46686	47626	}
46687	47627	#endif
46688	47628
46689	47629	__ai __attribute__((target("neon"))) uint64x1_t vcgez_f64(float64x1_t __p0) {
46690	47630	uint64x1_t __ret;
46691		__ret = (uint64x1_t) __builtin_neon_vcgez_v((int8x8_t)__p0, 19);
	47631	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vcgez_v(__builtin_bit_cast(int8x8_t, __p0), 10));
46692	47632	return __ret;
46693	47633	}
46694	47634	#ifdef __LITTLE_ENDIAN__
46695	47635	__ai __attribute__((target("neon"))) uint32x2_t vcgez_f32(float32x2_t __p0) {
46696	47636	uint32x2_t __ret;
46697		__ret = (uint32x2_t) __builtin_neon_vcgez_v((int8x8_t)__p0, 18);
	47637	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcgez_v(__builtin_bit_cast(int8x8_t, __p0), 9));
46698	47638	return __ret;
46699	47639	}
46700	47640	#else
46701	47641	__ai __attribute__((target("neon"))) uint32x2_t vcgez_f32(float32x2_t __p0) {
46702	47642	uint32x2_t __ret;
46703		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
46704		__ret = (uint32x2_t) __builtin_neon_vcgez_v((int8x8_t)__rev0, 18);
46705		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	47643	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	47644	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcgez_v(__builtin_bit_cast(int8x8_t, __rev0), 9));
	47645	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
46706	47646	return __ret;
46707	47647	}
46708	47648	#endif
...	...	@@ -46710,68 +47650,68 @@ __ai __attribute__((target("neon"))) uint32x2_t vcgez_f32(float32x2_t __p0) {
46710	47650	#ifdef __LITTLE_ENDIAN__
46711	47651	__ai __attribute__((target("neon"))) uint32x2_t vcgez_s32(int32x2_t __p0) {
46712	47652	uint32x2_t __ret;
46713		__ret = (uint32x2_t) __builtin_neon_vcgez_v((int8x8_t)__p0, 18);
	47653	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcgez_v(__builtin_bit_cast(int8x8_t, __p0), 2));
46714	47654	return __ret;
46715	47655	}
46716	47656	#else
46717	47657	__ai __attribute__((target("neon"))) uint32x2_t vcgez_s32(int32x2_t __p0) {
46718	47658	uint32x2_t __ret;
46719		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
46720		__ret = (uint32x2_t) __builtin_neon_vcgez_v((int8x8_t)__rev0, 18);
46721		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	47659	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	47660	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcgez_v(__builtin_bit_cast(int8x8_t, __rev0), 2));
	47661	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
46722	47662	return __ret;
46723	47663	}
46724	47664	#endif
46725	47665
46726	47666	__ai __attribute__((target("neon"))) uint64x1_t vcgez_s64(int64x1_t __p0) {
46727	47667	uint64x1_t __ret;
46728		__ret = (uint64x1_t) __builtin_neon_vcgez_v((int8x8_t)__p0, 19);
	47668	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vcgez_v(__builtin_bit_cast(int8x8_t, __p0), 3));
46729	47669	return __ret;
46730	47670	}
46731	47671	#ifdef __LITTLE_ENDIAN__
46732	47672	__ai __attribute__((target("neon"))) uint16x4_t vcgez_s16(int16x4_t __p0) {
46733	47673	uint16x4_t __ret;
46734		__ret = (uint16x4_t) __builtin_neon_vcgez_v((int8x8_t)__p0, 17);
	47674	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcgez_v(__builtin_bit_cast(int8x8_t, __p0), 1));
46735	47675	return __ret;
46736	47676	}
46737	47677	#else
46738	47678	__ai __attribute__((target("neon"))) uint16x4_t vcgez_s16(int16x4_t __p0) {
46739	47679	uint16x4_t __ret;
46740		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
46741		__ret = (uint16x4_t) __builtin_neon_vcgez_v((int8x8_t)__rev0, 17);
46742		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	47680	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	47681	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcgez_v(__builtin_bit_cast(int8x8_t, __rev0), 1));
	47682	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
46743	47683	return __ret;
46744	47684	}
46745	47685	#endif
46746	47686
46747	47687	__ai __attribute__((target("neon"))) uint64_t vcgezd_s64(int64_t __p0) {
46748	47688	uint64_t __ret;
46749		__ret = (uint64_t) __builtin_neon_vcgezd_s64(__p0);
	47689	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcgezd_s64(__p0));
46750	47690	return __ret;
46751	47691	}
46752	47692	__ai __attribute__((target("neon"))) uint64_t vcgezd_f64(float64_t __p0) {
46753	47693	uint64_t __ret;
46754		__ret = (uint64_t) __builtin_neon_vcgezd_f64(__p0);
	47694	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcgezd_f64(__p0));
46755	47695	return __ret;
46756	47696	}
46757	47697	__ai __attribute__((target("neon"))) uint32_t vcgezs_f32(float32_t __p0) {
46758	47698	uint32_t __ret;
46759		__ret = (uint32_t) __builtin_neon_vcgezs_f32(__p0);
	47699	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vcgezs_f32(__p0));
46760	47700	return __ret;
46761	47701	}
46762	47702	#ifdef __LITTLE_ENDIAN__
46763	47703	__ai __attribute__((target("neon"))) uint64x2_t vcgtq_u64(uint64x2_t __p0, uint64x2_t __p1) {
46764	47704	uint64x2_t __ret;
46765		__ret = (uint64x2_t)(__p0 > __p1);
	47705	__ret = __builtin_bit_cast(uint64x2_t, __p0 > __p1);
46766	47706	return __ret;
46767	47707	}
46768	47708	#else
46769	47709	__ai __attribute__((target("neon"))) uint64x2_t vcgtq_u64(uint64x2_t __p0, uint64x2_t __p1) {
46770	47710	uint64x2_t __ret;
46771		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
46772		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
46773		__ret = (uint64x2_t)(__rev0 > __rev1);
46774		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	47711	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	47712	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	47713	__ret = __builtin_bit_cast(uint64x2_t, __rev0 > __rev1);
	47714	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
46775	47715	return __ret;
46776	47716	}
46777	47717	#endif
...	...	@@ -46779,16 +47719,16 @@ __ai __attribute__((target("neon"))) uint64x2_t vcgtq_u64(uint64x2_t __p0, uint6
46779	47719	#ifdef __LITTLE_ENDIAN__
46780	47720	__ai __attribute__((target("neon"))) uint64x2_t vcgtq_f64(float64x2_t __p0, float64x2_t __p1) {
46781	47721	uint64x2_t __ret;
46782		__ret = (uint64x2_t)(__p0 > __p1);
	47722	__ret = __builtin_bit_cast(uint64x2_t, __p0 > __p1);
46783	47723	return __ret;
46784	47724	}
46785	47725	#else
46786	47726	__ai __attribute__((target("neon"))) uint64x2_t vcgtq_f64(float64x2_t __p0, float64x2_t __p1) {
46787	47727	uint64x2_t __ret;
46788		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
46789		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
46790		__ret = (uint64x2_t)(__rev0 > __rev1);
46791		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	47728	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	47729	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	47730	__ret = __builtin_bit_cast(uint64x2_t, __rev0 > __rev1);
	47731	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
46792	47732	return __ret;
46793	47733	}
46794	47734	#endif
...	...	@@ -46796,67 +47736,67 @@ __ai __attribute__((target("neon"))) uint64x2_t vcgtq_f64(float64x2_t __p0, floa
46796	47736	#ifdef __LITTLE_ENDIAN__
46797	47737	__ai __attribute__((target("neon"))) uint64x2_t vcgtq_s64(int64x2_t __p0, int64x2_t __p1) {
46798	47738	uint64x2_t __ret;
46799		__ret = (uint64x2_t)(__p0 > __p1);
	47739	__ret = __builtin_bit_cast(uint64x2_t, __p0 > __p1);
46800	47740	return __ret;
46801	47741	}
46802	47742	#else
46803	47743	__ai __attribute__((target("neon"))) uint64x2_t vcgtq_s64(int64x2_t __p0, int64x2_t __p1) {
46804	47744	uint64x2_t __ret;
46805		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
46806		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
46807		__ret = (uint64x2_t)(__rev0 > __rev1);
46808		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	47745	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	47746	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	47747	__ret = __builtin_bit_cast(uint64x2_t, __rev0 > __rev1);
	47748	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
46809	47749	return __ret;
46810	47750	}
46811	47751	#endif
46812	47752
46813	47753	__ai __attribute__((target("neon"))) uint64x1_t vcgt_u64(uint64x1_t __p0, uint64x1_t __p1) {
46814	47754	uint64x1_t __ret;
46815		__ret = (uint64x1_t)(__p0 > __p1);
	47755	__ret = __builtin_bit_cast(uint64x1_t, __p0 > __p1);
46816	47756	return __ret;
46817	47757	}
46818	47758	__ai __attribute__((target("neon"))) uint64x1_t vcgt_f64(float64x1_t __p0, float64x1_t __p1) {
46819	47759	uint64x1_t __ret;
46820		__ret = (uint64x1_t)(__p0 > __p1);
	47760	__ret = __builtin_bit_cast(uint64x1_t, __p0 > __p1);
46821	47761	return __ret;
46822	47762	}
46823	47763	__ai __attribute__((target("neon"))) uint64x1_t vcgt_s64(int64x1_t __p0, int64x1_t __p1) {
46824	47764	uint64x1_t __ret;
46825		__ret = (uint64x1_t)(__p0 > __p1);
	47765	__ret = __builtin_bit_cast(uint64x1_t, __p0 > __p1);
46826	47766	return __ret;
46827	47767	}
46828	47768	__ai __attribute__((target("neon"))) uint64_t vcgtd_s64(int64_t __p0, int64_t __p1) {
46829	47769	uint64_t __ret;
46830		__ret = (uint64_t) __builtin_neon_vcgtd_s64(__p0, __p1);
	47770	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcgtd_s64(__p0, __p1));
46831	47771	return __ret;
46832	47772	}
46833	47773	__ai __attribute__((target("neon"))) uint64_t vcgtd_u64(uint64_t __p0, uint64_t __p1) {
46834	47774	uint64_t __ret;
46835		__ret = (uint64_t) __builtin_neon_vcgtd_u64(__p0, __p1);
	47775	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcgtd_u64(__p0, __p1));
46836	47776	return __ret;
46837	47777	}
46838	47778	__ai __attribute__((target("neon"))) uint64_t vcgtd_f64(float64_t __p0, float64_t __p1) {
46839	47779	uint64_t __ret;
46840		__ret = (uint64_t) __builtin_neon_vcgtd_f64(__p0, __p1);
	47780	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcgtd_f64(__p0, __p1));
46841	47781	return __ret;
46842	47782	}
46843	47783	__ai __attribute__((target("neon"))) uint32_t vcgts_f32(float32_t __p0, float32_t __p1) {
46844	47784	uint32_t __ret;
46845		__ret = (uint32_t) __builtin_neon_vcgts_f32(__p0, __p1);
	47785	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vcgts_f32(__p0, __p1));
46846	47786	return __ret;
46847	47787	}
46848	47788	#ifdef __LITTLE_ENDIAN__
46849	47789	__ai __attribute__((target("neon"))) uint8x16_t vcgtzq_s8(int8x16_t __p0) {
46850	47790	uint8x16_t __ret;
46851		__ret = (uint8x16_t) __builtin_neon_vcgtzq_v((int8x16_t)__p0, 48);
	47791	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vcgtzq_v(__builtin_bit_cast(int8x16_t, __p0), 32));
46852	47792	return __ret;
46853	47793	}
46854	47794	#else
46855	47795	__ai __attribute__((target("neon"))) uint8x16_t vcgtzq_s8(int8x16_t __p0) {
46856	47796	uint8x16_t __ret;
46857		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
46858		__ret = (uint8x16_t) __builtin_neon_vcgtzq_v((int8x16_t)__rev0, 48);
46859		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	47797	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	47798	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vcgtzq_v(__builtin_bit_cast(int8x16_t, __rev0), 32));
	47799	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
46860	47800	return __ret;
46861	47801	}
46862	47802	#endif
...	...	@@ -46864,15 +47804,15 @@ __ai __attribute__((target("neon"))) uint8x16_t vcgtzq_s8(int8x16_t __p0) {
46864	47804	#ifdef __LITTLE_ENDIAN__
46865	47805	__ai __attribute__((target("neon"))) uint64x2_t vcgtzq_f64(float64x2_t __p0) {
46866	47806	uint64x2_t __ret;
46867		__ret = (uint64x2_t) __builtin_neon_vcgtzq_v((int8x16_t)__p0, 51);
	47807	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcgtzq_v(__builtin_bit_cast(int8x16_t, __p0), 42));
46868	47808	return __ret;
46869	47809	}
46870	47810	#else
46871	47811	__ai __attribute__((target("neon"))) uint64x2_t vcgtzq_f64(float64x2_t __p0) {
46872	47812	uint64x2_t __ret;
46873		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
46874		__ret = (uint64x2_t) __builtin_neon_vcgtzq_v((int8x16_t)__rev0, 51);
46875		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	47813	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	47814	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcgtzq_v(__builtin_bit_cast(int8x16_t, __rev0), 42));
	47815	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
46876	47816	return __ret;
46877	47817	}
46878	47818	#endif
...	...	@@ -46880,15 +47820,15 @@ __ai __attribute__((target("neon"))) uint64x2_t vcgtzq_f64(float64x2_t __p0) {
46880	47820	#ifdef __LITTLE_ENDIAN__
46881	47821	__ai __attribute__((target("neon"))) uint32x4_t vcgtzq_f32(float32x4_t __p0) {
46882	47822	uint32x4_t __ret;
46883		__ret = (uint32x4_t) __builtin_neon_vcgtzq_v((int8x16_t)__p0, 50);
	47823	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcgtzq_v(__builtin_bit_cast(int8x16_t, __p0), 41));
46884	47824	return __ret;
46885	47825	}
46886	47826	#else
46887	47827	__ai __attribute__((target("neon"))) uint32x4_t vcgtzq_f32(float32x4_t __p0) {
46888	47828	uint32x4_t __ret;
46889		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
46890		__ret = (uint32x4_t) __builtin_neon_vcgtzq_v((int8x16_t)__rev0, 50);
46891		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	47829	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	47830	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcgtzq_v(__builtin_bit_cast(int8x16_t, __rev0), 41));
	47831	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
46892	47832	return __ret;
46893	47833	}
46894	47834	#endif
...	...	@@ -46896,15 +47836,15 @@ __ai __attribute__((target("neon"))) uint32x4_t vcgtzq_f32(float32x4_t __p0) {
46896	47836	#ifdef __LITTLE_ENDIAN__
46897	47837	__ai __attribute__((target("neon"))) uint32x4_t vcgtzq_s32(int32x4_t __p0) {
46898	47838	uint32x4_t __ret;
46899		__ret = (uint32x4_t) __builtin_neon_vcgtzq_v((int8x16_t)__p0, 50);
	47839	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcgtzq_v(__builtin_bit_cast(int8x16_t, __p0), 34));
46900	47840	return __ret;
46901	47841	}
46902	47842	#else
46903	47843	__ai __attribute__((target("neon"))) uint32x4_t vcgtzq_s32(int32x4_t __p0) {
46904	47844	uint32x4_t __ret;
46905		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
46906		__ret = (uint32x4_t) __builtin_neon_vcgtzq_v((int8x16_t)__rev0, 50);
46907		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	47845	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	47846	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcgtzq_v(__builtin_bit_cast(int8x16_t, __rev0), 34));
	47847	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
46908	47848	return __ret;
46909	47849	}
46910	47850	#endif
...	...	@@ -46912,15 +47852,15 @@ __ai __attribute__((target("neon"))) uint32x4_t vcgtzq_s32(int32x4_t __p0) {
46912	47852	#ifdef __LITTLE_ENDIAN__
46913	47853	__ai __attribute__((target("neon"))) uint64x2_t vcgtzq_s64(int64x2_t __p0) {
46914	47854	uint64x2_t __ret;
46915		__ret = (uint64x2_t) __builtin_neon_vcgtzq_v((int8x16_t)__p0, 51);
	47855	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcgtzq_v(__builtin_bit_cast(int8x16_t, __p0), 35));
46916	47856	return __ret;
46917	47857	}
46918	47858	#else
46919	47859	__ai __attribute__((target("neon"))) uint64x2_t vcgtzq_s64(int64x2_t __p0) {
46920	47860	uint64x2_t __ret;
46921		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
46922		__ret = (uint64x2_t) __builtin_neon_vcgtzq_v((int8x16_t)__rev0, 51);
46923		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	47861	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	47862	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcgtzq_v(__builtin_bit_cast(int8x16_t, __rev0), 35));
	47863	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
46924	47864	return __ret;
46925	47865	}
46926	47866	#endif
...	...	@@ -46928,15 +47868,15 @@ __ai __attribute__((target("neon"))) uint64x2_t vcgtzq_s64(int64x2_t __p0) {
46928	47868	#ifdef __LITTLE_ENDIAN__
46929	47869	__ai __attribute__((target("neon"))) uint16x8_t vcgtzq_s16(int16x8_t __p0) {
46930	47870	uint16x8_t __ret;
46931		__ret = (uint16x8_t) __builtin_neon_vcgtzq_v((int8x16_t)__p0, 49);
	47871	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcgtzq_v(__builtin_bit_cast(int8x16_t, __p0), 33));
46932	47872	return __ret;
46933	47873	}
46934	47874	#else
46935	47875	__ai __attribute__((target("neon"))) uint16x8_t vcgtzq_s16(int16x8_t __p0) {
46936	47876	uint16x8_t __ret;
46937		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
46938		__ret = (uint16x8_t) __builtin_neon_vcgtzq_v((int8x16_t)__rev0, 49);
46939		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	47877	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	47878	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcgtzq_v(__builtin_bit_cast(int8x16_t, __rev0), 33));
	47879	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
46940	47880	return __ret;
46941	47881	}
46942	47882	#endif
...	...	@@ -46944,36 +47884,36 @@ __ai __attribute__((target("neon"))) uint16x8_t vcgtzq_s16(int16x8_t __p0) {
46944	47884	#ifdef __LITTLE_ENDIAN__
46945	47885	__ai __attribute__((target("neon"))) uint8x8_t vcgtz_s8(int8x8_t __p0) {
46946	47886	uint8x8_t __ret;
46947		__ret = (uint8x8_t) __builtin_neon_vcgtz_v((int8x8_t)__p0, 16);
	47887	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vcgtz_v(__builtin_bit_cast(int8x8_t, __p0), 0));
46948	47888	return __ret;
46949	47889	}
46950	47890	#else
46951	47891	__ai __attribute__((target("neon"))) uint8x8_t vcgtz_s8(int8x8_t __p0) {
46952	47892	uint8x8_t __ret;
46953		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
46954		__ret = (uint8x8_t) __builtin_neon_vcgtz_v((int8x8_t)__rev0, 16);
46955		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	47893	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	47894	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vcgtz_v(__builtin_bit_cast(int8x8_t, __rev0), 0));
	47895	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
46956	47896	return __ret;
46957	47897	}
46958	47898	#endif
46959	47899
46960	47900	__ai __attribute__((target("neon"))) uint64x1_t vcgtz_f64(float64x1_t __p0) {
46961	47901	uint64x1_t __ret;
46962		__ret = (uint64x1_t) __builtin_neon_vcgtz_v((int8x8_t)__p0, 19);
	47902	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vcgtz_v(__builtin_bit_cast(int8x8_t, __p0), 10));
46963	47903	return __ret;
46964	47904	}
46965	47905	#ifdef __LITTLE_ENDIAN__
46966	47906	__ai __attribute__((target("neon"))) uint32x2_t vcgtz_f32(float32x2_t __p0) {
46967	47907	uint32x2_t __ret;
46968		__ret = (uint32x2_t) __builtin_neon_vcgtz_v((int8x8_t)__p0, 18);
	47908	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcgtz_v(__builtin_bit_cast(int8x8_t, __p0), 9));
46969	47909	return __ret;
46970	47910	}
46971	47911	#else
46972	47912	__ai __attribute__((target("neon"))) uint32x2_t vcgtz_f32(float32x2_t __p0) {
46973	47913	uint32x2_t __ret;
46974		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
46975		__ret = (uint32x2_t) __builtin_neon_vcgtz_v((int8x8_t)__rev0, 18);
46976		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	47914	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	47915	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcgtz_v(__builtin_bit_cast(int8x8_t, __rev0), 9));
	47916	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
46977	47917	return __ret;
46978	47918	}
46979	47919	#endif
...	...	@@ -46981,68 +47921,68 @@ __ai __attribute__((target("neon"))) uint32x2_t vcgtz_f32(float32x2_t __p0) {
46981	47921	#ifdef __LITTLE_ENDIAN__
46982	47922	__ai __attribute__((target("neon"))) uint32x2_t vcgtz_s32(int32x2_t __p0) {
46983	47923	uint32x2_t __ret;
46984		__ret = (uint32x2_t) __builtin_neon_vcgtz_v((int8x8_t)__p0, 18);
	47924	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcgtz_v(__builtin_bit_cast(int8x8_t, __p0), 2));
46985	47925	return __ret;
46986	47926	}
46987	47927	#else
46988	47928	__ai __attribute__((target("neon"))) uint32x2_t vcgtz_s32(int32x2_t __p0) {
46989	47929	uint32x2_t __ret;
46990		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
46991		__ret = (uint32x2_t) __builtin_neon_vcgtz_v((int8x8_t)__rev0, 18);
46992		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	47930	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	47931	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcgtz_v(__builtin_bit_cast(int8x8_t, __rev0), 2));
	47932	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
46993	47933	return __ret;
46994	47934	}
46995	47935	#endif
46996	47936
46997	47937	__ai __attribute__((target("neon"))) uint64x1_t vcgtz_s64(int64x1_t __p0) {
46998	47938	uint64x1_t __ret;
46999		__ret = (uint64x1_t) __builtin_neon_vcgtz_v((int8x8_t)__p0, 19);
	47939	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vcgtz_v(__builtin_bit_cast(int8x8_t, __p0), 3));
47000	47940	return __ret;
47001	47941	}
47002	47942	#ifdef __LITTLE_ENDIAN__
47003	47943	__ai __attribute__((target("neon"))) uint16x4_t vcgtz_s16(int16x4_t __p0) {
47004	47944	uint16x4_t __ret;
47005		__ret = (uint16x4_t) __builtin_neon_vcgtz_v((int8x8_t)__p0, 17);
	47945	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcgtz_v(__builtin_bit_cast(int8x8_t, __p0), 1));
47006	47946	return __ret;
47007	47947	}
47008	47948	#else
47009	47949	__ai __attribute__((target("neon"))) uint16x4_t vcgtz_s16(int16x4_t __p0) {
47010	47950	uint16x4_t __ret;
47011		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
47012		__ret = (uint16x4_t) __builtin_neon_vcgtz_v((int8x8_t)__rev0, 17);
47013		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	47951	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	47952	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcgtz_v(__builtin_bit_cast(int8x8_t, __rev0), 1));
	47953	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
47014	47954	return __ret;
47015	47955	}
47016	47956	#endif
47017	47957
47018	47958	__ai __attribute__((target("neon"))) uint64_t vcgtzd_s64(int64_t __p0) {
47019	47959	uint64_t __ret;
47020		__ret = (uint64_t) __builtin_neon_vcgtzd_s64(__p0);
	47960	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcgtzd_s64(__p0));
47021	47961	return __ret;
47022	47962	}
47023	47963	__ai __attribute__((target("neon"))) uint64_t vcgtzd_f64(float64_t __p0) {
47024	47964	uint64_t __ret;
47025		__ret = (uint64_t) __builtin_neon_vcgtzd_f64(__p0);
	47965	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcgtzd_f64(__p0));
47026	47966	return __ret;
47027	47967	}
47028	47968	__ai __attribute__((target("neon"))) uint32_t vcgtzs_f32(float32_t __p0) {
47029	47969	uint32_t __ret;
47030		__ret = (uint32_t) __builtin_neon_vcgtzs_f32(__p0);
	47970	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vcgtzs_f32(__p0));
47031	47971	return __ret;
47032	47972	}
47033	47973	#ifdef __LITTLE_ENDIAN__
47034	47974	__ai __attribute__((target("neon"))) uint64x2_t vcleq_u64(uint64x2_t __p0, uint64x2_t __p1) {
47035	47975	uint64x2_t __ret;
47036		__ret = (uint64x2_t)(__p0 <= __p1);
	47976	__ret = __builtin_bit_cast(uint64x2_t, __p0 <= __p1);
47037	47977	return __ret;
47038	47978	}
47039	47979	#else
47040	47980	__ai __attribute__((target("neon"))) uint64x2_t vcleq_u64(uint64x2_t __p0, uint64x2_t __p1) {
47041	47981	uint64x2_t __ret;
47042		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
47043		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
47044		__ret = (uint64x2_t)(__rev0 <= __rev1);
47045		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	47982	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	47983	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	47984	__ret = __builtin_bit_cast(uint64x2_t, __rev0 <= __rev1);
	47985	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
47046	47986	return __ret;
47047	47987	}
47048	47988	#endif
...	...	@@ -47050,16 +47990,16 @@ __ai __attribute__((target("neon"))) uint64x2_t vcleq_u64(uint64x2_t __p0, uint6
47050	47990	#ifdef __LITTLE_ENDIAN__
47051	47991	__ai __attribute__((target("neon"))) uint64x2_t vcleq_f64(float64x2_t __p0, float64x2_t __p1) {
47052	47992	uint64x2_t __ret;
47053		__ret = (uint64x2_t)(__p0 <= __p1);
	47993	__ret = __builtin_bit_cast(uint64x2_t, __p0 <= __p1);
47054	47994	return __ret;
47055	47995	}
47056	47996	#else
47057	47997	__ai __attribute__((target("neon"))) uint64x2_t vcleq_f64(float64x2_t __p0, float64x2_t __p1) {
47058	47998	uint64x2_t __ret;
47059		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
47060		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
47061		__ret = (uint64x2_t)(__rev0 <= __rev1);
47062		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	47999	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	48000	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	48001	__ret = __builtin_bit_cast(uint64x2_t, __rev0 <= __rev1);
	48002	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
47063	48003	return __ret;
47064	48004	}
47065	48005	#endif
...	...	@@ -47067,67 +48007,67 @@ __ai __attribute__((target("neon"))) uint64x2_t vcleq_f64(float64x2_t __p0, floa
47067	48007	#ifdef __LITTLE_ENDIAN__
47068	48008	__ai __attribute__((target("neon"))) uint64x2_t vcleq_s64(int64x2_t __p0, int64x2_t __p1) {
47069	48009	uint64x2_t __ret;
47070		__ret = (uint64x2_t)(__p0 <= __p1);
	48010	__ret = __builtin_bit_cast(uint64x2_t, __p0 <= __p1);
47071	48011	return __ret;
47072	48012	}
47073	48013	#else
47074	48014	__ai __attribute__((target("neon"))) uint64x2_t vcleq_s64(int64x2_t __p0, int64x2_t __p1) {
47075	48015	uint64x2_t __ret;
47076		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
47077		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
47078		__ret = (uint64x2_t)(__rev0 <= __rev1);
47079		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	48016	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	48017	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	48018	__ret = __builtin_bit_cast(uint64x2_t, __rev0 <= __rev1);
	48019	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
47080	48020	return __ret;
47081	48021	}
47082	48022	#endif
47083	48023
47084	48024	__ai __attribute__((target("neon"))) uint64x1_t vcle_u64(uint64x1_t __p0, uint64x1_t __p1) {
47085	48025	uint64x1_t __ret;
47086		__ret = (uint64x1_t)(__p0 <= __p1);
	48026	__ret = __builtin_bit_cast(uint64x1_t, __p0 <= __p1);
47087	48027	return __ret;
47088	48028	}
47089	48029	__ai __attribute__((target("neon"))) uint64x1_t vcle_f64(float64x1_t __p0, float64x1_t __p1) {
47090	48030	uint64x1_t __ret;
47091		__ret = (uint64x1_t)(__p0 <= __p1);
	48031	__ret = __builtin_bit_cast(uint64x1_t, __p0 <= __p1);
47092	48032	return __ret;
47093	48033	}
47094	48034	__ai __attribute__((target("neon"))) uint64x1_t vcle_s64(int64x1_t __p0, int64x1_t __p1) {
47095	48035	uint64x1_t __ret;
47096		__ret = (uint64x1_t)(__p0 <= __p1);
	48036	__ret = __builtin_bit_cast(uint64x1_t, __p0 <= __p1);
47097	48037	return __ret;
47098	48038	}
47099	48039	__ai __attribute__((target("neon"))) uint64_t vcled_u64(uint64_t __p0, uint64_t __p1) {
47100	48040	uint64_t __ret;
47101		__ret = (uint64_t) __builtin_neon_vcled_u64(__p0, __p1);
	48041	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcled_u64(__p0, __p1));
47102	48042	return __ret;
47103	48043	}
47104	48044	__ai __attribute__((target("neon"))) uint64_t vcled_s64(int64_t __p0, int64_t __p1) {
47105	48045	uint64_t __ret;
47106		__ret = (uint64_t) __builtin_neon_vcled_s64(__p0, __p1);
	48046	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcled_s64(__p0, __p1));
47107	48047	return __ret;
47108	48048	}
47109	48049	__ai __attribute__((target("neon"))) uint64_t vcled_f64(float64_t __p0, float64_t __p1) {
47110	48050	uint64_t __ret;
47111		__ret = (uint64_t) __builtin_neon_vcled_f64(__p0, __p1);
	48051	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcled_f64(__p0, __p1));
47112	48052	return __ret;
47113	48053	}
47114	48054	__ai __attribute__((target("neon"))) uint32_t vcles_f32(float32_t __p0, float32_t __p1) {
47115	48055	uint32_t __ret;
47116		__ret = (uint32_t) __builtin_neon_vcles_f32(__p0, __p1);
	48056	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vcles_f32(__p0, __p1));
47117	48057	return __ret;
47118	48058	}
47119	48059	#ifdef __LITTLE_ENDIAN__
47120	48060	__ai __attribute__((target("neon"))) uint8x16_t vclezq_s8(int8x16_t __p0) {
47121	48061	uint8x16_t __ret;
47122		__ret = (uint8x16_t) __builtin_neon_vclezq_v((int8x16_t)__p0, 48);
	48062	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vclezq_v(__builtin_bit_cast(int8x16_t, __p0), 32));
47123	48063	return __ret;
47124	48064	}
47125	48065	#else
47126	48066	__ai __attribute__((target("neon"))) uint8x16_t vclezq_s8(int8x16_t __p0) {
47127	48067	uint8x16_t __ret;
47128		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
47129		__ret = (uint8x16_t) __builtin_neon_vclezq_v((int8x16_t)__rev0, 48);
47130		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	48068	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	48069	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vclezq_v(__builtin_bit_cast(int8x16_t, __rev0), 32));
	48070	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
47131	48071	return __ret;
47132	48072	}
47133	48073	#endif
...	...	@@ -47135,15 +48075,15 @@ __ai __attribute__((target("neon"))) uint8x16_t vclezq_s8(int8x16_t __p0) {
47135	48075	#ifdef __LITTLE_ENDIAN__
47136	48076	__ai __attribute__((target("neon"))) uint64x2_t vclezq_f64(float64x2_t __p0) {
47137	48077	uint64x2_t __ret;
47138		__ret = (uint64x2_t) __builtin_neon_vclezq_v((int8x16_t)__p0, 51);
	48078	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vclezq_v(__builtin_bit_cast(int8x16_t, __p0), 42));
47139	48079	return __ret;
47140	48080	}
47141	48081	#else
47142	48082	__ai __attribute__((target("neon"))) uint64x2_t vclezq_f64(float64x2_t __p0) {
47143	48083	uint64x2_t __ret;
47144		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
47145		__ret = (uint64x2_t) __builtin_neon_vclezq_v((int8x16_t)__rev0, 51);
47146		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	48084	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	48085	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vclezq_v(__builtin_bit_cast(int8x16_t, __rev0), 42));
	48086	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
47147	48087	return __ret;
47148	48088	}
47149	48089	#endif
...	...	@@ -47151,15 +48091,15 @@ __ai __attribute__((target("neon"))) uint64x2_t vclezq_f64(float64x2_t __p0) {
47151	48091	#ifdef __LITTLE_ENDIAN__
47152	48092	__ai __attribute__((target("neon"))) uint32x4_t vclezq_f32(float32x4_t __p0) {
47153	48093	uint32x4_t __ret;
47154		__ret = (uint32x4_t) __builtin_neon_vclezq_v((int8x16_t)__p0, 50);
	48094	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vclezq_v(__builtin_bit_cast(int8x16_t, __p0), 41));
47155	48095	return __ret;
47156	48096	}
47157	48097	#else
47158	48098	__ai __attribute__((target("neon"))) uint32x4_t vclezq_f32(float32x4_t __p0) {
47159	48099	uint32x4_t __ret;
47160		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
47161		__ret = (uint32x4_t) __builtin_neon_vclezq_v((int8x16_t)__rev0, 50);
47162		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	48100	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	48101	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vclezq_v(__builtin_bit_cast(int8x16_t, __rev0), 41));
	48102	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
47163	48103	return __ret;
47164	48104	}
47165	48105	#endif
...	...	@@ -47167,15 +48107,15 @@ __ai __attribute__((target("neon"))) uint32x4_t vclezq_f32(float32x4_t __p0) {
47167	48107	#ifdef __LITTLE_ENDIAN__
47168	48108	__ai __attribute__((target("neon"))) uint32x4_t vclezq_s32(int32x4_t __p0) {
47169	48109	uint32x4_t __ret;
47170		__ret = (uint32x4_t) __builtin_neon_vclezq_v((int8x16_t)__p0, 50);
	48110	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vclezq_v(__builtin_bit_cast(int8x16_t, __p0), 34));
47171	48111	return __ret;
47172	48112	}
47173	48113	#else
47174	48114	__ai __attribute__((target("neon"))) uint32x4_t vclezq_s32(int32x4_t __p0) {
47175	48115	uint32x4_t __ret;
47176		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
47177		__ret = (uint32x4_t) __builtin_neon_vclezq_v((int8x16_t)__rev0, 50);
47178		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	48116	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	48117	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vclezq_v(__builtin_bit_cast(int8x16_t, __rev0), 34));
	48118	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
47179	48119	return __ret;
47180	48120	}
47181	48121	#endif
...	...	@@ -47183,15 +48123,15 @@ __ai __attribute__((target("neon"))) uint32x4_t vclezq_s32(int32x4_t __p0) {
47183	48123	#ifdef __LITTLE_ENDIAN__
47184	48124	__ai __attribute__((target("neon"))) uint64x2_t vclezq_s64(int64x2_t __p0) {
47185	48125	uint64x2_t __ret;
47186		__ret = (uint64x2_t) __builtin_neon_vclezq_v((int8x16_t)__p0, 51);
	48126	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vclezq_v(__builtin_bit_cast(int8x16_t, __p0), 35));
47187	48127	return __ret;
47188	48128	}
47189	48129	#else
47190	48130	__ai __attribute__((target("neon"))) uint64x2_t vclezq_s64(int64x2_t __p0) {
47191	48131	uint64x2_t __ret;
47192		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
47193		__ret = (uint64x2_t) __builtin_neon_vclezq_v((int8x16_t)__rev0, 51);
47194		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	48132	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	48133	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vclezq_v(__builtin_bit_cast(int8x16_t, __rev0), 35));
	48134	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
47195	48135	return __ret;
47196	48136	}
47197	48137	#endif
...	...	@@ -47199,15 +48139,15 @@ __ai __attribute__((target("neon"))) uint64x2_t vclezq_s64(int64x2_t __p0) {
47199	48139	#ifdef __LITTLE_ENDIAN__
47200	48140	__ai __attribute__((target("neon"))) uint16x8_t vclezq_s16(int16x8_t __p0) {
47201	48141	uint16x8_t __ret;
47202		__ret = (uint16x8_t) __builtin_neon_vclezq_v((int8x16_t)__p0, 49);
	48142	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vclezq_v(__builtin_bit_cast(int8x16_t, __p0), 33));
47203	48143	return __ret;
47204	48144	}
47205	48145	#else
47206	48146	__ai __attribute__((target("neon"))) uint16x8_t vclezq_s16(int16x8_t __p0) {
47207	48147	uint16x8_t __ret;
47208		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
47209		__ret = (uint16x8_t) __builtin_neon_vclezq_v((int8x16_t)__rev0, 49);
47210		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	48148	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	48149	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vclezq_v(__builtin_bit_cast(int8x16_t, __rev0), 33));
	48150	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
47211	48151	return __ret;
47212	48152	}
47213	48153	#endif
...	...	@@ -47215,36 +48155,36 @@ __ai __attribute__((target("neon"))) uint16x8_t vclezq_s16(int16x8_t __p0) {
47215	48155	#ifdef __LITTLE_ENDIAN__
47216	48156	__ai __attribute__((target("neon"))) uint8x8_t vclez_s8(int8x8_t __p0) {
47217	48157	uint8x8_t __ret;
47218		__ret = (uint8x8_t) __builtin_neon_vclez_v((int8x8_t)__p0, 16);
	48158	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vclez_v(__builtin_bit_cast(int8x8_t, __p0), 0));
47219	48159	return __ret;
47220	48160	}
47221	48161	#else
47222	48162	__ai __attribute__((target("neon"))) uint8x8_t vclez_s8(int8x8_t __p0) {
47223	48163	uint8x8_t __ret;
47224		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
47225		__ret = (uint8x8_t) __builtin_neon_vclez_v((int8x8_t)__rev0, 16);
47226		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	48164	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	48165	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vclez_v(__builtin_bit_cast(int8x8_t, __rev0), 0));
	48166	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
47227	48167	return __ret;
47228	48168	}
47229	48169	#endif
47230	48170
47231	48171	__ai __attribute__((target("neon"))) uint64x1_t vclez_f64(float64x1_t __p0) {
47232	48172	uint64x1_t __ret;
47233		__ret = (uint64x1_t) __builtin_neon_vclez_v((int8x8_t)__p0, 19);
	48173	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vclez_v(__builtin_bit_cast(int8x8_t, __p0), 10));
47234	48174	return __ret;
47235	48175	}
47236	48176	#ifdef __LITTLE_ENDIAN__
47237	48177	__ai __attribute__((target("neon"))) uint32x2_t vclez_f32(float32x2_t __p0) {
47238	48178	uint32x2_t __ret;
47239		__ret = (uint32x2_t) __builtin_neon_vclez_v((int8x8_t)__p0, 18);
	48179	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vclez_v(__builtin_bit_cast(int8x8_t, __p0), 9));
47240	48180	return __ret;
47241	48181	}
47242	48182	#else
47243	48183	__ai __attribute__((target("neon"))) uint32x2_t vclez_f32(float32x2_t __p0) {
47244	48184	uint32x2_t __ret;
47245		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
47246		__ret = (uint32x2_t) __builtin_neon_vclez_v((int8x8_t)__rev0, 18);
47247		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	48185	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	48186	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vclez_v(__builtin_bit_cast(int8x8_t, __rev0), 9));
	48187	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
47248	48188	return __ret;
47249	48189	}
47250	48190	#endif
...	...	@@ -47252,68 +48192,68 @@ __ai __attribute__((target("neon"))) uint32x2_t vclez_f32(float32x2_t __p0) {
47252	48192	#ifdef __LITTLE_ENDIAN__
47253	48193	__ai __attribute__((target("neon"))) uint32x2_t vclez_s32(int32x2_t __p0) {
47254	48194	uint32x2_t __ret;
47255		__ret = (uint32x2_t) __builtin_neon_vclez_v((int8x8_t)__p0, 18);
	48195	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vclez_v(__builtin_bit_cast(int8x8_t, __p0), 2));
47256	48196	return __ret;
47257	48197	}
47258	48198	#else
47259	48199	__ai __attribute__((target("neon"))) uint32x2_t vclez_s32(int32x2_t __p0) {
47260	48200	uint32x2_t __ret;
47261		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
47262		__ret = (uint32x2_t) __builtin_neon_vclez_v((int8x8_t)__rev0, 18);
47263		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	48201	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	48202	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vclez_v(__builtin_bit_cast(int8x8_t, __rev0), 2));
	48203	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
47264	48204	return __ret;
47265	48205	}
47266	48206	#endif
47267	48207
47268	48208	__ai __attribute__((target("neon"))) uint64x1_t vclez_s64(int64x1_t __p0) {
47269	48209	uint64x1_t __ret;
47270		__ret = (uint64x1_t) __builtin_neon_vclez_v((int8x8_t)__p0, 19);
	48210	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vclez_v(__builtin_bit_cast(int8x8_t, __p0), 3));
47271	48211	return __ret;
47272	48212	}
47273	48213	#ifdef __LITTLE_ENDIAN__
47274	48214	__ai __attribute__((target("neon"))) uint16x4_t vclez_s16(int16x4_t __p0) {
47275	48215	uint16x4_t __ret;
47276		__ret = (uint16x4_t) __builtin_neon_vclez_v((int8x8_t)__p0, 17);
	48216	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vclez_v(__builtin_bit_cast(int8x8_t, __p0), 1));
47277	48217	return __ret;
47278	48218	}
47279	48219	#else
47280	48220	__ai __attribute__((target("neon"))) uint16x4_t vclez_s16(int16x4_t __p0) {
47281	48221	uint16x4_t __ret;
47282		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
47283		__ret = (uint16x4_t) __builtin_neon_vclez_v((int8x8_t)__rev0, 17);
47284		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	48222	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	48223	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vclez_v(__builtin_bit_cast(int8x8_t, __rev0), 1));
	48224	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
47285	48225	return __ret;
47286	48226	}
47287	48227	#endif
47288	48228
47289	48229	__ai __attribute__((target("neon"))) uint64_t vclezd_s64(int64_t __p0) {
47290	48230	uint64_t __ret;
47291		__ret = (uint64_t) __builtin_neon_vclezd_s64(__p0);
	48231	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vclezd_s64(__p0));
47292	48232	return __ret;
47293	48233	}
47294	48234	__ai __attribute__((target("neon"))) uint64_t vclezd_f64(float64_t __p0) {
47295	48235	uint64_t __ret;
47296		__ret = (uint64_t) __builtin_neon_vclezd_f64(__p0);
	48236	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vclezd_f64(__p0));
47297	48237	return __ret;
47298	48238	}
47299	48239	__ai __attribute__((target("neon"))) uint32_t vclezs_f32(float32_t __p0) {
47300	48240	uint32_t __ret;
47301		__ret = (uint32_t) __builtin_neon_vclezs_f32(__p0);
	48241	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vclezs_f32(__p0));
47302	48242	return __ret;
47303	48243	}
47304	48244	#ifdef __LITTLE_ENDIAN__
47305	48245	__ai __attribute__((target("neon"))) uint64x2_t vcltq_u64(uint64x2_t __p0, uint64x2_t __p1) {
47306	48246	uint64x2_t __ret;
47307		__ret = (uint64x2_t)(__p0 < __p1);
	48247	__ret = __builtin_bit_cast(uint64x2_t, __p0 < __p1);
47308	48248	return __ret;
47309	48249	}
47310	48250	#else
47311	48251	__ai __attribute__((target("neon"))) uint64x2_t vcltq_u64(uint64x2_t __p0, uint64x2_t __p1) {
47312	48252	uint64x2_t __ret;
47313		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
47314		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
47315		__ret = (uint64x2_t)(__rev0 < __rev1);
47316		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	48253	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	48254	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	48255	__ret = __builtin_bit_cast(uint64x2_t, __rev0 < __rev1);
	48256	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
47317	48257	return __ret;
47318	48258	}
47319	48259	#endif
...	...	@@ -47321,16 +48261,16 @@ __ai __attribute__((target("neon"))) uint64x2_t vcltq_u64(uint64x2_t __p0, uint6
47321	48261	#ifdef __LITTLE_ENDIAN__
47322	48262	__ai __attribute__((target("neon"))) uint64x2_t vcltq_f64(float64x2_t __p0, float64x2_t __p1) {
47323	48263	uint64x2_t __ret;
47324		__ret = (uint64x2_t)(__p0 < __p1);
	48264	__ret = __builtin_bit_cast(uint64x2_t, __p0 < __p1);
47325	48265	return __ret;
47326	48266	}
47327	48267	#else
47328	48268	__ai __attribute__((target("neon"))) uint64x2_t vcltq_f64(float64x2_t __p0, float64x2_t __p1) {
47329	48269	uint64x2_t __ret;
47330		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
47331		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
47332		__ret = (uint64x2_t)(__rev0 < __rev1);
47333		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	48270	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	48271	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	48272	__ret = __builtin_bit_cast(uint64x2_t, __rev0 < __rev1);
	48273	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
47334	48274	return __ret;
47335	48275	}
47336	48276	#endif
...	...	@@ -47338,67 +48278,67 @@ __ai __attribute__((target("neon"))) uint64x2_t vcltq_f64(float64x2_t __p0, floa
47338	48278	#ifdef __LITTLE_ENDIAN__
47339	48279	__ai __attribute__((target("neon"))) uint64x2_t vcltq_s64(int64x2_t __p0, int64x2_t __p1) {
47340	48280	uint64x2_t __ret;
47341		__ret = (uint64x2_t)(__p0 < __p1);
	48281	__ret = __builtin_bit_cast(uint64x2_t, __p0 < __p1);
47342	48282	return __ret;
47343	48283	}
47344	48284	#else
47345	48285	__ai __attribute__((target("neon"))) uint64x2_t vcltq_s64(int64x2_t __p0, int64x2_t __p1) {
47346	48286	uint64x2_t __ret;
47347		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
47348		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
47349		__ret = (uint64x2_t)(__rev0 < __rev1);
47350		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	48287	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	48288	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	48289	__ret = __builtin_bit_cast(uint64x2_t, __rev0 < __rev1);
	48290	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
47351	48291	return __ret;
47352	48292	}
47353	48293	#endif
47354	48294
47355	48295	__ai __attribute__((target("neon"))) uint64x1_t vclt_u64(uint64x1_t __p0, uint64x1_t __p1) {
47356	48296	uint64x1_t __ret;
47357		__ret = (uint64x1_t)(__p0 < __p1);
	48297	__ret = __builtin_bit_cast(uint64x1_t, __p0 < __p1);
47358	48298	return __ret;
47359	48299	}
47360	48300	__ai __attribute__((target("neon"))) uint64x1_t vclt_f64(float64x1_t __p0, float64x1_t __p1) {
47361	48301	uint64x1_t __ret;
47362		__ret = (uint64x1_t)(__p0 < __p1);
	48302	__ret = __builtin_bit_cast(uint64x1_t, __p0 < __p1);
47363	48303	return __ret;
47364	48304	}
47365	48305	__ai __attribute__((target("neon"))) uint64x1_t vclt_s64(int64x1_t __p0, int64x1_t __p1) {
47366	48306	uint64x1_t __ret;
47367		__ret = (uint64x1_t)(__p0 < __p1);
	48307	__ret = __builtin_bit_cast(uint64x1_t, __p0 < __p1);
47368	48308	return __ret;
47369	48309	}
47370	48310	__ai __attribute__((target("neon"))) uint64_t vcltd_u64(uint64_t __p0, uint64_t __p1) {
47371	48311	uint64_t __ret;
47372		__ret = (uint64_t) __builtin_neon_vcltd_u64(__p0, __p1);
	48312	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcltd_u64(__p0, __p1));
47373	48313	return __ret;
47374	48314	}
47375	48315	__ai __attribute__((target("neon"))) uint64_t vcltd_s64(int64_t __p0, int64_t __p1) {
47376	48316	uint64_t __ret;
47377		__ret = (uint64_t) __builtin_neon_vcltd_s64(__p0, __p1);
	48317	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcltd_s64(__p0, __p1));
47378	48318	return __ret;
47379	48319	}
47380	48320	__ai __attribute__((target("neon"))) uint64_t vcltd_f64(float64_t __p0, float64_t __p1) {
47381	48321	uint64_t __ret;
47382		__ret = (uint64_t) __builtin_neon_vcltd_f64(__p0, __p1);
	48322	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcltd_f64(__p0, __p1));
47383	48323	return __ret;
47384	48324	}
47385	48325	__ai __attribute__((target("neon"))) uint32_t vclts_f32(float32_t __p0, float32_t __p1) {
47386	48326	uint32_t __ret;
47387		__ret = (uint32_t) __builtin_neon_vclts_f32(__p0, __p1);
	48327	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vclts_f32(__p0, __p1));
47388	48328	return __ret;
47389	48329	}
47390	48330	#ifdef __LITTLE_ENDIAN__
47391	48331	__ai __attribute__((target("neon"))) uint8x16_t vcltzq_s8(int8x16_t __p0) {
47392	48332	uint8x16_t __ret;
47393		__ret = (uint8x16_t) __builtin_neon_vcltzq_v((int8x16_t)__p0, 48);
	48333	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vcltzq_v(__builtin_bit_cast(int8x16_t, __p0), 32));
47394	48334	return __ret;
47395	48335	}
47396	48336	#else
47397	48337	__ai __attribute__((target("neon"))) uint8x16_t vcltzq_s8(int8x16_t __p0) {
47398	48338	uint8x16_t __ret;
47399		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
47400		__ret = (uint8x16_t) __builtin_neon_vcltzq_v((int8x16_t)__rev0, 48);
47401		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	48339	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	48340	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vcltzq_v(__builtin_bit_cast(int8x16_t, __rev0), 32));
	48341	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
47402	48342	return __ret;
47403	48343	}
47404	48344	#endif
...	...	@@ -47406,15 +48346,15 @@ __ai __attribute__((target("neon"))) uint8x16_t vcltzq_s8(int8x16_t __p0) {
47406	48346	#ifdef __LITTLE_ENDIAN__
47407	48347	__ai __attribute__((target("neon"))) uint64x2_t vcltzq_f64(float64x2_t __p0) {
47408	48348	uint64x2_t __ret;
47409		__ret = (uint64x2_t) __builtin_neon_vcltzq_v((int8x16_t)__p0, 51);
	48349	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcltzq_v(__builtin_bit_cast(int8x16_t, __p0), 42));
47410	48350	return __ret;
47411	48351	}
47412	48352	#else
47413	48353	__ai __attribute__((target("neon"))) uint64x2_t vcltzq_f64(float64x2_t __p0) {
47414	48354	uint64x2_t __ret;
47415		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
47416		__ret = (uint64x2_t) __builtin_neon_vcltzq_v((int8x16_t)__rev0, 51);
47417		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	48355	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	48356	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcltzq_v(__builtin_bit_cast(int8x16_t, __rev0), 42));
	48357	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
47418	48358	return __ret;
47419	48359	}
47420	48360	#endif
...	...	@@ -47422,15 +48362,15 @@ __ai __attribute__((target("neon"))) uint64x2_t vcltzq_f64(float64x2_t __p0) {
47422	48362	#ifdef __LITTLE_ENDIAN__
47423	48363	__ai __attribute__((target("neon"))) uint32x4_t vcltzq_f32(float32x4_t __p0) {
47424	48364	uint32x4_t __ret;
47425		__ret = (uint32x4_t) __builtin_neon_vcltzq_v((int8x16_t)__p0, 50);
	48365	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcltzq_v(__builtin_bit_cast(int8x16_t, __p0), 41));
47426	48366	return __ret;
47427	48367	}
47428	48368	#else
47429	48369	__ai __attribute__((target("neon"))) uint32x4_t vcltzq_f32(float32x4_t __p0) {
47430	48370	uint32x4_t __ret;
47431		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
47432		__ret = (uint32x4_t) __builtin_neon_vcltzq_v((int8x16_t)__rev0, 50);
47433		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	48371	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	48372	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcltzq_v(__builtin_bit_cast(int8x16_t, __rev0), 41));
	48373	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
47434	48374	return __ret;
47435	48375	}
47436	48376	#endif
...	...	@@ -47438,15 +48378,15 @@ __ai __attribute__((target("neon"))) uint32x4_t vcltzq_f32(float32x4_t __p0) {
47438	48378	#ifdef __LITTLE_ENDIAN__
47439	48379	__ai __attribute__((target("neon"))) uint32x4_t vcltzq_s32(int32x4_t __p0) {
47440	48380	uint32x4_t __ret;
47441		__ret = (uint32x4_t) __builtin_neon_vcltzq_v((int8x16_t)__p0, 50);
	48381	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcltzq_v(__builtin_bit_cast(int8x16_t, __p0), 34));
47442	48382	return __ret;
47443	48383	}
47444	48384	#else
47445	48385	__ai __attribute__((target("neon"))) uint32x4_t vcltzq_s32(int32x4_t __p0) {
47446	48386	uint32x4_t __ret;
47447		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
47448		__ret = (uint32x4_t) __builtin_neon_vcltzq_v((int8x16_t)__rev0, 50);
47449		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	48387	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	48388	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vcltzq_v(__builtin_bit_cast(int8x16_t, __rev0), 34));
	48389	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
47450	48390	return __ret;
47451	48391	}
47452	48392	#endif
...	...	@@ -47454,15 +48394,15 @@ __ai __attribute__((target("neon"))) uint32x4_t vcltzq_s32(int32x4_t __p0) {
47454	48394	#ifdef __LITTLE_ENDIAN__
47455	48395	__ai __attribute__((target("neon"))) uint64x2_t vcltzq_s64(int64x2_t __p0) {
47456	48396	uint64x2_t __ret;
47457		__ret = (uint64x2_t) __builtin_neon_vcltzq_v((int8x16_t)__p0, 51);
	48397	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcltzq_v(__builtin_bit_cast(int8x16_t, __p0), 35));
47458	48398	return __ret;
47459	48399	}
47460	48400	#else
47461	48401	__ai __attribute__((target("neon"))) uint64x2_t vcltzq_s64(int64x2_t __p0) {
47462	48402	uint64x2_t __ret;
47463		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
47464		__ret = (uint64x2_t) __builtin_neon_vcltzq_v((int8x16_t)__rev0, 51);
47465		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	48403	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	48404	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcltzq_v(__builtin_bit_cast(int8x16_t, __rev0), 35));
	48405	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
47466	48406	return __ret;
47467	48407	}
47468	48408	#endif
...	...	@@ -47470,15 +48410,15 @@ __ai __attribute__((target("neon"))) uint64x2_t vcltzq_s64(int64x2_t __p0) {
47470	48410	#ifdef __LITTLE_ENDIAN__
47471	48411	__ai __attribute__((target("neon"))) uint16x8_t vcltzq_s16(int16x8_t __p0) {
47472	48412	uint16x8_t __ret;
47473		__ret = (uint16x8_t) __builtin_neon_vcltzq_v((int8x16_t)__p0, 49);
	48413	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcltzq_v(__builtin_bit_cast(int8x16_t, __p0), 33));
47474	48414	return __ret;
47475	48415	}
47476	48416	#else
47477	48417	__ai __attribute__((target("neon"))) uint16x8_t vcltzq_s16(int16x8_t __p0) {
47478	48418	uint16x8_t __ret;
47479		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
47480		__ret = (uint16x8_t) __builtin_neon_vcltzq_v((int8x16_t)__rev0, 49);
47481		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	48419	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	48420	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vcltzq_v(__builtin_bit_cast(int8x16_t, __rev0), 33));
	48421	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
47482	48422	return __ret;
47483	48423	}
47484	48424	#endif
...	...	@@ -47486,36 +48426,36 @@ __ai __attribute__((target("neon"))) uint16x8_t vcltzq_s16(int16x8_t __p0) {
47486	48426	#ifdef __LITTLE_ENDIAN__
47487	48427	__ai __attribute__((target("neon"))) uint8x8_t vcltz_s8(int8x8_t __p0) {
47488	48428	uint8x8_t __ret;
47489		__ret = (uint8x8_t) __builtin_neon_vcltz_v((int8x8_t)__p0, 16);
	48429	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vcltz_v(__builtin_bit_cast(int8x8_t, __p0), 0));
47490	48430	return __ret;
47491	48431	}
47492	48432	#else
47493	48433	__ai __attribute__((target("neon"))) uint8x8_t vcltz_s8(int8x8_t __p0) {
47494	48434	uint8x8_t __ret;
47495		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
47496		__ret = (uint8x8_t) __builtin_neon_vcltz_v((int8x8_t)__rev0, 16);
47497		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	48435	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	48436	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vcltz_v(__builtin_bit_cast(int8x8_t, __rev0), 0));
	48437	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
47498	48438	return __ret;
47499	48439	}
47500	48440	#endif
47501	48441
47502	48442	__ai __attribute__((target("neon"))) uint64x1_t vcltz_f64(float64x1_t __p0) {
47503	48443	uint64x1_t __ret;
47504		__ret = (uint64x1_t) __builtin_neon_vcltz_v((int8x8_t)__p0, 19);
	48444	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vcltz_v(__builtin_bit_cast(int8x8_t, __p0), 10));
47505	48445	return __ret;
47506	48446	}
47507	48447	#ifdef __LITTLE_ENDIAN__
47508	48448	__ai __attribute__((target("neon"))) uint32x2_t vcltz_f32(float32x2_t __p0) {
47509	48449	uint32x2_t __ret;
47510		__ret = (uint32x2_t) __builtin_neon_vcltz_v((int8x8_t)__p0, 18);
	48450	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcltz_v(__builtin_bit_cast(int8x8_t, __p0), 9));
47511	48451	return __ret;
47512	48452	}
47513	48453	#else
47514	48454	__ai __attribute__((target("neon"))) uint32x2_t vcltz_f32(float32x2_t __p0) {
47515	48455	uint32x2_t __ret;
47516		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
47517		__ret = (uint32x2_t) __builtin_neon_vcltz_v((int8x8_t)__rev0, 18);
47518		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	48456	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	48457	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcltz_v(__builtin_bit_cast(int8x8_t, __rev0), 9));
	48458	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
47519	48459	return __ret;
47520	48460	}
47521	48461	#endif
...	...	@@ -47523,53 +48463,53 @@ __ai __attribute__((target("neon"))) uint32x2_t vcltz_f32(float32x2_t __p0) {
47523	48463	#ifdef __LITTLE_ENDIAN__
47524	48464	__ai __attribute__((target("neon"))) uint32x2_t vcltz_s32(int32x2_t __p0) {
47525	48465	uint32x2_t __ret;
47526		__ret = (uint32x2_t) __builtin_neon_vcltz_v((int8x8_t)__p0, 18);
	48466	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcltz_v(__builtin_bit_cast(int8x8_t, __p0), 2));
47527	48467	return __ret;
47528	48468	}
47529	48469	#else
47530	48470	__ai __attribute__((target("neon"))) uint32x2_t vcltz_s32(int32x2_t __p0) {
47531	48471	uint32x2_t __ret;
47532		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
47533		__ret = (uint32x2_t) __builtin_neon_vcltz_v((int8x8_t)__rev0, 18);
47534		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	48472	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	48473	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vcltz_v(__builtin_bit_cast(int8x8_t, __rev0), 2));
	48474	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
47535	48475	return __ret;
47536	48476	}
47537	48477	#endif
47538	48478
47539	48479	__ai __attribute__((target("neon"))) uint64x1_t vcltz_s64(int64x1_t __p0) {
47540	48480	uint64x1_t __ret;
47541		__ret = (uint64x1_t) __builtin_neon_vcltz_v((int8x8_t)__p0, 19);
	48481	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vcltz_v(__builtin_bit_cast(int8x8_t, __p0), 3));
47542	48482	return __ret;
47543	48483	}
47544	48484	#ifdef __LITTLE_ENDIAN__
47545	48485	__ai __attribute__((target("neon"))) uint16x4_t vcltz_s16(int16x4_t __p0) {
47546	48486	uint16x4_t __ret;
47547		__ret = (uint16x4_t) __builtin_neon_vcltz_v((int8x8_t)__p0, 17);
	48487	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcltz_v(__builtin_bit_cast(int8x8_t, __p0), 1));
47548	48488	return __ret;
47549	48489	}
47550	48490	#else
47551	48491	__ai __attribute__((target("neon"))) uint16x4_t vcltz_s16(int16x4_t __p0) {
47552	48492	uint16x4_t __ret;
47553		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
47554		__ret = (uint16x4_t) __builtin_neon_vcltz_v((int8x8_t)__rev0, 17);
47555		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	48493	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	48494	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vcltz_v(__builtin_bit_cast(int8x8_t, __rev0), 1));
	48495	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
47556	48496	return __ret;
47557	48497	}
47558	48498	#endif
47559	48499
47560	48500	__ai __attribute__((target("neon"))) uint64_t vcltzd_s64(int64_t __p0) {
47561	48501	uint64_t __ret;
47562		__ret = (uint64_t) __builtin_neon_vcltzd_s64(__p0);
	48502	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcltzd_s64(__p0));
47563	48503	return __ret;
47564	48504	}
47565	48505	__ai __attribute__((target("neon"))) uint64_t vcltzd_f64(float64_t __p0) {
47566	48506	uint64_t __ret;
47567		__ret = (uint64_t) __builtin_neon_vcltzd_f64(__p0);
	48507	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcltzd_f64(__p0));
47568	48508	return __ret;
47569	48509	}
47570	48510	__ai __attribute__((target("neon"))) uint32_t vcltzs_f32(float32_t __p0) {
47571	48511	uint32_t __ret;
47572		__ret = (uint32_t) __builtin_neon_vcltzs_f32(__p0);
	48512	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vcltzs_f32(__p0));
47573	48513	return __ret;
47574	48514	}
47575	48515	#ifdef __LITTLE_ENDIAN__
...	...	@@ -47582,7 +48522,7 @@ __ai __attribute__((target("neon"))) poly64x2_t vcombine_p64(poly64x1_t __p0, po
47582	48522	__ai __attribute__((target("neon"))) poly64x2_t vcombine_p64(poly64x1_t __p0, poly64x1_t __p1) {
47583	48523	poly64x2_t __ret;
47584	48524	__ret = __builtin_shufflevector(__p0, __p1, 0, 1);
47585		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	48525	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
47586	48526	return __ret;
47587	48527	}
47588	48528	#endif
...	...	@@ -47597,966 +48537,1050 @@ __ai __attribute__((target("neon"))) float64x2_t vcombine_f64(float64x1_t __p0,
47597	48537	__ai __attribute__((target("neon"))) float64x2_t vcombine_f64(float64x1_t __p0, float64x1_t __p1) {
47598	48538	float64x2_t __ret;
47599	48539	__ret = __builtin_shufflevector(__p0, __p1, 0, 1);
47600		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	48540	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
47601	48541	return __ret;
47602	48542	}
47603	48543	#endif
47604	48544
47605	48545	#ifdef __LITTLE_ENDIAN__
47606		#define vcopyq_lane_p8(__p0_278, __p1_278, __p2_278, __p3_278) __extension__ ({ \
47607		poly8x16_t __ret_278; \
47608		poly8x16_t __s0_278 = __p0_278; \
47609		poly8x8_t __s2_278 = __p2_278; \
47610		__ret_278 = vsetq_lane_p8(vget_lane_p8(__s2_278, __p3_278), __s0_278, __p1_278); \
47611		__ret_278; \
47612		})
47613		#else
47614		#define vcopyq_lane_p8(__p0_279, __p1_279, __p2_279, __p3_279) __extension__ ({ \
47615		poly8x16_t __ret_279; \
47616		poly8x16_t __s0_279 = __p0_279; \
47617		poly8x8_t __s2_279 = __p2_279; \
47618		poly8x16_t __rev0_279; __rev0_279 = __builtin_shufflevector(__s0_279, __s0_279, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
47619		poly8x8_t __rev2_279; __rev2_279 = __builtin_shufflevector(__s2_279, __s2_279, 7, 6, 5, 4, 3, 2, 1, 0); \
47620		__ret_279 = __noswap_vsetq_lane_p8(__noswap_vget_lane_p8(__rev2_279, __p3_279), __rev0_279, __p1_279); \
47621		__ret_279 = __builtin_shufflevector(__ret_279, __ret_279, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
47622		__ret_279; \
47623		})
47624		#endif
47625
47626		#ifdef __LITTLE_ENDIAN__
47627		#define vcopyq_lane_p16(__p0_280, __p1_280, __p2_280, __p3_280) __extension__ ({ \
47628		poly16x8_t __ret_280; \
47629		poly16x8_t __s0_280 = __p0_280; \
47630		poly16x4_t __s2_280 = __p2_280; \
47631		__ret_280 = vsetq_lane_p16(vget_lane_p16(__s2_280, __p3_280), __s0_280, __p1_280); \
47632		__ret_280; \
47633		})
47634		#else
47635		#define vcopyq_lane_p16(__p0_281, __p1_281, __p2_281, __p3_281) __extension__ ({ \
47636		poly16x8_t __ret_281; \
47637		poly16x8_t __s0_281 = __p0_281; \
47638		poly16x4_t __s2_281 = __p2_281; \
47639		poly16x8_t __rev0_281; __rev0_281 = __builtin_shufflevector(__s0_281, __s0_281, 7, 6, 5, 4, 3, 2, 1, 0); \
47640		poly16x4_t __rev2_281; __rev2_281 = __builtin_shufflevector(__s2_281, __s2_281, 3, 2, 1, 0); \
47641		__ret_281 = __noswap_vsetq_lane_p16(__noswap_vget_lane_p16(__rev2_281, __p3_281), __rev0_281, __p1_281); \
47642		__ret_281 = __builtin_shufflevector(__ret_281, __ret_281, 7, 6, 5, 4, 3, 2, 1, 0); \
47643		__ret_281; \
47644		})
47645		#endif
47646
47647		#ifdef __LITTLE_ENDIAN__
47648		#define vcopyq_lane_u8(__p0_282, __p1_282, __p2_282, __p3_282) __extension__ ({ \
47649		uint8x16_t __ret_282; \
47650		uint8x16_t __s0_282 = __p0_282; \
47651		uint8x8_t __s2_282 = __p2_282; \
47652		__ret_282 = vsetq_lane_u8(vget_lane_u8(__s2_282, __p3_282), __s0_282, __p1_282); \
	48546	#define vcopyq_lane_p8(__p0_282, __p1_282, __p2_282, __p3_282) __extension__ ({ \
	48547	poly8x16_t __ret_282; \
	48548	poly8x16_t __s0_282 = __p0_282; \
	48549	poly8x8_t __s2_282 = __p2_282; \
	48550	__ret_282 = vsetq_lane_p8(vget_lane_p8(__s2_282, __p3_282), __s0_282, __p1_282); \
47653	48551	__ret_282; \
47654	48552	})
47655	48553	#else
47656		#define vcopyq_lane_u8(__p0_283, __p1_283, __p2_283, __p3_283) __extension__ ({ \
47657		uint8x16_t __ret_283; \
47658		uint8x16_t __s0_283 = __p0_283; \
47659		uint8x8_t __s2_283 = __p2_283; \
47660		uint8x16_t __rev0_283; __rev0_283 = __builtin_shufflevector(__s0_283, __s0_283, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
47661		uint8x8_t __rev2_283; __rev2_283 = __builtin_shufflevector(__s2_283, __s2_283, 7, 6, 5, 4, 3, 2, 1, 0); \
47662		__ret_283 = __noswap_vsetq_lane_u8(__noswap_vget_lane_u8(__rev2_283, __p3_283), __rev0_283, __p1_283); \
47663		__ret_283 = __builtin_shufflevector(__ret_283, __ret_283, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	48554	#define vcopyq_lane_p8(__p0_283, __p1_283, __p2_283, __p3_283) __extension__ ({ \
	48555	poly8x16_t __ret_283; \
	48556	poly8x16_t __s0_283 = __p0_283; \
	48557	poly8x8_t __s2_283 = __p2_283; \
	48558	poly8x16_t __rev0_283; __rev0_283 = __builtin_shufflevector(__s0_283, __s0_283, __lane_reverse_128_8); \
	48559	poly8x8_t __rev2_283; __rev2_283 = __builtin_shufflevector(__s2_283, __s2_283, __lane_reverse_64_8); \
	48560	__ret_283 = __noswap_vsetq_lane_p8(__noswap_vget_lane_p8(__rev2_283, __p3_283), __rev0_283, __p1_283); \
	48561	__ret_283 = __builtin_shufflevector(__ret_283, __ret_283, __lane_reverse_128_8); \
47664	48562	__ret_283; \
47665	48563	})
47666	48564	#endif
47667	48565
47668	48566	#ifdef __LITTLE_ENDIAN__
47669		#define vcopyq_lane_u32(__p0_284, __p1_284, __p2_284, __p3_284) __extension__ ({ \
47670		uint32x4_t __ret_284; \
47671		uint32x4_t __s0_284 = __p0_284; \
47672		uint32x2_t __s2_284 = __p2_284; \
47673		__ret_284 = vsetq_lane_u32(vget_lane_u32(__s2_284, __p3_284), __s0_284, __p1_284); \
	48567	#define vcopyq_lane_p16(__p0_284, __p1_284, __p2_284, __p3_284) __extension__ ({ \
	48568	poly16x8_t __ret_284; \
	48569	poly16x8_t __s0_284 = __p0_284; \
	48570	poly16x4_t __s2_284 = __p2_284; \
	48571	__ret_284 = vsetq_lane_p16(vget_lane_p16(__s2_284, __p3_284), __s0_284, __p1_284); \
47674	48572	__ret_284; \
47675	48573	})
47676	48574	#else
47677		#define vcopyq_lane_u32(__p0_285, __p1_285, __p2_285, __p3_285) __extension__ ({ \
47678		uint32x4_t __ret_285; \
47679		uint32x4_t __s0_285 = __p0_285; \
47680		uint32x2_t __s2_285 = __p2_285; \
47681		uint32x4_t __rev0_285; __rev0_285 = __builtin_shufflevector(__s0_285, __s0_285, 3, 2, 1, 0); \
47682		uint32x2_t __rev2_285; __rev2_285 = __builtin_shufflevector(__s2_285, __s2_285, 1, 0); \
47683		__ret_285 = __noswap_vsetq_lane_u32(__noswap_vget_lane_u32(__rev2_285, __p3_285), __rev0_285, __p1_285); \
47684		__ret_285 = __builtin_shufflevector(__ret_285, __ret_285, 3, 2, 1, 0); \
	48575	#define vcopyq_lane_p16(__p0_285, __p1_285, __p2_285, __p3_285) __extension__ ({ \
	48576	poly16x8_t __ret_285; \
	48577	poly16x8_t __s0_285 = __p0_285; \
	48578	poly16x4_t __s2_285 = __p2_285; \
	48579	poly16x8_t __rev0_285; __rev0_285 = __builtin_shufflevector(__s0_285, __s0_285, __lane_reverse_128_16); \
	48580	poly16x4_t __rev2_285; __rev2_285 = __builtin_shufflevector(__s2_285, __s2_285, __lane_reverse_64_16); \
	48581	__ret_285 = __noswap_vsetq_lane_p16(__noswap_vget_lane_p16(__rev2_285, __p3_285), __rev0_285, __p1_285); \
	48582	__ret_285 = __builtin_shufflevector(__ret_285, __ret_285, __lane_reverse_128_16); \
47685	48583	__ret_285; \
47686	48584	})
47687	48585	#endif
47688	48586
47689	48587	#ifdef __LITTLE_ENDIAN__
47690		#define vcopyq_lane_u64(__p0_286, __p1_286, __p2_286, __p3_286) __extension__ ({ \
47691		uint64x2_t __ret_286; \
47692		uint64x2_t __s0_286 = __p0_286; \
47693		uint64x1_t __s2_286 = __p2_286; \
47694		__ret_286 = vsetq_lane_u64(vget_lane_u64(__s2_286, __p3_286), __s0_286, __p1_286); \
	48588	#define vcopyq_lane_u8(__p0_286, __p1_286, __p2_286, __p3_286) __extension__ ({ \
	48589	uint8x16_t __ret_286; \
	48590	uint8x16_t __s0_286 = __p0_286; \
	48591	uint8x8_t __s2_286 = __p2_286; \
	48592	__ret_286 = vsetq_lane_u8(vget_lane_u8(__s2_286, __p3_286), __s0_286, __p1_286); \
47695	48593	__ret_286; \
47696	48594	})
47697	48595	#else
47698		#define vcopyq_lane_u64(__p0_287, __p1_287, __p2_287, __p3_287) __extension__ ({ \
47699		uint64x2_t __ret_287; \
47700		uint64x2_t __s0_287 = __p0_287; \
47701		uint64x1_t __s2_287 = __p2_287; \
47702		uint64x2_t __rev0_287; __rev0_287 = __builtin_shufflevector(__s0_287, __s0_287, 1, 0); \
47703		__ret_287 = __noswap_vsetq_lane_u64(vget_lane_u64(__s2_287, __p3_287), __rev0_287, __p1_287); \
47704		__ret_287 = __builtin_shufflevector(__ret_287, __ret_287, 1, 0); \
	48596	#define vcopyq_lane_u8(__p0_287, __p1_287, __p2_287, __p3_287) __extension__ ({ \
	48597	uint8x16_t __ret_287; \
	48598	uint8x16_t __s0_287 = __p0_287; \
	48599	uint8x8_t __s2_287 = __p2_287; \
	48600	uint8x16_t __rev0_287; __rev0_287 = __builtin_shufflevector(__s0_287, __s0_287, __lane_reverse_128_8); \
	48601	uint8x8_t __rev2_287; __rev2_287 = __builtin_shufflevector(__s2_287, __s2_287, __lane_reverse_64_8); \
	48602	__ret_287 = __noswap_vsetq_lane_u8(__noswap_vget_lane_u8(__rev2_287, __p3_287), __rev0_287, __p1_287); \
	48603	__ret_287 = __builtin_shufflevector(__ret_287, __ret_287, __lane_reverse_128_8); \
47705	48604	__ret_287; \
47706	48605	})
47707	48606	#endif
47708	48607
47709	48608	#ifdef __LITTLE_ENDIAN__
47710		#define vcopyq_lane_u16(__p0_288, __p1_288, __p2_288, __p3_288) __extension__ ({ \
47711		uint16x8_t __ret_288; \
47712		uint16x8_t __s0_288 = __p0_288; \
47713		uint16x4_t __s2_288 = __p2_288; \
47714		__ret_288 = vsetq_lane_u16(vget_lane_u16(__s2_288, __p3_288), __s0_288, __p1_288); \
	48609	#define vcopyq_lane_u32(__p0_288, __p1_288, __p2_288, __p3_288) __extension__ ({ \
	48610	uint32x4_t __ret_288; \
	48611	uint32x4_t __s0_288 = __p0_288; \
	48612	uint32x2_t __s2_288 = __p2_288; \
	48613	__ret_288 = vsetq_lane_u32(vget_lane_u32(__s2_288, __p3_288), __s0_288, __p1_288); \
47715	48614	__ret_288; \
47716	48615	})
47717	48616	#else
47718		#define vcopyq_lane_u16(__p0_289, __p1_289, __p2_289, __p3_289) __extension__ ({ \
47719		uint16x8_t __ret_289; \
47720		uint16x8_t __s0_289 = __p0_289; \
47721		uint16x4_t __s2_289 = __p2_289; \
47722		uint16x8_t __rev0_289; __rev0_289 = __builtin_shufflevector(__s0_289, __s0_289, 7, 6, 5, 4, 3, 2, 1, 0); \
47723		uint16x4_t __rev2_289; __rev2_289 = __builtin_shufflevector(__s2_289, __s2_289, 3, 2, 1, 0); \
47724		__ret_289 = __noswap_vsetq_lane_u16(__noswap_vget_lane_u16(__rev2_289, __p3_289), __rev0_289, __p1_289); \
47725		__ret_289 = __builtin_shufflevector(__ret_289, __ret_289, 7, 6, 5, 4, 3, 2, 1, 0); \
	48617	#define vcopyq_lane_u32(__p0_289, __p1_289, __p2_289, __p3_289) __extension__ ({ \
	48618	uint32x4_t __ret_289; \
	48619	uint32x4_t __s0_289 = __p0_289; \
	48620	uint32x2_t __s2_289 = __p2_289; \
	48621	uint32x4_t __rev0_289; __rev0_289 = __builtin_shufflevector(__s0_289, __s0_289, __lane_reverse_128_32); \
	48622	uint32x2_t __rev2_289; __rev2_289 = __builtin_shufflevector(__s2_289, __s2_289, __lane_reverse_64_32); \
	48623	__ret_289 = __noswap_vsetq_lane_u32(__noswap_vget_lane_u32(__rev2_289, __p3_289), __rev0_289, __p1_289); \
	48624	__ret_289 = __builtin_shufflevector(__ret_289, __ret_289, __lane_reverse_128_32); \
47726	48625	__ret_289; \
47727	48626	})
47728	48627	#endif
47729	48628
47730	48629	#ifdef __LITTLE_ENDIAN__
47731		#define vcopyq_lane_s8(__p0_290, __p1_290, __p2_290, __p3_290) __extension__ ({ \
47732		int8x16_t __ret_290; \
47733		int8x16_t __s0_290 = __p0_290; \
47734		int8x8_t __s2_290 = __p2_290; \
47735		__ret_290 = vsetq_lane_s8(vget_lane_s8(__s2_290, __p3_290), __s0_290, __p1_290); \
	48630	#define vcopyq_lane_u64(__p0_290, __p1_290, __p2_290, __p3_290) __extension__ ({ \
	48631	uint64x2_t __ret_290; \
	48632	uint64x2_t __s0_290 = __p0_290; \
	48633	uint64x1_t __s2_290 = __p2_290; \
	48634	__ret_290 = vsetq_lane_u64(vget_lane_u64(__s2_290, __p3_290), __s0_290, __p1_290); \
47736	48635	__ret_290; \
47737	48636	})
47738	48637	#else
47739		#define vcopyq_lane_s8(__p0_291, __p1_291, __p2_291, __p3_291) __extension__ ({ \
47740		int8x16_t __ret_291; \
47741		int8x16_t __s0_291 = __p0_291; \
47742		int8x8_t __s2_291 = __p2_291; \
47743		int8x16_t __rev0_291; __rev0_291 = __builtin_shufflevector(__s0_291, __s0_291, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
47744		int8x8_t __rev2_291; __rev2_291 = __builtin_shufflevector(__s2_291, __s2_291, 7, 6, 5, 4, 3, 2, 1, 0); \
47745		__ret_291 = __noswap_vsetq_lane_s8(__noswap_vget_lane_s8(__rev2_291, __p3_291), __rev0_291, __p1_291); \
47746		__ret_291 = __builtin_shufflevector(__ret_291, __ret_291, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	48638	#define vcopyq_lane_u64(__p0_291, __p1_291, __p2_291, __p3_291) __extension__ ({ \
	48639	uint64x2_t __ret_291; \
	48640	uint64x2_t __s0_291 = __p0_291; \
	48641	uint64x1_t __s2_291 = __p2_291; \
	48642	uint64x2_t __rev0_291; __rev0_291 = __builtin_shufflevector(__s0_291, __s0_291, __lane_reverse_128_64); \
	48643	__ret_291 = __noswap_vsetq_lane_u64(vget_lane_u64(__s2_291, __p3_291), __rev0_291, __p1_291); \
	48644	__ret_291 = __builtin_shufflevector(__ret_291, __ret_291, __lane_reverse_128_64); \
47747	48645	__ret_291; \
47748	48646	})
47749	48647	#endif
47750	48648
47751	48649	#ifdef __LITTLE_ENDIAN__
47752		#define vcopyq_lane_f32(__p0_292, __p1_292, __p2_292, __p3_292) __extension__ ({ \
47753		float32x4_t __ret_292; \
47754		float32x4_t __s0_292 = __p0_292; \
47755		float32x2_t __s2_292 = __p2_292; \
47756		__ret_292 = vsetq_lane_f32(vget_lane_f32(__s2_292, __p3_292), __s0_292, __p1_292); \
	48650	#define vcopyq_lane_u16(__p0_292, __p1_292, __p2_292, __p3_292) __extension__ ({ \
	48651	uint16x8_t __ret_292; \
	48652	uint16x8_t __s0_292 = __p0_292; \
	48653	uint16x4_t __s2_292 = __p2_292; \
	48654	__ret_292 = vsetq_lane_u16(vget_lane_u16(__s2_292, __p3_292), __s0_292, __p1_292); \
47757	48655	__ret_292; \
47758	48656	})
47759	48657	#else
47760		#define vcopyq_lane_f32(__p0_293, __p1_293, __p2_293, __p3_293) __extension__ ({ \
47761		float32x4_t __ret_293; \
47762		float32x4_t __s0_293 = __p0_293; \
47763		float32x2_t __s2_293 = __p2_293; \
47764		float32x4_t __rev0_293; __rev0_293 = __builtin_shufflevector(__s0_293, __s0_293, 3, 2, 1, 0); \
47765		float32x2_t __rev2_293; __rev2_293 = __builtin_shufflevector(__s2_293, __s2_293, 1, 0); \
47766		__ret_293 = __noswap_vsetq_lane_f32(__noswap_vget_lane_f32(__rev2_293, __p3_293), __rev0_293, __p1_293); \
47767		__ret_293 = __builtin_shufflevector(__ret_293, __ret_293, 3, 2, 1, 0); \
	48658	#define vcopyq_lane_u16(__p0_293, __p1_293, __p2_293, __p3_293) __extension__ ({ \
	48659	uint16x8_t __ret_293; \
	48660	uint16x8_t __s0_293 = __p0_293; \
	48661	uint16x4_t __s2_293 = __p2_293; \
	48662	uint16x8_t __rev0_293; __rev0_293 = __builtin_shufflevector(__s0_293, __s0_293, __lane_reverse_128_16); \
	48663	uint16x4_t __rev2_293; __rev2_293 = __builtin_shufflevector(__s2_293, __s2_293, __lane_reverse_64_16); \
	48664	__ret_293 = __noswap_vsetq_lane_u16(__noswap_vget_lane_u16(__rev2_293, __p3_293), __rev0_293, __p1_293); \
	48665	__ret_293 = __builtin_shufflevector(__ret_293, __ret_293, __lane_reverse_128_16); \
47768	48666	__ret_293; \
47769	48667	})
47770	48668	#endif
47771	48669
47772	48670	#ifdef __LITTLE_ENDIAN__
47773		#define vcopyq_lane_s32(__p0_294, __p1_294, __p2_294, __p3_294) __extension__ ({ \
47774		int32x4_t __ret_294; \
47775		int32x4_t __s0_294 = __p0_294; \
47776		int32x2_t __s2_294 = __p2_294; \
47777		__ret_294 = vsetq_lane_s32(vget_lane_s32(__s2_294, __p3_294), __s0_294, __p1_294); \
	48671	#define vcopyq_lane_s8(__p0_294, __p1_294, __p2_294, __p3_294) __extension__ ({ \
	48672	int8x16_t __ret_294; \
	48673	int8x16_t __s0_294 = __p0_294; \
	48674	int8x8_t __s2_294 = __p2_294; \
	48675	__ret_294 = vsetq_lane_s8(vget_lane_s8(__s2_294, __p3_294), __s0_294, __p1_294); \
47778	48676	__ret_294; \
47779	48677	})
47780	48678	#else
47781		#define vcopyq_lane_s32(__p0_295, __p1_295, __p2_295, __p3_295) __extension__ ({ \
47782		int32x4_t __ret_295; \
47783		int32x4_t __s0_295 = __p0_295; \
47784		int32x2_t __s2_295 = __p2_295; \
47785		int32x4_t __rev0_295; __rev0_295 = __builtin_shufflevector(__s0_295, __s0_295, 3, 2, 1, 0); \
47786		int32x2_t __rev2_295; __rev2_295 = __builtin_shufflevector(__s2_295, __s2_295, 1, 0); \
47787		__ret_295 = __noswap_vsetq_lane_s32(__noswap_vget_lane_s32(__rev2_295, __p3_295), __rev0_295, __p1_295); \
47788		__ret_295 = __builtin_shufflevector(__ret_295, __ret_295, 3, 2, 1, 0); \
	48679	#define vcopyq_lane_s8(__p0_295, __p1_295, __p2_295, __p3_295) __extension__ ({ \
	48680	int8x16_t __ret_295; \
	48681	int8x16_t __s0_295 = __p0_295; \
	48682	int8x8_t __s2_295 = __p2_295; \
	48683	int8x16_t __rev0_295; __rev0_295 = __builtin_shufflevector(__s0_295, __s0_295, __lane_reverse_128_8); \
	48684	int8x8_t __rev2_295; __rev2_295 = __builtin_shufflevector(__s2_295, __s2_295, __lane_reverse_64_8); \
	48685	__ret_295 = __noswap_vsetq_lane_s8(__noswap_vget_lane_s8(__rev2_295, __p3_295), __rev0_295, __p1_295); \
	48686	__ret_295 = __builtin_shufflevector(__ret_295, __ret_295, __lane_reverse_128_8); \
47789	48687	__ret_295; \
47790	48688	})
47791	48689	#endif
47792	48690
47793	48691	#ifdef __LITTLE_ENDIAN__
47794		#define vcopyq_lane_s64(__p0_296, __p1_296, __p2_296, __p3_296) __extension__ ({ \
47795		int64x2_t __ret_296; \
47796		int64x2_t __s0_296 = __p0_296; \
47797		int64x1_t __s2_296 = __p2_296; \
47798		__ret_296 = vsetq_lane_s64(vget_lane_s64(__s2_296, __p3_296), __s0_296, __p1_296); \
	48692	#define vcopyq_lane_f32(__p0_296, __p1_296, __p2_296, __p3_296) __extension__ ({ \
	48693	float32x4_t __ret_296; \
	48694	float32x4_t __s0_296 = __p0_296; \
	48695	float32x2_t __s2_296 = __p2_296; \
	48696	__ret_296 = vsetq_lane_f32(vget_lane_f32(__s2_296, __p3_296), __s0_296, __p1_296); \
47799	48697	__ret_296; \
47800	48698	})
47801	48699	#else
47802		#define vcopyq_lane_s64(__p0_297, __p1_297, __p2_297, __p3_297) __extension__ ({ \
47803		int64x2_t __ret_297; \
47804		int64x2_t __s0_297 = __p0_297; \
47805		int64x1_t __s2_297 = __p2_297; \
47806		int64x2_t __rev0_297; __rev0_297 = __builtin_shufflevector(__s0_297, __s0_297, 1, 0); \
47807		__ret_297 = __noswap_vsetq_lane_s64(vget_lane_s64(__s2_297, __p3_297), __rev0_297, __p1_297); \
47808		__ret_297 = __builtin_shufflevector(__ret_297, __ret_297, 1, 0); \
	48700	#define vcopyq_lane_f32(__p0_297, __p1_297, __p2_297, __p3_297) __extension__ ({ \
	48701	float32x4_t __ret_297; \
	48702	float32x4_t __s0_297 = __p0_297; \
	48703	float32x2_t __s2_297 = __p2_297; \
	48704	float32x4_t __rev0_297; __rev0_297 = __builtin_shufflevector(__s0_297, __s0_297, __lane_reverse_128_32); \
	48705	float32x2_t __rev2_297; __rev2_297 = __builtin_shufflevector(__s2_297, __s2_297, __lane_reverse_64_32); \
	48706	__ret_297 = __noswap_vsetq_lane_f32(__noswap_vget_lane_f32(__rev2_297, __p3_297), __rev0_297, __p1_297); \
	48707	__ret_297 = __builtin_shufflevector(__ret_297, __ret_297, __lane_reverse_128_32); \
47809	48708	__ret_297; \
47810	48709	})
47811	48710	#endif
47812	48711
47813	48712	#ifdef __LITTLE_ENDIAN__
47814		#define vcopyq_lane_s16(__p0_298, __p1_298, __p2_298, __p3_298) __extension__ ({ \
47815		int16x8_t __ret_298; \
47816		int16x8_t __s0_298 = __p0_298; \
47817		int16x4_t __s2_298 = __p2_298; \
47818		__ret_298 = vsetq_lane_s16(vget_lane_s16(__s2_298, __p3_298), __s0_298, __p1_298); \
	48713	#define vcopyq_lane_s32(__p0_298, __p1_298, __p2_298, __p3_298) __extension__ ({ \
	48714	int32x4_t __ret_298; \
	48715	int32x4_t __s0_298 = __p0_298; \
	48716	int32x2_t __s2_298 = __p2_298; \
	48717	__ret_298 = vsetq_lane_s32(vget_lane_s32(__s2_298, __p3_298), __s0_298, __p1_298); \
47819	48718	__ret_298; \
47820	48719	})
47821	48720	#else
47822		#define vcopyq_lane_s16(__p0_299, __p1_299, __p2_299, __p3_299) __extension__ ({ \
47823		int16x8_t __ret_299; \
47824		int16x8_t __s0_299 = __p0_299; \
47825		int16x4_t __s2_299 = __p2_299; \
47826		int16x8_t __rev0_299; __rev0_299 = __builtin_shufflevector(__s0_299, __s0_299, 7, 6, 5, 4, 3, 2, 1, 0); \
47827		int16x4_t __rev2_299; __rev2_299 = __builtin_shufflevector(__s2_299, __s2_299, 3, 2, 1, 0); \
47828		__ret_299 = __noswap_vsetq_lane_s16(__noswap_vget_lane_s16(__rev2_299, __p3_299), __rev0_299, __p1_299); \
47829		__ret_299 = __builtin_shufflevector(__ret_299, __ret_299, 7, 6, 5, 4, 3, 2, 1, 0); \
	48721	#define vcopyq_lane_s32(__p0_299, __p1_299, __p2_299, __p3_299) __extension__ ({ \
	48722	int32x4_t __ret_299; \
	48723	int32x4_t __s0_299 = __p0_299; \
	48724	int32x2_t __s2_299 = __p2_299; \
	48725	int32x4_t __rev0_299; __rev0_299 = __builtin_shufflevector(__s0_299, __s0_299, __lane_reverse_128_32); \
	48726	int32x2_t __rev2_299; __rev2_299 = __builtin_shufflevector(__s2_299, __s2_299, __lane_reverse_64_32); \
	48727	__ret_299 = __noswap_vsetq_lane_s32(__noswap_vget_lane_s32(__rev2_299, __p3_299), __rev0_299, __p1_299); \
	48728	__ret_299 = __builtin_shufflevector(__ret_299, __ret_299, __lane_reverse_128_32); \
47830	48729	__ret_299; \
47831	48730	})
47832	48731	#endif
47833	48732
47834	48733	#ifdef __LITTLE_ENDIAN__
47835		#define vcopy_lane_p8(__p0_300, __p1_300, __p2_300, __p3_300) __extension__ ({ \
47836		poly8x8_t __ret_300; \
47837		poly8x8_t __s0_300 = __p0_300; \
47838		poly8x8_t __s2_300 = __p2_300; \
47839		__ret_300 = vset_lane_p8(vget_lane_p8(__s2_300, __p3_300), __s0_300, __p1_300); \
	48734	#define vcopyq_lane_s64(__p0_300, __p1_300, __p2_300, __p3_300) __extension__ ({ \
	48735	int64x2_t __ret_300; \
	48736	int64x2_t __s0_300 = __p0_300; \
	48737	int64x1_t __s2_300 = __p2_300; \
	48738	__ret_300 = vsetq_lane_s64(vget_lane_s64(__s2_300, __p3_300), __s0_300, __p1_300); \
47840	48739	__ret_300; \
47841	48740	})
47842	48741	#else
47843		#define vcopy_lane_p8(__p0_301, __p1_301, __p2_301, __p3_301) __extension__ ({ \
47844		poly8x8_t __ret_301; \
47845		poly8x8_t __s0_301 = __p0_301; \
47846		poly8x8_t __s2_301 = __p2_301; \
47847		poly8x8_t __rev0_301; __rev0_301 = __builtin_shufflevector(__s0_301, __s0_301, 7, 6, 5, 4, 3, 2, 1, 0); \
47848		poly8x8_t __rev2_301; __rev2_301 = __builtin_shufflevector(__s2_301, __s2_301, 7, 6, 5, 4, 3, 2, 1, 0); \
47849		__ret_301 = __noswap_vset_lane_p8(__noswap_vget_lane_p8(__rev2_301, __p3_301), __rev0_301, __p1_301); \
47850		__ret_301 = __builtin_shufflevector(__ret_301, __ret_301, 7, 6, 5, 4, 3, 2, 1, 0); \
	48742	#define vcopyq_lane_s64(__p0_301, __p1_301, __p2_301, __p3_301) __extension__ ({ \
	48743	int64x2_t __ret_301; \
	48744	int64x2_t __s0_301 = __p0_301; \
	48745	int64x1_t __s2_301 = __p2_301; \
	48746	int64x2_t __rev0_301; __rev0_301 = __builtin_shufflevector(__s0_301, __s0_301, __lane_reverse_128_64); \
	48747	__ret_301 = __noswap_vsetq_lane_s64(vget_lane_s64(__s2_301, __p3_301), __rev0_301, __p1_301); \
	48748	__ret_301 = __builtin_shufflevector(__ret_301, __ret_301, __lane_reverse_128_64); \
47851	48749	__ret_301; \
47852	48750	})
47853	48751	#endif
47854	48752
47855	48753	#ifdef __LITTLE_ENDIAN__
47856		#define vcopy_lane_p16(__p0_302, __p1_302, __p2_302, __p3_302) __extension__ ({ \
47857		poly16x4_t __ret_302; \
47858		poly16x4_t __s0_302 = __p0_302; \
47859		poly16x4_t __s2_302 = __p2_302; \
47860		__ret_302 = vset_lane_p16(vget_lane_p16(__s2_302, __p3_302), __s0_302, __p1_302); \
	48754	#define vcopyq_lane_mf8(__p0_302, __p1_302, __p2_302, __p3_302) __extension__ ({ \
	48755	mfloat8x16_t __ret_302; \
	48756	mfloat8x16_t __s0_302 = __p0_302; \
	48757	mfloat8x8_t __s2_302 = __p2_302; \
	48758	__ret_302 = vsetq_lane_mf8(vget_lane_mf8(__s2_302, __p3_302), __s0_302, __p1_302); \
47861	48759	__ret_302; \
47862	48760	})
47863	48761	#else
47864		#define vcopy_lane_p16(__p0_303, __p1_303, __p2_303, __p3_303) __extension__ ({ \
47865		poly16x4_t __ret_303; \
47866		poly16x4_t __s0_303 = __p0_303; \
47867		poly16x4_t __s2_303 = __p2_303; \
47868		poly16x4_t __rev0_303; __rev0_303 = __builtin_shufflevector(__s0_303, __s0_303, 3, 2, 1, 0); \
47869		poly16x4_t __rev2_303; __rev2_303 = __builtin_shufflevector(__s2_303, __s2_303, 3, 2, 1, 0); \
47870		__ret_303 = __noswap_vset_lane_p16(__noswap_vget_lane_p16(__rev2_303, __p3_303), __rev0_303, __p1_303); \
47871		__ret_303 = __builtin_shufflevector(__ret_303, __ret_303, 3, 2, 1, 0); \
	48762	#define vcopyq_lane_mf8(__p0_303, __p1_303, __p2_303, __p3_303) __extension__ ({ \
	48763	mfloat8x16_t __ret_303; \
	48764	mfloat8x16_t __s0_303 = __p0_303; \
	48765	mfloat8x8_t __s2_303 = __p2_303; \
	48766	mfloat8x16_t __rev0_303; __rev0_303 = __builtin_shufflevector(__s0_303, __s0_303, __lane_reverse_128_8); \
	48767	mfloat8x8_t __rev2_303; __rev2_303 = __builtin_shufflevector(__s2_303, __s2_303, __lane_reverse_64_8); \
	48768	__ret_303 = __noswap_vsetq_lane_mf8(__noswap_vget_lane_mf8(__rev2_303, __p3_303), __rev0_303, __p1_303); \
	48769	__ret_303 = __builtin_shufflevector(__ret_303, __ret_303, __lane_reverse_128_8); \
47872	48770	__ret_303; \
47873	48771	})
47874	48772	#endif
47875	48773
47876	48774	#ifdef __LITTLE_ENDIAN__
47877		#define vcopy_lane_u8(__p0_304, __p1_304, __p2_304, __p3_304) __extension__ ({ \
47878		uint8x8_t __ret_304; \
47879		uint8x8_t __s0_304 = __p0_304; \
47880		uint8x8_t __s2_304 = __p2_304; \
47881		__ret_304 = vset_lane_u8(vget_lane_u8(__s2_304, __p3_304), __s0_304, __p1_304); \
	48775	#define vcopyq_lane_s16(__p0_304, __p1_304, __p2_304, __p3_304) __extension__ ({ \
	48776	int16x8_t __ret_304; \
	48777	int16x8_t __s0_304 = __p0_304; \
	48778	int16x4_t __s2_304 = __p2_304; \
	48779	__ret_304 = vsetq_lane_s16(vget_lane_s16(__s2_304, __p3_304), __s0_304, __p1_304); \
47882	48780	__ret_304; \
47883	48781	})
47884	48782	#else
47885		#define vcopy_lane_u8(__p0_305, __p1_305, __p2_305, __p3_305) __extension__ ({ \
47886		uint8x8_t __ret_305; \
47887		uint8x8_t __s0_305 = __p0_305; \
47888		uint8x8_t __s2_305 = __p2_305; \
47889		uint8x8_t __rev0_305; __rev0_305 = __builtin_shufflevector(__s0_305, __s0_305, 7, 6, 5, 4, 3, 2, 1, 0); \
47890		uint8x8_t __rev2_305; __rev2_305 = __builtin_shufflevector(__s2_305, __s2_305, 7, 6, 5, 4, 3, 2, 1, 0); \
47891		__ret_305 = __noswap_vset_lane_u8(__noswap_vget_lane_u8(__rev2_305, __p3_305), __rev0_305, __p1_305); \
47892		__ret_305 = __builtin_shufflevector(__ret_305, __ret_305, 7, 6, 5, 4, 3, 2, 1, 0); \
	48783	#define vcopyq_lane_s16(__p0_305, __p1_305, __p2_305, __p3_305) __extension__ ({ \
	48784	int16x8_t __ret_305; \
	48785	int16x8_t __s0_305 = __p0_305; \
	48786	int16x4_t __s2_305 = __p2_305; \
	48787	int16x8_t __rev0_305; __rev0_305 = __builtin_shufflevector(__s0_305, __s0_305, __lane_reverse_128_16); \
	48788	int16x4_t __rev2_305; __rev2_305 = __builtin_shufflevector(__s2_305, __s2_305, __lane_reverse_64_16); \
	48789	__ret_305 = __noswap_vsetq_lane_s16(__noswap_vget_lane_s16(__rev2_305, __p3_305), __rev0_305, __p1_305); \
	48790	__ret_305 = __builtin_shufflevector(__ret_305, __ret_305, __lane_reverse_128_16); \
47893	48791	__ret_305; \
47894	48792	})
47895	48793	#endif
47896	48794
47897	48795	#ifdef __LITTLE_ENDIAN__
47898		#define vcopy_lane_u32(__p0_306, __p1_306, __p2_306, __p3_306) __extension__ ({ \
47899		uint32x2_t __ret_306; \
47900		uint32x2_t __s0_306 = __p0_306; \
47901		uint32x2_t __s2_306 = __p2_306; \
47902		__ret_306 = vset_lane_u32(vget_lane_u32(__s2_306, __p3_306), __s0_306, __p1_306); \
	48796	#define vcopy_lane_p8(__p0_306, __p1_306, __p2_306, __p3_306) __extension__ ({ \
	48797	poly8x8_t __ret_306; \
	48798	poly8x8_t __s0_306 = __p0_306; \
	48799	poly8x8_t __s2_306 = __p2_306; \
	48800	__ret_306 = vset_lane_p8(vget_lane_p8(__s2_306, __p3_306), __s0_306, __p1_306); \
47903	48801	__ret_306; \
47904	48802	})
47905	48803	#else
47906		#define vcopy_lane_u32(__p0_307, __p1_307, __p2_307, __p3_307) __extension__ ({ \
47907		uint32x2_t __ret_307; \
47908		uint32x2_t __s0_307 = __p0_307; \
47909		uint32x2_t __s2_307 = __p2_307; \
47910		uint32x2_t __rev0_307; __rev0_307 = __builtin_shufflevector(__s0_307, __s0_307, 1, 0); \
47911		uint32x2_t __rev2_307; __rev2_307 = __builtin_shufflevector(__s2_307, __s2_307, 1, 0); \
47912		__ret_307 = __noswap_vset_lane_u32(__noswap_vget_lane_u32(__rev2_307, __p3_307), __rev0_307, __p1_307); \
47913		__ret_307 = __builtin_shufflevector(__ret_307, __ret_307, 1, 0); \
	48804	#define vcopy_lane_p8(__p0_307, __p1_307, __p2_307, __p3_307) __extension__ ({ \
	48805	poly8x8_t __ret_307; \
	48806	poly8x8_t __s0_307 = __p0_307; \
	48807	poly8x8_t __s2_307 = __p2_307; \
	48808	poly8x8_t __rev0_307; __rev0_307 = __builtin_shufflevector(__s0_307, __s0_307, __lane_reverse_64_8); \
	48809	poly8x8_t __rev2_307; __rev2_307 = __builtin_shufflevector(__s2_307, __s2_307, __lane_reverse_64_8); \
	48810	__ret_307 = __noswap_vset_lane_p8(__noswap_vget_lane_p8(__rev2_307, __p3_307), __rev0_307, __p1_307); \
	48811	__ret_307 = __builtin_shufflevector(__ret_307, __ret_307, __lane_reverse_64_8); \
47914	48812	__ret_307; \
47915	48813	})
47916	48814	#endif
47917	48815
47918		#define vcopy_lane_u64(__p0_308, __p1_308, __p2_308, __p3_308) __extension__ ({ \
47919		uint64x1_t __ret_308; \
47920		uint64x1_t __s0_308 = __p0_308; \
47921		uint64x1_t __s2_308 = __p2_308; \
47922		__ret_308 = vset_lane_u64(vget_lane_u64(__s2_308, __p3_308), __s0_308, __p1_308); \
47923		__ret_308; \
47924		})
47925	48816	#ifdef __LITTLE_ENDIAN__
47926		#define vcopy_lane_u16(__p0_309, __p1_309, __p2_309, __p3_309) __extension__ ({ \
47927		uint16x4_t __ret_309; \
47928		uint16x4_t __s0_309 = __p0_309; \
47929		uint16x4_t __s2_309 = __p2_309; \
47930		__ret_309 = vset_lane_u16(vget_lane_u16(__s2_309, __p3_309), __s0_309, __p1_309); \
47931		__ret_309; \
	48817	#define vcopy_lane_p16(__p0_308, __p1_308, __p2_308, __p3_308) __extension__ ({ \
	48818	poly16x4_t __ret_308; \
	48819	poly16x4_t __s0_308 = __p0_308; \
	48820	poly16x4_t __s2_308 = __p2_308; \
	48821	__ret_308 = vset_lane_p16(vget_lane_p16(__s2_308, __p3_308), __s0_308, __p1_308); \
	48822	__ret_308; \
47932	48823	})
47933	48824	#else
47934		#define vcopy_lane_u16(__p0_310, __p1_310, __p2_310, __p3_310) __extension__ ({ \
47935		uint16x4_t __ret_310; \
47936		uint16x4_t __s0_310 = __p0_310; \
47937		uint16x4_t __s2_310 = __p2_310; \
47938		uint16x4_t __rev0_310; __rev0_310 = __builtin_shufflevector(__s0_310, __s0_310, 3, 2, 1, 0); \
47939		uint16x4_t __rev2_310; __rev2_310 = __builtin_shufflevector(__s2_310, __s2_310, 3, 2, 1, 0); \
47940		__ret_310 = __noswap_vset_lane_u16(__noswap_vget_lane_u16(__rev2_310, __p3_310), __rev0_310, __p1_310); \
47941		__ret_310 = __builtin_shufflevector(__ret_310, __ret_310, 3, 2, 1, 0); \
47942		__ret_310; \
	48825	#define vcopy_lane_p16(__p0_309, __p1_309, __p2_309, __p3_309) __extension__ ({ \
	48826	poly16x4_t __ret_309; \
	48827	poly16x4_t __s0_309 = __p0_309; \
	48828	poly16x4_t __s2_309 = __p2_309; \
	48829	poly16x4_t __rev0_309; __rev0_309 = __builtin_shufflevector(__s0_309, __s0_309, __lane_reverse_64_16); \
	48830	poly16x4_t __rev2_309; __rev2_309 = __builtin_shufflevector(__s2_309, __s2_309, __lane_reverse_64_16); \
	48831	__ret_309 = __noswap_vset_lane_p16(__noswap_vget_lane_p16(__rev2_309, __p3_309), __rev0_309, __p1_309); \
	48832	__ret_309 = __builtin_shufflevector(__ret_309, __ret_309, __lane_reverse_64_16); \
	48833	__ret_309; \
47943	48834	})
47944	48835	#endif
47945	48836
47946	48837	#ifdef __LITTLE_ENDIAN__
47947		#define vcopy_lane_s8(__p0_311, __p1_311, __p2_311, __p3_311) __extension__ ({ \
47948		int8x8_t __ret_311; \
47949		int8x8_t __s0_311 = __p0_311; \
47950		int8x8_t __s2_311 = __p2_311; \
47951		__ret_311 = vset_lane_s8(vget_lane_s8(__s2_311, __p3_311), __s0_311, __p1_311); \
47952		__ret_311; \
	48838	#define vcopy_lane_u8(__p0_310, __p1_310, __p2_310, __p3_310) __extension__ ({ \
	48839	uint8x8_t __ret_310; \
	48840	uint8x8_t __s0_310 = __p0_310; \
	48841	uint8x8_t __s2_310 = __p2_310; \
	48842	__ret_310 = vset_lane_u8(vget_lane_u8(__s2_310, __p3_310), __s0_310, __p1_310); \
	48843	__ret_310; \
47953	48844	})
47954	48845	#else
47955		#define vcopy_lane_s8(__p0_312, __p1_312, __p2_312, __p3_312) __extension__ ({ \
47956		int8x8_t __ret_312; \
47957		int8x8_t __s0_312 = __p0_312; \
47958		int8x8_t __s2_312 = __p2_312; \
47959		int8x8_t __rev0_312; __rev0_312 = __builtin_shufflevector(__s0_312, __s0_312, 7, 6, 5, 4, 3, 2, 1, 0); \
47960		int8x8_t __rev2_312; __rev2_312 = __builtin_shufflevector(__s2_312, __s2_312, 7, 6, 5, 4, 3, 2, 1, 0); \
47961		__ret_312 = __noswap_vset_lane_s8(__noswap_vget_lane_s8(__rev2_312, __p3_312), __rev0_312, __p1_312); \
47962		__ret_312 = __builtin_shufflevector(__ret_312, __ret_312, 7, 6, 5, 4, 3, 2, 1, 0); \
47963		__ret_312; \
	48846	#define vcopy_lane_u8(__p0_311, __p1_311, __p2_311, __p3_311) __extension__ ({ \
	48847	uint8x8_t __ret_311; \
	48848	uint8x8_t __s0_311 = __p0_311; \
	48849	uint8x8_t __s2_311 = __p2_311; \
	48850	uint8x8_t __rev0_311; __rev0_311 = __builtin_shufflevector(__s0_311, __s0_311, __lane_reverse_64_8); \
	48851	uint8x8_t __rev2_311; __rev2_311 = __builtin_shufflevector(__s2_311, __s2_311, __lane_reverse_64_8); \
	48852	__ret_311 = __noswap_vset_lane_u8(__noswap_vget_lane_u8(__rev2_311, __p3_311), __rev0_311, __p1_311); \
	48853	__ret_311 = __builtin_shufflevector(__ret_311, __ret_311, __lane_reverse_64_8); \
	48854	__ret_311; \
47964	48855	})
47965	48856	#endif
47966	48857
47967	48858	#ifdef __LITTLE_ENDIAN__
47968		#define vcopy_lane_f32(__p0_313, __p1_313, __p2_313, __p3_313) __extension__ ({ \
47969		float32x2_t __ret_313; \
47970		float32x2_t __s0_313 = __p0_313; \
47971		float32x2_t __s2_313 = __p2_313; \
47972		__ret_313 = vset_lane_f32(vget_lane_f32(__s2_313, __p3_313), __s0_313, __p1_313); \
47973		__ret_313; \
	48859	#define vcopy_lane_u32(__p0_312, __p1_312, __p2_312, __p3_312) __extension__ ({ \
	48860	uint32x2_t __ret_312; \
	48861	uint32x2_t __s0_312 = __p0_312; \
	48862	uint32x2_t __s2_312 = __p2_312; \
	48863	__ret_312 = vset_lane_u32(vget_lane_u32(__s2_312, __p3_312), __s0_312, __p1_312); \
	48864	__ret_312; \
47974	48865	})
47975	48866	#else
47976		#define vcopy_lane_f32(__p0_314, __p1_314, __p2_314, __p3_314) __extension__ ({ \
47977		float32x2_t __ret_314; \
47978		float32x2_t __s0_314 = __p0_314; \
47979		float32x2_t __s2_314 = __p2_314; \
47980		float32x2_t __rev0_314; __rev0_314 = __builtin_shufflevector(__s0_314, __s0_314, 1, 0); \
47981		float32x2_t __rev2_314; __rev2_314 = __builtin_shufflevector(__s2_314, __s2_314, 1, 0); \
47982		__ret_314 = __noswap_vset_lane_f32(__noswap_vget_lane_f32(__rev2_314, __p3_314), __rev0_314, __p1_314); \
47983		__ret_314 = __builtin_shufflevector(__ret_314, __ret_314, 1, 0); \
47984		__ret_314; \
	48867	#define vcopy_lane_u32(__p0_313, __p1_313, __p2_313, __p3_313) __extension__ ({ \
	48868	uint32x2_t __ret_313; \
	48869	uint32x2_t __s0_313 = __p0_313; \
	48870	uint32x2_t __s2_313 = __p2_313; \
	48871	uint32x2_t __rev0_313; __rev0_313 = __builtin_shufflevector(__s0_313, __s0_313, __lane_reverse_64_32); \
	48872	uint32x2_t __rev2_313; __rev2_313 = __builtin_shufflevector(__s2_313, __s2_313, __lane_reverse_64_32); \
	48873	__ret_313 = __noswap_vset_lane_u32(__noswap_vget_lane_u32(__rev2_313, __p3_313), __rev0_313, __p1_313); \
	48874	__ret_313 = __builtin_shufflevector(__ret_313, __ret_313, __lane_reverse_64_32); \
	48875	__ret_313; \
47985	48876	})
47986	48877	#endif
47987	48878
	48879	#define vcopy_lane_u64(__p0_314, __p1_314, __p2_314, __p3_314) __extension__ ({ \
	48880	uint64x1_t __ret_314; \
	48881	uint64x1_t __s0_314 = __p0_314; \
	48882	uint64x1_t __s2_314 = __p2_314; \
	48883	__ret_314 = vset_lane_u64(vget_lane_u64(__s2_314, __p3_314), __s0_314, __p1_314); \
	48884	__ret_314; \
	48885	})
47988	48886	#ifdef __LITTLE_ENDIAN__
47989		#define vcopy_lane_s32(__p0_315, __p1_315, __p2_315, __p3_315) __extension__ ({ \
47990		int32x2_t __ret_315; \
47991		int32x2_t __s0_315 = __p0_315; \
47992		int32x2_t __s2_315 = __p2_315; \
47993		__ret_315 = vset_lane_s32(vget_lane_s32(__s2_315, __p3_315), __s0_315, __p1_315); \
	48887	#define vcopy_lane_u16(__p0_315, __p1_315, __p2_315, __p3_315) __extension__ ({ \
	48888	uint16x4_t __ret_315; \
	48889	uint16x4_t __s0_315 = __p0_315; \
	48890	uint16x4_t __s2_315 = __p2_315; \
	48891	__ret_315 = vset_lane_u16(vget_lane_u16(__s2_315, __p3_315), __s0_315, __p1_315); \
47994	48892	__ret_315; \
47995	48893	})
47996	48894	#else
47997		#define vcopy_lane_s32(__p0_316, __p1_316, __p2_316, __p3_316) __extension__ ({ \
47998		int32x2_t __ret_316; \
47999		int32x2_t __s0_316 = __p0_316; \
48000		int32x2_t __s2_316 = __p2_316; \
48001		int32x2_t __rev0_316; __rev0_316 = __builtin_shufflevector(__s0_316, __s0_316, 1, 0); \
48002		int32x2_t __rev2_316; __rev2_316 = __builtin_shufflevector(__s2_316, __s2_316, 1, 0); \
48003		__ret_316 = __noswap_vset_lane_s32(__noswap_vget_lane_s32(__rev2_316, __p3_316), __rev0_316, __p1_316); \
48004		__ret_316 = __builtin_shufflevector(__ret_316, __ret_316, 1, 0); \
	48895	#define vcopy_lane_u16(__p0_316, __p1_316, __p2_316, __p3_316) __extension__ ({ \
	48896	uint16x4_t __ret_316; \
	48897	uint16x4_t __s0_316 = __p0_316; \
	48898	uint16x4_t __s2_316 = __p2_316; \
	48899	uint16x4_t __rev0_316; __rev0_316 = __builtin_shufflevector(__s0_316, __s0_316, __lane_reverse_64_16); \
	48900	uint16x4_t __rev2_316; __rev2_316 = __builtin_shufflevector(__s2_316, __s2_316, __lane_reverse_64_16); \
	48901	__ret_316 = __noswap_vset_lane_u16(__noswap_vget_lane_u16(__rev2_316, __p3_316), __rev0_316, __p1_316); \
	48902	__ret_316 = __builtin_shufflevector(__ret_316, __ret_316, __lane_reverse_64_16); \
48005	48903	__ret_316; \
48006	48904	})
48007	48905	#endif
48008	48906
48009		#define vcopy_lane_s64(__p0_317, __p1_317, __p2_317, __p3_317) __extension__ ({ \
48010		int64x1_t __ret_317; \
48011		int64x1_t __s0_317 = __p0_317; \
48012		int64x1_t __s2_317 = __p2_317; \
48013		__ret_317 = vset_lane_s64(vget_lane_s64(__s2_317, __p3_317), __s0_317, __p1_317); \
48014		__ret_317; \
48015		})
48016	48907	#ifdef __LITTLE_ENDIAN__
48017		#define vcopy_lane_s16(__p0_318, __p1_318, __p2_318, __p3_318) __extension__ ({ \
48018		int16x4_t __ret_318; \
48019		int16x4_t __s0_318 = __p0_318; \
48020		int16x4_t __s2_318 = __p2_318; \
48021		__ret_318 = vset_lane_s16(vget_lane_s16(__s2_318, __p3_318), __s0_318, __p1_318); \
48022		__ret_318; \
	48908	#define vcopy_lane_s8(__p0_317, __p1_317, __p2_317, __p3_317) __extension__ ({ \
	48909	int8x8_t __ret_317; \
	48910	int8x8_t __s0_317 = __p0_317; \
	48911	int8x8_t __s2_317 = __p2_317; \
	48912	__ret_317 = vset_lane_s8(vget_lane_s8(__s2_317, __p3_317), __s0_317, __p1_317); \
	48913	__ret_317; \
48023	48914	})
48024	48915	#else
48025		#define vcopy_lane_s16(__p0_319, __p1_319, __p2_319, __p3_319) __extension__ ({ \
48026		int16x4_t __ret_319; \
48027		int16x4_t __s0_319 = __p0_319; \
48028		int16x4_t __s2_319 = __p2_319; \
48029		int16x4_t __rev0_319; __rev0_319 = __builtin_shufflevector(__s0_319, __s0_319, 3, 2, 1, 0); \
48030		int16x4_t __rev2_319; __rev2_319 = __builtin_shufflevector(__s2_319, __s2_319, 3, 2, 1, 0); \
48031		__ret_319 = __noswap_vset_lane_s16(__noswap_vget_lane_s16(__rev2_319, __p3_319), __rev0_319, __p1_319); \
48032		__ret_319 = __builtin_shufflevector(__ret_319, __ret_319, 3, 2, 1, 0); \
48033		__ret_319; \
	48916	#define vcopy_lane_s8(__p0_318, __p1_318, __p2_318, __p3_318) __extension__ ({ \
	48917	int8x8_t __ret_318; \
	48918	int8x8_t __s0_318 = __p0_318; \
	48919	int8x8_t __s2_318 = __p2_318; \
	48920	int8x8_t __rev0_318; __rev0_318 = __builtin_shufflevector(__s0_318, __s0_318, __lane_reverse_64_8); \
	48921	int8x8_t __rev2_318; __rev2_318 = __builtin_shufflevector(__s2_318, __s2_318, __lane_reverse_64_8); \
	48922	__ret_318 = __noswap_vset_lane_s8(__noswap_vget_lane_s8(__rev2_318, __p3_318), __rev0_318, __p1_318); \
	48923	__ret_318 = __builtin_shufflevector(__ret_318, __ret_318, __lane_reverse_64_8); \
	48924	__ret_318; \
48034	48925	})
48035	48926	#endif
48036	48927
48037	48928	#ifdef __LITTLE_ENDIAN__
48038		#define vcopyq_laneq_p8(__p0_320, __p1_320, __p2_320, __p3_320) __extension__ ({ \
48039		poly8x16_t __ret_320; \
48040		poly8x16_t __s0_320 = __p0_320; \
48041		poly8x16_t __s2_320 = __p2_320; \
48042		__ret_320 = vsetq_lane_p8(vgetq_lane_p8(__s2_320, __p3_320), __s0_320, __p1_320); \
48043		__ret_320; \
	48929	#define vcopy_lane_f32(__p0_319, __p1_319, __p2_319, __p3_319) __extension__ ({ \
	48930	float32x2_t __ret_319; \
	48931	float32x2_t __s0_319 = __p0_319; \
	48932	float32x2_t __s2_319 = __p2_319; \
	48933	__ret_319 = vset_lane_f32(vget_lane_f32(__s2_319, __p3_319), __s0_319, __p1_319); \
	48934	__ret_319; \
48044	48935	})
48045	48936	#else
48046		#define vcopyq_laneq_p8(__p0_321, __p1_321, __p2_321, __p3_321) __extension__ ({ \
48047		poly8x16_t __ret_321; \
48048		poly8x16_t __s0_321 = __p0_321; \
48049		poly8x16_t __s2_321 = __p2_321; \
48050		poly8x16_t __rev0_321; __rev0_321 = __builtin_shufflevector(__s0_321, __s0_321, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
48051		poly8x16_t __rev2_321; __rev2_321 = __builtin_shufflevector(__s2_321, __s2_321, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
48052		__ret_321 = __noswap_vsetq_lane_p8(__noswap_vgetq_lane_p8(__rev2_321, __p3_321), __rev0_321, __p1_321); \
48053		__ret_321 = __builtin_shufflevector(__ret_321, __ret_321, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
48054		__ret_321; \
	48937	#define vcopy_lane_f32(__p0_320, __p1_320, __p2_320, __p3_320) __extension__ ({ \
	48938	float32x2_t __ret_320; \
	48939	float32x2_t __s0_320 = __p0_320; \
	48940	float32x2_t __s2_320 = __p2_320; \
	48941	float32x2_t __rev0_320; __rev0_320 = __builtin_shufflevector(__s0_320, __s0_320, __lane_reverse_64_32); \
	48942	float32x2_t __rev2_320; __rev2_320 = __builtin_shufflevector(__s2_320, __s2_320, __lane_reverse_64_32); \
	48943	__ret_320 = __noswap_vset_lane_f32(__noswap_vget_lane_f32(__rev2_320, __p3_320), __rev0_320, __p1_320); \
	48944	__ret_320 = __builtin_shufflevector(__ret_320, __ret_320, __lane_reverse_64_32); \
	48945	__ret_320; \
48055	48946	})
48056	48947	#endif
48057	48948
48058	48949	#ifdef __LITTLE_ENDIAN__
48059		#define vcopyq_laneq_p16(__p0_322, __p1_322, __p2_322, __p3_322) __extension__ ({ \
48060		poly16x8_t __ret_322; \
48061		poly16x8_t __s0_322 = __p0_322; \
48062		poly16x8_t __s2_322 = __p2_322; \
48063		__ret_322 = vsetq_lane_p16(vgetq_lane_p16(__s2_322, __p3_322), __s0_322, __p1_322); \
48064		__ret_322; \
	48950	#define vcopy_lane_s32(__p0_321, __p1_321, __p2_321, __p3_321) __extension__ ({ \
	48951	int32x2_t __ret_321; \
	48952	int32x2_t __s0_321 = __p0_321; \
	48953	int32x2_t __s2_321 = __p2_321; \
	48954	__ret_321 = vset_lane_s32(vget_lane_s32(__s2_321, __p3_321), __s0_321, __p1_321); \
	48955	__ret_321; \
48065	48956	})
48066	48957	#else
48067		#define vcopyq_laneq_p16(__p0_323, __p1_323, __p2_323, __p3_323) __extension__ ({ \
48068		poly16x8_t __ret_323; \
48069		poly16x8_t __s0_323 = __p0_323; \
48070		poly16x8_t __s2_323 = __p2_323; \
48071		poly16x8_t __rev0_323; __rev0_323 = __builtin_shufflevector(__s0_323, __s0_323, 7, 6, 5, 4, 3, 2, 1, 0); \
48072		poly16x8_t __rev2_323; __rev2_323 = __builtin_shufflevector(__s2_323, __s2_323, 7, 6, 5, 4, 3, 2, 1, 0); \
48073		__ret_323 = __noswap_vsetq_lane_p16(__noswap_vgetq_lane_p16(__rev2_323, __p3_323), __rev0_323, __p1_323); \
48074		__ret_323 = __builtin_shufflevector(__ret_323, __ret_323, 7, 6, 5, 4, 3, 2, 1, 0); \
48075		__ret_323; \
	48958	#define vcopy_lane_s32(__p0_322, __p1_322, __p2_322, __p3_322) __extension__ ({ \
	48959	int32x2_t __ret_322; \
	48960	int32x2_t __s0_322 = __p0_322; \
	48961	int32x2_t __s2_322 = __p2_322; \
	48962	int32x2_t __rev0_322; __rev0_322 = __builtin_shufflevector(__s0_322, __s0_322, __lane_reverse_64_32); \
	48963	int32x2_t __rev2_322; __rev2_322 = __builtin_shufflevector(__s2_322, __s2_322, __lane_reverse_64_32); \
	48964	__ret_322 = __noswap_vset_lane_s32(__noswap_vget_lane_s32(__rev2_322, __p3_322), __rev0_322, __p1_322); \
	48965	__ret_322 = __builtin_shufflevector(__ret_322, __ret_322, __lane_reverse_64_32); \
	48966	__ret_322; \
48076	48967	})
48077	48968	#endif
48078	48969
	48970	#define vcopy_lane_s64(__p0_323, __p1_323, __p2_323, __p3_323) __extension__ ({ \
	48971	int64x1_t __ret_323; \
	48972	int64x1_t __s0_323 = __p0_323; \
	48973	int64x1_t __s2_323 = __p2_323; \
	48974	__ret_323 = vset_lane_s64(vget_lane_s64(__s2_323, __p3_323), __s0_323, __p1_323); \
	48975	__ret_323; \
	48976	})
48079	48977	#ifdef __LITTLE_ENDIAN__
48080		#define vcopyq_laneq_u8(__p0_324, __p1_324, __p2_324, __p3_324) __extension__ ({ \
48081		uint8x16_t __ret_324; \
48082		uint8x16_t __s0_324 = __p0_324; \
48083		uint8x16_t __s2_324 = __p2_324; \
48084		__ret_324 = vsetq_lane_u8(vgetq_lane_u8(__s2_324, __p3_324), __s0_324, __p1_324); \
	48978	#define vcopy_lane_mf8(__p0_324, __p1_324, __p2_324, __p3_324) __extension__ ({ \
	48979	mfloat8x8_t __ret_324; \
	48980	mfloat8x8_t __s0_324 = __p0_324; \
	48981	mfloat8x8_t __s2_324 = __p2_324; \
	48982	__ret_324 = vset_lane_mf8(vget_lane_mf8(__s2_324, __p3_324), __s0_324, __p1_324); \
48085	48983	__ret_324; \
48086	48984	})
48087	48985	#else
48088		#define vcopyq_laneq_u8(__p0_325, __p1_325, __p2_325, __p3_325) __extension__ ({ \
48089		uint8x16_t __ret_325; \
48090		uint8x16_t __s0_325 = __p0_325; \
48091		uint8x16_t __s2_325 = __p2_325; \
48092		uint8x16_t __rev0_325; __rev0_325 = __builtin_shufflevector(__s0_325, __s0_325, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
48093		uint8x16_t __rev2_325; __rev2_325 = __builtin_shufflevector(__s2_325, __s2_325, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
48094		__ret_325 = __noswap_vsetq_lane_u8(__noswap_vgetq_lane_u8(__rev2_325, __p3_325), __rev0_325, __p1_325); \
48095		__ret_325 = __builtin_shufflevector(__ret_325, __ret_325, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	48986	#define vcopy_lane_mf8(__p0_325, __p1_325, __p2_325, __p3_325) __extension__ ({ \
	48987	mfloat8x8_t __ret_325; \
	48988	mfloat8x8_t __s0_325 = __p0_325; \
	48989	mfloat8x8_t __s2_325 = __p2_325; \
	48990	mfloat8x8_t __rev0_325; __rev0_325 = __builtin_shufflevector(__s0_325, __s0_325, __lane_reverse_64_8); \
	48991	mfloat8x8_t __rev2_325; __rev2_325 = __builtin_shufflevector(__s2_325, __s2_325, __lane_reverse_64_8); \
	48992	__ret_325 = __noswap_vset_lane_mf8(__noswap_vget_lane_mf8(__rev2_325, __p3_325), __rev0_325, __p1_325); \
	48993	__ret_325 = __builtin_shufflevector(__ret_325, __ret_325, __lane_reverse_64_8); \
48096	48994	__ret_325; \
48097	48995	})
48098	48996	#endif
48099	48997
48100	48998	#ifdef __LITTLE_ENDIAN__
48101		#define vcopyq_laneq_u32(__p0_326, __p1_326, __p2_326, __p3_326) __extension__ ({ \
48102		uint32x4_t __ret_326; \
48103		uint32x4_t __s0_326 = __p0_326; \
48104		uint32x4_t __s2_326 = __p2_326; \
48105		__ret_326 = vsetq_lane_u32(vgetq_lane_u32(__s2_326, __p3_326), __s0_326, __p1_326); \
	48999	#define vcopy_lane_s16(__p0_326, __p1_326, __p2_326, __p3_326) __extension__ ({ \
	49000	int16x4_t __ret_326; \
	49001	int16x4_t __s0_326 = __p0_326; \
	49002	int16x4_t __s2_326 = __p2_326; \
	49003	__ret_326 = vset_lane_s16(vget_lane_s16(__s2_326, __p3_326), __s0_326, __p1_326); \
48106	49004	__ret_326; \
48107	49005	})
48108	49006	#else
48109		#define vcopyq_laneq_u32(__p0_327, __p1_327, __p2_327, __p3_327) __extension__ ({ \
48110		uint32x4_t __ret_327; \
48111		uint32x4_t __s0_327 = __p0_327; \
48112		uint32x4_t __s2_327 = __p2_327; \
48113		uint32x4_t __rev0_327; __rev0_327 = __builtin_shufflevector(__s0_327, __s0_327, 3, 2, 1, 0); \
48114		uint32x4_t __rev2_327; __rev2_327 = __builtin_shufflevector(__s2_327, __s2_327, 3, 2, 1, 0); \
48115		__ret_327 = __noswap_vsetq_lane_u32(__noswap_vgetq_lane_u32(__rev2_327, __p3_327), __rev0_327, __p1_327); \
48116		__ret_327 = __builtin_shufflevector(__ret_327, __ret_327, 3, 2, 1, 0); \
	49007	#define vcopy_lane_s16(__p0_327, __p1_327, __p2_327, __p3_327) __extension__ ({ \
	49008	int16x4_t __ret_327; \
	49009	int16x4_t __s0_327 = __p0_327; \
	49010	int16x4_t __s2_327 = __p2_327; \
	49011	int16x4_t __rev0_327; __rev0_327 = __builtin_shufflevector(__s0_327, __s0_327, __lane_reverse_64_16); \
	49012	int16x4_t __rev2_327; __rev2_327 = __builtin_shufflevector(__s2_327, __s2_327, __lane_reverse_64_16); \
	49013	__ret_327 = __noswap_vset_lane_s16(__noswap_vget_lane_s16(__rev2_327, __p3_327), __rev0_327, __p1_327); \
	49014	__ret_327 = __builtin_shufflevector(__ret_327, __ret_327, __lane_reverse_64_16); \
48117	49015	__ret_327; \
48118	49016	})
48119	49017	#endif
48120	49018
48121	49019	#ifdef __LITTLE_ENDIAN__
48122		#define vcopyq_laneq_u64(__p0_328, __p1_328, __p2_328, __p3_328) __extension__ ({ \
48123		uint64x2_t __ret_328; \
48124		uint64x2_t __s0_328 = __p0_328; \
48125		uint64x2_t __s2_328 = __p2_328; \
48126		__ret_328 = vsetq_lane_u64(vgetq_lane_u64(__s2_328, __p3_328), __s0_328, __p1_328); \
	49020	#define vcopyq_laneq_p8(__p0_328, __p1_328, __p2_328, __p3_328) __extension__ ({ \
	49021	poly8x16_t __ret_328; \
	49022	poly8x16_t __s0_328 = __p0_328; \
	49023	poly8x16_t __s2_328 = __p2_328; \
	49024	__ret_328 = vsetq_lane_p8(vgetq_lane_p8(__s2_328, __p3_328), __s0_328, __p1_328); \
48127	49025	__ret_328; \
48128	49026	})
48129	49027	#else
48130		#define vcopyq_laneq_u64(__p0_329, __p1_329, __p2_329, __p3_329) __extension__ ({ \
48131		uint64x2_t __ret_329; \
48132		uint64x2_t __s0_329 = __p0_329; \
48133		uint64x2_t __s2_329 = __p2_329; \
48134		uint64x2_t __rev0_329; __rev0_329 = __builtin_shufflevector(__s0_329, __s0_329, 1, 0); \
48135		uint64x2_t __rev2_329; __rev2_329 = __builtin_shufflevector(__s2_329, __s2_329, 1, 0); \
48136		__ret_329 = __noswap_vsetq_lane_u64(__noswap_vgetq_lane_u64(__rev2_329, __p3_329), __rev0_329, __p1_329); \
48137		__ret_329 = __builtin_shufflevector(__ret_329, __ret_329, 1, 0); \
	49028	#define vcopyq_laneq_p8(__p0_329, __p1_329, __p2_329, __p3_329) __extension__ ({ \
	49029	poly8x16_t __ret_329; \
	49030	poly8x16_t __s0_329 = __p0_329; \
	49031	poly8x16_t __s2_329 = __p2_329; \
	49032	poly8x16_t __rev0_329; __rev0_329 = __builtin_shufflevector(__s0_329, __s0_329, __lane_reverse_128_8); \
	49033	poly8x16_t __rev2_329; __rev2_329 = __builtin_shufflevector(__s2_329, __s2_329, __lane_reverse_128_8); \
	49034	__ret_329 = __noswap_vsetq_lane_p8(__noswap_vgetq_lane_p8(__rev2_329, __p3_329), __rev0_329, __p1_329); \
	49035	__ret_329 = __builtin_shufflevector(__ret_329, __ret_329, __lane_reverse_128_8); \
48138	49036	__ret_329; \
48139	49037	})
48140	49038	#endif
48141	49039
48142	49040	#ifdef __LITTLE_ENDIAN__
48143		#define vcopyq_laneq_u16(__p0_330, __p1_330, __p2_330, __p3_330) __extension__ ({ \
48144		uint16x8_t __ret_330; \
48145		uint16x8_t __s0_330 = __p0_330; \
48146		uint16x8_t __s2_330 = __p2_330; \
48147		__ret_330 = vsetq_lane_u16(vgetq_lane_u16(__s2_330, __p3_330), __s0_330, __p1_330); \
	49041	#define vcopyq_laneq_p16(__p0_330, __p1_330, __p2_330, __p3_330) __extension__ ({ \
	49042	poly16x8_t __ret_330; \
	49043	poly16x8_t __s0_330 = __p0_330; \
	49044	poly16x8_t __s2_330 = __p2_330; \
	49045	__ret_330 = vsetq_lane_p16(vgetq_lane_p16(__s2_330, __p3_330), __s0_330, __p1_330); \
48148	49046	__ret_330; \
48149	49047	})
48150	49048	#else
48151		#define vcopyq_laneq_u16(__p0_331, __p1_331, __p2_331, __p3_331) __extension__ ({ \
48152		uint16x8_t __ret_331; \
48153		uint16x8_t __s0_331 = __p0_331; \
48154		uint16x8_t __s2_331 = __p2_331; \
48155		uint16x8_t __rev0_331; __rev0_331 = __builtin_shufflevector(__s0_331, __s0_331, 7, 6, 5, 4, 3, 2, 1, 0); \
48156		uint16x8_t __rev2_331; __rev2_331 = __builtin_shufflevector(__s2_331, __s2_331, 7, 6, 5, 4, 3, 2, 1, 0); \
48157		__ret_331 = __noswap_vsetq_lane_u16(__noswap_vgetq_lane_u16(__rev2_331, __p3_331), __rev0_331, __p1_331); \
48158		__ret_331 = __builtin_shufflevector(__ret_331, __ret_331, 7, 6, 5, 4, 3, 2, 1, 0); \
	49049	#define vcopyq_laneq_p16(__p0_331, __p1_331, __p2_331, __p3_331) __extension__ ({ \
	49050	poly16x8_t __ret_331; \
	49051	poly16x8_t __s0_331 = __p0_331; \
	49052	poly16x8_t __s2_331 = __p2_331; \
	49053	poly16x8_t __rev0_331; __rev0_331 = __builtin_shufflevector(__s0_331, __s0_331, __lane_reverse_128_16); \
	49054	poly16x8_t __rev2_331; __rev2_331 = __builtin_shufflevector(__s2_331, __s2_331, __lane_reverse_128_16); \
	49055	__ret_331 = __noswap_vsetq_lane_p16(__noswap_vgetq_lane_p16(__rev2_331, __p3_331), __rev0_331, __p1_331); \
	49056	__ret_331 = __builtin_shufflevector(__ret_331, __ret_331, __lane_reverse_128_16); \
48159	49057	__ret_331; \
48160	49058	})
48161	49059	#endif
48162	49060
48163	49061	#ifdef __LITTLE_ENDIAN__
48164		#define vcopyq_laneq_s8(__p0_332, __p1_332, __p2_332, __p3_332) __extension__ ({ \
48165		int8x16_t __ret_332; \
48166		int8x16_t __s0_332 = __p0_332; \
48167		int8x16_t __s2_332 = __p2_332; \
48168		__ret_332 = vsetq_lane_s8(vgetq_lane_s8(__s2_332, __p3_332), __s0_332, __p1_332); \
	49062	#define vcopyq_laneq_u8(__p0_332, __p1_332, __p2_332, __p3_332) __extension__ ({ \
	49063	uint8x16_t __ret_332; \
	49064	uint8x16_t __s0_332 = __p0_332; \
	49065	uint8x16_t __s2_332 = __p2_332; \
	49066	__ret_332 = vsetq_lane_u8(vgetq_lane_u8(__s2_332, __p3_332), __s0_332, __p1_332); \
48169	49067	__ret_332; \
48170	49068	})
48171	49069	#else
48172		#define vcopyq_laneq_s8(__p0_333, __p1_333, __p2_333, __p3_333) __extension__ ({ \
48173		int8x16_t __ret_333; \
48174		int8x16_t __s0_333 = __p0_333; \
48175		int8x16_t __s2_333 = __p2_333; \
48176		int8x16_t __rev0_333; __rev0_333 = __builtin_shufflevector(__s0_333, __s0_333, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
48177		int8x16_t __rev2_333; __rev2_333 = __builtin_shufflevector(__s2_333, __s2_333, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
48178		__ret_333 = __noswap_vsetq_lane_s8(__noswap_vgetq_lane_s8(__rev2_333, __p3_333), __rev0_333, __p1_333); \
48179		__ret_333 = __builtin_shufflevector(__ret_333, __ret_333, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	49070	#define vcopyq_laneq_u8(__p0_333, __p1_333, __p2_333, __p3_333) __extension__ ({ \
	49071	uint8x16_t __ret_333; \
	49072	uint8x16_t __s0_333 = __p0_333; \
	49073	uint8x16_t __s2_333 = __p2_333; \
	49074	uint8x16_t __rev0_333; __rev0_333 = __builtin_shufflevector(__s0_333, __s0_333, __lane_reverse_128_8); \
	49075	uint8x16_t __rev2_333; __rev2_333 = __builtin_shufflevector(__s2_333, __s2_333, __lane_reverse_128_8); \
	49076	__ret_333 = __noswap_vsetq_lane_u8(__noswap_vgetq_lane_u8(__rev2_333, __p3_333), __rev0_333, __p1_333); \
	49077	__ret_333 = __builtin_shufflevector(__ret_333, __ret_333, __lane_reverse_128_8); \
48180	49078	__ret_333; \
48181	49079	})
48182	49080	#endif
48183	49081
48184	49082	#ifdef __LITTLE_ENDIAN__
48185		#define vcopyq_laneq_f32(__p0_334, __p1_334, __p2_334, __p3_334) __extension__ ({ \
48186		float32x4_t __ret_334; \
48187		float32x4_t __s0_334 = __p0_334; \
48188		float32x4_t __s2_334 = __p2_334; \
48189		__ret_334 = vsetq_lane_f32(vgetq_lane_f32(__s2_334, __p3_334), __s0_334, __p1_334); \
	49083	#define vcopyq_laneq_u32(__p0_334, __p1_334, __p2_334, __p3_334) __extension__ ({ \
	49084	uint32x4_t __ret_334; \
	49085	uint32x4_t __s0_334 = __p0_334; \
	49086	uint32x4_t __s2_334 = __p2_334; \
	49087	__ret_334 = vsetq_lane_u32(vgetq_lane_u32(__s2_334, __p3_334), __s0_334, __p1_334); \
48190	49088	__ret_334; \
48191	49089	})
48192	49090	#else
48193		#define vcopyq_laneq_f32(__p0_335, __p1_335, __p2_335, __p3_335) __extension__ ({ \
48194		float32x4_t __ret_335; \
48195		float32x4_t __s0_335 = __p0_335; \
48196		float32x4_t __s2_335 = __p2_335; \
48197		float32x4_t __rev0_335; __rev0_335 = __builtin_shufflevector(__s0_335, __s0_335, 3, 2, 1, 0); \
48198		float32x4_t __rev2_335; __rev2_335 = __builtin_shufflevector(__s2_335, __s2_335, 3, 2, 1, 0); \
48199		__ret_335 = __noswap_vsetq_lane_f32(__noswap_vgetq_lane_f32(__rev2_335, __p3_335), __rev0_335, __p1_335); \
48200		__ret_335 = __builtin_shufflevector(__ret_335, __ret_335, 3, 2, 1, 0); \
	49091	#define vcopyq_laneq_u32(__p0_335, __p1_335, __p2_335, __p3_335) __extension__ ({ \
	49092	uint32x4_t __ret_335; \
	49093	uint32x4_t __s0_335 = __p0_335; \
	49094	uint32x4_t __s2_335 = __p2_335; \
	49095	uint32x4_t __rev0_335; __rev0_335 = __builtin_shufflevector(__s0_335, __s0_335, __lane_reverse_128_32); \
	49096	uint32x4_t __rev2_335; __rev2_335 = __builtin_shufflevector(__s2_335, __s2_335, __lane_reverse_128_32); \
	49097	__ret_335 = __noswap_vsetq_lane_u32(__noswap_vgetq_lane_u32(__rev2_335, __p3_335), __rev0_335, __p1_335); \
	49098	__ret_335 = __builtin_shufflevector(__ret_335, __ret_335, __lane_reverse_128_32); \
48201	49099	__ret_335; \
48202	49100	})
48203	49101	#endif
48204	49102
48205	49103	#ifdef __LITTLE_ENDIAN__
48206		#define vcopyq_laneq_s32(__p0_336, __p1_336, __p2_336, __p3_336) __extension__ ({ \
48207		int32x4_t __ret_336; \
48208		int32x4_t __s0_336 = __p0_336; \
48209		int32x4_t __s2_336 = __p2_336; \
48210		__ret_336 = vsetq_lane_s32(vgetq_lane_s32(__s2_336, __p3_336), __s0_336, __p1_336); \
	49104	#define vcopyq_laneq_u64(__p0_336, __p1_336, __p2_336, __p3_336) __extension__ ({ \
	49105	uint64x2_t __ret_336; \
	49106	uint64x2_t __s0_336 = __p0_336; \
	49107	uint64x2_t __s2_336 = __p2_336; \
	49108	__ret_336 = vsetq_lane_u64(vgetq_lane_u64(__s2_336, __p3_336), __s0_336, __p1_336); \
48211	49109	__ret_336; \
48212	49110	})
48213	49111	#else
48214		#define vcopyq_laneq_s32(__p0_337, __p1_337, __p2_337, __p3_337) __extension__ ({ \
48215		int32x4_t __ret_337; \
48216		int32x4_t __s0_337 = __p0_337; \
48217		int32x4_t __s2_337 = __p2_337; \
48218		int32x4_t __rev0_337; __rev0_337 = __builtin_shufflevector(__s0_337, __s0_337, 3, 2, 1, 0); \
48219		int32x4_t __rev2_337; __rev2_337 = __builtin_shufflevector(__s2_337, __s2_337, 3, 2, 1, 0); \
48220		__ret_337 = __noswap_vsetq_lane_s32(__noswap_vgetq_lane_s32(__rev2_337, __p3_337), __rev0_337, __p1_337); \
48221		__ret_337 = __builtin_shufflevector(__ret_337, __ret_337, 3, 2, 1, 0); \
	49112	#define vcopyq_laneq_u64(__p0_337, __p1_337, __p2_337, __p3_337) __extension__ ({ \
	49113	uint64x2_t __ret_337; \
	49114	uint64x2_t __s0_337 = __p0_337; \
	49115	uint64x2_t __s2_337 = __p2_337; \
	49116	uint64x2_t __rev0_337; __rev0_337 = __builtin_shufflevector(__s0_337, __s0_337, __lane_reverse_128_64); \
	49117	uint64x2_t __rev2_337; __rev2_337 = __builtin_shufflevector(__s2_337, __s2_337, __lane_reverse_128_64); \
	49118	__ret_337 = __noswap_vsetq_lane_u64(__noswap_vgetq_lane_u64(__rev2_337, __p3_337), __rev0_337, __p1_337); \
	49119	__ret_337 = __builtin_shufflevector(__ret_337, __ret_337, __lane_reverse_128_64); \
48222	49120	__ret_337; \
48223	49121	})
48224	49122	#endif
48225	49123
48226	49124	#ifdef __LITTLE_ENDIAN__
48227		#define vcopyq_laneq_s64(__p0_338, __p1_338, __p2_338, __p3_338) __extension__ ({ \
48228		int64x2_t __ret_338; \
48229		int64x2_t __s0_338 = __p0_338; \
48230		int64x2_t __s2_338 = __p2_338; \
48231		__ret_338 = vsetq_lane_s64(vgetq_lane_s64(__s2_338, __p3_338), __s0_338, __p1_338); \
	49125	#define vcopyq_laneq_u16(__p0_338, __p1_338, __p2_338, __p3_338) __extension__ ({ \
	49126	uint16x8_t __ret_338; \
	49127	uint16x8_t __s0_338 = __p0_338; \
	49128	uint16x8_t __s2_338 = __p2_338; \
	49129	__ret_338 = vsetq_lane_u16(vgetq_lane_u16(__s2_338, __p3_338), __s0_338, __p1_338); \
48232	49130	__ret_338; \
48233	49131	})
48234	49132	#else
48235		#define vcopyq_laneq_s64(__p0_339, __p1_339, __p2_339, __p3_339) __extension__ ({ \
48236		int64x2_t __ret_339; \
48237		int64x2_t __s0_339 = __p0_339; \
48238		int64x2_t __s2_339 = __p2_339; \
48239		int64x2_t __rev0_339; __rev0_339 = __builtin_shufflevector(__s0_339, __s0_339, 1, 0); \
48240		int64x2_t __rev2_339; __rev2_339 = __builtin_shufflevector(__s2_339, __s2_339, 1, 0); \
48241		__ret_339 = __noswap_vsetq_lane_s64(__noswap_vgetq_lane_s64(__rev2_339, __p3_339), __rev0_339, __p1_339); \
48242		__ret_339 = __builtin_shufflevector(__ret_339, __ret_339, 1, 0); \
	49133	#define vcopyq_laneq_u16(__p0_339, __p1_339, __p2_339, __p3_339) __extension__ ({ \
	49134	uint16x8_t __ret_339; \
	49135	uint16x8_t __s0_339 = __p0_339; \
	49136	uint16x8_t __s2_339 = __p2_339; \
	49137	uint16x8_t __rev0_339; __rev0_339 = __builtin_shufflevector(__s0_339, __s0_339, __lane_reverse_128_16); \
	49138	uint16x8_t __rev2_339; __rev2_339 = __builtin_shufflevector(__s2_339, __s2_339, __lane_reverse_128_16); \
	49139	__ret_339 = __noswap_vsetq_lane_u16(__noswap_vgetq_lane_u16(__rev2_339, __p3_339), __rev0_339, __p1_339); \
	49140	__ret_339 = __builtin_shufflevector(__ret_339, __ret_339, __lane_reverse_128_16); \
48243	49141	__ret_339; \
48244	49142	})
48245	49143	#endif
48246	49144
48247	49145	#ifdef __LITTLE_ENDIAN__
48248		#define vcopyq_laneq_s16(__p0_340, __p1_340, __p2_340, __p3_340) __extension__ ({ \
48249		int16x8_t __ret_340; \
48250		int16x8_t __s0_340 = __p0_340; \
48251		int16x8_t __s2_340 = __p2_340; \
48252		__ret_340 = vsetq_lane_s16(vgetq_lane_s16(__s2_340, __p3_340), __s0_340, __p1_340); \
	49146	#define vcopyq_laneq_s8(__p0_340, __p1_340, __p2_340, __p3_340) __extension__ ({ \
	49147	int8x16_t __ret_340; \
	49148	int8x16_t __s0_340 = __p0_340; \
	49149	int8x16_t __s2_340 = __p2_340; \
	49150	__ret_340 = vsetq_lane_s8(vgetq_lane_s8(__s2_340, __p3_340), __s0_340, __p1_340); \
48253	49151	__ret_340; \
48254	49152	})
48255	49153	#else
48256		#define vcopyq_laneq_s16(__p0_341, __p1_341, __p2_341, __p3_341) __extension__ ({ \
48257		int16x8_t __ret_341; \
48258		int16x8_t __s0_341 = __p0_341; \
48259		int16x8_t __s2_341 = __p2_341; \
48260		int16x8_t __rev0_341; __rev0_341 = __builtin_shufflevector(__s0_341, __s0_341, 7, 6, 5, 4, 3, 2, 1, 0); \
48261		int16x8_t __rev2_341; __rev2_341 = __builtin_shufflevector(__s2_341, __s2_341, 7, 6, 5, 4, 3, 2, 1, 0); \
48262		__ret_341 = __noswap_vsetq_lane_s16(__noswap_vgetq_lane_s16(__rev2_341, __p3_341), __rev0_341, __p1_341); \
48263		__ret_341 = __builtin_shufflevector(__ret_341, __ret_341, 7, 6, 5, 4, 3, 2, 1, 0); \
	49154	#define vcopyq_laneq_s8(__p0_341, __p1_341, __p2_341, __p3_341) __extension__ ({ \
	49155	int8x16_t __ret_341; \
	49156	int8x16_t __s0_341 = __p0_341; \
	49157	int8x16_t __s2_341 = __p2_341; \
	49158	int8x16_t __rev0_341; __rev0_341 = __builtin_shufflevector(__s0_341, __s0_341, __lane_reverse_128_8); \
	49159	int8x16_t __rev2_341; __rev2_341 = __builtin_shufflevector(__s2_341, __s2_341, __lane_reverse_128_8); \
	49160	__ret_341 = __noswap_vsetq_lane_s8(__noswap_vgetq_lane_s8(__rev2_341, __p3_341), __rev0_341, __p1_341); \
	49161	__ret_341 = __builtin_shufflevector(__ret_341, __ret_341, __lane_reverse_128_8); \
48264	49162	__ret_341; \
48265	49163	})
48266	49164	#endif
48267	49165
48268	49166	#ifdef __LITTLE_ENDIAN__
48269		#define vcopy_laneq_p8(__p0_342, __p1_342, __p2_342, __p3_342) __extension__ ({ \
48270		poly8x8_t __ret_342; \
48271		poly8x8_t __s0_342 = __p0_342; \
48272		poly8x16_t __s2_342 = __p2_342; \
48273		__ret_342 = vset_lane_p8(vgetq_lane_p8(__s2_342, __p3_342), __s0_342, __p1_342); \
	49167	#define vcopyq_laneq_f32(__p0_342, __p1_342, __p2_342, __p3_342) __extension__ ({ \
	49168	float32x4_t __ret_342; \
	49169	float32x4_t __s0_342 = __p0_342; \
	49170	float32x4_t __s2_342 = __p2_342; \
	49171	__ret_342 = vsetq_lane_f32(vgetq_lane_f32(__s2_342, __p3_342), __s0_342, __p1_342); \
48274	49172	__ret_342; \
48275	49173	})
48276	49174	#else
48277		#define vcopy_laneq_p8(__p0_343, __p1_343, __p2_343, __p3_343) __extension__ ({ \
48278		poly8x8_t __ret_343; \
48279		poly8x8_t __s0_343 = __p0_343; \
48280		poly8x16_t __s2_343 = __p2_343; \
48281		poly8x8_t __rev0_343; __rev0_343 = __builtin_shufflevector(__s0_343, __s0_343, 7, 6, 5, 4, 3, 2, 1, 0); \
48282		poly8x16_t __rev2_343; __rev2_343 = __builtin_shufflevector(__s2_343, __s2_343, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
48283		__ret_343 = __noswap_vset_lane_p8(__noswap_vgetq_lane_p8(__rev2_343, __p3_343), __rev0_343, __p1_343); \
48284		__ret_343 = __builtin_shufflevector(__ret_343, __ret_343, 7, 6, 5, 4, 3, 2, 1, 0); \
	49175	#define vcopyq_laneq_f32(__p0_343, __p1_343, __p2_343, __p3_343) __extension__ ({ \
	49176	float32x4_t __ret_343; \
	49177	float32x4_t __s0_343 = __p0_343; \
	49178	float32x4_t __s2_343 = __p2_343; \
	49179	float32x4_t __rev0_343; __rev0_343 = __builtin_shufflevector(__s0_343, __s0_343, __lane_reverse_128_32); \
	49180	float32x4_t __rev2_343; __rev2_343 = __builtin_shufflevector(__s2_343, __s2_343, __lane_reverse_128_32); \
	49181	__ret_343 = __noswap_vsetq_lane_f32(__noswap_vgetq_lane_f32(__rev2_343, __p3_343), __rev0_343, __p1_343); \
	49182	__ret_343 = __builtin_shufflevector(__ret_343, __ret_343, __lane_reverse_128_32); \
48285	49183	__ret_343; \
48286	49184	})
48287	49185	#endif
48288	49186
48289	49187	#ifdef __LITTLE_ENDIAN__
48290		#define vcopy_laneq_p16(__p0_344, __p1_344, __p2_344, __p3_344) __extension__ ({ \
48291		poly16x4_t __ret_344; \
48292		poly16x4_t __s0_344 = __p0_344; \
48293		poly16x8_t __s2_344 = __p2_344; \
48294		__ret_344 = vset_lane_p16(vgetq_lane_p16(__s2_344, __p3_344), __s0_344, __p1_344); \
	49188	#define vcopyq_laneq_s32(__p0_344, __p1_344, __p2_344, __p3_344) __extension__ ({ \
	49189	int32x4_t __ret_344; \
	49190	int32x4_t __s0_344 = __p0_344; \
	49191	int32x4_t __s2_344 = __p2_344; \
	49192	__ret_344 = vsetq_lane_s32(vgetq_lane_s32(__s2_344, __p3_344), __s0_344, __p1_344); \
48295	49193	__ret_344; \
48296	49194	})
48297	49195	#else
48298		#define vcopy_laneq_p16(__p0_345, __p1_345, __p2_345, __p3_345) __extension__ ({ \
48299		poly16x4_t __ret_345; \
48300		poly16x4_t __s0_345 = __p0_345; \
48301		poly16x8_t __s2_345 = __p2_345; \
48302		poly16x4_t __rev0_345; __rev0_345 = __builtin_shufflevector(__s0_345, __s0_345, 3, 2, 1, 0); \
48303		poly16x8_t __rev2_345; __rev2_345 = __builtin_shufflevector(__s2_345, __s2_345, 7, 6, 5, 4, 3, 2, 1, 0); \
48304		__ret_345 = __noswap_vset_lane_p16(__noswap_vgetq_lane_p16(__rev2_345, __p3_345), __rev0_345, __p1_345); \
48305		__ret_345 = __builtin_shufflevector(__ret_345, __ret_345, 3, 2, 1, 0); \
	49196	#define vcopyq_laneq_s32(__p0_345, __p1_345, __p2_345, __p3_345) __extension__ ({ \
	49197	int32x4_t __ret_345; \
	49198	int32x4_t __s0_345 = __p0_345; \
	49199	int32x4_t __s2_345 = __p2_345; \
	49200	int32x4_t __rev0_345; __rev0_345 = __builtin_shufflevector(__s0_345, __s0_345, __lane_reverse_128_32); \
	49201	int32x4_t __rev2_345; __rev2_345 = __builtin_shufflevector(__s2_345, __s2_345, __lane_reverse_128_32); \
	49202	__ret_345 = __noswap_vsetq_lane_s32(__noswap_vgetq_lane_s32(__rev2_345, __p3_345), __rev0_345, __p1_345); \
	49203	__ret_345 = __builtin_shufflevector(__ret_345, __ret_345, __lane_reverse_128_32); \
48306	49204	__ret_345; \
48307	49205	})
48308	49206	#endif
48309	49207
48310	49208	#ifdef __LITTLE_ENDIAN__
48311		#define vcopy_laneq_u8(__p0_346, __p1_346, __p2_346, __p3_346) __extension__ ({ \
48312		uint8x8_t __ret_346; \
48313		uint8x8_t __s0_346 = __p0_346; \
48314		uint8x16_t __s2_346 = __p2_346; \
48315		__ret_346 = vset_lane_u8(vgetq_lane_u8(__s2_346, __p3_346), __s0_346, __p1_346); \
	49209	#define vcopyq_laneq_s64(__p0_346, __p1_346, __p2_346, __p3_346) __extension__ ({ \
	49210	int64x2_t __ret_346; \
	49211	int64x2_t __s0_346 = __p0_346; \
	49212	int64x2_t __s2_346 = __p2_346; \
	49213	__ret_346 = vsetq_lane_s64(vgetq_lane_s64(__s2_346, __p3_346), __s0_346, __p1_346); \
48316	49214	__ret_346; \
48317	49215	})
48318	49216	#else
48319		#define vcopy_laneq_u8(__p0_347, __p1_347, __p2_347, __p3_347) __extension__ ({ \
48320		uint8x8_t __ret_347; \
48321		uint8x8_t __s0_347 = __p0_347; \
48322		uint8x16_t __s2_347 = __p2_347; \
48323		uint8x8_t __rev0_347; __rev0_347 = __builtin_shufflevector(__s0_347, __s0_347, 7, 6, 5, 4, 3, 2, 1, 0); \
48324		uint8x16_t __rev2_347; __rev2_347 = __builtin_shufflevector(__s2_347, __s2_347, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
48325		__ret_347 = __noswap_vset_lane_u8(__noswap_vgetq_lane_u8(__rev2_347, __p3_347), __rev0_347, __p1_347); \
48326		__ret_347 = __builtin_shufflevector(__ret_347, __ret_347, 7, 6, 5, 4, 3, 2, 1, 0); \
	49217	#define vcopyq_laneq_s64(__p0_347, __p1_347, __p2_347, __p3_347) __extension__ ({ \
	49218	int64x2_t __ret_347; \
	49219	int64x2_t __s0_347 = __p0_347; \
	49220	int64x2_t __s2_347 = __p2_347; \
	49221	int64x2_t __rev0_347; __rev0_347 = __builtin_shufflevector(__s0_347, __s0_347, __lane_reverse_128_64); \
	49222	int64x2_t __rev2_347; __rev2_347 = __builtin_shufflevector(__s2_347, __s2_347, __lane_reverse_128_64); \
	49223	__ret_347 = __noswap_vsetq_lane_s64(__noswap_vgetq_lane_s64(__rev2_347, __p3_347), __rev0_347, __p1_347); \
	49224	__ret_347 = __builtin_shufflevector(__ret_347, __ret_347, __lane_reverse_128_64); \
48327	49225	__ret_347; \
48328	49226	})
48329	49227	#endif
48330	49228
48331	49229	#ifdef __LITTLE_ENDIAN__
48332		#define vcopy_laneq_u32(__p0_348, __p1_348, __p2_348, __p3_348) __extension__ ({ \
48333		uint32x2_t __ret_348; \
48334		uint32x2_t __s0_348 = __p0_348; \
48335		uint32x4_t __s2_348 = __p2_348; \
48336		__ret_348 = vset_lane_u32(vgetq_lane_u32(__s2_348, __p3_348), __s0_348, __p1_348); \
	49230	#define vcopyq_laneq_mf8(__p0_348, __p1_348, __p2_348, __p3_348) __extension__ ({ \
	49231	mfloat8x16_t __ret_348; \
	49232	mfloat8x16_t __s0_348 = __p0_348; \
	49233	mfloat8x16_t __s2_348 = __p2_348; \
	49234	__ret_348 = vsetq_lane_mf8(vgetq_lane_mf8(__s2_348, __p3_348), __s0_348, __p1_348); \
48337	49235	__ret_348; \
48338	49236	})
48339	49237	#else
48340		#define vcopy_laneq_u32(__p0_349, __p1_349, __p2_349, __p3_349) __extension__ ({ \
48341		uint32x2_t __ret_349; \
48342		uint32x2_t __s0_349 = __p0_349; \
48343		uint32x4_t __s2_349 = __p2_349; \
48344		uint32x2_t __rev0_349; __rev0_349 = __builtin_shufflevector(__s0_349, __s0_349, 1, 0); \
48345		uint32x4_t __rev2_349; __rev2_349 = __builtin_shufflevector(__s2_349, __s2_349, 3, 2, 1, 0); \
48346		__ret_349 = __noswap_vset_lane_u32(__noswap_vgetq_lane_u32(__rev2_349, __p3_349), __rev0_349, __p1_349); \
48347		__ret_349 = __builtin_shufflevector(__ret_349, __ret_349, 1, 0); \
	49238	#define vcopyq_laneq_mf8(__p0_349, __p1_349, __p2_349, __p3_349) __extension__ ({ \
	49239	mfloat8x16_t __ret_349; \
	49240	mfloat8x16_t __s0_349 = __p0_349; \
	49241	mfloat8x16_t __s2_349 = __p2_349; \
	49242	mfloat8x16_t __rev0_349; __rev0_349 = __builtin_shufflevector(__s0_349, __s0_349, __lane_reverse_128_8); \
	49243	mfloat8x16_t __rev2_349; __rev2_349 = __builtin_shufflevector(__s2_349, __s2_349, __lane_reverse_128_8); \
	49244	__ret_349 = __noswap_vsetq_lane_mf8(__noswap_vgetq_lane_mf8(__rev2_349, __p3_349), __rev0_349, __p1_349); \
	49245	__ret_349 = __builtin_shufflevector(__ret_349, __ret_349, __lane_reverse_128_8); \
48348	49246	__ret_349; \
48349	49247	})
48350	49248	#endif
48351	49249
48352	49250	#ifdef __LITTLE_ENDIAN__
48353		#define vcopy_laneq_u64(__p0_350, __p1_350, __p2_350, __p3_350) __extension__ ({ \
48354		uint64x1_t __ret_350; \
48355		uint64x1_t __s0_350 = __p0_350; \
48356		uint64x2_t __s2_350 = __p2_350; \
48357		__ret_350 = vset_lane_u64(vgetq_lane_u64(__s2_350, __p3_350), __s0_350, __p1_350); \
	49251	#define vcopyq_laneq_s16(__p0_350, __p1_350, __p2_350, __p3_350) __extension__ ({ \
	49252	int16x8_t __ret_350; \
	49253	int16x8_t __s0_350 = __p0_350; \
	49254	int16x8_t __s2_350 = __p2_350; \
	49255	__ret_350 = vsetq_lane_s16(vgetq_lane_s16(__s2_350, __p3_350), __s0_350, __p1_350); \
48358	49256	__ret_350; \
48359	49257	})
48360	49258	#else
48361		#define vcopy_laneq_u64(__p0_351, __p1_351, __p2_351, __p3_351) __extension__ ({ \
48362		uint64x1_t __ret_351; \
48363		uint64x1_t __s0_351 = __p0_351; \
48364		uint64x2_t __s2_351 = __p2_351; \
48365		uint64x2_t __rev2_351; __rev2_351 = __builtin_shufflevector(__s2_351, __s2_351, 1, 0); \
48366		__ret_351 = vset_lane_u64(__noswap_vgetq_lane_u64(__rev2_351, __p3_351), __s0_351, __p1_351); \
	49259	#define vcopyq_laneq_s16(__p0_351, __p1_351, __p2_351, __p3_351) __extension__ ({ \
	49260	int16x8_t __ret_351; \
	49261	int16x8_t __s0_351 = __p0_351; \
	49262	int16x8_t __s2_351 = __p2_351; \
	49263	int16x8_t __rev0_351; __rev0_351 = __builtin_shufflevector(__s0_351, __s0_351, __lane_reverse_128_16); \
	49264	int16x8_t __rev2_351; __rev2_351 = __builtin_shufflevector(__s2_351, __s2_351, __lane_reverse_128_16); \
	49265	__ret_351 = __noswap_vsetq_lane_s16(__noswap_vgetq_lane_s16(__rev2_351, __p3_351), __rev0_351, __p1_351); \
	49266	__ret_351 = __builtin_shufflevector(__ret_351, __ret_351, __lane_reverse_128_16); \
48367	49267	__ret_351; \
48368	49268	})
48369	49269	#endif
48370	49270
48371	49271	#ifdef __LITTLE_ENDIAN__
48372		#define vcopy_laneq_u16(__p0_352, __p1_352, __p2_352, __p3_352) __extension__ ({ \
48373		uint16x4_t __ret_352; \
48374		uint16x4_t __s0_352 = __p0_352; \
48375		uint16x8_t __s2_352 = __p2_352; \
48376		__ret_352 = vset_lane_u16(vgetq_lane_u16(__s2_352, __p3_352), __s0_352, __p1_352); \
	49272	#define vcopy_laneq_p8(__p0_352, __p1_352, __p2_352, __p3_352) __extension__ ({ \
	49273	poly8x8_t __ret_352; \
	49274	poly8x8_t __s0_352 = __p0_352; \
	49275	poly8x16_t __s2_352 = __p2_352; \
	49276	__ret_352 = vset_lane_p8(vgetq_lane_p8(__s2_352, __p3_352), __s0_352, __p1_352); \
48377	49277	__ret_352; \
48378	49278	})
48379	49279	#else
48380		#define vcopy_laneq_u16(__p0_353, __p1_353, __p2_353, __p3_353) __extension__ ({ \
48381		uint16x4_t __ret_353; \
48382		uint16x4_t __s0_353 = __p0_353; \
48383		uint16x8_t __s2_353 = __p2_353; \
48384		uint16x4_t __rev0_353; __rev0_353 = __builtin_shufflevector(__s0_353, __s0_353, 3, 2, 1, 0); \
48385		uint16x8_t __rev2_353; __rev2_353 = __builtin_shufflevector(__s2_353, __s2_353, 7, 6, 5, 4, 3, 2, 1, 0); \
48386		__ret_353 = __noswap_vset_lane_u16(__noswap_vgetq_lane_u16(__rev2_353, __p3_353), __rev0_353, __p1_353); \
48387		__ret_353 = __builtin_shufflevector(__ret_353, __ret_353, 3, 2, 1, 0); \
	49280	#define vcopy_laneq_p8(__p0_353, __p1_353, __p2_353, __p3_353) __extension__ ({ \
	49281	poly8x8_t __ret_353; \
	49282	poly8x8_t __s0_353 = __p0_353; \
	49283	poly8x16_t __s2_353 = __p2_353; \
	49284	poly8x8_t __rev0_353; __rev0_353 = __builtin_shufflevector(__s0_353, __s0_353, __lane_reverse_64_8); \
	49285	poly8x16_t __rev2_353; __rev2_353 = __builtin_shufflevector(__s2_353, __s2_353, __lane_reverse_128_8); \
	49286	__ret_353 = __noswap_vset_lane_p8(__noswap_vgetq_lane_p8(__rev2_353, __p3_353), __rev0_353, __p1_353); \
	49287	__ret_353 = __builtin_shufflevector(__ret_353, __ret_353, __lane_reverse_64_8); \
48388	49288	__ret_353; \
48389	49289	})
48390	49290	#endif
48391	49291
48392	49292	#ifdef __LITTLE_ENDIAN__
48393		#define vcopy_laneq_s8(__p0_354, __p1_354, __p2_354, __p3_354) __extension__ ({ \
48394		int8x8_t __ret_354; \
48395		int8x8_t __s0_354 = __p0_354; \
48396		int8x16_t __s2_354 = __p2_354; \
48397		__ret_354 = vset_lane_s8(vgetq_lane_s8(__s2_354, __p3_354), __s0_354, __p1_354); \
	49293	#define vcopy_laneq_p16(__p0_354, __p1_354, __p2_354, __p3_354) __extension__ ({ \
	49294	poly16x4_t __ret_354; \
	49295	poly16x4_t __s0_354 = __p0_354; \
	49296	poly16x8_t __s2_354 = __p2_354; \
	49297	__ret_354 = vset_lane_p16(vgetq_lane_p16(__s2_354, __p3_354), __s0_354, __p1_354); \
48398	49298	__ret_354; \
48399	49299	})
48400	49300	#else
48401		#define vcopy_laneq_s8(__p0_355, __p1_355, __p2_355, __p3_355) __extension__ ({ \
48402		int8x8_t __ret_355; \
48403		int8x8_t __s0_355 = __p0_355; \
48404		int8x16_t __s2_355 = __p2_355; \
48405		int8x8_t __rev0_355; __rev0_355 = __builtin_shufflevector(__s0_355, __s0_355, 7, 6, 5, 4, 3, 2, 1, 0); \
48406		int8x16_t __rev2_355; __rev2_355 = __builtin_shufflevector(__s2_355, __s2_355, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
48407		__ret_355 = __noswap_vset_lane_s8(__noswap_vgetq_lane_s8(__rev2_355, __p3_355), __rev0_355, __p1_355); \
48408		__ret_355 = __builtin_shufflevector(__ret_355, __ret_355, 7, 6, 5, 4, 3, 2, 1, 0); \
	49301	#define vcopy_laneq_p16(__p0_355, __p1_355, __p2_355, __p3_355) __extension__ ({ \
	49302	poly16x4_t __ret_355; \
	49303	poly16x4_t __s0_355 = __p0_355; \
	49304	poly16x8_t __s2_355 = __p2_355; \
	49305	poly16x4_t __rev0_355; __rev0_355 = __builtin_shufflevector(__s0_355, __s0_355, __lane_reverse_64_16); \
	49306	poly16x8_t __rev2_355; __rev2_355 = __builtin_shufflevector(__s2_355, __s2_355, __lane_reverse_128_16); \
	49307	__ret_355 = __noswap_vset_lane_p16(__noswap_vgetq_lane_p16(__rev2_355, __p3_355), __rev0_355, __p1_355); \
	49308	__ret_355 = __builtin_shufflevector(__ret_355, __ret_355, __lane_reverse_64_16); \
48409	49309	__ret_355; \
48410	49310	})
48411	49311	#endif
48412	49312
48413	49313	#ifdef __LITTLE_ENDIAN__
48414		#define vcopy_laneq_f32(__p0_356, __p1_356, __p2_356, __p3_356) __extension__ ({ \
48415		float32x2_t __ret_356; \
48416		float32x2_t __s0_356 = __p0_356; \
48417		float32x4_t __s2_356 = __p2_356; \
48418		__ret_356 = vset_lane_f32(vgetq_lane_f32(__s2_356, __p3_356), __s0_356, __p1_356); \
	49314	#define vcopy_laneq_u8(__p0_356, __p1_356, __p2_356, __p3_356) __extension__ ({ \
	49315	uint8x8_t __ret_356; \
	49316	uint8x8_t __s0_356 = __p0_356; \
	49317	uint8x16_t __s2_356 = __p2_356; \
	49318	__ret_356 = vset_lane_u8(vgetq_lane_u8(__s2_356, __p3_356), __s0_356, __p1_356); \
48419	49319	__ret_356; \
48420	49320	})
48421	49321	#else
48422		#define vcopy_laneq_f32(__p0_357, __p1_357, __p2_357, __p3_357) __extension__ ({ \
48423		float32x2_t __ret_357; \
48424		float32x2_t __s0_357 = __p0_357; \
48425		float32x4_t __s2_357 = __p2_357; \
48426		float32x2_t __rev0_357; __rev0_357 = __builtin_shufflevector(__s0_357, __s0_357, 1, 0); \
48427		float32x4_t __rev2_357; __rev2_357 = __builtin_shufflevector(__s2_357, __s2_357, 3, 2, 1, 0); \
48428		__ret_357 = __noswap_vset_lane_f32(__noswap_vgetq_lane_f32(__rev2_357, __p3_357), __rev0_357, __p1_357); \
48429		__ret_357 = __builtin_shufflevector(__ret_357, __ret_357, 1, 0); \
	49322	#define vcopy_laneq_u8(__p0_357, __p1_357, __p2_357, __p3_357) __extension__ ({ \
	49323	uint8x8_t __ret_357; \
	49324	uint8x8_t __s0_357 = __p0_357; \
	49325	uint8x16_t __s2_357 = __p2_357; \
	49326	uint8x8_t __rev0_357; __rev0_357 = __builtin_shufflevector(__s0_357, __s0_357, __lane_reverse_64_8); \
	49327	uint8x16_t __rev2_357; __rev2_357 = __builtin_shufflevector(__s2_357, __s2_357, __lane_reverse_128_8); \
	49328	__ret_357 = __noswap_vset_lane_u8(__noswap_vgetq_lane_u8(__rev2_357, __p3_357), __rev0_357, __p1_357); \
	49329	__ret_357 = __builtin_shufflevector(__ret_357, __ret_357, __lane_reverse_64_8); \
48430	49330	__ret_357; \
48431	49331	})
48432	49332	#endif
48433	49333
48434	49334	#ifdef __LITTLE_ENDIAN__
48435		#define vcopy_laneq_s32(__p0_358, __p1_358, __p2_358, __p3_358) __extension__ ({ \
48436		int32x2_t __ret_358; \
48437		int32x2_t __s0_358 = __p0_358; \
48438		int32x4_t __s2_358 = __p2_358; \
48439		__ret_358 = vset_lane_s32(vgetq_lane_s32(__s2_358, __p3_358), __s0_358, __p1_358); \
	49335	#define vcopy_laneq_u32(__p0_358, __p1_358, __p2_358, __p3_358) __extension__ ({ \
	49336	uint32x2_t __ret_358; \
	49337	uint32x2_t __s0_358 = __p0_358; \
	49338	uint32x4_t __s2_358 = __p2_358; \
	49339	__ret_358 = vset_lane_u32(vgetq_lane_u32(__s2_358, __p3_358), __s0_358, __p1_358); \
48440	49340	__ret_358; \
48441	49341	})
48442	49342	#else
48443		#define vcopy_laneq_s32(__p0_359, __p1_359, __p2_359, __p3_359) __extension__ ({ \
48444		int32x2_t __ret_359; \
48445		int32x2_t __s0_359 = __p0_359; \
48446		int32x4_t __s2_359 = __p2_359; \
48447		int32x2_t __rev0_359; __rev0_359 = __builtin_shufflevector(__s0_359, __s0_359, 1, 0); \
48448		int32x4_t __rev2_359; __rev2_359 = __builtin_shufflevector(__s2_359, __s2_359, 3, 2, 1, 0); \
48449		__ret_359 = __noswap_vset_lane_s32(__noswap_vgetq_lane_s32(__rev2_359, __p3_359), __rev0_359, __p1_359); \
48450		__ret_359 = __builtin_shufflevector(__ret_359, __ret_359, 1, 0); \
	49343	#define vcopy_laneq_u32(__p0_359, __p1_359, __p2_359, __p3_359) __extension__ ({ \
	49344	uint32x2_t __ret_359; \
	49345	uint32x2_t __s0_359 = __p0_359; \
	49346	uint32x4_t __s2_359 = __p2_359; \
	49347	uint32x2_t __rev0_359; __rev0_359 = __builtin_shufflevector(__s0_359, __s0_359, __lane_reverse_64_32); \
	49348	uint32x4_t __rev2_359; __rev2_359 = __builtin_shufflevector(__s2_359, __s2_359, __lane_reverse_128_32); \
	49349	__ret_359 = __noswap_vset_lane_u32(__noswap_vgetq_lane_u32(__rev2_359, __p3_359), __rev0_359, __p1_359); \
	49350	__ret_359 = __builtin_shufflevector(__ret_359, __ret_359, __lane_reverse_64_32); \
48451	49351	__ret_359; \
48452	49352	})
48453	49353	#endif
48454	49354
48455	49355	#ifdef __LITTLE_ENDIAN__
48456		#define vcopy_laneq_s64(__p0_360, __p1_360, __p2_360, __p3_360) __extension__ ({ \
48457		int64x1_t __ret_360; \
48458		int64x1_t __s0_360 = __p0_360; \
48459		int64x2_t __s2_360 = __p2_360; \
48460		__ret_360 = vset_lane_s64(vgetq_lane_s64(__s2_360, __p3_360), __s0_360, __p1_360); \
	49356	#define vcopy_laneq_u64(__p0_360, __p1_360, __p2_360, __p3_360) __extension__ ({ \
	49357	uint64x1_t __ret_360; \
	49358	uint64x1_t __s0_360 = __p0_360; \
	49359	uint64x2_t __s2_360 = __p2_360; \
	49360	__ret_360 = vset_lane_u64(vgetq_lane_u64(__s2_360, __p3_360), __s0_360, __p1_360); \
48461	49361	__ret_360; \
48462	49362	})
48463	49363	#else
48464		#define vcopy_laneq_s64(__p0_361, __p1_361, __p2_361, __p3_361) __extension__ ({ \
48465		int64x1_t __ret_361; \
48466		int64x1_t __s0_361 = __p0_361; \
48467		int64x2_t __s2_361 = __p2_361; \
48468		int64x2_t __rev2_361; __rev2_361 = __builtin_shufflevector(__s2_361, __s2_361, 1, 0); \
48469		__ret_361 = vset_lane_s64(__noswap_vgetq_lane_s64(__rev2_361, __p3_361), __s0_361, __p1_361); \
	49364	#define vcopy_laneq_u64(__p0_361, __p1_361, __p2_361, __p3_361) __extension__ ({ \
	49365	uint64x1_t __ret_361; \
	49366	uint64x1_t __s0_361 = __p0_361; \
	49367	uint64x2_t __s2_361 = __p2_361; \
	49368	uint64x2_t __rev2_361; __rev2_361 = __builtin_shufflevector(__s2_361, __s2_361, __lane_reverse_128_64); \
	49369	__ret_361 = vset_lane_u64(__noswap_vgetq_lane_u64(__rev2_361, __p3_361), __s0_361, __p1_361); \
48470	49370	__ret_361; \
48471	49371	})
48472	49372	#endif
48473	49373
48474	49374	#ifdef __LITTLE_ENDIAN__
48475		#define vcopy_laneq_s16(__p0_362, __p1_362, __p2_362, __p3_362) __extension__ ({ \
48476		int16x4_t __ret_362; \
48477		int16x4_t __s0_362 = __p0_362; \
48478		int16x8_t __s2_362 = __p2_362; \
48479		__ret_362 = vset_lane_s16(vgetq_lane_s16(__s2_362, __p3_362), __s0_362, __p1_362); \
	49375	#define vcopy_laneq_u16(__p0_362, __p1_362, __p2_362, __p3_362) __extension__ ({ \
	49376	uint16x4_t __ret_362; \
	49377	uint16x4_t __s0_362 = __p0_362; \
	49378	uint16x8_t __s2_362 = __p2_362; \
	49379	__ret_362 = vset_lane_u16(vgetq_lane_u16(__s2_362, __p3_362), __s0_362, __p1_362); \
48480	49380	__ret_362; \
48481	49381	})
48482	49382	#else
48483		#define vcopy_laneq_s16(__p0_363, __p1_363, __p2_363, __p3_363) __extension__ ({ \
48484		int16x4_t __ret_363; \
48485		int16x4_t __s0_363 = __p0_363; \
48486		int16x8_t __s2_363 = __p2_363; \
48487		int16x4_t __rev0_363; __rev0_363 = __builtin_shufflevector(__s0_363, __s0_363, 3, 2, 1, 0); \
48488		int16x8_t __rev2_363; __rev2_363 = __builtin_shufflevector(__s2_363, __s2_363, 7, 6, 5, 4, 3, 2, 1, 0); \
48489		__ret_363 = __noswap_vset_lane_s16(__noswap_vgetq_lane_s16(__rev2_363, __p3_363), __rev0_363, __p1_363); \
48490		__ret_363 = __builtin_shufflevector(__ret_363, __ret_363, 3, 2, 1, 0); \
	49383	#define vcopy_laneq_u16(__p0_363, __p1_363, __p2_363, __p3_363) __extension__ ({ \
	49384	uint16x4_t __ret_363; \
	49385	uint16x4_t __s0_363 = __p0_363; \
	49386	uint16x8_t __s2_363 = __p2_363; \
	49387	uint16x4_t __rev0_363; __rev0_363 = __builtin_shufflevector(__s0_363, __s0_363, __lane_reverse_64_16); \
	49388	uint16x8_t __rev2_363; __rev2_363 = __builtin_shufflevector(__s2_363, __s2_363, __lane_reverse_128_16); \
	49389	__ret_363 = __noswap_vset_lane_u16(__noswap_vgetq_lane_u16(__rev2_363, __p3_363), __rev0_363, __p1_363); \
	49390	__ret_363 = __builtin_shufflevector(__ret_363, __ret_363, __lane_reverse_64_16); \
48491	49391	__ret_363; \
48492	49392	})
48493	49393	#endif
48494	49394
	49395	#ifdef __LITTLE_ENDIAN__
	49396	#define vcopy_laneq_s8(__p0_364, __p1_364, __p2_364, __p3_364) __extension__ ({ \
	49397	int8x8_t __ret_364; \
	49398	int8x8_t __s0_364 = __p0_364; \
	49399	int8x16_t __s2_364 = __p2_364; \
	49400	__ret_364 = vset_lane_s8(vgetq_lane_s8(__s2_364, __p3_364), __s0_364, __p1_364); \
	49401	__ret_364; \
	49402	})
	49403	#else
	49404	#define vcopy_laneq_s8(__p0_365, __p1_365, __p2_365, __p3_365) __extension__ ({ \
	49405	int8x8_t __ret_365; \
	49406	int8x8_t __s0_365 = __p0_365; \
	49407	int8x16_t __s2_365 = __p2_365; \
	49408	int8x8_t __rev0_365; __rev0_365 = __builtin_shufflevector(__s0_365, __s0_365, __lane_reverse_64_8); \
	49409	int8x16_t __rev2_365; __rev2_365 = __builtin_shufflevector(__s2_365, __s2_365, __lane_reverse_128_8); \
	49410	__ret_365 = __noswap_vset_lane_s8(__noswap_vgetq_lane_s8(__rev2_365, __p3_365), __rev0_365, __p1_365); \
	49411	__ret_365 = __builtin_shufflevector(__ret_365, __ret_365, __lane_reverse_64_8); \
	49412	__ret_365; \
	49413	})
	49414	#endif
	49415
	49416	#ifdef __LITTLE_ENDIAN__
	49417	#define vcopy_laneq_f32(__p0_366, __p1_366, __p2_366, __p3_366) __extension__ ({ \
	49418	float32x2_t __ret_366; \
	49419	float32x2_t __s0_366 = __p0_366; \
	49420	float32x4_t __s2_366 = __p2_366; \
	49421	__ret_366 = vset_lane_f32(vgetq_lane_f32(__s2_366, __p3_366), __s0_366, __p1_366); \
	49422	__ret_366; \
	49423	})
	49424	#else
	49425	#define vcopy_laneq_f32(__p0_367, __p1_367, __p2_367, __p3_367) __extension__ ({ \
	49426	float32x2_t __ret_367; \
	49427	float32x2_t __s0_367 = __p0_367; \
	49428	float32x4_t __s2_367 = __p2_367; \
	49429	float32x2_t __rev0_367; __rev0_367 = __builtin_shufflevector(__s0_367, __s0_367, __lane_reverse_64_32); \
	49430	float32x4_t __rev2_367; __rev2_367 = __builtin_shufflevector(__s2_367, __s2_367, __lane_reverse_128_32); \
	49431	__ret_367 = __noswap_vset_lane_f32(__noswap_vgetq_lane_f32(__rev2_367, __p3_367), __rev0_367, __p1_367); \
	49432	__ret_367 = __builtin_shufflevector(__ret_367, __ret_367, __lane_reverse_64_32); \
	49433	__ret_367; \
	49434	})
	49435	#endif
	49436
	49437	#ifdef __LITTLE_ENDIAN__
	49438	#define vcopy_laneq_s32(__p0_368, __p1_368, __p2_368, __p3_368) __extension__ ({ \
	49439	int32x2_t __ret_368; \
	49440	int32x2_t __s0_368 = __p0_368; \
	49441	int32x4_t __s2_368 = __p2_368; \
	49442	__ret_368 = vset_lane_s32(vgetq_lane_s32(__s2_368, __p3_368), __s0_368, __p1_368); \
	49443	__ret_368; \
	49444	})
	49445	#else
	49446	#define vcopy_laneq_s32(__p0_369, __p1_369, __p2_369, __p3_369) __extension__ ({ \
	49447	int32x2_t __ret_369; \
	49448	int32x2_t __s0_369 = __p0_369; \
	49449	int32x4_t __s2_369 = __p2_369; \
	49450	int32x2_t __rev0_369; __rev0_369 = __builtin_shufflevector(__s0_369, __s0_369, __lane_reverse_64_32); \
	49451	int32x4_t __rev2_369; __rev2_369 = __builtin_shufflevector(__s2_369, __s2_369, __lane_reverse_128_32); \
	49452	__ret_369 = __noswap_vset_lane_s32(__noswap_vgetq_lane_s32(__rev2_369, __p3_369), __rev0_369, __p1_369); \
	49453	__ret_369 = __builtin_shufflevector(__ret_369, __ret_369, __lane_reverse_64_32); \
	49454	__ret_369; \
	49455	})
	49456	#endif
	49457
	49458	#ifdef __LITTLE_ENDIAN__
	49459	#define vcopy_laneq_s64(__p0_370, __p1_370, __p2_370, __p3_370) __extension__ ({ \
	49460	int64x1_t __ret_370; \
	49461	int64x1_t __s0_370 = __p0_370; \
	49462	int64x2_t __s2_370 = __p2_370; \
	49463	__ret_370 = vset_lane_s64(vgetq_lane_s64(__s2_370, __p3_370), __s0_370, __p1_370); \
	49464	__ret_370; \
	49465	})
	49466	#else
	49467	#define vcopy_laneq_s64(__p0_371, __p1_371, __p2_371, __p3_371) __extension__ ({ \
	49468	int64x1_t __ret_371; \
	49469	int64x1_t __s0_371 = __p0_371; \
	49470	int64x2_t __s2_371 = __p2_371; \
	49471	int64x2_t __rev2_371; __rev2_371 = __builtin_shufflevector(__s2_371, __s2_371, __lane_reverse_128_64); \
	49472	__ret_371 = vset_lane_s64(__noswap_vgetq_lane_s64(__rev2_371, __p3_371), __s0_371, __p1_371); \
	49473	__ret_371; \
	49474	})
	49475	#endif
	49476
	49477	#ifdef __LITTLE_ENDIAN__
	49478	#define vcopy_laneq_mf8(__p0_372, __p1_372, __p2_372, __p3_372) __extension__ ({ \
	49479	mfloat8x8_t __ret_372; \
	49480	mfloat8x8_t __s0_372 = __p0_372; \
	49481	mfloat8x16_t __s2_372 = __p2_372; \
	49482	__ret_372 = vset_lane_mf8(vgetq_lane_mf8(__s2_372, __p3_372), __s0_372, __p1_372); \
	49483	__ret_372; \
	49484	})
	49485	#else
	49486	#define vcopy_laneq_mf8(__p0_373, __p1_373, __p2_373, __p3_373) __extension__ ({ \
	49487	mfloat8x8_t __ret_373; \
	49488	mfloat8x8_t __s0_373 = __p0_373; \
	49489	mfloat8x16_t __s2_373 = __p2_373; \
	49490	mfloat8x8_t __rev0_373; __rev0_373 = __builtin_shufflevector(__s0_373, __s0_373, __lane_reverse_64_8); \
	49491	mfloat8x16_t __rev2_373; __rev2_373 = __builtin_shufflevector(__s2_373, __s2_373, __lane_reverse_128_8); \
	49492	__ret_373 = __noswap_vset_lane_mf8(__noswap_vgetq_lane_mf8(__rev2_373, __p3_373), __rev0_373, __p1_373); \
	49493	__ret_373 = __builtin_shufflevector(__ret_373, __ret_373, __lane_reverse_64_8); \
	49494	__ret_373; \
	49495	})
	49496	#endif
	49497
	49498	#ifdef __LITTLE_ENDIAN__
	49499	#define vcopy_laneq_s16(__p0_374, __p1_374, __p2_374, __p3_374) __extension__ ({ \
	49500	int16x4_t __ret_374; \
	49501	int16x4_t __s0_374 = __p0_374; \
	49502	int16x8_t __s2_374 = __p2_374; \
	49503	__ret_374 = vset_lane_s16(vgetq_lane_s16(__s2_374, __p3_374), __s0_374, __p1_374); \
	49504	__ret_374; \
	49505	})
	49506	#else
	49507	#define vcopy_laneq_s16(__p0_375, __p1_375, __p2_375, __p3_375) __extension__ ({ \
	49508	int16x4_t __ret_375; \
	49509	int16x4_t __s0_375 = __p0_375; \
	49510	int16x8_t __s2_375 = __p2_375; \
	49511	int16x4_t __rev0_375; __rev0_375 = __builtin_shufflevector(__s0_375, __s0_375, __lane_reverse_64_16); \
	49512	int16x8_t __rev2_375; __rev2_375 = __builtin_shufflevector(__s2_375, __s2_375, __lane_reverse_128_16); \
	49513	__ret_375 = __noswap_vset_lane_s16(__noswap_vgetq_lane_s16(__rev2_375, __p3_375), __rev0_375, __p1_375); \
	49514	__ret_375 = __builtin_shufflevector(__ret_375, __ret_375, __lane_reverse_64_16); \
	49515	__ret_375; \
	49516	})
	49517	#endif
	49518
48495	49519	#define vcreate_p64(__p0) __extension__ ({ \
48496	49520	poly64x1_t __ret; \
48497	49521	uint64_t __promote = __p0; \
48498		__ret = (poly64x1_t)(__promote); \
	49522	__ret = __builtin_bit_cast(poly64x1_t, __promote); \
48499	49523	__ret; \
48500	49524	})
48501	49525	#define vcreate_f64(__p0) __extension__ ({ \
48502	49526	float64x1_t __ret; \
48503	49527	uint64_t __promote = __p0; \
48504		__ret = (float64x1_t)(__promote); \
	49528	__ret = __builtin_bit_cast(float64x1_t, __promote); \
48505	49529	__ret; \
48506	49530	})
48507	49531	__ai __attribute__((target("neon"))) float32_t vcvts_f32_s32(int32_t __p0) {
48508	49532	float32_t __ret;
48509		__ret = (float32_t) __builtin_neon_vcvts_f32_s32(__p0);
	49533	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vcvts_f32_s32(__p0));
48510	49534	return __ret;
48511	49535	}
48512	49536	__ai __attribute__((target("neon"))) float32_t vcvts_f32_u32(uint32_t __p0) {
48513	49537	float32_t __ret;
48514		__ret = (float32_t) __builtin_neon_vcvts_f32_u32(__p0);
	49538	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vcvts_f32_u32(__p0));
48515	49539	return __ret;
48516	49540	}
48517	49541	#ifdef __LITTLE_ENDIAN__
48518	49542	__ai __attribute__((target("neon"))) float32x2_t vcvt_f32_f64(float64x2_t __p0) {
48519	49543	float32x2_t __ret;
48520		__ret = (float32x2_t) __builtin_neon_vcvt_f32_f64((int8x16_t)__p0, 9);
	49544	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vcvt_f32_f64(__builtin_bit_cast(int8x16_t, __p0), 9));
48521	49545	return __ret;
48522	49546	}
48523	49547	#else
48524	49548	__ai __attribute__((target("neon"))) float32x2_t vcvt_f32_f64(float64x2_t __p0) {
48525	49549	float32x2_t __ret;
48526		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
48527		__ret = (float32x2_t) __builtin_neon_vcvt_f32_f64((int8x16_t)__rev0, 9);
48528		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	49550	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	49551	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vcvt_f32_f64(__builtin_bit_cast(int8x16_t, __rev0), 9));
	49552	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
48529	49553	return __ret;
48530	49554	}
48531	49555	__ai __attribute__((target("neon"))) float32x2_t __noswap_vcvt_f32_f64(float64x2_t __p0) {
48532	49556	float32x2_t __ret;
48533		__ret = (float32x2_t) __builtin_neon_vcvt_f32_f64((int8x16_t)__p0, 9);
	49557	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vcvt_f32_f64(__builtin_bit_cast(int8x16_t, __p0), 9));
48534	49558	return __ret;
48535	49559	}
48536	49560	#endif
48537	49561
48538	49562	__ai __attribute__((target("neon"))) float64_t vcvtd_f64_s64(int64_t __p0) {
48539	49563	float64_t __ret;
48540		__ret = (float64_t) __builtin_neon_vcvtd_f64_s64(__p0);
	49564	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vcvtd_f64_s64(__p0));
48541	49565	return __ret;
48542	49566	}
48543	49567	__ai __attribute__((target("neon"))) float64_t vcvtd_f64_u64(uint64_t __p0) {
48544	49568	float64_t __ret;
48545		__ret = (float64_t) __builtin_neon_vcvtd_f64_u64(__p0);
	49569	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vcvtd_f64_u64(__p0));
48546	49570	return __ret;
48547	49571	}
48548	49572	#ifdef __LITTLE_ENDIAN__
48549	49573	__ai __attribute__((target("neon"))) float64x2_t vcvtq_f64_u64(uint64x2_t __p0) {
48550	49574	float64x2_t __ret;
48551		__ret = (float64x2_t) __builtin_neon_vcvtq_f64_v((int8x16_t)__p0, 51);
	49575	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vcvtq_f64_v(__builtin_bit_cast(int8x16_t, __p0), 51));
48552	49576	return __ret;
48553	49577	}
48554	49578	#else
48555	49579	__ai __attribute__((target("neon"))) float64x2_t vcvtq_f64_u64(uint64x2_t __p0) {
48556	49580	float64x2_t __ret;
48557		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
48558		__ret = (float64x2_t) __builtin_neon_vcvtq_f64_v((int8x16_t)__rev0, 51);
48559		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	49581	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	49582	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vcvtq_f64_v(__builtin_bit_cast(int8x16_t, __rev0), 51));
	49583	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
48560	49584	return __ret;
48561	49585	}
48562	49586	#endif
...	...	@@ -48564,46 +49588,46 @@ __ai __attribute__((target("neon"))) float64x2_t vcvtq_f64_u64(uint64x2_t __p0)
48564	49588	#ifdef __LITTLE_ENDIAN__
48565	49589	__ai __attribute__((target("neon"))) float64x2_t vcvtq_f64_s64(int64x2_t __p0) {
48566	49590	float64x2_t __ret;
48567		__ret = (float64x2_t) __builtin_neon_vcvtq_f64_v((int8x16_t)__p0, 35);
	49591	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vcvtq_f64_v(__builtin_bit_cast(int8x16_t, __p0), 35));
48568	49592	return __ret;
48569	49593	}
48570	49594	#else
48571	49595	__ai __attribute__((target("neon"))) float64x2_t vcvtq_f64_s64(int64x2_t __p0) {
48572	49596	float64x2_t __ret;
48573		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
48574		__ret = (float64x2_t) __builtin_neon_vcvtq_f64_v((int8x16_t)__rev0, 35);
48575		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	49597	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	49598	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vcvtq_f64_v(__builtin_bit_cast(int8x16_t, __rev0), 35));
	49599	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
48576	49600	return __ret;
48577	49601	}
48578	49602	#endif
48579	49603
48580	49604	__ai __attribute__((target("neon"))) float64x1_t vcvt_f64_u64(uint64x1_t __p0) {
48581	49605	float64x1_t __ret;
48582		__ret = (float64x1_t) __builtin_neon_vcvt_f64_v((int8x8_t)__p0, 19);
	49606	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vcvt_f64_v(__builtin_bit_cast(int8x8_t, __p0), 19));
48583	49607	return __ret;
48584	49608	}
48585	49609	__ai __attribute__((target("neon"))) float64x1_t vcvt_f64_s64(int64x1_t __p0) {
48586	49610	float64x1_t __ret;
48587		__ret = (float64x1_t) __builtin_neon_vcvt_f64_v((int8x8_t)__p0, 3);
	49611	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vcvt_f64_v(__builtin_bit_cast(int8x8_t, __p0), 3));
48588	49612	return __ret;
48589	49613	}
48590	49614	#ifdef __LITTLE_ENDIAN__
48591	49615	__ai __attribute__((target("neon"))) float64x2_t vcvt_f64_f32(float32x2_t __p0) {
48592	49616	float64x2_t __ret;
48593		__ret = (float64x2_t) __builtin_neon_vcvt_f64_f32((int8x8_t)__p0, 42);
	49617	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vcvt_f64_f32(__builtin_bit_cast(int8x8_t, __p0), 42));
48594	49618	return __ret;
48595	49619	}
48596	49620	#else
48597	49621	__ai __attribute__((target("neon"))) float64x2_t vcvt_f64_f32(float32x2_t __p0) {
48598	49622	float64x2_t __ret;
48599		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
48600		__ret = (float64x2_t) __builtin_neon_vcvt_f64_f32((int8x8_t)__rev0, 42);
48601		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	49623	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	49624	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vcvt_f64_f32(__builtin_bit_cast(int8x8_t, __rev0), 42));
	49625	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
48602	49626	return __ret;
48603	49627	}
48604	49628	__ai __attribute__((target("neon"))) float64x2_t __noswap_vcvt_f64_f32(float32x2_t __p0) {
48605	49629	float64x2_t __ret;
48606		__ret = (float64x2_t) __builtin_neon_vcvt_f64_f32((int8x8_t)__p0, 42);
	49630	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vcvt_f64_f32(__builtin_bit_cast(int8x8_t, __p0), 42));
48607	49631	return __ret;
48608	49632	}
48609	49633	#endif
...	...	@@ -48617,10 +49641,10 @@ __ai __attribute__((target("neon"))) float16x8_t vcvt_high_f16_f32(float16x4_t _
48617	49641	#else
48618	49642	__ai __attribute__((target("neon"))) float16x8_t vcvt_high_f16_f32(float16x4_t __p0, float32x4_t __p1) {
48619	49643	float16x8_t __ret;
48620		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
48621		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	49644	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	49645	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
48622	49646	__ret = __noswap_vcombine_f16(__rev0, __noswap_vcvt_f16_f32(__rev1));
48623		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	49647	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
48624	49648	return __ret;
48625	49649	}
48626	49650	#endif
...	...	@@ -48634,9 +49658,9 @@ __ai __attribute__((target("neon"))) float32x4_t vcvt_high_f32_f16(float16x8_t _
48634	49658	#else
48635	49659	__ai __attribute__((target("neon"))) float32x4_t vcvt_high_f32_f16(float16x8_t __p0) {
48636	49660	float32x4_t __ret;
48637		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	49661	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
48638	49662	__ret = __noswap_vcvt_f32_f16(__noswap_vget_high_f16(__rev0));
48639		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	49663	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
48640	49664	return __ret;
48641	49665	}
48642	49666	#endif
...	...	@@ -48650,10 +49674,10 @@ __ai __attribute__((target("neon"))) float32x4_t vcvt_high_f32_f64(float32x2_t _
48650	49674	#else
48651	49675	__ai __attribute__((target("neon"))) float32x4_t vcvt_high_f32_f64(float32x2_t __p0, float64x2_t __p1) {
48652	49676	float32x4_t __ret;
48653		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
48654		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	49677	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	49678	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
48655	49679	__ret = __noswap_vcombine_f32(__rev0, __noswap_vcvt_f32_f64(__rev1));
48656		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	49680	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
48657	49681	return __ret;
48658	49682	}
48659	49683	#endif
...	...	@@ -48667,9 +49691,9 @@ __ai __attribute__((target("neon"))) float64x2_t vcvt_high_f64_f32(float32x4_t _
48667	49691	#else
48668	49692	__ai __attribute__((target("neon"))) float64x2_t vcvt_high_f64_f32(float32x4_t __p0) {
48669	49693	float64x2_t __ret;
48670		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	49694	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
48671	49695	__ret = __noswap_vcvt_f64_f32(__noswap_vget_high_f32(__rev0));
48672		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	49696	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
48673	49697	return __ret;
48674	49698	}
48675	49699	#endif
...	...	@@ -48677,29 +49701,29 @@ __ai __attribute__((target("neon"))) float64x2_t vcvt_high_f64_f32(float32x4_t _
48677	49701	#define vcvts_n_f32_u32(__p0, __p1) __extension__ ({ \
48678	49702	float32_t __ret; \
48679	49703	uint32_t __s0 = __p0; \
48680		__ret = (float32_t) __builtin_neon_vcvts_n_f32_u32(__s0, __p1); \
	49704	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vcvts_n_f32_u32(__s0, __p1)); \
48681	49705	__ret; \
48682	49706	})
48683	49707	#define vcvts_n_f32_s32(__p0, __p1) __extension__ ({ \
48684	49708	float32_t __ret; \
48685	49709	int32_t __s0 = __p0; \
48686		__ret = (float32_t) __builtin_neon_vcvts_n_f32_s32(__s0, __p1); \
	49710	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vcvts_n_f32_s32(__s0, __p1)); \
48687	49711	__ret; \
48688	49712	})
48689	49713	#ifdef __LITTLE_ENDIAN__
48690	49714	#define vcvtq_n_f64_u64(__p0, __p1) __extension__ ({ \
48691	49715	float64x2_t __ret; \
48692	49716	uint64x2_t __s0 = __p0; \
48693		__ret = (float64x2_t) __builtin_neon_vcvtq_n_f64_v((int8x16_t)__s0, __p1, 51); \
	49717	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vcvtq_n_f64_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 51)); \
48694	49718	__ret; \
48695	49719	})
48696	49720	#else
48697	49721	#define vcvtq_n_f64_u64(__p0, __p1) __extension__ ({ \
48698	49722	float64x2_t __ret; \
48699	49723	uint64x2_t __s0 = __p0; \
48700		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
48701		__ret = (float64x2_t) __builtin_neon_vcvtq_n_f64_v((int8x16_t)__rev0, __p1, 51); \
48702		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	49724	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	49725	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vcvtq_n_f64_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 51)); \
	49726	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
48703	49727	__ret; \
48704	49728	})
48705	49729	#endif
...	...	@@ -48708,16 +49732,16 @@ __ai __attribute__((target("neon"))) float64x2_t vcvt_high_f64_f32(float32x4_t _
48708	49732	#define vcvtq_n_f64_s64(__p0, __p1) __extension__ ({ \
48709	49733	float64x2_t __ret; \
48710	49734	int64x2_t __s0 = __p0; \
48711		__ret = (float64x2_t) __builtin_neon_vcvtq_n_f64_v((int8x16_t)__s0, __p1, 35); \
	49735	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vcvtq_n_f64_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 35)); \
48712	49736	__ret; \
48713	49737	})
48714	49738	#else
48715	49739	#define vcvtq_n_f64_s64(__p0, __p1) __extension__ ({ \
48716	49740	float64x2_t __ret; \
48717	49741	int64x2_t __s0 = __p0; \
48718		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
48719		__ret = (float64x2_t) __builtin_neon_vcvtq_n_f64_v((int8x16_t)__rev0, __p1, 35); \
48720		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	49742	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	49743	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vcvtq_n_f64_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 35)); \
	49744	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
48721	49745	__ret; \
48722	49746	})
48723	49747	#endif
...	...	@@ -48725,47 +49749,47 @@ __ai __attribute__((target("neon"))) float64x2_t vcvt_high_f64_f32(float32x4_t _
48725	49749	#define vcvt_n_f64_u64(__p0, __p1) __extension__ ({ \
48726	49750	float64x1_t __ret; \
48727	49751	uint64x1_t __s0 = __p0; \
48728		__ret = (float64x1_t) __builtin_neon_vcvt_n_f64_v((int8x8_t)__s0, __p1, 19); \
	49752	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vcvt_n_f64_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 19)); \
48729	49753	__ret; \
48730	49754	})
48731	49755	#define vcvt_n_f64_s64(__p0, __p1) __extension__ ({ \
48732	49756	float64x1_t __ret; \
48733	49757	int64x1_t __s0 = __p0; \
48734		__ret = (float64x1_t) __builtin_neon_vcvt_n_f64_v((int8x8_t)__s0, __p1, 3); \
	49758	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vcvt_n_f64_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 3)); \
48735	49759	__ret; \
48736	49760	})
48737	49761	#define vcvtd_n_f64_u64(__p0, __p1) __extension__ ({ \
48738	49762	float64_t __ret; \
48739	49763	uint64_t __s0 = __p0; \
48740		__ret = (float64_t) __builtin_neon_vcvtd_n_f64_u64(__s0, __p1); \
	49764	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vcvtd_n_f64_u64(__s0, __p1)); \
48741	49765	__ret; \
48742	49766	})
48743	49767	#define vcvtd_n_f64_s64(__p0, __p1) __extension__ ({ \
48744	49768	float64_t __ret; \
48745	49769	int64_t __s0 = __p0; \
48746		__ret = (float64_t) __builtin_neon_vcvtd_n_f64_s64(__s0, __p1); \
	49770	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vcvtd_n_f64_s64(__s0, __p1)); \
48747	49771	__ret; \
48748	49772	})
48749	49773	#define vcvts_n_s32_f32(__p0, __p1) __extension__ ({ \
48750	49774	int32_t __ret; \
48751	49775	float32_t __s0 = __p0; \
48752		__ret = (int32_t) __builtin_neon_vcvts_n_s32_f32(__s0, __p1); \
	49776	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vcvts_n_s32_f32(__s0, __p1)); \
48753	49777	__ret; \
48754	49778	})
48755	49779	#ifdef __LITTLE_ENDIAN__
48756	49780	#define vcvtq_n_s64_f64(__p0, __p1) __extension__ ({ \
48757	49781	int64x2_t __ret; \
48758	49782	float64x2_t __s0 = __p0; \
48759		__ret = (int64x2_t) __builtin_neon_vcvtq_n_s64_v((int8x16_t)__s0, __p1, 35); \
	49783	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vcvtq_n_s64_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 35)); \
48760	49784	__ret; \
48761	49785	})
48762	49786	#else
48763	49787	#define vcvtq_n_s64_f64(__p0, __p1) __extension__ ({ \
48764	49788	int64x2_t __ret; \
48765	49789	float64x2_t __s0 = __p0; \
48766		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
48767		__ret = (int64x2_t) __builtin_neon_vcvtq_n_s64_v((int8x16_t)__rev0, __p1, 35); \
48768		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	49790	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	49791	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vcvtq_n_s64_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 35)); \
	49792	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
48769	49793	__ret; \
48770	49794	})
48771	49795	#endif
...	...	@@ -48773,35 +49797,35 @@ __ai __attribute__((target("neon"))) float64x2_t vcvt_high_f64_f32(float32x4_t _
48773	49797	#define vcvt_n_s64_f64(__p0, __p1) __extension__ ({ \
48774	49798	int64x1_t __ret; \
48775	49799	float64x1_t __s0 = __p0; \
48776		__ret = (int64x1_t) __builtin_neon_vcvt_n_s64_v((int8x8_t)__s0, __p1, 3); \
	49800	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vcvt_n_s64_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 3)); \
48777	49801	__ret; \
48778	49802	})
48779	49803	#define vcvtd_n_s64_f64(__p0, __p1) __extension__ ({ \
48780	49804	int64_t __ret; \
48781	49805	float64_t __s0 = __p0; \
48782		__ret = (int64_t) __builtin_neon_vcvtd_n_s64_f64(__s0, __p1); \
	49806	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vcvtd_n_s64_f64(__s0, __p1)); \
48783	49807	__ret; \
48784	49808	})
48785	49809	#define vcvts_n_u32_f32(__p0, __p1) __extension__ ({ \
48786	49810	uint32_t __ret; \
48787	49811	float32_t __s0 = __p0; \
48788		__ret = (uint32_t) __builtin_neon_vcvts_n_u32_f32(__s0, __p1); \
	49812	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vcvts_n_u32_f32(__s0, __p1)); \
48789	49813	__ret; \
48790	49814	})
48791	49815	#ifdef __LITTLE_ENDIAN__
48792	49816	#define vcvtq_n_u64_f64(__p0, __p1) __extension__ ({ \
48793	49817	uint64x2_t __ret; \
48794	49818	float64x2_t __s0 = __p0; \
48795		__ret = (uint64x2_t) __builtin_neon_vcvtq_n_u64_v((int8x16_t)__s0, __p1, 51); \
	49819	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcvtq_n_u64_v(__builtin_bit_cast(int8x16_t, __s0), __p1, 51)); \
48796	49820	__ret; \
48797	49821	})
48798	49822	#else
48799	49823	#define vcvtq_n_u64_f64(__p0, __p1) __extension__ ({ \
48800	49824	uint64x2_t __ret; \
48801	49825	float64x2_t __s0 = __p0; \
48802		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
48803		__ret = (uint64x2_t) __builtin_neon_vcvtq_n_u64_v((int8x16_t)__rev0, __p1, 51); \
48804		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	49826	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	49827	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcvtq_n_u64_v(__builtin_bit_cast(int8x16_t, __rev0), __p1, 51)); \
	49828	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
48805	49829	__ret; \
48806	49830	})
48807	49831	#endif
...	...	@@ -48809,347 +49833,347 @@ __ai __attribute__((target("neon"))) float64x2_t vcvt_high_f64_f32(float32x4_t _
48809	49833	#define vcvt_n_u64_f64(__p0, __p1) __extension__ ({ \
48810	49834	uint64x1_t __ret; \
48811	49835	float64x1_t __s0 = __p0; \
48812		__ret = (uint64x1_t) __builtin_neon_vcvt_n_u64_v((int8x8_t)__s0, __p1, 19); \
	49836	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vcvt_n_u64_v(__builtin_bit_cast(int8x8_t, __s0), __p1, 19)); \
48813	49837	__ret; \
48814	49838	})
48815	49839	#define vcvtd_n_u64_f64(__p0, __p1) __extension__ ({ \
48816	49840	uint64_t __ret; \
48817	49841	float64_t __s0 = __p0; \
48818		__ret = (uint64_t) __builtin_neon_vcvtd_n_u64_f64(__s0, __p1); \
	49842	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcvtd_n_u64_f64(__s0, __p1)); \
48819	49843	__ret; \
48820	49844	})
48821	49845	__ai __attribute__((target("neon"))) int32_t vcvts_s32_f32(float32_t __p0) {
48822	49846	int32_t __ret;
48823		__ret = (int32_t) __builtin_neon_vcvts_s32_f32(__p0);
	49847	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vcvts_s32_f32(__p0));
48824	49848	return __ret;
48825	49849	}
48826	49850	__ai __attribute__((target("neon"))) int64_t vcvtd_s64_f64(float64_t __p0) {
48827	49851	int64_t __ret;
48828		__ret = (int64_t) __builtin_neon_vcvtd_s64_f64(__p0);
	49852	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vcvtd_s64_f64(__p0));
48829	49853	return __ret;
48830	49854	}
48831	49855	#ifdef __LITTLE_ENDIAN__
48832	49856	__ai __attribute__((target("neon"))) int64x2_t vcvtq_s64_f64(float64x2_t __p0) {
48833	49857	int64x2_t __ret;
48834		__ret = (int64x2_t) __builtin_neon_vcvtq_s64_v((int8x16_t)__p0, 35);
	49858	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vcvtq_s64_v(__builtin_bit_cast(int8x16_t, __p0), 35));
48835	49859	return __ret;
48836	49860	}
48837	49861	#else
48838	49862	__ai __attribute__((target("neon"))) int64x2_t vcvtq_s64_f64(float64x2_t __p0) {
48839	49863	int64x2_t __ret;
48840		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
48841		__ret = (int64x2_t) __builtin_neon_vcvtq_s64_v((int8x16_t)__rev0, 35);
48842		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	49864	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	49865	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vcvtq_s64_v(__builtin_bit_cast(int8x16_t, __rev0), 35));
	49866	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
48843	49867	return __ret;
48844	49868	}
48845	49869	#endif
48846	49870
48847	49871	__ai __attribute__((target("neon"))) int64x1_t vcvt_s64_f64(float64x1_t __p0) {
48848	49872	int64x1_t __ret;
48849		__ret = (int64x1_t) __builtin_neon_vcvt_s64_v((int8x8_t)__p0, 3);
	49873	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vcvt_s64_v(__builtin_bit_cast(int8x8_t, __p0), 3));
48850	49874	return __ret;
48851	49875	}
48852	49876	__ai __attribute__((target("neon"))) uint32_t vcvts_u32_f32(float32_t __p0) {
48853	49877	uint32_t __ret;
48854		__ret = (uint32_t) __builtin_neon_vcvts_u32_f32(__p0);
	49878	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vcvts_u32_f32(__p0));
48855	49879	return __ret;
48856	49880	}
48857	49881	__ai __attribute__((target("neon"))) uint64_t vcvtd_u64_f64(float64_t __p0) {
48858	49882	uint64_t __ret;
48859		__ret = (uint64_t) __builtin_neon_vcvtd_u64_f64(__p0);
	49883	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcvtd_u64_f64(__p0));
48860	49884	return __ret;
48861	49885	}
48862	49886	#ifdef __LITTLE_ENDIAN__
48863	49887	__ai __attribute__((target("neon"))) uint64x2_t vcvtq_u64_f64(float64x2_t __p0) {
48864	49888	uint64x2_t __ret;
48865		__ret = (uint64x2_t) __builtin_neon_vcvtq_u64_v((int8x16_t)__p0, 51);
	49889	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcvtq_u64_v(__builtin_bit_cast(int8x16_t, __p0), 51));
48866	49890	return __ret;
48867	49891	}
48868	49892	#else
48869	49893	__ai __attribute__((target("neon"))) uint64x2_t vcvtq_u64_f64(float64x2_t __p0) {
48870	49894	uint64x2_t __ret;
48871		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
48872		__ret = (uint64x2_t) __builtin_neon_vcvtq_u64_v((int8x16_t)__rev0, 51);
48873		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	49895	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	49896	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcvtq_u64_v(__builtin_bit_cast(int8x16_t, __rev0), 51));
	49897	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
48874	49898	return __ret;
48875	49899	}
48876	49900	#endif
48877	49901
48878	49902	__ai __attribute__((target("neon"))) uint64x1_t vcvt_u64_f64(float64x1_t __p0) {
48879	49903	uint64x1_t __ret;
48880		__ret = (uint64x1_t) __builtin_neon_vcvt_u64_v((int8x8_t)__p0, 19);
	49904	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vcvt_u64_v(__builtin_bit_cast(int8x8_t, __p0), 19));
48881	49905	return __ret;
48882	49906	}
48883	49907	__ai __attribute__((target("neon"))) int32_t vcvtas_s32_f32(float32_t __p0) {
48884	49908	int32_t __ret;
48885		__ret = (int32_t) __builtin_neon_vcvtas_s32_f32(__p0);
	49909	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vcvtas_s32_f32(__p0));
48886	49910	return __ret;
48887	49911	}
48888	49912	#ifdef __LITTLE_ENDIAN__
48889	49913	__ai __attribute__((target("neon"))) int64x2_t vcvtaq_s64_f64(float64x2_t __p0) {
48890	49914	int64x2_t __ret;
48891		__ret = (int64x2_t) __builtin_neon_vcvtaq_s64_v((int8x16_t)__p0, 35);
	49915	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vcvtaq_s64_v(__builtin_bit_cast(int8x16_t, __p0), 35));
48892	49916	return __ret;
48893	49917	}
48894	49918	#else
48895	49919	__ai __attribute__((target("neon"))) int64x2_t vcvtaq_s64_f64(float64x2_t __p0) {
48896	49920	int64x2_t __ret;
48897		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
48898		__ret = (int64x2_t) __builtin_neon_vcvtaq_s64_v((int8x16_t)__rev0, 35);
48899		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	49921	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	49922	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vcvtaq_s64_v(__builtin_bit_cast(int8x16_t, __rev0), 35));
	49923	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
48900	49924	return __ret;
48901	49925	}
48902	49926	#endif
48903	49927
48904	49928	__ai __attribute__((target("neon"))) int64x1_t vcvta_s64_f64(float64x1_t __p0) {
48905	49929	int64x1_t __ret;
48906		__ret = (int64x1_t) __builtin_neon_vcvta_s64_v((int8x8_t)__p0, 3);
	49930	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vcvta_s64_v(__builtin_bit_cast(int8x8_t, __p0), 3));
48907	49931	return __ret;
48908	49932	}
48909	49933	__ai __attribute__((target("neon"))) int64_t vcvtad_s64_f64(float64_t __p0) {
48910	49934	int64_t __ret;
48911		__ret = (int64_t) __builtin_neon_vcvtad_s64_f64(__p0);
	49935	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vcvtad_s64_f64(__p0));
48912	49936	return __ret;
48913	49937	}
48914	49938	__ai __attribute__((target("neon"))) uint32_t vcvtas_u32_f32(float32_t __p0) {
48915	49939	uint32_t __ret;
48916		__ret = (uint32_t) __builtin_neon_vcvtas_u32_f32(__p0);
	49940	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vcvtas_u32_f32(__p0));
48917	49941	return __ret;
48918	49942	}
48919	49943	#ifdef __LITTLE_ENDIAN__
48920	49944	__ai __attribute__((target("neon"))) uint64x2_t vcvtaq_u64_f64(float64x2_t __p0) {
48921	49945	uint64x2_t __ret;
48922		__ret = (uint64x2_t) __builtin_neon_vcvtaq_u64_v((int8x16_t)__p0, 51);
	49946	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcvtaq_u64_v(__builtin_bit_cast(int8x16_t, __p0), 51));
48923	49947	return __ret;
48924	49948	}
48925	49949	#else
48926	49950	__ai __attribute__((target("neon"))) uint64x2_t vcvtaq_u64_f64(float64x2_t __p0) {
48927	49951	uint64x2_t __ret;
48928		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
48929		__ret = (uint64x2_t) __builtin_neon_vcvtaq_u64_v((int8x16_t)__rev0, 51);
48930		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	49952	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	49953	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcvtaq_u64_v(__builtin_bit_cast(int8x16_t, __rev0), 51));
	49954	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
48931	49955	return __ret;
48932	49956	}
48933	49957	#endif
48934	49958
48935	49959	__ai __attribute__((target("neon"))) uint64x1_t vcvta_u64_f64(float64x1_t __p0) {
48936	49960	uint64x1_t __ret;
48937		__ret = (uint64x1_t) __builtin_neon_vcvta_u64_v((int8x8_t)__p0, 19);
	49961	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vcvta_u64_v(__builtin_bit_cast(int8x8_t, __p0), 19));
48938	49962	return __ret;
48939	49963	}
48940	49964	__ai __attribute__((target("neon"))) uint64_t vcvtad_u64_f64(float64_t __p0) {
48941	49965	uint64_t __ret;
48942		__ret = (uint64_t) __builtin_neon_vcvtad_u64_f64(__p0);
	49966	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcvtad_u64_f64(__p0));
48943	49967	return __ret;
48944	49968	}
48945	49969	__ai __attribute__((target("neon"))) int32_t vcvtms_s32_f32(float32_t __p0) {
48946	49970	int32_t __ret;
48947		__ret = (int32_t) __builtin_neon_vcvtms_s32_f32(__p0);
	49971	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vcvtms_s32_f32(__p0));
48948	49972	return __ret;
48949	49973	}
48950	49974	#ifdef __LITTLE_ENDIAN__
48951	49975	__ai __attribute__((target("neon"))) int64x2_t vcvtmq_s64_f64(float64x2_t __p0) {
48952	49976	int64x2_t __ret;
48953		__ret = (int64x2_t) __builtin_neon_vcvtmq_s64_v((int8x16_t)__p0, 35);
	49977	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vcvtmq_s64_v(__builtin_bit_cast(int8x16_t, __p0), 35));
48954	49978	return __ret;
48955	49979	}
48956	49980	#else
48957	49981	__ai __attribute__((target("neon"))) int64x2_t vcvtmq_s64_f64(float64x2_t __p0) {
48958	49982	int64x2_t __ret;
48959		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
48960		__ret = (int64x2_t) __builtin_neon_vcvtmq_s64_v((int8x16_t)__rev0, 35);
48961		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	49983	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	49984	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vcvtmq_s64_v(__builtin_bit_cast(int8x16_t, __rev0), 35));
	49985	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
48962	49986	return __ret;
48963	49987	}
48964	49988	#endif
48965	49989
48966	49990	__ai __attribute__((target("neon"))) int64x1_t vcvtm_s64_f64(float64x1_t __p0) {
48967	49991	int64x1_t __ret;
48968		__ret = (int64x1_t) __builtin_neon_vcvtm_s64_v((int8x8_t)__p0, 3);
	49992	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vcvtm_s64_v(__builtin_bit_cast(int8x8_t, __p0), 3));
48969	49993	return __ret;
48970	49994	}
48971	49995	__ai __attribute__((target("neon"))) int64_t vcvtmd_s64_f64(float64_t __p0) {
48972	49996	int64_t __ret;
48973		__ret = (int64_t) __builtin_neon_vcvtmd_s64_f64(__p0);
	49997	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vcvtmd_s64_f64(__p0));
48974	49998	return __ret;
48975	49999	}
48976	50000	__ai __attribute__((target("neon"))) uint32_t vcvtms_u32_f32(float32_t __p0) {
48977	50001	uint32_t __ret;
48978		__ret = (uint32_t) __builtin_neon_vcvtms_u32_f32(__p0);
	50002	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vcvtms_u32_f32(__p0));
48979	50003	return __ret;
48980	50004	}
48981	50005	#ifdef __LITTLE_ENDIAN__
48982	50006	__ai __attribute__((target("neon"))) uint64x2_t vcvtmq_u64_f64(float64x2_t __p0) {
48983	50007	uint64x2_t __ret;
48984		__ret = (uint64x2_t) __builtin_neon_vcvtmq_u64_v((int8x16_t)__p0, 51);
	50008	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcvtmq_u64_v(__builtin_bit_cast(int8x16_t, __p0), 51));
48985	50009	return __ret;
48986	50010	}
48987	50011	#else
48988	50012	__ai __attribute__((target("neon"))) uint64x2_t vcvtmq_u64_f64(float64x2_t __p0) {
48989	50013	uint64x2_t __ret;
48990		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
48991		__ret = (uint64x2_t) __builtin_neon_vcvtmq_u64_v((int8x16_t)__rev0, 51);
48992		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	50014	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	50015	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcvtmq_u64_v(__builtin_bit_cast(int8x16_t, __rev0), 51));
	50016	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
48993	50017	return __ret;
48994	50018	}
48995	50019	#endif
48996	50020
48997	50021	__ai __attribute__((target("neon"))) uint64x1_t vcvtm_u64_f64(float64x1_t __p0) {
48998	50022	uint64x1_t __ret;
48999		__ret = (uint64x1_t) __builtin_neon_vcvtm_u64_v((int8x8_t)__p0, 19);
	50023	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vcvtm_u64_v(__builtin_bit_cast(int8x8_t, __p0), 19));
49000	50024	return __ret;
49001	50025	}
49002	50026	__ai __attribute__((target("neon"))) uint64_t vcvtmd_u64_f64(float64_t __p0) {
49003	50027	uint64_t __ret;
49004		__ret = (uint64_t) __builtin_neon_vcvtmd_u64_f64(__p0);
	50028	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcvtmd_u64_f64(__p0));
49005	50029	return __ret;
49006	50030	}
49007	50031	__ai __attribute__((target("neon"))) int32_t vcvtns_s32_f32(float32_t __p0) {
49008	50032	int32_t __ret;
49009		__ret = (int32_t) __builtin_neon_vcvtns_s32_f32(__p0);
	50033	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vcvtns_s32_f32(__p0));
49010	50034	return __ret;
49011	50035	}
49012	50036	#ifdef __LITTLE_ENDIAN__
49013	50037	__ai __attribute__((target("neon"))) int64x2_t vcvtnq_s64_f64(float64x2_t __p0) {
49014	50038	int64x2_t __ret;
49015		__ret = (int64x2_t) __builtin_neon_vcvtnq_s64_v((int8x16_t)__p0, 35);
	50039	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vcvtnq_s64_v(__builtin_bit_cast(int8x16_t, __p0), 35));
49016	50040	return __ret;
49017	50041	}
49018	50042	#else
49019	50043	__ai __attribute__((target("neon"))) int64x2_t vcvtnq_s64_f64(float64x2_t __p0) {
49020	50044	int64x2_t __ret;
49021		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
49022		__ret = (int64x2_t) __builtin_neon_vcvtnq_s64_v((int8x16_t)__rev0, 35);
49023		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	50045	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	50046	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vcvtnq_s64_v(__builtin_bit_cast(int8x16_t, __rev0), 35));
	50047	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
49024	50048	return __ret;
49025	50049	}
49026	50050	#endif
49027	50051
49028	50052	__ai __attribute__((target("neon"))) int64x1_t vcvtn_s64_f64(float64x1_t __p0) {
49029	50053	int64x1_t __ret;
49030		__ret = (int64x1_t) __builtin_neon_vcvtn_s64_v((int8x8_t)__p0, 3);
	50054	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vcvtn_s64_v(__builtin_bit_cast(int8x8_t, __p0), 3));
49031	50055	return __ret;
49032	50056	}
49033	50057	__ai __attribute__((target("neon"))) int64_t vcvtnd_s64_f64(float64_t __p0) {
49034	50058	int64_t __ret;
49035		__ret = (int64_t) __builtin_neon_vcvtnd_s64_f64(__p0);
	50059	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vcvtnd_s64_f64(__p0));
49036	50060	return __ret;
49037	50061	}
49038	50062	__ai __attribute__((target("neon"))) uint32_t vcvtns_u32_f32(float32_t __p0) {
49039	50063	uint32_t __ret;
49040		__ret = (uint32_t) __builtin_neon_vcvtns_u32_f32(__p0);
	50064	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vcvtns_u32_f32(__p0));
49041	50065	return __ret;
49042	50066	}
49043	50067	#ifdef __LITTLE_ENDIAN__
49044	50068	__ai __attribute__((target("neon"))) uint64x2_t vcvtnq_u64_f64(float64x2_t __p0) {
49045	50069	uint64x2_t __ret;
49046		__ret = (uint64x2_t) __builtin_neon_vcvtnq_u64_v((int8x16_t)__p0, 51);
	50070	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcvtnq_u64_v(__builtin_bit_cast(int8x16_t, __p0), 51));
49047	50071	return __ret;
49048	50072	}
49049	50073	#else
49050	50074	__ai __attribute__((target("neon"))) uint64x2_t vcvtnq_u64_f64(float64x2_t __p0) {
49051	50075	uint64x2_t __ret;
49052		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
49053		__ret = (uint64x2_t) __builtin_neon_vcvtnq_u64_v((int8x16_t)__rev0, 51);
49054		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	50076	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	50077	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcvtnq_u64_v(__builtin_bit_cast(int8x16_t, __rev0), 51));
	50078	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
49055	50079	return __ret;
49056	50080	}
49057	50081	#endif
49058	50082
49059	50083	__ai __attribute__((target("neon"))) uint64x1_t vcvtn_u64_f64(float64x1_t __p0) {
49060	50084	uint64x1_t __ret;
49061		__ret = (uint64x1_t) __builtin_neon_vcvtn_u64_v((int8x8_t)__p0, 19);
	50085	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vcvtn_u64_v(__builtin_bit_cast(int8x8_t, __p0), 19));
49062	50086	return __ret;
49063	50087	}
49064	50088	__ai __attribute__((target("neon"))) uint64_t vcvtnd_u64_f64(float64_t __p0) {
49065	50089	uint64_t __ret;
49066		__ret = (uint64_t) __builtin_neon_vcvtnd_u64_f64(__p0);
	50090	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcvtnd_u64_f64(__p0));
49067	50091	return __ret;
49068	50092	}
49069	50093	__ai __attribute__((target("neon"))) int32_t vcvtps_s32_f32(float32_t __p0) {
49070	50094	int32_t __ret;
49071		__ret = (int32_t) __builtin_neon_vcvtps_s32_f32(__p0);
	50095	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vcvtps_s32_f32(__p0));
49072	50096	return __ret;
49073	50097	}
49074	50098	#ifdef __LITTLE_ENDIAN__
49075	50099	__ai __attribute__((target("neon"))) int64x2_t vcvtpq_s64_f64(float64x2_t __p0) {
49076	50100	int64x2_t __ret;
49077		__ret = (int64x2_t) __builtin_neon_vcvtpq_s64_v((int8x16_t)__p0, 35);
	50101	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vcvtpq_s64_v(__builtin_bit_cast(int8x16_t, __p0), 35));
49078	50102	return __ret;
49079	50103	}
49080	50104	#else
49081	50105	__ai __attribute__((target("neon"))) int64x2_t vcvtpq_s64_f64(float64x2_t __p0) {
49082	50106	int64x2_t __ret;
49083		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
49084		__ret = (int64x2_t) __builtin_neon_vcvtpq_s64_v((int8x16_t)__rev0, 35);
49085		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	50107	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	50108	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vcvtpq_s64_v(__builtin_bit_cast(int8x16_t, __rev0), 35));
	50109	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
49086	50110	return __ret;
49087	50111	}
49088	50112	#endif
49089	50113
49090	50114	__ai __attribute__((target("neon"))) int64x1_t vcvtp_s64_f64(float64x1_t __p0) {
49091	50115	int64x1_t __ret;
49092		__ret = (int64x1_t) __builtin_neon_vcvtp_s64_v((int8x8_t)__p0, 3);
	50116	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vcvtp_s64_v(__builtin_bit_cast(int8x8_t, __p0), 3));
49093	50117	return __ret;
49094	50118	}
49095	50119	__ai __attribute__((target("neon"))) int64_t vcvtpd_s64_f64(float64_t __p0) {
49096	50120	int64_t __ret;
49097		__ret = (int64_t) __builtin_neon_vcvtpd_s64_f64(__p0);
	50121	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vcvtpd_s64_f64(__p0));
49098	50122	return __ret;
49099	50123	}
49100	50124	__ai __attribute__((target("neon"))) uint32_t vcvtps_u32_f32(float32_t __p0) {
49101	50125	uint32_t __ret;
49102		__ret = (uint32_t) __builtin_neon_vcvtps_u32_f32(__p0);
	50126	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vcvtps_u32_f32(__p0));
49103	50127	return __ret;
49104	50128	}
49105	50129	#ifdef __LITTLE_ENDIAN__
49106	50130	__ai __attribute__((target("neon"))) uint64x2_t vcvtpq_u64_f64(float64x2_t __p0) {
49107	50131	uint64x2_t __ret;
49108		__ret = (uint64x2_t) __builtin_neon_vcvtpq_u64_v((int8x16_t)__p0, 51);
	50132	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcvtpq_u64_v(__builtin_bit_cast(int8x16_t, __p0), 51));
49109	50133	return __ret;
49110	50134	}
49111	50135	#else
49112	50136	__ai __attribute__((target("neon"))) uint64x2_t vcvtpq_u64_f64(float64x2_t __p0) {
49113	50137	uint64x2_t __ret;
49114		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
49115		__ret = (uint64x2_t) __builtin_neon_vcvtpq_u64_v((int8x16_t)__rev0, 51);
49116		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	50138	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	50139	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vcvtpq_u64_v(__builtin_bit_cast(int8x16_t, __rev0), 51));
	50140	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
49117	50141	return __ret;
49118	50142	}
49119	50143	#endif
49120	50144
49121	50145	__ai __attribute__((target("neon"))) uint64x1_t vcvtp_u64_f64(float64x1_t __p0) {
49122	50146	uint64x1_t __ret;
49123		__ret = (uint64x1_t) __builtin_neon_vcvtp_u64_v((int8x8_t)__p0, 19);
	50147	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vcvtp_u64_v(__builtin_bit_cast(int8x8_t, __p0), 19));
49124	50148	return __ret;
49125	50149	}
49126	50150	__ai __attribute__((target("neon"))) uint64_t vcvtpd_u64_f64(float64_t __p0) {
49127	50151	uint64_t __ret;
49128		__ret = (uint64_t) __builtin_neon_vcvtpd_u64_f64(__p0);
	50152	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vcvtpd_u64_f64(__p0));
49129	50153	return __ret;
49130	50154	}
49131	50155	__ai __attribute__((target("neon"))) float32_t vcvtxd_f32_f64(float64_t __p0) {
49132	50156	float32_t __ret;
49133		__ret = (float32_t) __builtin_neon_vcvtxd_f32_f64(__p0);
	50157	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vcvtxd_f32_f64(__p0));
49134	50158	return __ret;
49135	50159	}
49136	50160	#ifdef __LITTLE_ENDIAN__
49137	50161	__ai __attribute__((target("neon"))) float32x2_t vcvtx_f32_f64(float64x2_t __p0) {
49138	50162	float32x2_t __ret;
49139		__ret = (float32x2_t) __builtin_neon_vcvtx_f32_v((int8x16_t)__p0, 42);
	50163	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vcvtx_f32_v(__builtin_bit_cast(int8x16_t, __p0), 42));
49140	50164	return __ret;
49141	50165	}
49142	50166	#else
49143	50167	__ai __attribute__((target("neon"))) float32x2_t vcvtx_f32_f64(float64x2_t __p0) {
49144	50168	float32x2_t __ret;
49145		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
49146		__ret = (float32x2_t) __builtin_neon_vcvtx_f32_v((int8x16_t)__rev0, 42);
49147		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	50169	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	50170	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vcvtx_f32_v(__builtin_bit_cast(int8x16_t, __rev0), 42));
	50171	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
49148	50172	return __ret;
49149	50173	}
49150	50174	__ai __attribute__((target("neon"))) float32x2_t __noswap_vcvtx_f32_f64(float64x2_t __p0) {
49151	50175	float32x2_t __ret;
49152		__ret = (float32x2_t) __builtin_neon_vcvtx_f32_v((int8x16_t)__p0, 42);
	50176	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vcvtx_f32_v(__builtin_bit_cast(int8x16_t, __p0), 42));
49153	50177	return __ret;
49154	50178	}
49155	50179	#endif
...	...	@@ -49163,10 +50187,10 @@ __ai __attribute__((target("neon"))) float32x4_t vcvtx_high_f32_f64(float32x2_t
49163	50187	#else
49164	50188	__ai __attribute__((target("neon"))) float32x4_t vcvtx_high_f32_f64(float32x2_t __p0, float64x2_t __p1) {
49165	50189	float32x4_t __ret;
49166		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
49167		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	50190	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	50191	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
49168	50192	__ret = __noswap_vcombine_f32(__rev0, __noswap_vcvtx_f32_f64(__rev1));
49169		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	50193	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
49170	50194	return __ret;
49171	50195	}
49172	50196	#endif
...	...	@@ -49180,10 +50204,10 @@ __ai __attribute__((target("neon"))) float64x2_t vdivq_f64(float64x2_t __p0, flo
49180	50204	#else
49181	50205	__ai __attribute__((target("neon"))) float64x2_t vdivq_f64(float64x2_t __p0, float64x2_t __p1) {
49182	50206	float64x2_t __ret;
49183		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
49184		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	50207	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	50208	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
49185	50209	__ret = __rev0 / __rev1;
49186		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	50210	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
49187	50211	return __ret;
49188	50212	}
49189	50213	#endif
...	...	@@ -49197,10 +50221,10 @@ __ai __attribute__((target("neon"))) float32x4_t vdivq_f32(float32x4_t __p0, flo
49197	50221	#else
49198	50222	__ai __attribute__((target("neon"))) float32x4_t vdivq_f32(float32x4_t __p0, float32x4_t __p1) {
49199	50223	float32x4_t __ret;
49200		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
49201		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	50224	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	50225	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
49202	50226	__ret = __rev0 / __rev1;
49203		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	50227	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
49204	50228	return __ret;
49205	50229	}
49206	50230	#endif
...	...	@@ -49219,10 +50243,10 @@ __ai __attribute__((target("neon"))) float32x2_t vdiv_f32(float32x2_t __p0, floa
49219	50243	#else
49220	50244	__ai __attribute__((target("neon"))) float32x2_t vdiv_f32(float32x2_t __p0, float32x2_t __p1) {
49221	50245	float32x2_t __ret;
49222		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
49223		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	50246	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	50247	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
49224	50248	__ret = __rev0 / __rev1;
49225		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	50249	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
49226	50250	return __ret;
49227	50251	}
49228	50252	#endif
...	...	@@ -49231,15 +50255,15 @@ __ai __attribute__((target("neon"))) float32x2_t vdiv_f32(float32x2_t __p0, floa
49231	50255	#define vdupb_lane_p8(__p0, __p1) __extension__ ({ \
49232	50256	poly8_t __ret; \
49233	50257	poly8x8_t __s0 = __p0; \
49234		__ret = (poly8_t) __builtin_neon_vdupb_lane_i8((poly8x8_t)__s0, __p1); \
	50258	__ret = __builtin_bit_cast(poly8_t, __builtin_neon_vdupb_lane_i8(__s0, __p1)); \
49235	50259	__ret; \
49236	50260	})
49237	50261	#else
49238	50262	#define vdupb_lane_p8(__p0, __p1) __extension__ ({ \
49239	50263	poly8_t __ret; \
49240	50264	poly8x8_t __s0 = __p0; \
49241		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
49242		__ret = (poly8_t) __builtin_neon_vdupb_lane_i8((poly8x8_t)__rev0, __p1); \
	50265	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	50266	__ret = __builtin_bit_cast(poly8_t, __builtin_neon_vdupb_lane_i8(__rev0, __p1)); \
49243	50267	__ret; \
49244	50268	})
49245	50269	#endif
...	...	@@ -49248,15 +50272,15 @@ __ai __attribute__((target("neon"))) float32x2_t vdiv_f32(float32x2_t __p0, floa
49248	50272	#define vduph_lane_p16(__p0, __p1) __extension__ ({ \
49249	50273	poly16_t __ret; \
49250	50274	poly16x4_t __s0 = __p0; \
49251		__ret = (poly16_t) __builtin_neon_vduph_lane_i16((poly16x4_t)__s0, __p1); \
	50275	__ret = __builtin_bit_cast(poly16_t, __builtin_neon_vduph_lane_i16(__s0, __p1)); \
49252	50276	__ret; \
49253	50277	})
49254	50278	#else
49255	50279	#define vduph_lane_p16(__p0, __p1) __extension__ ({ \
49256	50280	poly16_t __ret; \
49257	50281	poly16x4_t __s0 = __p0; \
49258		poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
49259		__ret = (poly16_t) __builtin_neon_vduph_lane_i16((poly16x4_t)__rev0, __p1); \
	50282	poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	50283	__ret = __builtin_bit_cast(poly16_t, __builtin_neon_vduph_lane_i16(__rev0, __p1)); \
49260	50284	__ret; \
49261	50285	})
49262	50286	#endif
...	...	@@ -49265,15 +50289,15 @@ __ai __attribute__((target("neon"))) float32x2_t vdiv_f32(float32x2_t __p0, floa
49265	50289	#define vdupb_lane_u8(__p0, __p1) __extension__ ({ \
49266	50290	uint8_t __ret; \
49267	50291	uint8x8_t __s0 = __p0; \
49268		__ret = (uint8_t) __builtin_neon_vdupb_lane_i8((int8x8_t)__s0, __p1); \
	50292	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vdupb_lane_i8(__builtin_bit_cast(int8x8_t, __s0), __p1)); \
49269	50293	__ret; \
49270	50294	})
49271	50295	#else
49272	50296	#define vdupb_lane_u8(__p0, __p1) __extension__ ({ \
49273	50297	uint8_t __ret; \
49274	50298	uint8x8_t __s0 = __p0; \
49275		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
49276		__ret = (uint8_t) __builtin_neon_vdupb_lane_i8((int8x8_t)__rev0, __p1); \
	50299	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	50300	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vdupb_lane_i8(__builtin_bit_cast(int8x8_t, __rev0), __p1)); \
49277	50301	__ret; \
49278	50302	})
49279	50303	#endif
...	...	@@ -49282,15 +50306,15 @@ __ai __attribute__((target("neon"))) float32x2_t vdiv_f32(float32x2_t __p0, floa
49282	50306	#define vdups_lane_u32(__p0, __p1) __extension__ ({ \
49283	50307	uint32_t __ret; \
49284	50308	uint32x2_t __s0 = __p0; \
49285		__ret = (uint32_t) __builtin_neon_vdups_lane_i32((int32x2_t)__s0, __p1); \
	50309	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vdups_lane_i32(__builtin_bit_cast(int32x2_t, __s0), __p1)); \
49286	50310	__ret; \
49287	50311	})
49288	50312	#else
49289	50313	#define vdups_lane_u32(__p0, __p1) __extension__ ({ \
49290	50314	uint32_t __ret; \
49291	50315	uint32x2_t __s0 = __p0; \
49292		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
49293		__ret = (uint32_t) __builtin_neon_vdups_lane_i32((int32x2_t)__rev0, __p1); \
	50316	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	50317	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vdups_lane_i32(__builtin_bit_cast(int32x2_t, __rev0), __p1)); \
49294	50318	__ret; \
49295	50319	})
49296	50320	#endif
...	...	@@ -49298,22 +50322,22 @@ __ai __attribute__((target("neon"))) float32x2_t vdiv_f32(float32x2_t __p0, floa
49298	50322	#define vdupd_lane_u64(__p0, __p1) __extension__ ({ \
49299	50323	uint64_t __ret; \
49300	50324	uint64x1_t __s0 = __p0; \
49301		__ret = (uint64_t) __builtin_neon_vdupd_lane_i64((int64x1_t)__s0, __p1); \
	50325	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vdupd_lane_i64(__builtin_bit_cast(int64x1_t, __s0), __p1)); \
49302	50326	__ret; \
49303	50327	})
49304	50328	#ifdef __LITTLE_ENDIAN__
49305	50329	#define vduph_lane_u16(__p0, __p1) __extension__ ({ \
49306	50330	uint16_t __ret; \
49307	50331	uint16x4_t __s0 = __p0; \
49308		__ret = (uint16_t) __builtin_neon_vduph_lane_i16((int16x4_t)__s0, __p1); \
	50332	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vduph_lane_i16(__builtin_bit_cast(int16x4_t, __s0), __p1)); \
49309	50333	__ret; \
49310	50334	})
49311	50335	#else
49312	50336	#define vduph_lane_u16(__p0, __p1) __extension__ ({ \
49313	50337	uint16_t __ret; \
49314	50338	uint16x4_t __s0 = __p0; \
49315		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
49316		__ret = (uint16_t) __builtin_neon_vduph_lane_i16((int16x4_t)__rev0, __p1); \
	50339	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	50340	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vduph_lane_i16(__builtin_bit_cast(int16x4_t, __rev0), __p1)); \
49317	50341	__ret; \
49318	50342	})
49319	50343	#endif
...	...	@@ -49322,15 +50346,15 @@ __ai __attribute__((target("neon"))) float32x2_t vdiv_f32(float32x2_t __p0, floa
49322	50346	#define vdupb_lane_s8(__p0, __p1) __extension__ ({ \
49323	50347	int8_t __ret; \
49324	50348	int8x8_t __s0 = __p0; \
49325		__ret = (int8_t) __builtin_neon_vdupb_lane_i8((int8x8_t)__s0, __p1); \
	50349	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vdupb_lane_i8(__builtin_bit_cast(int8x8_t, __s0), __p1)); \
49326	50350	__ret; \
49327	50351	})
49328	50352	#else
49329	50353	#define vdupb_lane_s8(__p0, __p1) __extension__ ({ \
49330	50354	int8_t __ret; \
49331	50355	int8x8_t __s0 = __p0; \
49332		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
49333		__ret = (int8_t) __builtin_neon_vdupb_lane_i8((int8x8_t)__rev0, __p1); \
	50356	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	50357	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vdupb_lane_i8(__builtin_bit_cast(int8x8_t, __rev0), __p1)); \
49334	50358	__ret; \
49335	50359	})
49336	50360	#endif
...	...	@@ -49338,22 +50362,22 @@ __ai __attribute__((target("neon"))) float32x2_t vdiv_f32(float32x2_t __p0, floa
49338	50362	#define vdupd_lane_f64(__p0, __p1) __extension__ ({ \
49339	50363	float64_t __ret; \
49340	50364	float64x1_t __s0 = __p0; \
49341		__ret = (float64_t) __builtin_neon_vdupd_lane_f64((float64x1_t)__s0, __p1); \
	50365	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vdupd_lane_f64(__s0, __p1)); \
49342	50366	__ret; \
49343	50367	})
49344	50368	#ifdef __LITTLE_ENDIAN__
49345	50369	#define vdups_lane_f32(__p0, __p1) __extension__ ({ \
49346	50370	float32_t __ret; \
49347	50371	float32x2_t __s0 = __p0; \
49348		__ret = (float32_t) __builtin_neon_vdups_lane_f32((float32x2_t)__s0, __p1); \
	50372	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vdups_lane_f32(__s0, __p1)); \
49349	50373	__ret; \
49350	50374	})
49351	50375	#else
49352	50376	#define vdups_lane_f32(__p0, __p1) __extension__ ({ \
49353	50377	float32_t __ret; \
49354	50378	float32x2_t __s0 = __p0; \
49355		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
49356		__ret = (float32_t) __builtin_neon_vdups_lane_f32((float32x2_t)__rev0, __p1); \
	50379	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	50380	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vdups_lane_f32(__rev0, __p1)); \
49357	50381	__ret; \
49358	50382	})
49359	50383	#endif
...	...	@@ -49362,15 +50386,15 @@ __ai __attribute__((target("neon"))) float32x2_t vdiv_f32(float32x2_t __p0, floa
49362	50386	#define vdups_lane_s32(__p0, __p1) __extension__ ({ \
49363	50387	int32_t __ret; \
49364	50388	int32x2_t __s0 = __p0; \
49365		__ret = (int32_t) __builtin_neon_vdups_lane_i32((int32x2_t)__s0, __p1); \
	50389	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vdups_lane_i32(__builtin_bit_cast(int32x2_t, __s0), __p1)); \
49366	50390	__ret; \
49367	50391	})
49368	50392	#else
49369	50393	#define vdups_lane_s32(__p0, __p1) __extension__ ({ \
49370	50394	int32_t __ret; \
49371	50395	int32x2_t __s0 = __p0; \
49372		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
49373		__ret = (int32_t) __builtin_neon_vdups_lane_i32((int32x2_t)__rev0, __p1); \
	50396	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	50397	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vdups_lane_i32(__builtin_bit_cast(int32x2_t, __rev0), __p1)); \
49374	50398	__ret; \
49375	50399	})
49376	50400	#endif
...	...	@@ -49378,22 +50402,39 @@ __ai __attribute__((target("neon"))) float32x2_t vdiv_f32(float32x2_t __p0, floa
49378	50402	#define vdupd_lane_s64(__p0, __p1) __extension__ ({ \
49379	50403	int64_t __ret; \
49380	50404	int64x1_t __s0 = __p0; \
49381		__ret = (int64_t) __builtin_neon_vdupd_lane_i64((int64x1_t)__s0, __p1); \
	50405	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vdupd_lane_i64(__builtin_bit_cast(int64x1_t, __s0), __p1)); \
49382	50406	__ret; \
49383	50407	})
	50408	#ifdef __LITTLE_ENDIAN__
	50409	#define vdupb_lane_mf8(__p0, __p1) __extension__ ({ \
	50410	mfloat8_t __ret; \
	50411	mfloat8x8_t __s0 = __p0; \
	50412	__ret = __builtin_bit_cast(mfloat8_t, __builtin_neon_vdupb_lane_mf8(__s0, __p1)); \
	50413	__ret; \
	50414	})
	50415	#else
	50416	#define vdupb_lane_mf8(__p0, __p1) __extension__ ({ \
	50417	mfloat8_t __ret; \
	50418	mfloat8x8_t __s0 = __p0; \
	50419	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_8); \
	50420	__ret = __builtin_bit_cast(mfloat8_t, __builtin_neon_vdupb_lane_mf8(__rev0, __p1)); \
	50421	__ret; \
	50422	})
	50423	#endif
	50424
49384	50425	#ifdef __LITTLE_ENDIAN__
49385	50426	#define vduph_lane_s16(__p0, __p1) __extension__ ({ \
49386	50427	int16_t __ret; \
49387	50428	int16x4_t __s0 = __p0; \
49388		__ret = (int16_t) __builtin_neon_vduph_lane_i16((int16x4_t)__s0, __p1); \
	50429	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vduph_lane_i16(__builtin_bit_cast(int16x4_t, __s0), __p1)); \
49389	50430	__ret; \
49390	50431	})
49391	50432	#else
49392	50433	#define vduph_lane_s16(__p0, __p1) __extension__ ({ \
49393	50434	int16_t __ret; \
49394	50435	int16x4_t __s0 = __p0; \
49395		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
49396		__ret = (int16_t) __builtin_neon_vduph_lane_i16((int16x4_t)__rev0, __p1); \
	50436	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	50437	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vduph_lane_i16(__builtin_bit_cast(int16x4_t, __rev0), __p1)); \
49397	50438	__ret; \
49398	50439	})
49399	50440	#endif
...	...	@@ -49402,78 +50443,78 @@ __ai __attribute__((target("neon"))) float32x2_t vdiv_f32(float32x2_t __p0, floa
49402	50443	#define vduph_lane_f16(__p0, __p1) __extension__ ({ \
49403	50444	float16_t __ret; \
49404	50445	float16x4_t __s0 = __p0; \
49405		__ret = (float16_t) __builtin_neon_vduph_lane_f16((float16x4_t)__s0, __p1); \
	50446	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vduph_lane_f16(__s0, __p1)); \
49406	50447	__ret; \
49407	50448	})
49408	50449	#else
49409	50450	#define vduph_lane_f16(__p0, __p1) __extension__ ({ \
49410	50451	float16_t __ret; \
49411	50452	float16x4_t __s0 = __p0; \
49412		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
49413		__ret = (float16_t) __builtin_neon_vduph_lane_f16((float16x4_t)__rev0, __p1); \
	50453	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	50454	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vduph_lane_f16(__rev0, __p1)); \
49414	50455	__ret; \
49415	50456	})
49416	50457	#endif
49417	50458
49418		#define vdup_lane_p64(__p0_364, __p1_364) __extension__ ({ \
49419		poly64x1_t __ret_364; \
49420		poly64x1_t __s0_364 = __p0_364; \
49421		__ret_364 = splat_lane_p64(__s0_364, __p1_364); \
49422		__ret_364; \
	50459	#define vdup_lane_p64(__p0_376, __p1_376) __extension__ ({ \
	50460	poly64x1_t __ret_376; \
	50461	poly64x1_t __s0_376 = __p0_376; \
	50462	__ret_376 = splat_lane_p64(__s0_376, __p1_376); \
	50463	__ret_376; \
49423	50464	})
49424	50465	#ifdef __LITTLE_ENDIAN__
49425		#define vdupq_lane_p64(__p0_365, __p1_365) __extension__ ({ \
49426		poly64x2_t __ret_365; \
49427		poly64x1_t __s0_365 = __p0_365; \
49428		__ret_365 = splatq_lane_p64(__s0_365, __p1_365); \
49429		__ret_365; \
	50466	#define vdupq_lane_p64(__p0_377, __p1_377) __extension__ ({ \
	50467	poly64x2_t __ret_377; \
	50468	poly64x1_t __s0_377 = __p0_377; \
	50469	__ret_377 = splatq_lane_p64(__s0_377, __p1_377); \
	50470	__ret_377; \
49430	50471	})
49431	50472	#else
49432		#define vdupq_lane_p64(__p0_366, __p1_366) __extension__ ({ \
49433		poly64x2_t __ret_366; \
49434		poly64x1_t __s0_366 = __p0_366; \
49435		__ret_366 = __noswap_splatq_lane_p64(__s0_366, __p1_366); \
49436		__ret_366 = __builtin_shufflevector(__ret_366, __ret_366, 1, 0); \
49437		__ret_366; \
	50473	#define vdupq_lane_p64(__p0_378, __p1_378) __extension__ ({ \
	50474	poly64x2_t __ret_378; \
	50475	poly64x1_t __s0_378 = __p0_378; \
	50476	__ret_378 = __noswap_splatq_lane_p64(__s0_378, __p1_378); \
	50477	__ret_378 = __builtin_shufflevector(__ret_378, __ret_378, __lane_reverse_128_64); \
	50478	__ret_378; \
49438	50479	})
49439	50480	#endif
49440	50481
49441	50482	#ifdef __LITTLE_ENDIAN__
49442		#define vdupq_lane_f64(__p0_367, __p1_367) __extension__ ({ \
49443		float64x2_t __ret_367; \
49444		float64x1_t __s0_367 = __p0_367; \
49445		__ret_367 = splatq_lane_f64(__s0_367, __p1_367); \
49446		__ret_367; \
	50483	#define vdupq_lane_f64(__p0_379, __p1_379) __extension__ ({ \
	50484	float64x2_t __ret_379; \
	50485	float64x1_t __s0_379 = __p0_379; \
	50486	__ret_379 = splatq_lane_f64(__s0_379, __p1_379); \
	50487	__ret_379; \
49447	50488	})
49448	50489	#else
49449		#define vdupq_lane_f64(__p0_368, __p1_368) __extension__ ({ \
49450		float64x2_t __ret_368; \
49451		float64x1_t __s0_368 = __p0_368; \
49452		__ret_368 = __noswap_splatq_lane_f64(__s0_368, __p1_368); \
49453		__ret_368 = __builtin_shufflevector(__ret_368, __ret_368, 1, 0); \
49454		__ret_368; \
	50490	#define vdupq_lane_f64(__p0_380, __p1_380) __extension__ ({ \
	50491	float64x2_t __ret_380; \
	50492	float64x1_t __s0_380 = __p0_380; \
	50493	__ret_380 = __noswap_splatq_lane_f64(__s0_380, __p1_380); \
	50494	__ret_380 = __builtin_shufflevector(__ret_380, __ret_380, __lane_reverse_128_64); \
	50495	__ret_380; \
49455	50496	})
49456	50497	#endif
49457	50498
49458		#define vdup_lane_f64(__p0_369, __p1_369) __extension__ ({ \
49459		float64x1_t __ret_369; \
49460		float64x1_t __s0_369 = __p0_369; \
49461		__ret_369 = splat_lane_f64(__s0_369, __p1_369); \
49462		__ret_369; \
	50499	#define vdup_lane_f64(__p0_381, __p1_381) __extension__ ({ \
	50500	float64x1_t __ret_381; \
	50501	float64x1_t __s0_381 = __p0_381; \
	50502	__ret_381 = splat_lane_f64(__s0_381, __p1_381); \
	50503	__ret_381; \
49463	50504	})
49464	50505	#ifdef __LITTLE_ENDIAN__
49465	50506	#define vdupb_laneq_p8(__p0, __p1) __extension__ ({ \
49466	50507	poly8_t __ret; \
49467	50508	poly8x16_t __s0 = __p0; \
49468		__ret = (poly8_t) __builtin_neon_vdupb_laneq_i8((poly8x16_t)__s0, __p1); \
	50509	__ret = __builtin_bit_cast(poly8_t, __builtin_neon_vdupb_laneq_i8(__s0, __p1)); \
49469	50510	__ret; \
49470	50511	})
49471	50512	#else
49472	50513	#define vdupb_laneq_p8(__p0, __p1) __extension__ ({ \
49473	50514	poly8_t __ret; \
49474	50515	poly8x16_t __s0 = __p0; \
49475		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
49476		__ret = (poly8_t) __builtin_neon_vdupb_laneq_i8((poly8x16_t)__rev0, __p1); \
	50516	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	50517	__ret = __builtin_bit_cast(poly8_t, __builtin_neon_vdupb_laneq_i8(__rev0, __p1)); \
49477	50518	__ret; \
49478	50519	})
49479	50520	#endif
...	...	@@ -49482,15 +50523,15 @@ __ai __attribute__((target("neon"))) float32x2_t vdiv_f32(float32x2_t __p0, floa
49482	50523	#define vduph_laneq_p16(__p0, __p1) __extension__ ({ \
49483	50524	poly16_t __ret; \
49484	50525	poly16x8_t __s0 = __p0; \
49485		__ret = (poly16_t) __builtin_neon_vduph_laneq_i16((poly16x8_t)__s0, __p1); \
	50526	__ret = __builtin_bit_cast(poly16_t, __builtin_neon_vduph_laneq_i16(__s0, __p1)); \
49486	50527	__ret; \
49487	50528	})
49488	50529	#else
49489	50530	#define vduph_laneq_p16(__p0, __p1) __extension__ ({ \
49490	50531	poly16_t __ret; \
49491	50532	poly16x8_t __s0 = __p0; \
49492		poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
49493		__ret = (poly16_t) __builtin_neon_vduph_laneq_i16((poly16x8_t)__rev0, __p1); \
	50533	poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	50534	__ret = __builtin_bit_cast(poly16_t, __builtin_neon_vduph_laneq_i16(__rev0, __p1)); \
49494	50535	__ret; \
49495	50536	})
49496	50537	#endif
...	...	@@ -49499,15 +50540,15 @@ __ai __attribute__((target("neon"))) float32x2_t vdiv_f32(float32x2_t __p0, floa
49499	50540	#define vdupb_laneq_u8(__p0, __p1) __extension__ ({ \
49500	50541	uint8_t __ret; \
49501	50542	uint8x16_t __s0 = __p0; \
49502		__ret = (uint8_t) __builtin_neon_vdupb_laneq_i8((int8x16_t)__s0, __p1); \
	50543	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vdupb_laneq_i8(__builtin_bit_cast(int8x16_t, __s0), __p1)); \
49503	50544	__ret; \
49504	50545	})
49505	50546	#else
49506	50547	#define vdupb_laneq_u8(__p0, __p1) __extension__ ({ \
49507	50548	uint8_t __ret; \
49508	50549	uint8x16_t __s0 = __p0; \
49509		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
49510		__ret = (uint8_t) __builtin_neon_vdupb_laneq_i8((int8x16_t)__rev0, __p1); \
	50550	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	50551	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vdupb_laneq_i8(__builtin_bit_cast(int8x16_t, __rev0), __p1)); \
49511	50552	__ret; \
49512	50553	})
49513	50554	#endif
...	...	@@ -49516,15 +50557,15 @@ __ai __attribute__((target("neon"))) float32x2_t vdiv_f32(float32x2_t __p0, floa
49516	50557	#define vdups_laneq_u32(__p0, __p1) __extension__ ({ \
49517	50558	uint32_t __ret; \
49518	50559	uint32x4_t __s0 = __p0; \
49519		__ret = (uint32_t) __builtin_neon_vdups_laneq_i32((int32x4_t)__s0, __p1); \
	50560	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vdups_laneq_i32(__builtin_bit_cast(int32x4_t, __s0), __p1)); \
49520	50561	__ret; \
49521	50562	})
49522	50563	#else
49523	50564	#define vdups_laneq_u32(__p0, __p1) __extension__ ({ \
49524	50565	uint32_t __ret; \
49525	50566	uint32x4_t __s0 = __p0; \
49526		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
49527		__ret = (uint32_t) __builtin_neon_vdups_laneq_i32((int32x4_t)__rev0, __p1); \
	50567	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	50568	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vdups_laneq_i32(__builtin_bit_cast(int32x4_t, __rev0), __p1)); \
49528	50569	__ret; \
49529	50570	})
49530	50571	#endif
...	...	@@ -49533,15 +50574,15 @@ __ai __attribute__((target("neon"))) float32x2_t vdiv_f32(float32x2_t __p0, floa
49533	50574	#define vdupd_laneq_u64(__p0, __p1) __extension__ ({ \
49534	50575	uint64_t __ret; \
49535	50576	uint64x2_t __s0 = __p0; \
49536		__ret = (uint64_t) __builtin_neon_vdupd_laneq_i64((int64x2_t)__s0, __p1); \
	50577	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vdupd_laneq_i64(__builtin_bit_cast(int64x2_t, __s0), __p1)); \
49537	50578	__ret; \
49538	50579	})
49539	50580	#else
49540	50581	#define vdupd_laneq_u64(__p0, __p1) __extension__ ({ \
49541	50582	uint64_t __ret; \
49542	50583	uint64x2_t __s0 = __p0; \
49543		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
49544		__ret = (uint64_t) __builtin_neon_vdupd_laneq_i64((int64x2_t)__rev0, __p1); \
	50584	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	50585	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vdupd_laneq_i64(__builtin_bit_cast(int64x2_t, __rev0), __p1)); \
49545	50586	__ret; \
49546	50587	})
49547	50588	#endif
...	...	@@ -49550,15 +50591,15 @@ __ai __attribute__((target("neon"))) float32x2_t vdiv_f32(float32x2_t __p0, floa
49550	50591	#define vduph_laneq_u16(__p0, __p1) __extension__ ({ \
49551	50592	uint16_t __ret; \
49552	50593	uint16x8_t __s0 = __p0; \
49553		__ret = (uint16_t) __builtin_neon_vduph_laneq_i16((int16x8_t)__s0, __p1); \
	50594	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vduph_laneq_i16(__builtin_bit_cast(int16x8_t, __s0), __p1)); \
49554	50595	__ret; \
49555	50596	})
49556	50597	#else
49557	50598	#define vduph_laneq_u16(__p0, __p1) __extension__ ({ \
49558	50599	uint16_t __ret; \
49559	50600	uint16x8_t __s0 = __p0; \
49560		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
49561		__ret = (uint16_t) __builtin_neon_vduph_laneq_i16((int16x8_t)__rev0, __p1); \
	50601	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	50602	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vduph_laneq_i16(__builtin_bit_cast(int16x8_t, __rev0), __p1)); \
49562	50603	__ret; \
49563	50604	})
49564	50605	#endif
...	...	@@ -49567,15 +50608,15 @@ __ai __attribute__((target("neon"))) float32x2_t vdiv_f32(float32x2_t __p0, floa
49567	50608	#define vdupb_laneq_s8(__p0, __p1) __extension__ ({ \
49568	50609	int8_t __ret; \
49569	50610	int8x16_t __s0 = __p0; \
49570		__ret = (int8_t) __builtin_neon_vdupb_laneq_i8((int8x16_t)__s0, __p1); \
	50611	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vdupb_laneq_i8(__builtin_bit_cast(int8x16_t, __s0), __p1)); \
49571	50612	__ret; \
49572	50613	})
49573	50614	#else
49574	50615	#define vdupb_laneq_s8(__p0, __p1) __extension__ ({ \
49575	50616	int8_t __ret; \
49576	50617	int8x16_t __s0 = __p0; \
49577		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
49578		__ret = (int8_t) __builtin_neon_vdupb_laneq_i8((int8x16_t)__rev0, __p1); \
	50618	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	50619	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vdupb_laneq_i8(__builtin_bit_cast(int8x16_t, __rev0), __p1)); \
49579	50620	__ret; \
49580	50621	})
49581	50622	#endif
...	...	@@ -49584,15 +50625,15 @@ __ai __attribute__((target("neon"))) float32x2_t vdiv_f32(float32x2_t __p0, floa
49584	50625	#define vdupd_laneq_f64(__p0, __p1) __extension__ ({ \
49585	50626	float64_t __ret; \
49586	50627	float64x2_t __s0 = __p0; \
49587		__ret = (float64_t) __builtin_neon_vdupd_laneq_f64((float64x2_t)__s0, __p1); \
	50628	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vdupd_laneq_f64(__s0, __p1)); \
49588	50629	__ret; \
49589	50630	})
49590	50631	#else
49591	50632	#define vdupd_laneq_f64(__p0, __p1) __extension__ ({ \
49592	50633	float64_t __ret; \
49593	50634	float64x2_t __s0 = __p0; \
49594		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
49595		__ret = (float64_t) __builtin_neon_vdupd_laneq_f64((float64x2_t)__rev0, __p1); \
	50635	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	50636	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vdupd_laneq_f64(__rev0, __p1)); \
49596	50637	__ret; \
49597	50638	})
49598	50639	#endif
...	...	@@ -49601,15 +50642,15 @@ __ai __attribute__((target("neon"))) float32x2_t vdiv_f32(float32x2_t __p0, floa
49601	50642	#define vdups_laneq_f32(__p0, __p1) __extension__ ({ \
49602	50643	float32_t __ret; \
49603	50644	float32x4_t __s0 = __p0; \
49604		__ret = (float32_t) __builtin_neon_vdups_laneq_f32((float32x4_t)__s0, __p1); \
	50645	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vdups_laneq_f32(__s0, __p1)); \
49605	50646	__ret; \
49606	50647	})
49607	50648	#else
49608	50649	#define vdups_laneq_f32(__p0, __p1) __extension__ ({ \
49609	50650	float32_t __ret; \
49610	50651	float32x4_t __s0 = __p0; \
49611		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
49612		__ret = (float32_t) __builtin_neon_vdups_laneq_f32((float32x4_t)__rev0, __p1); \
	50652	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	50653	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vdups_laneq_f32(__rev0, __p1)); \
49613	50654	__ret; \
49614	50655	})
49615	50656	#endif
...	...	@@ -49618,15 +50659,15 @@ __ai __attribute__((target("neon"))) float32x2_t vdiv_f32(float32x2_t __p0, floa
49618	50659	#define vdups_laneq_s32(__p0, __p1) __extension__ ({ \
49619	50660	int32_t __ret; \
49620	50661	int32x4_t __s0 = __p0; \
49621		__ret = (int32_t) __builtin_neon_vdups_laneq_i32((int32x4_t)__s0, __p1); \
	50662	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vdups_laneq_i32(__builtin_bit_cast(int32x4_t, __s0), __p1)); \
49622	50663	__ret; \
49623	50664	})
49624	50665	#else
49625	50666	#define vdups_laneq_s32(__p0, __p1) __extension__ ({ \
49626	50667	int32_t __ret; \
49627	50668	int32x4_t __s0 = __p0; \
49628		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
49629		__ret = (int32_t) __builtin_neon_vdups_laneq_i32((int32x4_t)__rev0, __p1); \
	50669	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	50670	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vdups_laneq_i32(__builtin_bit_cast(int32x4_t, __rev0), __p1)); \
49630	50671	__ret; \
49631	50672	})
49632	50673	#endif
...	...	@@ -49635,15 +50676,32 @@ __ai __attribute__((target("neon"))) float32x2_t vdiv_f32(float32x2_t __p0, floa
49635	50676	#define vdupd_laneq_s64(__p0, __p1) __extension__ ({ \
49636	50677	int64_t __ret; \
49637	50678	int64x2_t __s0 = __p0; \
49638		__ret = (int64_t) __builtin_neon_vdupd_laneq_i64((int64x2_t)__s0, __p1); \
	50679	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vdupd_laneq_i64(__builtin_bit_cast(int64x2_t, __s0), __p1)); \
49639	50680	__ret; \
49640	50681	})
49641	50682	#else
49642	50683	#define vdupd_laneq_s64(__p0, __p1) __extension__ ({ \
49643	50684	int64_t __ret; \
49644	50685	int64x2_t __s0 = __p0; \
49645		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
49646		__ret = (int64_t) __builtin_neon_vdupd_laneq_i64((int64x2_t)__rev0, __p1); \
	50686	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	50687	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vdupd_laneq_i64(__builtin_bit_cast(int64x2_t, __rev0), __p1)); \
	50688	__ret; \
	50689	})
	50690	#endif
	50691
	50692	#ifdef __LITTLE_ENDIAN__
	50693	#define vdupb_laneq_mf8(__p0, __p1) __extension__ ({ \
	50694	mfloat8_t __ret; \
	50695	mfloat8x16_t __s0 = __p0; \
	50696	__ret = __builtin_bit_cast(mfloat8_t, __builtin_neon_vdupb_laneq_mf8(__s0, __p1)); \
	50697	__ret; \
	50698	})
	50699	#else
	50700	#define vdupb_laneq_mf8(__p0, __p1) __extension__ ({ \
	50701	mfloat8_t __ret; \
	50702	mfloat8x16_t __s0 = __p0; \
	50703	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_8); \
	50704	__ret = __builtin_bit_cast(mfloat8_t, __builtin_neon_vdupb_laneq_mf8(__rev0, __p1)); \
49647	50705	__ret; \
49648	50706	})
49649	50707	#endif
...	...	@@ -49652,15 +50710,15 @@ __ai __attribute__((target("neon"))) float32x2_t vdiv_f32(float32x2_t __p0, floa
49652	50710	#define vduph_laneq_s16(__p0, __p1) __extension__ ({ \
49653	50711	int16_t __ret; \
49654	50712	int16x8_t __s0 = __p0; \
49655		__ret = (int16_t) __builtin_neon_vduph_laneq_i16((int16x8_t)__s0, __p1); \
	50713	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vduph_laneq_i16(__builtin_bit_cast(int16x8_t, __s0), __p1)); \
49656	50714	__ret; \
49657	50715	})
49658	50716	#else
49659	50717	#define vduph_laneq_s16(__p0, __p1) __extension__ ({ \
49660	50718	int16_t __ret; \
49661	50719	int16x8_t __s0 = __p0; \
49662		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
49663		__ret = (int16_t) __builtin_neon_vduph_laneq_i16((int16x8_t)__rev0, __p1); \
	50720	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	50721	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vduph_laneq_i16(__builtin_bit_cast(int16x8_t, __rev0), __p1)); \
49664	50722	__ret; \
49665	50723	})
49666	50724	#endif
...	...	@@ -49669,519 +50727,555 @@ __ai __attribute__((target("neon"))) float32x2_t vdiv_f32(float32x2_t __p0, floa
49669	50727	#define vduph_laneq_f16(__p0, __p1) __extension__ ({ \
49670	50728	float16_t __ret; \
49671	50729	float16x8_t __s0 = __p0; \
49672		__ret = (float16_t) __builtin_neon_vduph_laneq_f16((float16x8_t)__s0, __p1); \
	50730	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vduph_laneq_f16(__s0, __p1)); \
49673	50731	__ret; \
49674	50732	})
49675	50733	#else
49676	50734	#define vduph_laneq_f16(__p0, __p1) __extension__ ({ \
49677	50735	float16_t __ret; \
49678	50736	float16x8_t __s0 = __p0; \
49679		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
49680		__ret = (float16_t) __builtin_neon_vduph_laneq_f16((float16x8_t)__rev0, __p1); \
	50737	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	50738	__ret = __builtin_bit_cast(float16_t, __builtin_neon_vduph_laneq_f16(__rev0, __p1)); \
49681	50739	__ret; \
49682	50740	})
49683	50741	#endif
49684	50742
49685	50743	#ifdef __LITTLE_ENDIAN__
49686		#define vdup_laneq_p8(__p0_370, __p1_370) __extension__ ({ \
49687		poly8x8_t __ret_370; \
49688		poly8x16_t __s0_370 = __p0_370; \
49689		__ret_370 = splat_laneq_p8(__s0_370, __p1_370); \
49690		__ret_370; \
49691		})
49692		#else
49693		#define vdup_laneq_p8(__p0_371, __p1_371) __extension__ ({ \
49694		poly8x8_t __ret_371; \
49695		poly8x16_t __s0_371 = __p0_371; \
49696		poly8x16_t __rev0_371; __rev0_371 = __builtin_shufflevector(__s0_371, __s0_371, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
49697		__ret_371 = __noswap_splat_laneq_p8(__rev0_371, __p1_371); \
49698		__ret_371 = __builtin_shufflevector(__ret_371, __ret_371, 7, 6, 5, 4, 3, 2, 1, 0); \
49699		__ret_371; \
49700		})
49701		#endif
49702
49703		#ifdef __LITTLE_ENDIAN__
49704		#define vdup_laneq_p64(__p0_372, __p1_372) __extension__ ({ \
49705		poly64x1_t __ret_372; \
49706		poly64x2_t __s0_372 = __p0_372; \
49707		__ret_372 = splat_laneq_p64(__s0_372, __p1_372); \
49708		__ret_372; \
49709		})
49710		#else
49711		#define vdup_laneq_p64(__p0_373, __p1_373) __extension__ ({ \
49712		poly64x1_t __ret_373; \
49713		poly64x2_t __s0_373 = __p0_373; \
49714		poly64x2_t __rev0_373; __rev0_373 = __builtin_shufflevector(__s0_373, __s0_373, 1, 0); \
49715		__ret_373 = __noswap_splat_laneq_p64(__rev0_373, __p1_373); \
49716		__ret_373; \
49717		})
49718		#endif
49719
49720		#ifdef __LITTLE_ENDIAN__
49721		#define vdup_laneq_p16(__p0_374, __p1_374) __extension__ ({ \
49722		poly16x4_t __ret_374; \
49723		poly16x8_t __s0_374 = __p0_374; \
49724		__ret_374 = splat_laneq_p16(__s0_374, __p1_374); \
49725		__ret_374; \
49726		})
49727		#else
49728		#define vdup_laneq_p16(__p0_375, __p1_375) __extension__ ({ \
49729		poly16x4_t __ret_375; \
49730		poly16x8_t __s0_375 = __p0_375; \
49731		poly16x8_t __rev0_375; __rev0_375 = __builtin_shufflevector(__s0_375, __s0_375, 7, 6, 5, 4, 3, 2, 1, 0); \
49732		__ret_375 = __noswap_splat_laneq_p16(__rev0_375, __p1_375); \
49733		__ret_375 = __builtin_shufflevector(__ret_375, __ret_375, 3, 2, 1, 0); \
49734		__ret_375; \
49735		})
49736		#endif
49737
49738		#ifdef __LITTLE_ENDIAN__
49739		#define vdupq_laneq_p8(__p0_376, __p1_376) __extension__ ({ \
49740		poly8x16_t __ret_376; \
49741		poly8x16_t __s0_376 = __p0_376; \
49742		__ret_376 = splatq_laneq_p8(__s0_376, __p1_376); \
49743		__ret_376; \
49744		})
49745		#else
49746		#define vdupq_laneq_p8(__p0_377, __p1_377) __extension__ ({ \
49747		poly8x16_t __ret_377; \
49748		poly8x16_t __s0_377 = __p0_377; \
49749		poly8x16_t __rev0_377; __rev0_377 = __builtin_shufflevector(__s0_377, __s0_377, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
49750		__ret_377 = __noswap_splatq_laneq_p8(__rev0_377, __p1_377); \
49751		__ret_377 = __builtin_shufflevector(__ret_377, __ret_377, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
49752		__ret_377; \
49753		})
49754		#endif
49755
49756		#ifdef __LITTLE_ENDIAN__
49757		#define vdupq_laneq_p64(__p0_378, __p1_378) __extension__ ({ \
49758		poly64x2_t __ret_378; \
49759		poly64x2_t __s0_378 = __p0_378; \
49760		__ret_378 = splatq_laneq_p64(__s0_378, __p1_378); \
49761		__ret_378; \
49762		})
49763		#else
49764		#define vdupq_laneq_p64(__p0_379, __p1_379) __extension__ ({ \
49765		poly64x2_t __ret_379; \
49766		poly64x2_t __s0_379 = __p0_379; \
49767		poly64x2_t __rev0_379; __rev0_379 = __builtin_shufflevector(__s0_379, __s0_379, 1, 0); \
49768		__ret_379 = __noswap_splatq_laneq_p64(__rev0_379, __p1_379); \
49769		__ret_379 = __builtin_shufflevector(__ret_379, __ret_379, 1, 0); \
49770		__ret_379; \
49771		})
49772		#endif
49773
49774		#ifdef __LITTLE_ENDIAN__
49775		#define vdupq_laneq_p16(__p0_380, __p1_380) __extension__ ({ \
49776		poly16x8_t __ret_380; \
49777		poly16x8_t __s0_380 = __p0_380; \
49778		__ret_380 = splatq_laneq_p16(__s0_380, __p1_380); \
49779		__ret_380; \
49780		})
49781		#else
49782		#define vdupq_laneq_p16(__p0_381, __p1_381) __extension__ ({ \
49783		poly16x8_t __ret_381; \
49784		poly16x8_t __s0_381 = __p0_381; \
49785		poly16x8_t __rev0_381; __rev0_381 = __builtin_shufflevector(__s0_381, __s0_381, 7, 6, 5, 4, 3, 2, 1, 0); \
49786		__ret_381 = __noswap_splatq_laneq_p16(__rev0_381, __p1_381); \
49787		__ret_381 = __builtin_shufflevector(__ret_381, __ret_381, 7, 6, 5, 4, 3, 2, 1, 0); \
49788		__ret_381; \
49789		})
49790		#endif
49791
49792		#ifdef __LITTLE_ENDIAN__
49793		#define vdupq_laneq_u8(__p0_382, __p1_382) __extension__ ({ \
49794		uint8x16_t __ret_382; \
49795		uint8x16_t __s0_382 = __p0_382; \
49796		__ret_382 = splatq_laneq_u8(__s0_382, __p1_382); \
	50744	#define vdup_laneq_p8(__p0_382, __p1_382) __extension__ ({ \
	50745	poly8x8_t __ret_382; \
	50746	poly8x16_t __s0_382 = __p0_382; \
	50747	__ret_382 = splat_laneq_p8(__s0_382, __p1_382); \
49797	50748	__ret_382; \
49798	50749	})
49799	50750	#else
49800		#define vdupq_laneq_u8(__p0_383, __p1_383) __extension__ ({ \
49801		uint8x16_t __ret_383; \
49802		uint8x16_t __s0_383 = __p0_383; \
49803		uint8x16_t __rev0_383; __rev0_383 = __builtin_shufflevector(__s0_383, __s0_383, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
49804		__ret_383 = __noswap_splatq_laneq_u8(__rev0_383, __p1_383); \
49805		__ret_383 = __builtin_shufflevector(__ret_383, __ret_383, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	50751	#define vdup_laneq_p8(__p0_383, __p1_383) __extension__ ({ \
	50752	poly8x8_t __ret_383; \
	50753	poly8x16_t __s0_383 = __p0_383; \
	50754	poly8x16_t __rev0_383; __rev0_383 = __builtin_shufflevector(__s0_383, __s0_383, __lane_reverse_128_8); \
	50755	__ret_383 = __noswap_splat_laneq_p8(__rev0_383, __p1_383); \
	50756	__ret_383 = __builtin_shufflevector(__ret_383, __ret_383, __lane_reverse_64_8); \
49806	50757	__ret_383; \
49807	50758	})
49808	50759	#endif
49809	50760
49810	50761	#ifdef __LITTLE_ENDIAN__
49811		#define vdupq_laneq_u32(__p0_384, __p1_384) __extension__ ({ \
49812		uint32x4_t __ret_384; \
49813		uint32x4_t __s0_384 = __p0_384; \
49814		__ret_384 = splatq_laneq_u32(__s0_384, __p1_384); \
	50762	#define vdup_laneq_p64(__p0_384, __p1_384) __extension__ ({ \
	50763	poly64x1_t __ret_384; \
	50764	poly64x2_t __s0_384 = __p0_384; \
	50765	__ret_384 = splat_laneq_p64(__s0_384, __p1_384); \
49815	50766	__ret_384; \
49816	50767	})
49817	50768	#else
49818		#define vdupq_laneq_u32(__p0_385, __p1_385) __extension__ ({ \
49819		uint32x4_t __ret_385; \
49820		uint32x4_t __s0_385 = __p0_385; \
49821		uint32x4_t __rev0_385; __rev0_385 = __builtin_shufflevector(__s0_385, __s0_385, 3, 2, 1, 0); \
49822		__ret_385 = __noswap_splatq_laneq_u32(__rev0_385, __p1_385); \
49823		__ret_385 = __builtin_shufflevector(__ret_385, __ret_385, 3, 2, 1, 0); \
	50769	#define vdup_laneq_p64(__p0_385, __p1_385) __extension__ ({ \
	50770	poly64x1_t __ret_385; \
	50771	poly64x2_t __s0_385 = __p0_385; \
	50772	poly64x2_t __rev0_385; __rev0_385 = __builtin_shufflevector(__s0_385, __s0_385, __lane_reverse_128_64); \
	50773	__ret_385 = __noswap_splat_laneq_p64(__rev0_385, __p1_385); \
49824	50774	__ret_385; \
49825	50775	})
49826	50776	#endif
49827	50777
49828	50778	#ifdef __LITTLE_ENDIAN__
49829		#define vdupq_laneq_u64(__p0_386, __p1_386) __extension__ ({ \
49830		uint64x2_t __ret_386; \
49831		uint64x2_t __s0_386 = __p0_386; \
49832		__ret_386 = splatq_laneq_u64(__s0_386, __p1_386); \
	50779	#define vdup_laneq_p16(__p0_386, __p1_386) __extension__ ({ \
	50780	poly16x4_t __ret_386; \
	50781	poly16x8_t __s0_386 = __p0_386; \
	50782	__ret_386 = splat_laneq_p16(__s0_386, __p1_386); \
49833	50783	__ret_386; \
49834	50784	})
49835	50785	#else
49836		#define vdupq_laneq_u64(__p0_387, __p1_387) __extension__ ({ \
49837		uint64x2_t __ret_387; \
49838		uint64x2_t __s0_387 = __p0_387; \
49839		uint64x2_t __rev0_387; __rev0_387 = __builtin_shufflevector(__s0_387, __s0_387, 1, 0); \
49840		__ret_387 = __noswap_splatq_laneq_u64(__rev0_387, __p1_387); \
49841		__ret_387 = __builtin_shufflevector(__ret_387, __ret_387, 1, 0); \
	50786	#define vdup_laneq_p16(__p0_387, __p1_387) __extension__ ({ \
	50787	poly16x4_t __ret_387; \
	50788	poly16x8_t __s0_387 = __p0_387; \
	50789	poly16x8_t __rev0_387; __rev0_387 = __builtin_shufflevector(__s0_387, __s0_387, __lane_reverse_128_16); \
	50790	__ret_387 = __noswap_splat_laneq_p16(__rev0_387, __p1_387); \
	50791	__ret_387 = __builtin_shufflevector(__ret_387, __ret_387, __lane_reverse_64_16); \
49842	50792	__ret_387; \
49843	50793	})
49844	50794	#endif
49845	50795
49846	50796	#ifdef __LITTLE_ENDIAN__
49847		#define vdupq_laneq_u16(__p0_388, __p1_388) __extension__ ({ \
49848		uint16x8_t __ret_388; \
49849		uint16x8_t __s0_388 = __p0_388; \
49850		__ret_388 = splatq_laneq_u16(__s0_388, __p1_388); \
	50797	#define vdupq_laneq_p8(__p0_388, __p1_388) __extension__ ({ \
	50798	poly8x16_t __ret_388; \
	50799	poly8x16_t __s0_388 = __p0_388; \
	50800	__ret_388 = splatq_laneq_p8(__s0_388, __p1_388); \
49851	50801	__ret_388; \
49852	50802	})
49853	50803	#else
49854		#define vdupq_laneq_u16(__p0_389, __p1_389) __extension__ ({ \
49855		uint16x8_t __ret_389; \
49856		uint16x8_t __s0_389 = __p0_389; \
49857		uint16x8_t __rev0_389; __rev0_389 = __builtin_shufflevector(__s0_389, __s0_389, 7, 6, 5, 4, 3, 2, 1, 0); \
49858		__ret_389 = __noswap_splatq_laneq_u16(__rev0_389, __p1_389); \
49859		__ret_389 = __builtin_shufflevector(__ret_389, __ret_389, 7, 6, 5, 4, 3, 2, 1, 0); \
	50804	#define vdupq_laneq_p8(__p0_389, __p1_389) __extension__ ({ \
	50805	poly8x16_t __ret_389; \
	50806	poly8x16_t __s0_389 = __p0_389; \
	50807	poly8x16_t __rev0_389; __rev0_389 = __builtin_shufflevector(__s0_389, __s0_389, __lane_reverse_128_8); \
	50808	__ret_389 = __noswap_splatq_laneq_p8(__rev0_389, __p1_389); \
	50809	__ret_389 = __builtin_shufflevector(__ret_389, __ret_389, __lane_reverse_128_8); \
49860	50810	__ret_389; \
49861	50811	})
49862	50812	#endif
49863	50813
49864	50814	#ifdef __LITTLE_ENDIAN__
49865		#define vdupq_laneq_s8(__p0_390, __p1_390) __extension__ ({ \
49866		int8x16_t __ret_390; \
49867		int8x16_t __s0_390 = __p0_390; \
49868		__ret_390 = splatq_laneq_s8(__s0_390, __p1_390); \
	50815	#define vdupq_laneq_p64(__p0_390, __p1_390) __extension__ ({ \
	50816	poly64x2_t __ret_390; \
	50817	poly64x2_t __s0_390 = __p0_390; \
	50818	__ret_390 = splatq_laneq_p64(__s0_390, __p1_390); \
49869	50819	__ret_390; \
49870	50820	})
49871	50821	#else
49872		#define vdupq_laneq_s8(__p0_391, __p1_391) __extension__ ({ \
49873		int8x16_t __ret_391; \
49874		int8x16_t __s0_391 = __p0_391; \
49875		int8x16_t __rev0_391; __rev0_391 = __builtin_shufflevector(__s0_391, __s0_391, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
49876		__ret_391 = __noswap_splatq_laneq_s8(__rev0_391, __p1_391); \
49877		__ret_391 = __builtin_shufflevector(__ret_391, __ret_391, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	50822	#define vdupq_laneq_p64(__p0_391, __p1_391) __extension__ ({ \
	50823	poly64x2_t __ret_391; \
	50824	poly64x2_t __s0_391 = __p0_391; \
	50825	poly64x2_t __rev0_391; __rev0_391 = __builtin_shufflevector(__s0_391, __s0_391, __lane_reverse_128_64); \
	50826	__ret_391 = __noswap_splatq_laneq_p64(__rev0_391, __p1_391); \
	50827	__ret_391 = __builtin_shufflevector(__ret_391, __ret_391, __lane_reverse_128_64); \
49878	50828	__ret_391; \
49879	50829	})
49880	50830	#endif
49881	50831
49882	50832	#ifdef __LITTLE_ENDIAN__
49883		#define vdupq_laneq_f64(__p0_392, __p1_392) __extension__ ({ \
49884		float64x2_t __ret_392; \
49885		float64x2_t __s0_392 = __p0_392; \
49886		__ret_392 = splatq_laneq_f64(__s0_392, __p1_392); \
	50833	#define vdupq_laneq_p16(__p0_392, __p1_392) __extension__ ({ \
	50834	poly16x8_t __ret_392; \
	50835	poly16x8_t __s0_392 = __p0_392; \
	50836	__ret_392 = splatq_laneq_p16(__s0_392, __p1_392); \
49887	50837	__ret_392; \
49888	50838	})
49889	50839	#else
49890		#define vdupq_laneq_f64(__p0_393, __p1_393) __extension__ ({ \
49891		float64x2_t __ret_393; \
49892		float64x2_t __s0_393 = __p0_393; \
49893		float64x2_t __rev0_393; __rev0_393 = __builtin_shufflevector(__s0_393, __s0_393, 1, 0); \
49894		__ret_393 = __noswap_splatq_laneq_f64(__rev0_393, __p1_393); \
49895		__ret_393 = __builtin_shufflevector(__ret_393, __ret_393, 1, 0); \
	50840	#define vdupq_laneq_p16(__p0_393, __p1_393) __extension__ ({ \
	50841	poly16x8_t __ret_393; \
	50842	poly16x8_t __s0_393 = __p0_393; \
	50843	poly16x8_t __rev0_393; __rev0_393 = __builtin_shufflevector(__s0_393, __s0_393, __lane_reverse_128_16); \
	50844	__ret_393 = __noswap_splatq_laneq_p16(__rev0_393, __p1_393); \
	50845	__ret_393 = __builtin_shufflevector(__ret_393, __ret_393, __lane_reverse_128_16); \
49896	50846	__ret_393; \
49897	50847	})
49898	50848	#endif
49899	50849
49900	50850	#ifdef __LITTLE_ENDIAN__
49901		#define vdupq_laneq_f32(__p0_394, __p1_394) __extension__ ({ \
49902		float32x4_t __ret_394; \
49903		float32x4_t __s0_394 = __p0_394; \
49904		__ret_394 = splatq_laneq_f32(__s0_394, __p1_394); \
	50851	#define vdupq_laneq_u8(__p0_394, __p1_394) __extension__ ({ \
	50852	uint8x16_t __ret_394; \
	50853	uint8x16_t __s0_394 = __p0_394; \
	50854	__ret_394 = splatq_laneq_u8(__s0_394, __p1_394); \
49905	50855	__ret_394; \
49906	50856	})
49907	50857	#else
49908		#define vdupq_laneq_f32(__p0_395, __p1_395) __extension__ ({ \
49909		float32x4_t __ret_395; \
49910		float32x4_t __s0_395 = __p0_395; \
49911		float32x4_t __rev0_395; __rev0_395 = __builtin_shufflevector(__s0_395, __s0_395, 3, 2, 1, 0); \
49912		__ret_395 = __noswap_splatq_laneq_f32(__rev0_395, __p1_395); \
49913		__ret_395 = __builtin_shufflevector(__ret_395, __ret_395, 3, 2, 1, 0); \
	50858	#define vdupq_laneq_u8(__p0_395, __p1_395) __extension__ ({ \
	50859	uint8x16_t __ret_395; \
	50860	uint8x16_t __s0_395 = __p0_395; \
	50861	uint8x16_t __rev0_395; __rev0_395 = __builtin_shufflevector(__s0_395, __s0_395, __lane_reverse_128_8); \
	50862	__ret_395 = __noswap_splatq_laneq_u8(__rev0_395, __p1_395); \
	50863	__ret_395 = __builtin_shufflevector(__ret_395, __ret_395, __lane_reverse_128_8); \
49914	50864	__ret_395; \
49915	50865	})
49916	50866	#endif
49917	50867
49918	50868	#ifdef __LITTLE_ENDIAN__
49919		#define vdupq_laneq_f16(__p0_396, __p1_396) __extension__ ({ \
49920		float16x8_t __ret_396; \
49921		float16x8_t __s0_396 = __p0_396; \
49922		__ret_396 = splatq_laneq_f16(__s0_396, __p1_396); \
	50869	#define vdupq_laneq_u32(__p0_396, __p1_396) __extension__ ({ \
	50870	uint32x4_t __ret_396; \
	50871	uint32x4_t __s0_396 = __p0_396; \
	50872	__ret_396 = splatq_laneq_u32(__s0_396, __p1_396); \
49923	50873	__ret_396; \
49924	50874	})
49925	50875	#else
49926		#define vdupq_laneq_f16(__p0_397, __p1_397) __extension__ ({ \
49927		float16x8_t __ret_397; \
49928		float16x8_t __s0_397 = __p0_397; \
49929		float16x8_t __rev0_397; __rev0_397 = __builtin_shufflevector(__s0_397, __s0_397, 7, 6, 5, 4, 3, 2, 1, 0); \
49930		__ret_397 = __noswap_splatq_laneq_f16(__rev0_397, __p1_397); \
49931		__ret_397 = __builtin_shufflevector(__ret_397, __ret_397, 7, 6, 5, 4, 3, 2, 1, 0); \
	50876	#define vdupq_laneq_u32(__p0_397, __p1_397) __extension__ ({ \
	50877	uint32x4_t __ret_397; \
	50878	uint32x4_t __s0_397 = __p0_397; \
	50879	uint32x4_t __rev0_397; __rev0_397 = __builtin_shufflevector(__s0_397, __s0_397, __lane_reverse_128_32); \
	50880	__ret_397 = __noswap_splatq_laneq_u32(__rev0_397, __p1_397); \
	50881	__ret_397 = __builtin_shufflevector(__ret_397, __ret_397, __lane_reverse_128_32); \
49932	50882	__ret_397; \
49933	50883	})
49934	50884	#endif
49935	50885
49936	50886	#ifdef __LITTLE_ENDIAN__
49937		#define vdupq_laneq_s32(__p0_398, __p1_398) __extension__ ({ \
49938		int32x4_t __ret_398; \
49939		int32x4_t __s0_398 = __p0_398; \
49940		__ret_398 = splatq_laneq_s32(__s0_398, __p1_398); \
	50887	#define vdupq_laneq_u64(__p0_398, __p1_398) __extension__ ({ \
	50888	uint64x2_t __ret_398; \
	50889	uint64x2_t __s0_398 = __p0_398; \
	50890	__ret_398 = splatq_laneq_u64(__s0_398, __p1_398); \
49941	50891	__ret_398; \
49942	50892	})
49943	50893	#else
49944		#define vdupq_laneq_s32(__p0_399, __p1_399) __extension__ ({ \
49945		int32x4_t __ret_399; \
49946		int32x4_t __s0_399 = __p0_399; \
49947		int32x4_t __rev0_399; __rev0_399 = __builtin_shufflevector(__s0_399, __s0_399, 3, 2, 1, 0); \
49948		__ret_399 = __noswap_splatq_laneq_s32(__rev0_399, __p1_399); \
49949		__ret_399 = __builtin_shufflevector(__ret_399, __ret_399, 3, 2, 1, 0); \
	50894	#define vdupq_laneq_u64(__p0_399, __p1_399) __extension__ ({ \
	50895	uint64x2_t __ret_399; \
	50896	uint64x2_t __s0_399 = __p0_399; \
	50897	uint64x2_t __rev0_399; __rev0_399 = __builtin_shufflevector(__s0_399, __s0_399, __lane_reverse_128_64); \
	50898	__ret_399 = __noswap_splatq_laneq_u64(__rev0_399, __p1_399); \
	50899	__ret_399 = __builtin_shufflevector(__ret_399, __ret_399, __lane_reverse_128_64); \
49950	50900	__ret_399; \
49951	50901	})
49952	50902	#endif
49953	50903
49954	50904	#ifdef __LITTLE_ENDIAN__
49955		#define vdupq_laneq_s64(__p0_400, __p1_400) __extension__ ({ \
49956		int64x2_t __ret_400; \
49957		int64x2_t __s0_400 = __p0_400; \
49958		__ret_400 = splatq_laneq_s64(__s0_400, __p1_400); \
	50905	#define vdupq_laneq_u16(__p0_400, __p1_400) __extension__ ({ \
	50906	uint16x8_t __ret_400; \
	50907	uint16x8_t __s0_400 = __p0_400; \
	50908	__ret_400 = splatq_laneq_u16(__s0_400, __p1_400); \
49959	50909	__ret_400; \
49960	50910	})
49961	50911	#else
49962		#define vdupq_laneq_s64(__p0_401, __p1_401) __extension__ ({ \
49963		int64x2_t __ret_401; \
49964		int64x2_t __s0_401 = __p0_401; \
49965		int64x2_t __rev0_401; __rev0_401 = __builtin_shufflevector(__s0_401, __s0_401, 1, 0); \
49966		__ret_401 = __noswap_splatq_laneq_s64(__rev0_401, __p1_401); \
49967		__ret_401 = __builtin_shufflevector(__ret_401, __ret_401, 1, 0); \
	50912	#define vdupq_laneq_u16(__p0_401, __p1_401) __extension__ ({ \
	50913	uint16x8_t __ret_401; \
	50914	uint16x8_t __s0_401 = __p0_401; \
	50915	uint16x8_t __rev0_401; __rev0_401 = __builtin_shufflevector(__s0_401, __s0_401, __lane_reverse_128_16); \
	50916	__ret_401 = __noswap_splatq_laneq_u16(__rev0_401, __p1_401); \
	50917	__ret_401 = __builtin_shufflevector(__ret_401, __ret_401, __lane_reverse_128_16); \
49968	50918	__ret_401; \
49969	50919	})
49970	50920	#endif
49971	50921
49972	50922	#ifdef __LITTLE_ENDIAN__
49973		#define vdupq_laneq_s16(__p0_402, __p1_402) __extension__ ({ \
49974		int16x8_t __ret_402; \
49975		int16x8_t __s0_402 = __p0_402; \
49976		__ret_402 = splatq_laneq_s16(__s0_402, __p1_402); \
	50923	#define vdupq_laneq_s8(__p0_402, __p1_402) __extension__ ({ \
	50924	int8x16_t __ret_402; \
	50925	int8x16_t __s0_402 = __p0_402; \
	50926	__ret_402 = splatq_laneq_s8(__s0_402, __p1_402); \
49977	50927	__ret_402; \
49978	50928	})
49979	50929	#else
49980		#define vdupq_laneq_s16(__p0_403, __p1_403) __extension__ ({ \
49981		int16x8_t __ret_403; \
49982		int16x8_t __s0_403 = __p0_403; \
49983		int16x8_t __rev0_403; __rev0_403 = __builtin_shufflevector(__s0_403, __s0_403, 7, 6, 5, 4, 3, 2, 1, 0); \
49984		__ret_403 = __noswap_splatq_laneq_s16(__rev0_403, __p1_403); \
49985		__ret_403 = __builtin_shufflevector(__ret_403, __ret_403, 7, 6, 5, 4, 3, 2, 1, 0); \
	50930	#define vdupq_laneq_s8(__p0_403, __p1_403) __extension__ ({ \
	50931	int8x16_t __ret_403; \
	50932	int8x16_t __s0_403 = __p0_403; \
	50933	int8x16_t __rev0_403; __rev0_403 = __builtin_shufflevector(__s0_403, __s0_403, __lane_reverse_128_8); \
	50934	__ret_403 = __noswap_splatq_laneq_s8(__rev0_403, __p1_403); \
	50935	__ret_403 = __builtin_shufflevector(__ret_403, __ret_403, __lane_reverse_128_8); \
49986	50936	__ret_403; \
49987	50937	})
49988	50938	#endif
49989	50939
49990	50940	#ifdef __LITTLE_ENDIAN__
49991		#define vdup_laneq_u8(__p0_404, __p1_404) __extension__ ({ \
49992		uint8x8_t __ret_404; \
49993		uint8x16_t __s0_404 = __p0_404; \
49994		__ret_404 = splat_laneq_u8(__s0_404, __p1_404); \
	50941	#define vdupq_laneq_f64(__p0_404, __p1_404) __extension__ ({ \
	50942	float64x2_t __ret_404; \
	50943	float64x2_t __s0_404 = __p0_404; \
	50944	__ret_404 = splatq_laneq_f64(__s0_404, __p1_404); \
49995	50945	__ret_404; \
49996	50946	})
49997	50947	#else
49998		#define vdup_laneq_u8(__p0_405, __p1_405) __extension__ ({ \
49999		uint8x8_t __ret_405; \
50000		uint8x16_t __s0_405 = __p0_405; \
50001		uint8x16_t __rev0_405; __rev0_405 = __builtin_shufflevector(__s0_405, __s0_405, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
50002		__ret_405 = __noswap_splat_laneq_u8(__rev0_405, __p1_405); \
50003		__ret_405 = __builtin_shufflevector(__ret_405, __ret_405, 7, 6, 5, 4, 3, 2, 1, 0); \
	50948	#define vdupq_laneq_f64(__p0_405, __p1_405) __extension__ ({ \
	50949	float64x2_t __ret_405; \
	50950	float64x2_t __s0_405 = __p0_405; \
	50951	float64x2_t __rev0_405; __rev0_405 = __builtin_shufflevector(__s0_405, __s0_405, __lane_reverse_128_64); \
	50952	__ret_405 = __noswap_splatq_laneq_f64(__rev0_405, __p1_405); \
	50953	__ret_405 = __builtin_shufflevector(__ret_405, __ret_405, __lane_reverse_128_64); \
50004	50954	__ret_405; \
50005	50955	})
50006	50956	#endif
50007	50957
50008	50958	#ifdef __LITTLE_ENDIAN__
50009		#define vdup_laneq_u32(__p0_406, __p1_406) __extension__ ({ \
50010		uint32x2_t __ret_406; \
50011		uint32x4_t __s0_406 = __p0_406; \
50012		__ret_406 = splat_laneq_u32(__s0_406, __p1_406); \
	50959	#define vdupq_laneq_f32(__p0_406, __p1_406) __extension__ ({ \
	50960	float32x4_t __ret_406; \
	50961	float32x4_t __s0_406 = __p0_406; \
	50962	__ret_406 = splatq_laneq_f32(__s0_406, __p1_406); \
50013	50963	__ret_406; \
50014	50964	})
50015	50965	#else
50016		#define vdup_laneq_u32(__p0_407, __p1_407) __extension__ ({ \
50017		uint32x2_t __ret_407; \
50018		uint32x4_t __s0_407 = __p0_407; \
50019		uint32x4_t __rev0_407; __rev0_407 = __builtin_shufflevector(__s0_407, __s0_407, 3, 2, 1, 0); \
50020		__ret_407 = __noswap_splat_laneq_u32(__rev0_407, __p1_407); \
50021		__ret_407 = __builtin_shufflevector(__ret_407, __ret_407, 1, 0); \
	50966	#define vdupq_laneq_f32(__p0_407, __p1_407) __extension__ ({ \
	50967	float32x4_t __ret_407; \
	50968	float32x4_t __s0_407 = __p0_407; \
	50969	float32x4_t __rev0_407; __rev0_407 = __builtin_shufflevector(__s0_407, __s0_407, __lane_reverse_128_32); \
	50970	__ret_407 = __noswap_splatq_laneq_f32(__rev0_407, __p1_407); \
	50971	__ret_407 = __builtin_shufflevector(__ret_407, __ret_407, __lane_reverse_128_32); \
50022	50972	__ret_407; \
50023	50973	})
50024	50974	#endif
50025	50975
50026	50976	#ifdef __LITTLE_ENDIAN__
50027		#define vdup_laneq_u64(__p0_408, __p1_408) __extension__ ({ \
50028		uint64x1_t __ret_408; \
50029		uint64x2_t __s0_408 = __p0_408; \
50030		__ret_408 = splat_laneq_u64(__s0_408, __p1_408); \
	50977	#define vdupq_laneq_f16(__p0_408, __p1_408) __extension__ ({ \
	50978	float16x8_t __ret_408; \
	50979	float16x8_t __s0_408 = __p0_408; \
	50980	__ret_408 = splatq_laneq_f16(__s0_408, __p1_408); \
50031	50981	__ret_408; \
50032	50982	})
50033	50983	#else
50034		#define vdup_laneq_u64(__p0_409, __p1_409) __extension__ ({ \
50035		uint64x1_t __ret_409; \
50036		uint64x2_t __s0_409 = __p0_409; \
50037		uint64x2_t __rev0_409; __rev0_409 = __builtin_shufflevector(__s0_409, __s0_409, 1, 0); \
50038		__ret_409 = __noswap_splat_laneq_u64(__rev0_409, __p1_409); \
	50984	#define vdupq_laneq_f16(__p0_409, __p1_409) __extension__ ({ \
	50985	float16x8_t __ret_409; \
	50986	float16x8_t __s0_409 = __p0_409; \
	50987	float16x8_t __rev0_409; __rev0_409 = __builtin_shufflevector(__s0_409, __s0_409, __lane_reverse_128_16); \
	50988	__ret_409 = __noswap_splatq_laneq_f16(__rev0_409, __p1_409); \
	50989	__ret_409 = __builtin_shufflevector(__ret_409, __ret_409, __lane_reverse_128_16); \
50039	50990	__ret_409; \
50040	50991	})
50041	50992	#endif
50042	50993
50043	50994	#ifdef __LITTLE_ENDIAN__
50044		#define vdup_laneq_u16(__p0_410, __p1_410) __extension__ ({ \
50045		uint16x4_t __ret_410; \
50046		uint16x8_t __s0_410 = __p0_410; \
50047		__ret_410 = splat_laneq_u16(__s0_410, __p1_410); \
	50995	#define vdupq_laneq_s32(__p0_410, __p1_410) __extension__ ({ \
	50996	int32x4_t __ret_410; \
	50997	int32x4_t __s0_410 = __p0_410; \
	50998	__ret_410 = splatq_laneq_s32(__s0_410, __p1_410); \
50048	50999	__ret_410; \
50049	51000	})
50050	51001	#else
50051		#define vdup_laneq_u16(__p0_411, __p1_411) __extension__ ({ \
50052		uint16x4_t __ret_411; \
50053		uint16x8_t __s0_411 = __p0_411; \
50054		uint16x8_t __rev0_411; __rev0_411 = __builtin_shufflevector(__s0_411, __s0_411, 7, 6, 5, 4, 3, 2, 1, 0); \
50055		__ret_411 = __noswap_splat_laneq_u16(__rev0_411, __p1_411); \
50056		__ret_411 = __builtin_shufflevector(__ret_411, __ret_411, 3, 2, 1, 0); \
	51002	#define vdupq_laneq_s32(__p0_411, __p1_411) __extension__ ({ \
	51003	int32x4_t __ret_411; \
	51004	int32x4_t __s0_411 = __p0_411; \
	51005	int32x4_t __rev0_411; __rev0_411 = __builtin_shufflevector(__s0_411, __s0_411, __lane_reverse_128_32); \
	51006	__ret_411 = __noswap_splatq_laneq_s32(__rev0_411, __p1_411); \
	51007	__ret_411 = __builtin_shufflevector(__ret_411, __ret_411, __lane_reverse_128_32); \
50057	51008	__ret_411; \
50058	51009	})
50059	51010	#endif
50060	51011
50061	51012	#ifdef __LITTLE_ENDIAN__
50062		#define vdup_laneq_s8(__p0_412, __p1_412) __extension__ ({ \
50063		int8x8_t __ret_412; \
50064		int8x16_t __s0_412 = __p0_412; \
50065		__ret_412 = splat_laneq_s8(__s0_412, __p1_412); \
	51013	#define vdupq_laneq_s64(__p0_412, __p1_412) __extension__ ({ \
	51014	int64x2_t __ret_412; \
	51015	int64x2_t __s0_412 = __p0_412; \
	51016	__ret_412 = splatq_laneq_s64(__s0_412, __p1_412); \
50066	51017	__ret_412; \
50067	51018	})
50068	51019	#else
50069		#define vdup_laneq_s8(__p0_413, __p1_413) __extension__ ({ \
50070		int8x8_t __ret_413; \
50071		int8x16_t __s0_413 = __p0_413; \
50072		int8x16_t __rev0_413; __rev0_413 = __builtin_shufflevector(__s0_413, __s0_413, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
50073		__ret_413 = __noswap_splat_laneq_s8(__rev0_413, __p1_413); \
50074		__ret_413 = __builtin_shufflevector(__ret_413, __ret_413, 7, 6, 5, 4, 3, 2, 1, 0); \
	51020	#define vdupq_laneq_s64(__p0_413, __p1_413) __extension__ ({ \
	51021	int64x2_t __ret_413; \
	51022	int64x2_t __s0_413 = __p0_413; \
	51023	int64x2_t __rev0_413; __rev0_413 = __builtin_shufflevector(__s0_413, __s0_413, __lane_reverse_128_64); \
	51024	__ret_413 = __noswap_splatq_laneq_s64(__rev0_413, __p1_413); \
	51025	__ret_413 = __builtin_shufflevector(__ret_413, __ret_413, __lane_reverse_128_64); \
50075	51026	__ret_413; \
50076	51027	})
50077	51028	#endif
50078	51029
50079	51030	#ifdef __LITTLE_ENDIAN__
50080		#define vdup_laneq_f64(__p0_414, __p1_414) __extension__ ({ \
50081		float64x1_t __ret_414; \
50082		float64x2_t __s0_414 = __p0_414; \
50083		__ret_414 = splat_laneq_f64(__s0_414, __p1_414); \
	51031	#define vdupq_laneq_mf8(__p0_414, __p1_414) __extension__ ({ \
	51032	mfloat8x16_t __ret_414; \
	51033	mfloat8x16_t __s0_414 = __p0_414; \
	51034	__ret_414 = splatq_laneq_mf8(__s0_414, __p1_414); \
50084	51035	__ret_414; \
50085	51036	})
50086	51037	#else
50087		#define vdup_laneq_f64(__p0_415, __p1_415) __extension__ ({ \
50088		float64x1_t __ret_415; \
50089		float64x2_t __s0_415 = __p0_415; \
50090		float64x2_t __rev0_415; __rev0_415 = __builtin_shufflevector(__s0_415, __s0_415, 1, 0); \
50091		__ret_415 = __noswap_splat_laneq_f64(__rev0_415, __p1_415); \
	51038	#define vdupq_laneq_mf8(__p0_415, __p1_415) __extension__ ({ \
	51039	mfloat8x16_t __ret_415; \
	51040	mfloat8x16_t __s0_415 = __p0_415; \
	51041	mfloat8x16_t __rev0_415; __rev0_415 = __builtin_shufflevector(__s0_415, __s0_415, __lane_reverse_128_8); \
	51042	__ret_415 = __noswap_splatq_laneq_mf8(__rev0_415, __p1_415); \
	51043	__ret_415 = __builtin_shufflevector(__ret_415, __ret_415, __lane_reverse_128_8); \
50092	51044	__ret_415; \
50093	51045	})
50094	51046	#endif
50095	51047
50096	51048	#ifdef __LITTLE_ENDIAN__
50097		#define vdup_laneq_f32(__p0_416, __p1_416) __extension__ ({ \
50098		float32x2_t __ret_416; \
50099		float32x4_t __s0_416 = __p0_416; \
50100		__ret_416 = splat_laneq_f32(__s0_416, __p1_416); \
	51049	#define vdupq_laneq_s16(__p0_416, __p1_416) __extension__ ({ \
	51050	int16x8_t __ret_416; \
	51051	int16x8_t __s0_416 = __p0_416; \
	51052	__ret_416 = splatq_laneq_s16(__s0_416, __p1_416); \
50101	51053	__ret_416; \
50102	51054	})
50103	51055	#else
50104		#define vdup_laneq_f32(__p0_417, __p1_417) __extension__ ({ \
50105		float32x2_t __ret_417; \
50106		float32x4_t __s0_417 = __p0_417; \
50107		float32x4_t __rev0_417; __rev0_417 = __builtin_shufflevector(__s0_417, __s0_417, 3, 2, 1, 0); \
50108		__ret_417 = __noswap_splat_laneq_f32(__rev0_417, __p1_417); \
50109		__ret_417 = __builtin_shufflevector(__ret_417, __ret_417, 1, 0); \
	51056	#define vdupq_laneq_s16(__p0_417, __p1_417) __extension__ ({ \
	51057	int16x8_t __ret_417; \
	51058	int16x8_t __s0_417 = __p0_417; \
	51059	int16x8_t __rev0_417; __rev0_417 = __builtin_shufflevector(__s0_417, __s0_417, __lane_reverse_128_16); \
	51060	__ret_417 = __noswap_splatq_laneq_s16(__rev0_417, __p1_417); \
	51061	__ret_417 = __builtin_shufflevector(__ret_417, __ret_417, __lane_reverse_128_16); \
50110	51062	__ret_417; \
50111	51063	})
50112	51064	#endif
50113	51065
50114	51066	#ifdef __LITTLE_ENDIAN__
50115		#define vdup_laneq_f16(__p0_418, __p1_418) __extension__ ({ \
50116		float16x4_t __ret_418; \
50117		float16x8_t __s0_418 = __p0_418; \
50118		__ret_418 = splat_laneq_f16(__s0_418, __p1_418); \
	51067	#define vdup_laneq_u8(__p0_418, __p1_418) __extension__ ({ \
	51068	uint8x8_t __ret_418; \
	51069	uint8x16_t __s0_418 = __p0_418; \
	51070	__ret_418 = splat_laneq_u8(__s0_418, __p1_418); \
50119	51071	__ret_418; \
50120	51072	})
50121	51073	#else
50122		#define vdup_laneq_f16(__p0_419, __p1_419) __extension__ ({ \
50123		float16x4_t __ret_419; \
50124		float16x8_t __s0_419 = __p0_419; \
50125		float16x8_t __rev0_419; __rev0_419 = __builtin_shufflevector(__s0_419, __s0_419, 7, 6, 5, 4, 3, 2, 1, 0); \
50126		__ret_419 = __noswap_splat_laneq_f16(__rev0_419, __p1_419); \
50127		__ret_419 = __builtin_shufflevector(__ret_419, __ret_419, 3, 2, 1, 0); \
	51074	#define vdup_laneq_u8(__p0_419, __p1_419) __extension__ ({ \
	51075	uint8x8_t __ret_419; \
	51076	uint8x16_t __s0_419 = __p0_419; \
	51077	uint8x16_t __rev0_419; __rev0_419 = __builtin_shufflevector(__s0_419, __s0_419, __lane_reverse_128_8); \
	51078	__ret_419 = __noswap_splat_laneq_u8(__rev0_419, __p1_419); \
	51079	__ret_419 = __builtin_shufflevector(__ret_419, __ret_419, __lane_reverse_64_8); \
50128	51080	__ret_419; \
50129	51081	})
50130	51082	#endif
50131	51083
50132	51084	#ifdef __LITTLE_ENDIAN__
50133		#define vdup_laneq_s32(__p0_420, __p1_420) __extension__ ({ \
50134		int32x2_t __ret_420; \
50135		int32x4_t __s0_420 = __p0_420; \
50136		__ret_420 = splat_laneq_s32(__s0_420, __p1_420); \
	51085	#define vdup_laneq_u32(__p0_420, __p1_420) __extension__ ({ \
	51086	uint32x2_t __ret_420; \
	51087	uint32x4_t __s0_420 = __p0_420; \
	51088	__ret_420 = splat_laneq_u32(__s0_420, __p1_420); \
50137	51089	__ret_420; \
50138	51090	})
50139	51091	#else
50140		#define vdup_laneq_s32(__p0_421, __p1_421) __extension__ ({ \
50141		int32x2_t __ret_421; \
50142		int32x4_t __s0_421 = __p0_421; \
50143		int32x4_t __rev0_421; __rev0_421 = __builtin_shufflevector(__s0_421, __s0_421, 3, 2, 1, 0); \
50144		__ret_421 = __noswap_splat_laneq_s32(__rev0_421, __p1_421); \
50145		__ret_421 = __builtin_shufflevector(__ret_421, __ret_421, 1, 0); \
	51092	#define vdup_laneq_u32(__p0_421, __p1_421) __extension__ ({ \
	51093	uint32x2_t __ret_421; \
	51094	uint32x4_t __s0_421 = __p0_421; \
	51095	uint32x4_t __rev0_421; __rev0_421 = __builtin_shufflevector(__s0_421, __s0_421, __lane_reverse_128_32); \
	51096	__ret_421 = __noswap_splat_laneq_u32(__rev0_421, __p1_421); \
	51097	__ret_421 = __builtin_shufflevector(__ret_421, __ret_421, __lane_reverse_64_32); \
50146	51098	__ret_421; \
50147	51099	})
50148	51100	#endif
50149	51101
50150	51102	#ifdef __LITTLE_ENDIAN__
50151		#define vdup_laneq_s64(__p0_422, __p1_422) __extension__ ({ \
50152		int64x1_t __ret_422; \
50153		int64x2_t __s0_422 = __p0_422; \
50154		__ret_422 = splat_laneq_s64(__s0_422, __p1_422); \
	51103	#define vdup_laneq_u64(__p0_422, __p1_422) __extension__ ({ \
	51104	uint64x1_t __ret_422; \
	51105	uint64x2_t __s0_422 = __p0_422; \
	51106	__ret_422 = splat_laneq_u64(__s0_422, __p1_422); \
50155	51107	__ret_422; \
50156	51108	})
50157	51109	#else
50158		#define vdup_laneq_s64(__p0_423, __p1_423) __extension__ ({ \
50159		int64x1_t __ret_423; \
50160		int64x2_t __s0_423 = __p0_423; \
50161		int64x2_t __rev0_423; __rev0_423 = __builtin_shufflevector(__s0_423, __s0_423, 1, 0); \
50162		__ret_423 = __noswap_splat_laneq_s64(__rev0_423, __p1_423); \
	51110	#define vdup_laneq_u64(__p0_423, __p1_423) __extension__ ({ \
	51111	uint64x1_t __ret_423; \
	51112	uint64x2_t __s0_423 = __p0_423; \
	51113	uint64x2_t __rev0_423; __rev0_423 = __builtin_shufflevector(__s0_423, __s0_423, __lane_reverse_128_64); \
	51114	__ret_423 = __noswap_splat_laneq_u64(__rev0_423, __p1_423); \
50163	51115	__ret_423; \
50164	51116	})
50165	51117	#endif
50166	51118
50167	51119	#ifdef __LITTLE_ENDIAN__
50168		#define vdup_laneq_s16(__p0_424, __p1_424) __extension__ ({ \
50169		int16x4_t __ret_424; \
50170		int16x8_t __s0_424 = __p0_424; \
50171		__ret_424 = splat_laneq_s16(__s0_424, __p1_424); \
	51120	#define vdup_laneq_u16(__p0_424, __p1_424) __extension__ ({ \
	51121	uint16x4_t __ret_424; \
	51122	uint16x8_t __s0_424 = __p0_424; \
	51123	__ret_424 = splat_laneq_u16(__s0_424, __p1_424); \
50172	51124	__ret_424; \
50173	51125	})
50174	51126	#else
50175		#define vdup_laneq_s16(__p0_425, __p1_425) __extension__ ({ \
50176		int16x4_t __ret_425; \
50177		int16x8_t __s0_425 = __p0_425; \
50178		int16x8_t __rev0_425; __rev0_425 = __builtin_shufflevector(__s0_425, __s0_425, 7, 6, 5, 4, 3, 2, 1, 0); \
50179		__ret_425 = __noswap_splat_laneq_s16(__rev0_425, __p1_425); \
50180		__ret_425 = __builtin_shufflevector(__ret_425, __ret_425, 3, 2, 1, 0); \
	51127	#define vdup_laneq_u16(__p0_425, __p1_425) __extension__ ({ \
	51128	uint16x4_t __ret_425; \
	51129	uint16x8_t __s0_425 = __p0_425; \
	51130	uint16x8_t __rev0_425; __rev0_425 = __builtin_shufflevector(__s0_425, __s0_425, __lane_reverse_128_16); \
	51131	__ret_425 = __noswap_splat_laneq_u16(__rev0_425, __p1_425); \
	51132	__ret_425 = __builtin_shufflevector(__ret_425, __ret_425, __lane_reverse_64_16); \
50181	51133	__ret_425; \
50182	51134	})
50183	51135	#endif
50184	51136
	51137	#ifdef __LITTLE_ENDIAN__
	51138	#define vdup_laneq_s8(__p0_426, __p1_426) __extension__ ({ \
	51139	int8x8_t __ret_426; \
	51140	int8x16_t __s0_426 = __p0_426; \
	51141	__ret_426 = splat_laneq_s8(__s0_426, __p1_426); \
	51142	__ret_426; \
	51143	})
	51144	#else
	51145	#define vdup_laneq_s8(__p0_427, __p1_427) __extension__ ({ \
	51146	int8x8_t __ret_427; \
	51147	int8x16_t __s0_427 = __p0_427; \
	51148	int8x16_t __rev0_427; __rev0_427 = __builtin_shufflevector(__s0_427, __s0_427, __lane_reverse_128_8); \
	51149	__ret_427 = __noswap_splat_laneq_s8(__rev0_427, __p1_427); \
	51150	__ret_427 = __builtin_shufflevector(__ret_427, __ret_427, __lane_reverse_64_8); \
	51151	__ret_427; \
	51152	})
	51153	#endif
	51154
	51155	#ifdef __LITTLE_ENDIAN__
	51156	#define vdup_laneq_f64(__p0_428, __p1_428) __extension__ ({ \
	51157	float64x1_t __ret_428; \
	51158	float64x2_t __s0_428 = __p0_428; \
	51159	__ret_428 = splat_laneq_f64(__s0_428, __p1_428); \
	51160	__ret_428; \
	51161	})
	51162	#else
	51163	#define vdup_laneq_f64(__p0_429, __p1_429) __extension__ ({ \
	51164	float64x1_t __ret_429; \
	51165	float64x2_t __s0_429 = __p0_429; \
	51166	float64x2_t __rev0_429; __rev0_429 = __builtin_shufflevector(__s0_429, __s0_429, __lane_reverse_128_64); \
	51167	__ret_429 = __noswap_splat_laneq_f64(__rev0_429, __p1_429); \
	51168	__ret_429; \
	51169	})
	51170	#endif
	51171
	51172	#ifdef __LITTLE_ENDIAN__
	51173	#define vdup_laneq_f32(__p0_430, __p1_430) __extension__ ({ \
	51174	float32x2_t __ret_430; \
	51175	float32x4_t __s0_430 = __p0_430; \
	51176	__ret_430 = splat_laneq_f32(__s0_430, __p1_430); \
	51177	__ret_430; \
	51178	})
	51179	#else
	51180	#define vdup_laneq_f32(__p0_431, __p1_431) __extension__ ({ \
	51181	float32x2_t __ret_431; \
	51182	float32x4_t __s0_431 = __p0_431; \
	51183	float32x4_t __rev0_431; __rev0_431 = __builtin_shufflevector(__s0_431, __s0_431, __lane_reverse_128_32); \
	51184	__ret_431 = __noswap_splat_laneq_f32(__rev0_431, __p1_431); \
	51185	__ret_431 = __builtin_shufflevector(__ret_431, __ret_431, __lane_reverse_64_32); \
	51186	__ret_431; \
	51187	})
	51188	#endif
	51189
	51190	#ifdef __LITTLE_ENDIAN__
	51191	#define vdup_laneq_f16(__p0_432, __p1_432) __extension__ ({ \
	51192	float16x4_t __ret_432; \
	51193	float16x8_t __s0_432 = __p0_432; \
	51194	__ret_432 = splat_laneq_f16(__s0_432, __p1_432); \
	51195	__ret_432; \
	51196	})
	51197	#else
	51198	#define vdup_laneq_f16(__p0_433, __p1_433) __extension__ ({ \
	51199	float16x4_t __ret_433; \
	51200	float16x8_t __s0_433 = __p0_433; \
	51201	float16x8_t __rev0_433; __rev0_433 = __builtin_shufflevector(__s0_433, __s0_433, __lane_reverse_128_16); \
	51202	__ret_433 = __noswap_splat_laneq_f16(__rev0_433, __p1_433); \
	51203	__ret_433 = __builtin_shufflevector(__ret_433, __ret_433, __lane_reverse_64_16); \
	51204	__ret_433; \
	51205	})
	51206	#endif
	51207
	51208	#ifdef __LITTLE_ENDIAN__
	51209	#define vdup_laneq_s32(__p0_434, __p1_434) __extension__ ({ \
	51210	int32x2_t __ret_434; \
	51211	int32x4_t __s0_434 = __p0_434; \
	51212	__ret_434 = splat_laneq_s32(__s0_434, __p1_434); \
	51213	__ret_434; \
	51214	})
	51215	#else
	51216	#define vdup_laneq_s32(__p0_435, __p1_435) __extension__ ({ \
	51217	int32x2_t __ret_435; \
	51218	int32x4_t __s0_435 = __p0_435; \
	51219	int32x4_t __rev0_435; __rev0_435 = __builtin_shufflevector(__s0_435, __s0_435, __lane_reverse_128_32); \
	51220	__ret_435 = __noswap_splat_laneq_s32(__rev0_435, __p1_435); \
	51221	__ret_435 = __builtin_shufflevector(__ret_435, __ret_435, __lane_reverse_64_32); \
	51222	__ret_435; \
	51223	})
	51224	#endif
	51225
	51226	#ifdef __LITTLE_ENDIAN__
	51227	#define vdup_laneq_s64(__p0_436, __p1_436) __extension__ ({ \
	51228	int64x1_t __ret_436; \
	51229	int64x2_t __s0_436 = __p0_436; \
	51230	__ret_436 = splat_laneq_s64(__s0_436, __p1_436); \
	51231	__ret_436; \
	51232	})
	51233	#else
	51234	#define vdup_laneq_s64(__p0_437, __p1_437) __extension__ ({ \
	51235	int64x1_t __ret_437; \
	51236	int64x2_t __s0_437 = __p0_437; \
	51237	int64x2_t __rev0_437; __rev0_437 = __builtin_shufflevector(__s0_437, __s0_437, __lane_reverse_128_64); \
	51238	__ret_437 = __noswap_splat_laneq_s64(__rev0_437, __p1_437); \
	51239	__ret_437; \
	51240	})
	51241	#endif
	51242
	51243	#ifdef __LITTLE_ENDIAN__
	51244	#define vdup_laneq_mf8(__p0_438, __p1_438) __extension__ ({ \
	51245	mfloat8x8_t __ret_438; \
	51246	mfloat8x16_t __s0_438 = __p0_438; \
	51247	__ret_438 = splat_laneq_mf8(__s0_438, __p1_438); \
	51248	__ret_438; \
	51249	})
	51250	#else
	51251	#define vdup_laneq_mf8(__p0_439, __p1_439) __extension__ ({ \
	51252	mfloat8x8_t __ret_439; \
	51253	mfloat8x16_t __s0_439 = __p0_439; \
	51254	mfloat8x16_t __rev0_439; __rev0_439 = __builtin_shufflevector(__s0_439, __s0_439, __lane_reverse_128_8); \
	51255	__ret_439 = __noswap_splat_laneq_mf8(__rev0_439, __p1_439); \
	51256	__ret_439 = __builtin_shufflevector(__ret_439, __ret_439, __lane_reverse_64_8); \
	51257	__ret_439; \
	51258	})
	51259	#endif
	51260
	51261	#ifdef __LITTLE_ENDIAN__
	51262	#define vdup_laneq_s16(__p0_440, __p1_440) __extension__ ({ \
	51263	int16x4_t __ret_440; \
	51264	int16x8_t __s0_440 = __p0_440; \
	51265	__ret_440 = splat_laneq_s16(__s0_440, __p1_440); \
	51266	__ret_440; \
	51267	})
	51268	#else
	51269	#define vdup_laneq_s16(__p0_441, __p1_441) __extension__ ({ \
	51270	int16x4_t __ret_441; \
	51271	int16x8_t __s0_441 = __p0_441; \
	51272	int16x8_t __rev0_441; __rev0_441 = __builtin_shufflevector(__s0_441, __s0_441, __lane_reverse_128_16); \
	51273	__ret_441 = __noswap_splat_laneq_s16(__rev0_441, __p1_441); \
	51274	__ret_441 = __builtin_shufflevector(__ret_441, __ret_441, __lane_reverse_64_16); \
	51275	__ret_441; \
	51276	})
	51277	#endif
	51278
50185	51279	__ai __attribute__((target("neon"))) poly64x1_t vdup_n_p64(poly64_t __p0) {
50186	51280	poly64x1_t __ret;
50187	51281	__ret = (poly64x1_t) {__p0};
...	...	@@ -50197,7 +51291,7 @@ __ai __attribute__((target("neon"))) poly64x2_t vdupq_n_p64(poly64_t __p0) {
50197	51291	__ai __attribute__((target("neon"))) poly64x2_t vdupq_n_p64(poly64_t __p0) {
50198	51292	poly64x2_t __ret;
50199	51293	__ret = (poly64x2_t) {__p0, __p0};
50200		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	51294	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
50201	51295	return __ret;
50202	51296	}
50203	51297	#endif
...	...	@@ -50212,7 +51306,7 @@ __ai __attribute__((target("neon"))) float64x2_t vdupq_n_f64(float64_t __p0) {
50212	51306	__ai __attribute__((target("neon"))) float64x2_t vdupq_n_f64(float64_t __p0) {
50213	51307	float64x2_t __ret;
50214	51308	__ret = (float64x2_t) {__p0, __p0};
50215		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	51309	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
50216	51310	return __ret;
50217	51311	}
50218	51312	#endif
...	...	@@ -50226,7 +51320,7 @@ __ai __attribute__((target("neon"))) float64x1_t vdup_n_f64(float64_t __p0) {
50226	51320	poly64x1_t __ret; \
50227	51321	poly64x1_t __s0 = __p0; \
50228	51322	poly64x1_t __s1 = __p1; \
50229		__ret = (poly64x1_t) __builtin_neon_vext_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 6); \
	51323	__ret = __builtin_bit_cast(poly64x1_t, __builtin_neon_vext_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 6)); \
50230	51324	__ret; \
50231	51325	})
50232	51326	#ifdef __LITTLE_ENDIAN__
...	...	@@ -50234,7 +51328,7 @@ __ai __attribute__((target("neon"))) float64x1_t vdup_n_f64(float64_t __p0) {
50234	51328	poly64x2_t __ret; \
50235	51329	poly64x2_t __s0 = __p0; \
50236	51330	poly64x2_t __s1 = __p1; \
50237		__ret = (poly64x2_t) __builtin_neon_vextq_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 38); \
	51331	__ret = __builtin_bit_cast(poly64x2_t, __builtin_neon_vextq_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 38)); \
50238	51332	__ret; \
50239	51333	})
50240	51334	#else
...	...	@@ -50242,10 +51336,10 @@ __ai __attribute__((target("neon"))) float64x1_t vdup_n_f64(float64_t __p0) {
50242	51336	poly64x2_t __ret; \
50243	51337	poly64x2_t __s0 = __p0; \
50244	51338	poly64x2_t __s1 = __p1; \
50245		poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
50246		poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
50247		__ret = (poly64x2_t) __builtin_neon_vextq_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 38); \
50248		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	51339	poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	51340	poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	51341	__ret = __builtin_bit_cast(poly64x2_t, __builtin_neon_vextq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 38)); \
	51342	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
50249	51343	__ret; \
50250	51344	})
50251	51345	#endif
...	...	@@ -50255,7 +51349,7 @@ __ai __attribute__((target("neon"))) float64x1_t vdup_n_f64(float64_t __p0) {
50255	51349	float64x2_t __ret; \
50256	51350	float64x2_t __s0 = __p0; \
50257	51351	float64x2_t __s1 = __p1; \
50258		__ret = (float64x2_t) __builtin_neon_vextq_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 42); \
	51352	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vextq_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 42)); \
50259	51353	__ret; \
50260	51354	})
50261	51355	#else
...	...	@@ -50263,10 +51357,10 @@ __ai __attribute__((target("neon"))) float64x1_t vdup_n_f64(float64_t __p0) {
50263	51357	float64x2_t __ret; \
50264	51358	float64x2_t __s0 = __p0; \
50265	51359	float64x2_t __s1 = __p1; \
50266		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
50267		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
50268		__ret = (float64x2_t) __builtin_neon_vextq_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 42); \
50269		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	51360	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	51361	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	51362	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vextq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 42)); \
	51363	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
50270	51364	__ret; \
50271	51365	})
50272	51366	#endif
...	...	@@ -50275,35 +51369,35 @@ __ai __attribute__((target("neon"))) float64x1_t vdup_n_f64(float64_t __p0) {
50275	51369	float64x1_t __ret; \
50276	51370	float64x1_t __s0 = __p0; \
50277	51371	float64x1_t __s1 = __p1; \
50278		__ret = (float64x1_t) __builtin_neon_vext_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 10); \
	51372	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vext_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 10)); \
50279	51373	__ret; \
50280	51374	})
50281	51375	#ifdef __LITTLE_ENDIAN__
50282	51376	__ai __attribute__((target("neon"))) float64x2_t vfmaq_f64(float64x2_t __p0, float64x2_t __p1, float64x2_t __p2) {
50283	51377	float64x2_t __ret;
50284		__ret = (float64x2_t) __builtin_neon_vfmaq_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 42);
	51378	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vfmaq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 42));
50285	51379	return __ret;
50286	51380	}
50287	51381	#else
50288	51382	__ai __attribute__((target("neon"))) float64x2_t vfmaq_f64(float64x2_t __p0, float64x2_t __p1, float64x2_t __p2) {
50289	51383	float64x2_t __ret;
50290		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
50291		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
50292		float64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
50293		__ret = (float64x2_t) __builtin_neon_vfmaq_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 42);
50294		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	51384	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	51385	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	51386	float64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_64);
	51387	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vfmaq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 42));
	51388	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
50295	51389	return __ret;
50296	51390	}
50297	51391	__ai __attribute__((target("neon"))) float64x2_t __noswap_vfmaq_f64(float64x2_t __p0, float64x2_t __p1, float64x2_t __p2) {
50298	51392	float64x2_t __ret;
50299		__ret = (float64x2_t) __builtin_neon_vfmaq_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 42);
	51393	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vfmaq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 42));
50300	51394	return __ret;
50301	51395	}
50302	51396	#endif
50303	51397
50304	51398	__ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, float64x1_t __p1, float64x1_t __p2) {
50305	51399	float64x1_t __ret;
50306		__ret = (float64x1_t) __builtin_neon_vfma_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 10);
	51400	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vfma_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 10));
50307	51401	return __ret;
50308	51402	}
50309	51403	#define vfmad_lane_f64(__p0, __p1, __p2, __p3) __extension__ ({ \
...	...	@@ -50311,7 +51405,7 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50311	51405	float64_t __s0 = __p0; \
50312	51406	float64_t __s1 = __p1; \
50313	51407	float64x1_t __s2 = __p2; \
50314		__ret = (float64_t) __builtin_neon_vfmad_lane_f64(__s0, __s1, (float64x1_t)__s2, __p3); \
	51408	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vfmad_lane_f64(__s0, __s1, __s2, __p3)); \
50315	51409	__ret; \
50316	51410	})
50317	51411	#ifdef __LITTLE_ENDIAN__
...	...	@@ -50320,7 +51414,7 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50320	51414	float32_t __s0 = __p0; \
50321	51415	float32_t __s1 = __p1; \
50322	51416	float32x2_t __s2 = __p2; \
50323		__ret = (float32_t) __builtin_neon_vfmas_lane_f32(__s0, __s1, (float32x2_t)__s2, __p3); \
	51417	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vfmas_lane_f32(__s0, __s1, __s2, __p3)); \
50324	51418	__ret; \
50325	51419	})
50326	51420	#else
...	...	@@ -50329,8 +51423,8 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50329	51423	float32_t __s0 = __p0; \
50330	51424	float32_t __s1 = __p1; \
50331	51425	float32x2_t __s2 = __p2; \
50332		float32x2_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 1, 0); \
50333		__ret = (float32_t) __builtin_neon_vfmas_lane_f32(__s0, __s1, (float32x2_t)__rev2, __p3); \
	51426	float32x2_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_64_32); \
	51427	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vfmas_lane_f32(__s0, __s1, __rev2, __p3)); \
50334	51428	__ret; \
50335	51429	})
50336	51430	#define __noswap_vfmas_lane_f32(__p0, __p1, __p2, __p3) __extension__ ({ \
...	...	@@ -50338,7 +51432,7 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50338	51432	float32_t __s0 = __p0; \
50339	51433	float32_t __s1 = __p1; \
50340	51434	float32x2_t __s2 = __p2; \
50341		__ret = (float32_t) __builtin_neon_vfmas_lane_f32(__s0, __s1, (float32x2_t)__s2, __p3); \
	51435	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vfmas_lane_f32(__s0, __s1, __s2, __p3)); \
50342	51436	__ret; \
50343	51437	})
50344	51438	#endif
...	...	@@ -50349,7 +51443,7 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50349	51443	float64x2_t __s0 = __p0; \
50350	51444	float64x2_t __s1 = __p1; \
50351	51445	float64x1_t __s2 = __p2; \
50352		__ret = (float64x2_t) __builtin_neon_vfmaq_lane_v((int8x16_t)__s0, (int8x16_t)__s1, (int8x8_t)__s2, __p3, 42); \
	51446	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vfmaq_lane_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __builtin_bit_cast(int8x8_t, __s2), __p3, 42)); \
50353	51447	__ret; \
50354	51448	})
50355	51449	#else
...	...	@@ -50358,10 +51452,10 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50358	51452	float64x2_t __s0 = __p0; \
50359	51453	float64x2_t __s1 = __p1; \
50360	51454	float64x1_t __s2 = __p2; \
50361		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
50362		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
50363		__ret = (float64x2_t) __builtin_neon_vfmaq_lane_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x8_t)__s2, __p3, 42); \
50364		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	51455	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	51456	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	51457	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vfmaq_lane_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x8_t, __s2), __p3, 42)); \
	51458	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
50365	51459	__ret; \
50366	51460	})
50367	51461	#define __noswap_vfmaq_lane_f64(__p0, __p1, __p2, __p3) __extension__ ({ \
...	...	@@ -50369,7 +51463,7 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50369	51463	float64x2_t __s0 = __p0; \
50370	51464	float64x2_t __s1 = __p1; \
50371	51465	float64x1_t __s2 = __p2; \
50372		__ret = (float64x2_t) __builtin_neon_vfmaq_lane_v((int8x16_t)__s0, (int8x16_t)__s1, (int8x8_t)__s2, __p3, 42); \
	51466	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vfmaq_lane_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __builtin_bit_cast(int8x8_t, __s2), __p3, 42)); \
50373	51467	__ret; \
50374	51468	})
50375	51469	#endif
...	...	@@ -50380,7 +51474,7 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50380	51474	float32x4_t __s0 = __p0; \
50381	51475	float32x4_t __s1 = __p1; \
50382	51476	float32x2_t __s2 = __p2; \
50383		__ret = (float32x4_t) __builtin_neon_vfmaq_lane_v((int8x16_t)__s0, (int8x16_t)__s1, (int8x8_t)__s2, __p3, 41); \
	51477	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vfmaq_lane_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __builtin_bit_cast(int8x8_t, __s2), __p3, 41)); \
50384	51478	__ret; \
50385	51479	})
50386	51480	#else
...	...	@@ -50389,11 +51483,11 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50389	51483	float32x4_t __s0 = __p0; \
50390	51484	float32x4_t __s1 = __p1; \
50391	51485	float32x2_t __s2 = __p2; \
50392		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
50393		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
50394		float32x2_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 1, 0); \
50395		__ret = (float32x4_t) __builtin_neon_vfmaq_lane_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x8_t)__rev2, __p3, 41); \
50396		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	51486	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	51487	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	51488	float32x2_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_64_32); \
	51489	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vfmaq_lane_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), __p3, 41)); \
	51490	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
50397	51491	__ret; \
50398	51492	})
50399	51493	#define __noswap_vfmaq_lane_f32(__p0, __p1, __p2, __p3) __extension__ ({ \
...	...	@@ -50401,7 +51495,7 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50401	51495	float32x4_t __s0 = __p0; \
50402	51496	float32x4_t __s1 = __p1; \
50403	51497	float32x2_t __s2 = __p2; \
50404		__ret = (float32x4_t) __builtin_neon_vfmaq_lane_v((int8x16_t)__s0, (int8x16_t)__s1, (int8x8_t)__s2, __p3, 41); \
	51498	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vfmaq_lane_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __builtin_bit_cast(int8x8_t, __s2), __p3, 41)); \
50405	51499	__ret; \
50406	51500	})
50407	51501	#endif
...	...	@@ -50411,7 +51505,7 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50411	51505	float64x1_t __s0 = __p0; \
50412	51506	float64x1_t __s1 = __p1; \
50413	51507	float64x1_t __s2 = __p2; \
50414		__ret = (float64x1_t) __builtin_neon_vfma_lane_v((int8x8_t)__s0, (int8x8_t)__s1, (int8x8_t)__s2, __p3, 10); \
	51508	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vfma_lane_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __builtin_bit_cast(int8x8_t, __s2), __p3, 10)); \
50415	51509	__ret; \
50416	51510	})
50417	51511	#ifdef __LITTLE_ENDIAN__
...	...	@@ -50420,7 +51514,7 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50420	51514	float32x2_t __s0 = __p0; \
50421	51515	float32x2_t __s1 = __p1; \
50422	51516	float32x2_t __s2 = __p2; \
50423		__ret = (float32x2_t) __builtin_neon_vfma_lane_v((int8x8_t)__s0, (int8x8_t)__s1, (int8x8_t)__s2, __p3, 9); \
	51517	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vfma_lane_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __builtin_bit_cast(int8x8_t, __s2), __p3, 9)); \
50424	51518	__ret; \
50425	51519	})
50426	51520	#else
...	...	@@ -50429,11 +51523,11 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50429	51523	float32x2_t __s0 = __p0; \
50430	51524	float32x2_t __s1 = __p1; \
50431	51525	float32x2_t __s2 = __p2; \
50432		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
50433		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
50434		float32x2_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 1, 0); \
50435		__ret = (float32x2_t) __builtin_neon_vfma_lane_v((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, __p3, 9); \
50436		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	51526	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	51527	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_32); \
	51528	float32x2_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_64_32); \
	51529	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vfma_lane_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), __p3, 9)); \
	51530	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
50437	51531	__ret; \
50438	51532	})
50439	51533	#define __noswap_vfma_lane_f32(__p0, __p1, __p2, __p3) __extension__ ({ \
...	...	@@ -50441,7 +51535,7 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50441	51535	float32x2_t __s0 = __p0; \
50442	51536	float32x2_t __s1 = __p1; \
50443	51537	float32x2_t __s2 = __p2; \
50444		__ret = (float32x2_t) __builtin_neon_vfma_lane_v((int8x8_t)__s0, (int8x8_t)__s1, (int8x8_t)__s2, __p3, 9); \
	51538	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vfma_lane_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __builtin_bit_cast(int8x8_t, __s2), __p3, 9)); \
50445	51539	__ret; \
50446	51540	})
50447	51541	#endif
...	...	@@ -50452,7 +51546,7 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50452	51546	float64_t __s0 = __p0; \
50453	51547	float64_t __s1 = __p1; \
50454	51548	float64x2_t __s2 = __p2; \
50455		__ret = (float64_t) __builtin_neon_vfmad_laneq_f64(__s0, __s1, (float64x2_t)__s2, __p3); \
	51549	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vfmad_laneq_f64(__s0, __s1, __s2, __p3)); \
50456	51550	__ret; \
50457	51551	})
50458	51552	#else
...	...	@@ -50461,8 +51555,8 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50461	51555	float64_t __s0 = __p0; \
50462	51556	float64_t __s1 = __p1; \
50463	51557	float64x2_t __s2 = __p2; \
50464		float64x2_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 1, 0); \
50465		__ret = (float64_t) __builtin_neon_vfmad_laneq_f64(__s0, __s1, (float64x2_t)__rev2, __p3); \
	51558	float64x2_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_128_64); \
	51559	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vfmad_laneq_f64(__s0, __s1, __rev2, __p3)); \
50466	51560	__ret; \
50467	51561	})
50468	51562	#define __noswap_vfmad_laneq_f64(__p0, __p1, __p2, __p3) __extension__ ({ \
...	...	@@ -50470,7 +51564,7 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50470	51564	float64_t __s0 = __p0; \
50471	51565	float64_t __s1 = __p1; \
50472	51566	float64x2_t __s2 = __p2; \
50473		__ret = (float64_t) __builtin_neon_vfmad_laneq_f64(__s0, __s1, (float64x2_t)__s2, __p3); \
	51567	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vfmad_laneq_f64(__s0, __s1, __s2, __p3)); \
50474	51568	__ret; \
50475	51569	})
50476	51570	#endif
...	...	@@ -50481,7 +51575,7 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50481	51575	float32_t __s0 = __p0; \
50482	51576	float32_t __s1 = __p1; \
50483	51577	float32x4_t __s2 = __p2; \
50484		__ret = (float32_t) __builtin_neon_vfmas_laneq_f32(__s0, __s1, (float32x4_t)__s2, __p3); \
	51578	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vfmas_laneq_f32(__s0, __s1, __s2, __p3)); \
50485	51579	__ret; \
50486	51580	})
50487	51581	#else
...	...	@@ -50490,8 +51584,8 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50490	51584	float32_t __s0 = __p0; \
50491	51585	float32_t __s1 = __p1; \
50492	51586	float32x4_t __s2 = __p2; \
50493		float32x4_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 3, 2, 1, 0); \
50494		__ret = (float32_t) __builtin_neon_vfmas_laneq_f32(__s0, __s1, (float32x4_t)__rev2, __p3); \
	51587	float32x4_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_128_32); \
	51588	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vfmas_laneq_f32(__s0, __s1, __rev2, __p3)); \
50495	51589	__ret; \
50496	51590	})
50497	51591	#define __noswap_vfmas_laneq_f32(__p0, __p1, __p2, __p3) __extension__ ({ \
...	...	@@ -50499,7 +51593,7 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50499	51593	float32_t __s0 = __p0; \
50500	51594	float32_t __s1 = __p1; \
50501	51595	float32x4_t __s2 = __p2; \
50502		__ret = (float32_t) __builtin_neon_vfmas_laneq_f32(__s0, __s1, (float32x4_t)__s2, __p3); \
	51596	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vfmas_laneq_f32(__s0, __s1, __s2, __p3)); \
50503	51597	__ret; \
50504	51598	})
50505	51599	#endif
...	...	@@ -50510,7 +51604,7 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50510	51604	float64x2_t __s0 = __p0; \
50511	51605	float64x2_t __s1 = __p1; \
50512	51606	float64x2_t __s2 = __p2; \
50513		__ret = (float64x2_t) __builtin_neon_vfmaq_laneq_v((int8x16_t)__s0, (int8x16_t)__s1, (int8x16_t)__s2, __p3, 42); \
	51607	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vfmaq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __builtin_bit_cast(int8x16_t, __s2), __p3, 42)); \
50514	51608	__ret; \
50515	51609	})
50516	51610	#else
...	...	@@ -50519,11 +51613,11 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50519	51613	float64x2_t __s0 = __p0; \
50520	51614	float64x2_t __s1 = __p1; \
50521	51615	float64x2_t __s2 = __p2; \
50522		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
50523		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
50524		float64x2_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 1, 0); \
50525		__ret = (float64x2_t) __builtin_neon_vfmaq_laneq_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, __p3, 42); \
50526		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	51616	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	51617	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	51618	float64x2_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_128_64); \
	51619	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vfmaq_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), __p3, 42)); \
	51620	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
50527	51621	__ret; \
50528	51622	})
50529	51623	#define __noswap_vfmaq_laneq_f64(__p0, __p1, __p2, __p3) __extension__ ({ \
...	...	@@ -50531,7 +51625,7 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50531	51625	float64x2_t __s0 = __p0; \
50532	51626	float64x2_t __s1 = __p1; \
50533	51627	float64x2_t __s2 = __p2; \
50534		__ret = (float64x2_t) __builtin_neon_vfmaq_laneq_v((int8x16_t)__s0, (int8x16_t)__s1, (int8x16_t)__s2, __p3, 42); \
	51628	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vfmaq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __builtin_bit_cast(int8x16_t, __s2), __p3, 42)); \
50535	51629	__ret; \
50536	51630	})
50537	51631	#endif
...	...	@@ -50542,7 +51636,7 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50542	51636	float32x4_t __s0 = __p0; \
50543	51637	float32x4_t __s1 = __p1; \
50544	51638	float32x4_t __s2 = __p2; \
50545		__ret = (float32x4_t) __builtin_neon_vfmaq_laneq_v((int8x16_t)__s0, (int8x16_t)__s1, (int8x16_t)__s2, __p3, 41); \
	51639	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vfmaq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __builtin_bit_cast(int8x16_t, __s2), __p3, 41)); \
50546	51640	__ret; \
50547	51641	})
50548	51642	#else
...	...	@@ -50551,11 +51645,11 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50551	51645	float32x4_t __s0 = __p0; \
50552	51646	float32x4_t __s1 = __p1; \
50553	51647	float32x4_t __s2 = __p2; \
50554		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
50555		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
50556		float32x4_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 3, 2, 1, 0); \
50557		__ret = (float32x4_t) __builtin_neon_vfmaq_laneq_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, __p3, 41); \
50558		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	51648	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	51649	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	51650	float32x4_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_128_32); \
	51651	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vfmaq_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), __p3, 41)); \
	51652	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
50559	51653	__ret; \
50560	51654	})
50561	51655	#define __noswap_vfmaq_laneq_f32(__p0, __p1, __p2, __p3) __extension__ ({ \
...	...	@@ -50563,7 +51657,7 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50563	51657	float32x4_t __s0 = __p0; \
50564	51658	float32x4_t __s1 = __p1; \
50565	51659	float32x4_t __s2 = __p2; \
50566		__ret = (float32x4_t) __builtin_neon_vfmaq_laneq_v((int8x16_t)__s0, (int8x16_t)__s1, (int8x16_t)__s2, __p3, 41); \
	51660	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vfmaq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __builtin_bit_cast(int8x16_t, __s2), __p3, 41)); \
50567	51661	__ret; \
50568	51662	})
50569	51663	#endif
...	...	@@ -50574,7 +51668,7 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50574	51668	float64x1_t __s0 = __p0; \
50575	51669	float64x1_t __s1 = __p1; \
50576	51670	float64x2_t __s2 = __p2; \
50577		__ret = (float64x1_t) __builtin_neon_vfma_laneq_v((int8x8_t)__s0, (int8x8_t)__s1, (int8x16_t)__s2, __p3, 10); \
	51671	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vfma_laneq_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __builtin_bit_cast(int8x16_t, __s2), __p3, 10)); \
50578	51672	__ret; \
50579	51673	})
50580	51674	#else
...	...	@@ -50583,8 +51677,8 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50583	51677	float64x1_t __s0 = __p0; \
50584	51678	float64x1_t __s1 = __p1; \
50585	51679	float64x2_t __s2 = __p2; \
50586		float64x2_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 1, 0); \
50587		__ret = (float64x1_t) __builtin_neon_vfma_laneq_v((int8x8_t)__s0, (int8x8_t)__s1, (int8x16_t)__rev2, __p3, 10); \
	51680	float64x2_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_128_64); \
	51681	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vfma_laneq_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __builtin_bit_cast(int8x16_t, __rev2), __p3, 10)); \
50588	51682	__ret; \
50589	51683	})
50590	51684	#define __noswap_vfma_laneq_f64(__p0, __p1, __p2, __p3) __extension__ ({ \
...	...	@@ -50592,7 +51686,7 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50592	51686	float64x1_t __s0 = __p0; \
50593	51687	float64x1_t __s1 = __p1; \
50594	51688	float64x2_t __s2 = __p2; \
50595		__ret = (float64x1_t) __builtin_neon_vfma_laneq_v((int8x8_t)__s0, (int8x8_t)__s1, (int8x16_t)__s2, __p3, 10); \
	51689	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vfma_laneq_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __builtin_bit_cast(int8x16_t, __s2), __p3, 10)); \
50596	51690	__ret; \
50597	51691	})
50598	51692	#endif
...	...	@@ -50603,7 +51697,7 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50603	51697	float32x2_t __s0 = __p0; \
50604	51698	float32x2_t __s1 = __p1; \
50605	51699	float32x4_t __s2 = __p2; \
50606		__ret = (float32x2_t) __builtin_neon_vfma_laneq_v((int8x8_t)__s0, (int8x8_t)__s1, (int8x16_t)__s2, __p3, 9); \
	51700	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vfma_laneq_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __builtin_bit_cast(int8x16_t, __s2), __p3, 9)); \
50607	51701	__ret; \
50608	51702	})
50609	51703	#else
...	...	@@ -50612,11 +51706,11 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50612	51706	float32x2_t __s0 = __p0; \
50613	51707	float32x2_t __s1 = __p1; \
50614	51708	float32x4_t __s2 = __p2; \
50615		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
50616		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
50617		float32x4_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 3, 2, 1, 0); \
50618		__ret = (float32x2_t) __builtin_neon_vfma_laneq_v((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x16_t)__rev2, __p3, 9); \
50619		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	51709	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	51710	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_32); \
	51711	float32x4_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_128_32); \
	51712	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vfma_laneq_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), __p3, 9)); \
	51713	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
50620	51714	__ret; \
50621	51715	})
50622	51716	#define __noswap_vfma_laneq_f32(__p0, __p1, __p2, __p3) __extension__ ({ \
...	...	@@ -50624,7 +51718,7 @@ __ai __attribute__((target("neon"))) float64x1_t vfma_f64(float64x1_t __p0, floa
50624	51718	float32x2_t __s0 = __p0; \
50625	51719	float32x2_t __s1 = __p1; \
50626	51720	float32x4_t __s2 = __p2; \
50627		__ret = (float32x2_t) __builtin_neon_vfma_laneq_v((int8x8_t)__s0, (int8x8_t)__s1, (int8x16_t)__s2, __p3, 9); \
	51721	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vfma_laneq_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __builtin_bit_cast(int8x16_t, __s2), __p3, 9)); \
50628	51722	__ret; \
50629	51723	})
50630	51724	#endif
...	...	@@ -50638,10 +51732,10 @@ __ai __attribute__((target("neon"))) float64x2_t vfmaq_n_f64(float64x2_t __p0, f
50638	51732	#else
50639	51733	__ai __attribute__((target("neon"))) float64x2_t vfmaq_n_f64(float64x2_t __p0, float64x2_t __p1, float64_t __p2) {
50640	51734	float64x2_t __ret;
50641		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
50642		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	51735	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	51736	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
50643	51737	__ret = __noswap_vfmaq_f64(__rev0, __rev1, (float64x2_t) {__p2, __p2});
50644		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	51738	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
50645	51739	return __ret;
50646	51740	}
50647	51741	#endif
...	...	@@ -50660,11 +51754,11 @@ __ai __attribute__((target("neon"))) float64x2_t vfmsq_f64(float64x2_t __p0, flo
50660	51754	#else
50661	51755	__ai __attribute__((target("neon"))) float64x2_t vfmsq_f64(float64x2_t __p0, float64x2_t __p1, float64x2_t __p2) {
50662	51756	float64x2_t __ret;
50663		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
50664		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
50665		float64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
	51757	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	51758	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	51759	float64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_64);
50666	51760	__ret = __noswap_vfmaq_f64(__rev0, -__rev1, __rev2);
50667		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	51761	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
50668	51762	return __ret;
50669	51763	}
50670	51764	#endif
...	...	@@ -50674,246 +51768,246 @@ __ai __attribute__((target("neon"))) float64x1_t vfms_f64(float64x1_t __p0, floa
50674	51768	__ret = vfma_f64(__p0, -__p1, __p2);
50675	51769	return __ret;
50676	51770	}
50677		#define vfmsd_lane_f64(__p0_426, __p1_426, __p2_426, __p3_426) __extension__ ({ \
50678		float64_t __ret_426; \
50679		float64_t __s0_426 = __p0_426; \
50680		float64_t __s1_426 = __p1_426; \
50681		float64x1_t __s2_426 = __p2_426; \
50682		__ret_426 = vfmad_lane_f64(__s0_426, -__s1_426, __s2_426, __p3_426); \
50683		__ret_426; \
	51771	#define vfmsd_lane_f64(__p0_442, __p1_442, __p2_442, __p3_442) __extension__ ({ \
	51772	float64_t __ret_442; \
	51773	float64_t __s0_442 = __p0_442; \
	51774	float64_t __s1_442 = __p1_442; \
	51775	float64x1_t __s2_442 = __p2_442; \
	51776	__ret_442 = vfmad_lane_f64(__s0_442, -__s1_442, __s2_442, __p3_442); \
	51777	__ret_442; \
50684	51778	})
50685	51779	#ifdef __LITTLE_ENDIAN__
50686		#define vfmss_lane_f32(__p0_427, __p1_427, __p2_427, __p3_427) __extension__ ({ \
50687		float32_t __ret_427; \
50688		float32_t __s0_427 = __p0_427; \
50689		float32_t __s1_427 = __p1_427; \
50690		float32x2_t __s2_427 = __p2_427; \
50691		__ret_427 = vfmas_lane_f32(__s0_427, -__s1_427, __s2_427, __p3_427); \
50692		__ret_427; \
	51780	#define vfmss_lane_f32(__p0_443, __p1_443, __p2_443, __p3_443) __extension__ ({ \
	51781	float32_t __ret_443; \
	51782	float32_t __s0_443 = __p0_443; \
	51783	float32_t __s1_443 = __p1_443; \
	51784	float32x2_t __s2_443 = __p2_443; \
	51785	__ret_443 = vfmas_lane_f32(__s0_443, -__s1_443, __s2_443, __p3_443); \
	51786	__ret_443; \
50693	51787	})
50694	51788	#else
50695		#define vfmss_lane_f32(__p0_428, __p1_428, __p2_428, __p3_428) __extension__ ({ \
50696		float32_t __ret_428; \
50697		float32_t __s0_428 = __p0_428; \
50698		float32_t __s1_428 = __p1_428; \
50699		float32x2_t __s2_428 = __p2_428; \
50700		float32x2_t __rev2_428; __rev2_428 = __builtin_shufflevector(__s2_428, __s2_428, 1, 0); \
50701		__ret_428 = __noswap_vfmas_lane_f32(__s0_428, -__s1_428, __rev2_428, __p3_428); \
50702		__ret_428; \
	51789	#define vfmss_lane_f32(__p0_444, __p1_444, __p2_444, __p3_444) __extension__ ({ \
	51790	float32_t __ret_444; \
	51791	float32_t __s0_444 = __p0_444; \
	51792	float32_t __s1_444 = __p1_444; \
	51793	float32x2_t __s2_444 = __p2_444; \
	51794	float32x2_t __rev2_444; __rev2_444 = __builtin_shufflevector(__s2_444, __s2_444, __lane_reverse_64_32); \
	51795	__ret_444 = __noswap_vfmas_lane_f32(__s0_444, -__s1_444, __rev2_444, __p3_444); \
	51796	__ret_444; \
50703	51797	})
50704	51798	#endif
50705	51799
50706	51800	#ifdef __LITTLE_ENDIAN__
50707		#define vfmsq_lane_f64(__p0_429, __p1_429, __p2_429, __p3_429) __extension__ ({ \
50708		float64x2_t __ret_429; \
50709		float64x2_t __s0_429 = __p0_429; \
50710		float64x2_t __s1_429 = __p1_429; \
50711		float64x1_t __s2_429 = __p2_429; \
50712		__ret_429 = vfmaq_lane_f64(__s0_429, -__s1_429, __s2_429, __p3_429); \
50713		__ret_429; \
	51801	#define vfmsq_lane_f64(__p0_445, __p1_445, __p2_445, __p3_445) __extension__ ({ \
	51802	float64x2_t __ret_445; \
	51803	float64x2_t __s0_445 = __p0_445; \
	51804	float64x2_t __s1_445 = __p1_445; \
	51805	float64x1_t __s2_445 = __p2_445; \
	51806	__ret_445 = vfmaq_lane_f64(__s0_445, -__s1_445, __s2_445, __p3_445); \
	51807	__ret_445; \
50714	51808	})
50715	51809	#else
50716		#define vfmsq_lane_f64(__p0_430, __p1_430, __p2_430, __p3_430) __extension__ ({ \
50717		float64x2_t __ret_430; \
50718		float64x2_t __s0_430 = __p0_430; \
50719		float64x2_t __s1_430 = __p1_430; \
50720		float64x1_t __s2_430 = __p2_430; \
50721		float64x2_t __rev0_430; __rev0_430 = __builtin_shufflevector(__s0_430, __s0_430, 1, 0); \
50722		float64x2_t __rev1_430; __rev1_430 = __builtin_shufflevector(__s1_430, __s1_430, 1, 0); \
50723		__ret_430 = __noswap_vfmaq_lane_f64(__rev0_430, -__rev1_430, __s2_430, __p3_430); \
50724		__ret_430 = __builtin_shufflevector(__ret_430, __ret_430, 1, 0); \
50725		__ret_430; \
	51810	#define vfmsq_lane_f64(__p0_446, __p1_446, __p2_446, __p3_446) __extension__ ({ \
	51811	float64x2_t __ret_446; \
	51812	float64x2_t __s0_446 = __p0_446; \
	51813	float64x2_t __s1_446 = __p1_446; \
	51814	float64x1_t __s2_446 = __p2_446; \
	51815	float64x2_t __rev0_446; __rev0_446 = __builtin_shufflevector(__s0_446, __s0_446, __lane_reverse_128_64); \
	51816	float64x2_t __rev1_446; __rev1_446 = __builtin_shufflevector(__s1_446, __s1_446, __lane_reverse_128_64); \
	51817	__ret_446 = __noswap_vfmaq_lane_f64(__rev0_446, -__rev1_446, __s2_446, __p3_446); \
	51818	__ret_446 = __builtin_shufflevector(__ret_446, __ret_446, __lane_reverse_128_64); \
	51819	__ret_446; \
50726	51820	})
50727	51821	#endif
50728	51822
50729	51823	#ifdef __LITTLE_ENDIAN__
50730		#define vfmsq_lane_f32(__p0_431, __p1_431, __p2_431, __p3_431) __extension__ ({ \
50731		float32x4_t __ret_431; \
50732		float32x4_t __s0_431 = __p0_431; \
50733		float32x4_t __s1_431 = __p1_431; \
50734		float32x2_t __s2_431 = __p2_431; \
50735		__ret_431 = vfmaq_lane_f32(__s0_431, -__s1_431, __s2_431, __p3_431); \
50736		__ret_431; \
	51824	#define vfmsq_lane_f32(__p0_447, __p1_447, __p2_447, __p3_447) __extension__ ({ \
	51825	float32x4_t __ret_447; \
	51826	float32x4_t __s0_447 = __p0_447; \
	51827	float32x4_t __s1_447 = __p1_447; \
	51828	float32x2_t __s2_447 = __p2_447; \
	51829	__ret_447 = vfmaq_lane_f32(__s0_447, -__s1_447, __s2_447, __p3_447); \
	51830	__ret_447; \
50737	51831	})
50738	51832	#else
50739		#define vfmsq_lane_f32(__p0_432, __p1_432, __p2_432, __p3_432) __extension__ ({ \
50740		float32x4_t __ret_432; \
50741		float32x4_t __s0_432 = __p0_432; \
50742		float32x4_t __s1_432 = __p1_432; \
50743		float32x2_t __s2_432 = __p2_432; \
50744		float32x4_t __rev0_432; __rev0_432 = __builtin_shufflevector(__s0_432, __s0_432, 3, 2, 1, 0); \
50745		float32x4_t __rev1_432; __rev1_432 = __builtin_shufflevector(__s1_432, __s1_432, 3, 2, 1, 0); \
50746		float32x2_t __rev2_432; __rev2_432 = __builtin_shufflevector(__s2_432, __s2_432, 1, 0); \
50747		__ret_432 = __noswap_vfmaq_lane_f32(__rev0_432, -__rev1_432, __rev2_432, __p3_432); \
50748		__ret_432 = __builtin_shufflevector(__ret_432, __ret_432, 3, 2, 1, 0); \
50749		__ret_432; \
	51833	#define vfmsq_lane_f32(__p0_448, __p1_448, __p2_448, __p3_448) __extension__ ({ \
	51834	float32x4_t __ret_448; \
	51835	float32x4_t __s0_448 = __p0_448; \
	51836	float32x4_t __s1_448 = __p1_448; \
	51837	float32x2_t __s2_448 = __p2_448; \
	51838	float32x4_t __rev0_448; __rev0_448 = __builtin_shufflevector(__s0_448, __s0_448, __lane_reverse_128_32); \
	51839	float32x4_t __rev1_448; __rev1_448 = __builtin_shufflevector(__s1_448, __s1_448, __lane_reverse_128_32); \
	51840	float32x2_t __rev2_448; __rev2_448 = __builtin_shufflevector(__s2_448, __s2_448, __lane_reverse_64_32); \
	51841	__ret_448 = __noswap_vfmaq_lane_f32(__rev0_448, -__rev1_448, __rev2_448, __p3_448); \
	51842	__ret_448 = __builtin_shufflevector(__ret_448, __ret_448, __lane_reverse_128_32); \
	51843	__ret_448; \
50750	51844	})
50751	51845	#endif
50752	51846
50753		#define vfms_lane_f64(__p0_433, __p1_433, __p2_433, __p3_433) __extension__ ({ \
50754		float64x1_t __ret_433; \
50755		float64x1_t __s0_433 = __p0_433; \
50756		float64x1_t __s1_433 = __p1_433; \
50757		float64x1_t __s2_433 = __p2_433; \
50758		__ret_433 = vfma_lane_f64(__s0_433, -__s1_433, __s2_433, __p3_433); \
50759		__ret_433; \
	51847	#define vfms_lane_f64(__p0_449, __p1_449, __p2_449, __p3_449) __extension__ ({ \
	51848	float64x1_t __ret_449; \
	51849	float64x1_t __s0_449 = __p0_449; \
	51850	float64x1_t __s1_449 = __p1_449; \
	51851	float64x1_t __s2_449 = __p2_449; \
	51852	__ret_449 = vfma_lane_f64(__s0_449, -__s1_449, __s2_449, __p3_449); \
	51853	__ret_449; \
50760	51854	})
50761	51855	#ifdef __LITTLE_ENDIAN__
50762		#define vfms_lane_f32(__p0_434, __p1_434, __p2_434, __p3_434) __extension__ ({ \
50763		float32x2_t __ret_434; \
50764		float32x2_t __s0_434 = __p0_434; \
50765		float32x2_t __s1_434 = __p1_434; \
50766		float32x2_t __s2_434 = __p2_434; \
50767		__ret_434 = vfma_lane_f32(__s0_434, -__s1_434, __s2_434, __p3_434); \
50768		__ret_434; \
	51856	#define vfms_lane_f32(__p0_450, __p1_450, __p2_450, __p3_450) __extension__ ({ \
	51857	float32x2_t __ret_450; \
	51858	float32x2_t __s0_450 = __p0_450; \
	51859	float32x2_t __s1_450 = __p1_450; \
	51860	float32x2_t __s2_450 = __p2_450; \
	51861	__ret_450 = vfma_lane_f32(__s0_450, -__s1_450, __s2_450, __p3_450); \
	51862	__ret_450; \
50769	51863	})
50770	51864	#else
50771		#define vfms_lane_f32(__p0_435, __p1_435, __p2_435, __p3_435) __extension__ ({ \
50772		float32x2_t __ret_435; \
50773		float32x2_t __s0_435 = __p0_435; \
50774		float32x2_t __s1_435 = __p1_435; \
50775		float32x2_t __s2_435 = __p2_435; \
50776		float32x2_t __rev0_435; __rev0_435 = __builtin_shufflevector(__s0_435, __s0_435, 1, 0); \
50777		float32x2_t __rev1_435; __rev1_435 = __builtin_shufflevector(__s1_435, __s1_435, 1, 0); \
50778		float32x2_t __rev2_435; __rev2_435 = __builtin_shufflevector(__s2_435, __s2_435, 1, 0); \
50779		__ret_435 = __noswap_vfma_lane_f32(__rev0_435, -__rev1_435, __rev2_435, __p3_435); \
50780		__ret_435 = __builtin_shufflevector(__ret_435, __ret_435, 1, 0); \
50781		__ret_435; \
	51865	#define vfms_lane_f32(__p0_451, __p1_451, __p2_451, __p3_451) __extension__ ({ \
	51866	float32x2_t __ret_451; \
	51867	float32x2_t __s0_451 = __p0_451; \
	51868	float32x2_t __s1_451 = __p1_451; \
	51869	float32x2_t __s2_451 = __p2_451; \
	51870	float32x2_t __rev0_451; __rev0_451 = __builtin_shufflevector(__s0_451, __s0_451, __lane_reverse_64_32); \
	51871	float32x2_t __rev1_451; __rev1_451 = __builtin_shufflevector(__s1_451, __s1_451, __lane_reverse_64_32); \
	51872	float32x2_t __rev2_451; __rev2_451 = __builtin_shufflevector(__s2_451, __s2_451, __lane_reverse_64_32); \
	51873	__ret_451 = __noswap_vfma_lane_f32(__rev0_451, -__rev1_451, __rev2_451, __p3_451); \
	51874	__ret_451 = __builtin_shufflevector(__ret_451, __ret_451, __lane_reverse_64_32); \
	51875	__ret_451; \
50782	51876	})
50783	51877	#endif
50784	51878
50785	51879	#ifdef __LITTLE_ENDIAN__
50786		#define vfmsd_laneq_f64(__p0_436, __p1_436, __p2_436, __p3_436) __extension__ ({ \
50787		float64_t __ret_436; \
50788		float64_t __s0_436 = __p0_436; \
50789		float64_t __s1_436 = __p1_436; \
50790		float64x2_t __s2_436 = __p2_436; \
50791		__ret_436 = vfmad_laneq_f64(__s0_436, -__s1_436, __s2_436, __p3_436); \
50792		__ret_436; \
	51880	#define vfmsd_laneq_f64(__p0_452, __p1_452, __p2_452, __p3_452) __extension__ ({ \
	51881	float64_t __ret_452; \
	51882	float64_t __s0_452 = __p0_452; \
	51883	float64_t __s1_452 = __p1_452; \
	51884	float64x2_t __s2_452 = __p2_452; \
	51885	__ret_452 = vfmad_laneq_f64(__s0_452, -__s1_452, __s2_452, __p3_452); \
	51886	__ret_452; \
50793	51887	})
50794	51888	#else
50795		#define vfmsd_laneq_f64(__p0_437, __p1_437, __p2_437, __p3_437) __extension__ ({ \
50796		float64_t __ret_437; \
50797		float64_t __s0_437 = __p0_437; \
50798		float64_t __s1_437 = __p1_437; \
50799		float64x2_t __s2_437 = __p2_437; \
50800		float64x2_t __rev2_437; __rev2_437 = __builtin_shufflevector(__s2_437, __s2_437, 1, 0); \
50801		__ret_437 = __noswap_vfmad_laneq_f64(__s0_437, -__s1_437, __rev2_437, __p3_437); \
50802		__ret_437; \
	51889	#define vfmsd_laneq_f64(__p0_453, __p1_453, __p2_453, __p3_453) __extension__ ({ \
	51890	float64_t __ret_453; \
	51891	float64_t __s0_453 = __p0_453; \
	51892	float64_t __s1_453 = __p1_453; \
	51893	float64x2_t __s2_453 = __p2_453; \
	51894	float64x2_t __rev2_453; __rev2_453 = __builtin_shufflevector(__s2_453, __s2_453, __lane_reverse_128_64); \
	51895	__ret_453 = __noswap_vfmad_laneq_f64(__s0_453, -__s1_453, __rev2_453, __p3_453); \
	51896	__ret_453; \
50803	51897	})
50804	51898	#endif
50805	51899
50806	51900	#ifdef __LITTLE_ENDIAN__
50807		#define vfmss_laneq_f32(__p0_438, __p1_438, __p2_438, __p3_438) __extension__ ({ \
50808		float32_t __ret_438; \
50809		float32_t __s0_438 = __p0_438; \
50810		float32_t __s1_438 = __p1_438; \
50811		float32x4_t __s2_438 = __p2_438; \
50812		__ret_438 = vfmas_laneq_f32(__s0_438, -__s1_438, __s2_438, __p3_438); \
50813		__ret_438; \
	51901	#define vfmss_laneq_f32(__p0_454, __p1_454, __p2_454, __p3_454) __extension__ ({ \
	51902	float32_t __ret_454; \
	51903	float32_t __s0_454 = __p0_454; \
	51904	float32_t __s1_454 = __p1_454; \
	51905	float32x4_t __s2_454 = __p2_454; \
	51906	__ret_454 = vfmas_laneq_f32(__s0_454, -__s1_454, __s2_454, __p3_454); \
	51907	__ret_454; \
50814	51908	})
50815	51909	#else
50816		#define vfmss_laneq_f32(__p0_439, __p1_439, __p2_439, __p3_439) __extension__ ({ \
50817		float32_t __ret_439; \
50818		float32_t __s0_439 = __p0_439; \
50819		float32_t __s1_439 = __p1_439; \
50820		float32x4_t __s2_439 = __p2_439; \
50821		float32x4_t __rev2_439; __rev2_439 = __builtin_shufflevector(__s2_439, __s2_439, 3, 2, 1, 0); \
50822		__ret_439 = __noswap_vfmas_laneq_f32(__s0_439, -__s1_439, __rev2_439, __p3_439); \
50823		__ret_439; \
	51910	#define vfmss_laneq_f32(__p0_455, __p1_455, __p2_455, __p3_455) __extension__ ({ \
	51911	float32_t __ret_455; \
	51912	float32_t __s0_455 = __p0_455; \
	51913	float32_t __s1_455 = __p1_455; \
	51914	float32x4_t __s2_455 = __p2_455; \
	51915	float32x4_t __rev2_455; __rev2_455 = __builtin_shufflevector(__s2_455, __s2_455, __lane_reverse_128_32); \
	51916	__ret_455 = __noswap_vfmas_laneq_f32(__s0_455, -__s1_455, __rev2_455, __p3_455); \
	51917	__ret_455; \
50824	51918	})
50825	51919	#endif
50826	51920
50827	51921	#ifdef __LITTLE_ENDIAN__
50828		#define vfmsq_laneq_f64(__p0_440, __p1_440, __p2_440, __p3_440) __extension__ ({ \
50829		float64x2_t __ret_440; \
50830		float64x2_t __s0_440 = __p0_440; \
50831		float64x2_t __s1_440 = __p1_440; \
50832		float64x2_t __s2_440 = __p2_440; \
50833		__ret_440 = vfmaq_laneq_f64(__s0_440, -__s1_440, __s2_440, __p3_440); \
50834		__ret_440; \
	51922	#define vfmsq_laneq_f64(__p0_456, __p1_456, __p2_456, __p3_456) __extension__ ({ \
	51923	float64x2_t __ret_456; \
	51924	float64x2_t __s0_456 = __p0_456; \
	51925	float64x2_t __s1_456 = __p1_456; \
	51926	float64x2_t __s2_456 = __p2_456; \
	51927	__ret_456 = vfmaq_laneq_f64(__s0_456, -__s1_456, __s2_456, __p3_456); \
	51928	__ret_456; \
50835	51929	})
50836	51930	#else
50837		#define vfmsq_laneq_f64(__p0_441, __p1_441, __p2_441, __p3_441) __extension__ ({ \
50838		float64x2_t __ret_441; \
50839		float64x2_t __s0_441 = __p0_441; \
50840		float64x2_t __s1_441 = __p1_441; \
50841		float64x2_t __s2_441 = __p2_441; \
50842		float64x2_t __rev0_441; __rev0_441 = __builtin_shufflevector(__s0_441, __s0_441, 1, 0); \
50843		float64x2_t __rev1_441; __rev1_441 = __builtin_shufflevector(__s1_441, __s1_441, 1, 0); \
50844		float64x2_t __rev2_441; __rev2_441 = __builtin_shufflevector(__s2_441, __s2_441, 1, 0); \
50845		__ret_441 = __noswap_vfmaq_laneq_f64(__rev0_441, -__rev1_441, __rev2_441, __p3_441); \
50846		__ret_441 = __builtin_shufflevector(__ret_441, __ret_441, 1, 0); \
50847		__ret_441; \
	51931	#define vfmsq_laneq_f64(__p0_457, __p1_457, __p2_457, __p3_457) __extension__ ({ \
	51932	float64x2_t __ret_457; \
	51933	float64x2_t __s0_457 = __p0_457; \
	51934	float64x2_t __s1_457 = __p1_457; \
	51935	float64x2_t __s2_457 = __p2_457; \
	51936	float64x2_t __rev0_457; __rev0_457 = __builtin_shufflevector(__s0_457, __s0_457, __lane_reverse_128_64); \
	51937	float64x2_t __rev1_457; __rev1_457 = __builtin_shufflevector(__s1_457, __s1_457, __lane_reverse_128_64); \
	51938	float64x2_t __rev2_457; __rev2_457 = __builtin_shufflevector(__s2_457, __s2_457, __lane_reverse_128_64); \
	51939	__ret_457 = __noswap_vfmaq_laneq_f64(__rev0_457, -__rev1_457, __rev2_457, __p3_457); \
	51940	__ret_457 = __builtin_shufflevector(__ret_457, __ret_457, __lane_reverse_128_64); \
	51941	__ret_457; \
50848	51942	})
50849	51943	#endif
50850	51944
50851	51945	#ifdef __LITTLE_ENDIAN__
50852		#define vfmsq_laneq_f32(__p0_442, __p1_442, __p2_442, __p3_442) __extension__ ({ \
50853		float32x4_t __ret_442; \
50854		float32x4_t __s0_442 = __p0_442; \
50855		float32x4_t __s1_442 = __p1_442; \
50856		float32x4_t __s2_442 = __p2_442; \
50857		__ret_442 = vfmaq_laneq_f32(__s0_442, -__s1_442, __s2_442, __p3_442); \
50858		__ret_442; \
	51946	#define vfmsq_laneq_f32(__p0_458, __p1_458, __p2_458, __p3_458) __extension__ ({ \
	51947	float32x4_t __ret_458; \
	51948	float32x4_t __s0_458 = __p0_458; \
	51949	float32x4_t __s1_458 = __p1_458; \
	51950	float32x4_t __s2_458 = __p2_458; \
	51951	__ret_458 = vfmaq_laneq_f32(__s0_458, -__s1_458, __s2_458, __p3_458); \
	51952	__ret_458; \
50859	51953	})
50860	51954	#else
50861		#define vfmsq_laneq_f32(__p0_443, __p1_443, __p2_443, __p3_443) __extension__ ({ \
50862		float32x4_t __ret_443; \
50863		float32x4_t __s0_443 = __p0_443; \
50864		float32x4_t __s1_443 = __p1_443; \
50865		float32x4_t __s2_443 = __p2_443; \
50866		float32x4_t __rev0_443; __rev0_443 = __builtin_shufflevector(__s0_443, __s0_443, 3, 2, 1, 0); \
50867		float32x4_t __rev1_443; __rev1_443 = __builtin_shufflevector(__s1_443, __s1_443, 3, 2, 1, 0); \
50868		float32x4_t __rev2_443; __rev2_443 = __builtin_shufflevector(__s2_443, __s2_443, 3, 2, 1, 0); \
50869		__ret_443 = __noswap_vfmaq_laneq_f32(__rev0_443, -__rev1_443, __rev2_443, __p3_443); \
50870		__ret_443 = __builtin_shufflevector(__ret_443, __ret_443, 3, 2, 1, 0); \
50871		__ret_443; \
	51955	#define vfmsq_laneq_f32(__p0_459, __p1_459, __p2_459, __p3_459) __extension__ ({ \
	51956	float32x4_t __ret_459; \
	51957	float32x4_t __s0_459 = __p0_459; \
	51958	float32x4_t __s1_459 = __p1_459; \
	51959	float32x4_t __s2_459 = __p2_459; \
	51960	float32x4_t __rev0_459; __rev0_459 = __builtin_shufflevector(__s0_459, __s0_459, __lane_reverse_128_32); \
	51961	float32x4_t __rev1_459; __rev1_459 = __builtin_shufflevector(__s1_459, __s1_459, __lane_reverse_128_32); \
	51962	float32x4_t __rev2_459; __rev2_459 = __builtin_shufflevector(__s2_459, __s2_459, __lane_reverse_128_32); \
	51963	__ret_459 = __noswap_vfmaq_laneq_f32(__rev0_459, -__rev1_459, __rev2_459, __p3_459); \
	51964	__ret_459 = __builtin_shufflevector(__ret_459, __ret_459, __lane_reverse_128_32); \
	51965	__ret_459; \
50872	51966	})
50873	51967	#endif
50874	51968
50875	51969	#ifdef __LITTLE_ENDIAN__
50876		#define vfms_laneq_f64(__p0_444, __p1_444, __p2_444, __p3_444) __extension__ ({ \
50877		float64x1_t __ret_444; \
50878		float64x1_t __s0_444 = __p0_444; \
50879		float64x1_t __s1_444 = __p1_444; \
50880		float64x2_t __s2_444 = __p2_444; \
50881		__ret_444 = vfma_laneq_f64(__s0_444, -__s1_444, __s2_444, __p3_444); \
50882		__ret_444; \
	51970	#define vfms_laneq_f64(__p0_460, __p1_460, __p2_460, __p3_460) __extension__ ({ \
	51971	float64x1_t __ret_460; \
	51972	float64x1_t __s0_460 = __p0_460; \
	51973	float64x1_t __s1_460 = __p1_460; \
	51974	float64x2_t __s2_460 = __p2_460; \
	51975	__ret_460 = vfma_laneq_f64(__s0_460, -__s1_460, __s2_460, __p3_460); \
	51976	__ret_460; \
50883	51977	})
50884	51978	#else
50885		#define vfms_laneq_f64(__p0_445, __p1_445, __p2_445, __p3_445) __extension__ ({ \
50886		float64x1_t __ret_445; \
50887		float64x1_t __s0_445 = __p0_445; \
50888		float64x1_t __s1_445 = __p1_445; \
50889		float64x2_t __s2_445 = __p2_445; \
50890		float64x2_t __rev2_445; __rev2_445 = __builtin_shufflevector(__s2_445, __s2_445, 1, 0); \
50891		__ret_445 = __noswap_vfma_laneq_f64(__s0_445, -__s1_445, __rev2_445, __p3_445); \
50892		__ret_445; \
	51979	#define vfms_laneq_f64(__p0_461, __p1_461, __p2_461, __p3_461) __extension__ ({ \
	51980	float64x1_t __ret_461; \
	51981	float64x1_t __s0_461 = __p0_461; \
	51982	float64x1_t __s1_461 = __p1_461; \
	51983	float64x2_t __s2_461 = __p2_461; \
	51984	float64x2_t __rev2_461; __rev2_461 = __builtin_shufflevector(__s2_461, __s2_461, __lane_reverse_128_64); \
	51985	__ret_461 = __noswap_vfma_laneq_f64(__s0_461, -__s1_461, __rev2_461, __p3_461); \
	51986	__ret_461; \
50893	51987	})
50894	51988	#endif
50895	51989
50896	51990	#ifdef __LITTLE_ENDIAN__
50897		#define vfms_laneq_f32(__p0_446, __p1_446, __p2_446, __p3_446) __extension__ ({ \
50898		float32x2_t __ret_446; \
50899		float32x2_t __s0_446 = __p0_446; \
50900		float32x2_t __s1_446 = __p1_446; \
50901		float32x4_t __s2_446 = __p2_446; \
50902		__ret_446 = vfma_laneq_f32(__s0_446, -__s1_446, __s2_446, __p3_446); \
50903		__ret_446; \
	51991	#define vfms_laneq_f32(__p0_462, __p1_462, __p2_462, __p3_462) __extension__ ({ \
	51992	float32x2_t __ret_462; \
	51993	float32x2_t __s0_462 = __p0_462; \
	51994	float32x2_t __s1_462 = __p1_462; \
	51995	float32x4_t __s2_462 = __p2_462; \
	51996	__ret_462 = vfma_laneq_f32(__s0_462, -__s1_462, __s2_462, __p3_462); \
	51997	__ret_462; \
50904	51998	})
50905	51999	#else
50906		#define vfms_laneq_f32(__p0_447, __p1_447, __p2_447, __p3_447) __extension__ ({ \
50907		float32x2_t __ret_447; \
50908		float32x2_t __s0_447 = __p0_447; \
50909		float32x2_t __s1_447 = __p1_447; \
50910		float32x4_t __s2_447 = __p2_447; \
50911		float32x2_t __rev0_447; __rev0_447 = __builtin_shufflevector(__s0_447, __s0_447, 1, 0); \
50912		float32x2_t __rev1_447; __rev1_447 = __builtin_shufflevector(__s1_447, __s1_447, 1, 0); \
50913		float32x4_t __rev2_447; __rev2_447 = __builtin_shufflevector(__s2_447, __s2_447, 3, 2, 1, 0); \
50914		__ret_447 = __noswap_vfma_laneq_f32(__rev0_447, -__rev1_447, __rev2_447, __p3_447); \
50915		__ret_447 = __builtin_shufflevector(__ret_447, __ret_447, 1, 0); \
50916		__ret_447; \
	52000	#define vfms_laneq_f32(__p0_463, __p1_463, __p2_463, __p3_463) __extension__ ({ \
	52001	float32x2_t __ret_463; \
	52002	float32x2_t __s0_463 = __p0_463; \
	52003	float32x2_t __s1_463 = __p1_463; \
	52004	float32x4_t __s2_463 = __p2_463; \
	52005	float32x2_t __rev0_463; __rev0_463 = __builtin_shufflevector(__s0_463, __s0_463, __lane_reverse_64_32); \
	52006	float32x2_t __rev1_463; __rev1_463 = __builtin_shufflevector(__s1_463, __s1_463, __lane_reverse_64_32); \
	52007	float32x4_t __rev2_463; __rev2_463 = __builtin_shufflevector(__s2_463, __s2_463, __lane_reverse_128_32); \
	52008	__ret_463 = __noswap_vfma_laneq_f32(__rev0_463, -__rev1_463, __rev2_463, __p3_463); \
	52009	__ret_463 = __builtin_shufflevector(__ret_463, __ret_463, __lane_reverse_64_32); \
	52010	__ret_463; \
50917	52011	})
50918	52012	#endif
50919	52013
...	...	@@ -50926,10 +52020,10 @@ __ai __attribute__((target("neon"))) float64x2_t vfmsq_n_f64(float64x2_t __p0, f
50926	52020	#else
50927	52021	__ai __attribute__((target("neon"))) float64x2_t vfmsq_n_f64(float64x2_t __p0, float64x2_t __p1, float64_t __p2) {
50928	52022	float64x2_t __ret;
50929		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
50930		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	52023	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	52024	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
50931	52025	__ret = __noswap_vfmaq_f64(__rev0, -__rev1, (float64x2_t) {__p2, __p2});
50932		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	52026	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
50933	52027	return __ret;
50934	52028	}
50935	52029	#endif
...	...	@@ -50943,10 +52037,10 @@ __ai __attribute__((target("neon"))) float32x4_t vfmsq_n_f32(float32x4_t __p0, f
50943	52037	#else
50944	52038	__ai __attribute__((target("neon"))) float32x4_t vfmsq_n_f32(float32x4_t __p0, float32x4_t __p1, float32_t __p2) {
50945	52039	float32x4_t __ret;
50946		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
50947		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	52040	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	52041	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
50948	52042	__ret = __noswap_vfmaq_f32(__rev0, -__rev1, (float32x4_t) {__p2, __p2, __p2, __p2});
50949		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	52043	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
50950	52044	return __ret;
50951	52045	}
50952	52046	#endif
...	...	@@ -50965,10 +52059,10 @@ __ai __attribute__((target("neon"))) float32x2_t vfms_n_f32(float32x2_t __p0, fl
50965	52059	#else
50966	52060	__ai __attribute__((target("neon"))) float32x2_t vfms_n_f32(float32x2_t __p0, float32x2_t __p1, float32_t __p2) {
50967	52061	float32x2_t __ret;
50968		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
50969		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	52062	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	52063	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
50970	52064	__ret = __noswap_vfma_f32(__rev0, -__rev1, (float32x2_t) {__p2, __p2});
50971		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	52065	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
50972	52066	return __ret;
50973	52067	}
50974	52068	#endif
...	...	@@ -50982,7 +52076,7 @@ __ai __attribute__((target("neon"))) poly64x1_t vget_high_p64(poly64x2_t __p0) {
50982	52076	#else
50983	52077	__ai __attribute__((target("neon"))) poly64x1_t vget_high_p64(poly64x2_t __p0) {
50984	52078	poly64x1_t __ret;
50985		poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
	52079	poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
50986	52080	__ret = __builtin_shufflevector(__rev0, __rev0, 1);
50987	52081	return __ret;
50988	52082	}
...	...	@@ -51002,7 +52096,7 @@ __ai __attribute__((target("neon"))) float64x1_t vget_high_f64(float64x2_t __p0)
51002	52096	#else
51003	52097	__ai __attribute__((target("neon"))) float64x1_t vget_high_f64(float64x2_t __p0) {
51004	52098	float64x1_t __ret;
51005		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
	52099	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
51006	52100	__ret = __builtin_shufflevector(__rev0, __rev0, 1);
51007	52101	return __ret;
51008	52102	}
...	...	@@ -51011,28 +52105,28 @@ __ai __attribute__((target("neon"))) float64x1_t vget_high_f64(float64x2_t __p0)
51011	52105	#define vget_lane_p64(__p0, __p1) __extension__ ({ \
51012	52106	poly64_t __ret; \
51013	52107	poly64x1_t __s0 = __p0; \
51014		__ret = (poly64_t) __builtin_neon_vget_lane_i64((poly64x1_t)__s0, __p1); \
	52108	__ret = __builtin_bit_cast(poly64_t, __builtin_neon_vget_lane_i64(__s0, __p1)); \
51015	52109	__ret; \
51016	52110	})
51017	52111	#ifdef __LITTLE_ENDIAN__
51018	52112	#define vgetq_lane_p64(__p0, __p1) __extension__ ({ \
51019	52113	poly64_t __ret; \
51020	52114	poly64x2_t __s0 = __p0; \
51021		__ret = (poly64_t) __builtin_neon_vgetq_lane_i64((poly64x2_t)__s0, __p1); \
	52115	__ret = __builtin_bit_cast(poly64_t, __builtin_neon_vgetq_lane_i64(__s0, __p1)); \
51022	52116	__ret; \
51023	52117	})
51024	52118	#else
51025	52119	#define vgetq_lane_p64(__p0, __p1) __extension__ ({ \
51026	52120	poly64_t __ret; \
51027	52121	poly64x2_t __s0 = __p0; \
51028		poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
51029		__ret = (poly64_t) __builtin_neon_vgetq_lane_i64((poly64x2_t)__rev0, __p1); \
	52122	poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	52123	__ret = __builtin_bit_cast(poly64_t, __builtin_neon_vgetq_lane_i64(__rev0, __p1)); \
51030	52124	__ret; \
51031	52125	})
51032	52126	#define __noswap_vgetq_lane_p64(__p0, __p1) __extension__ ({ \
51033	52127	poly64_t __ret; \
51034	52128	poly64x2_t __s0 = __p0; \
51035		__ret = (poly64_t) __builtin_neon_vgetq_lane_i64((poly64x2_t)__s0, __p1); \
	52129	__ret = __builtin_bit_cast(poly64_t, __builtin_neon_vgetq_lane_i64(__s0, __p1)); \
51036	52130	__ret; \
51037	52131	})
51038	52132	#endif
...	...	@@ -51041,21 +52135,21 @@ __ai __attribute__((target("neon"))) float64x1_t vget_high_f64(float64x2_t __p0)
51041	52135	#define vgetq_lane_f64(__p0, __p1) __extension__ ({ \
51042	52136	float64_t __ret; \
51043	52137	float64x2_t __s0 = __p0; \
51044		__ret = (float64_t) __builtin_neon_vgetq_lane_f64((float64x2_t)__s0, __p1); \
	52138	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vgetq_lane_f64(__s0, __p1)); \
51045	52139	__ret; \
51046	52140	})
51047	52141	#else
51048	52142	#define vgetq_lane_f64(__p0, __p1) __extension__ ({ \
51049	52143	float64_t __ret; \
51050	52144	float64x2_t __s0 = __p0; \
51051		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
51052		__ret = (float64_t) __builtin_neon_vgetq_lane_f64((float64x2_t)__rev0, __p1); \
	52145	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	52146	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vgetq_lane_f64(__rev0, __p1)); \
51053	52147	__ret; \
51054	52148	})
51055	52149	#define __noswap_vgetq_lane_f64(__p0, __p1) __extension__ ({ \
51056	52150	float64_t __ret; \
51057	52151	float64x2_t __s0 = __p0; \
51058		__ret = (float64_t) __builtin_neon_vgetq_lane_f64((float64x2_t)__s0, __p1); \
	52152	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vgetq_lane_f64(__s0, __p1)); \
51059	52153	__ret; \
51060	52154	})
51061	52155	#endif
...	...	@@ -51063,7 +52157,7 @@ __ai __attribute__((target("neon"))) float64x1_t vget_high_f64(float64x2_t __p0)
51063	52157	#define vget_lane_f64(__p0, __p1) __extension__ ({ \
51064	52158	float64_t __ret; \
51065	52159	float64x1_t __s0 = __p0; \
51066		__ret = (float64_t) __builtin_neon_vget_lane_f64((float64x1_t)__s0, __p1); \
	52160	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vget_lane_f64(__s0, __p1)); \
51067	52161	__ret; \
51068	52162	})
51069	52163	#ifdef __LITTLE_ENDIAN__
...	...	@@ -51075,7 +52169,7 @@ __ai __attribute__((target("neon"))) poly64x1_t vget_low_p64(poly64x2_t __p0) {
51075	52169	#else
51076	52170	__ai __attribute__((target("neon"))) poly64x1_t vget_low_p64(poly64x2_t __p0) {
51077	52171	poly64x1_t __ret;
51078		poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
	52172	poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
51079	52173	__ret = __builtin_shufflevector(__rev0, __rev0, 0);
51080	52174	return __ret;
51081	52175	}
...	...	@@ -51090,7 +52184,7 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51090	52184	#else
51091	52185	__ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0) {
51092	52186	float64x1_t __ret;
51093		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
	52187	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
51094	52188	__ret = __builtin_shufflevector(__rev0, __rev0, 0);
51095	52189	return __ret;
51096	52190	}
...	...	@@ -51098,20 +52192,20 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51098	52192
51099	52193	#define vld1_p64(__p0) __extension__ ({ \
51100	52194	poly64x1_t __ret; \
51101		__ret = (poly64x1_t) __builtin_neon_vld1_v(__p0, 6); \
	52195	__ret = __builtin_bit_cast(poly64x1_t, __builtin_neon_vld1_v(__p0, 6)); \
51102	52196	__ret; \
51103	52197	})
51104	52198	#ifdef __LITTLE_ENDIAN__
51105	52199	#define vld1q_p64(__p0) __extension__ ({ \
51106	52200	poly64x2_t __ret; \
51107		__ret = (poly64x2_t) __builtin_neon_vld1q_v(__p0, 38); \
	52201	__ret = __builtin_bit_cast(poly64x2_t, __builtin_neon_vld1q_v(__p0, 38)); \
51108	52202	__ret; \
51109	52203	})
51110	52204	#else
51111	52205	#define vld1q_p64(__p0) __extension__ ({ \
51112	52206	poly64x2_t __ret; \
51113		__ret = (poly64x2_t) __builtin_neon_vld1q_v(__p0, 38); \
51114		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	52207	__ret = __builtin_bit_cast(poly64x2_t, __builtin_neon_vld1q_v(__p0, 38)); \
	52208	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
51115	52209	__ret; \
51116	52210	})
51117	52211	#endif
...	...	@@ -51119,39 +52213,69 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51119	52213	#ifdef __LITTLE_ENDIAN__
51120	52214	#define vld1q_f64(__p0) __extension__ ({ \
51121	52215	float64x2_t __ret; \
51122		__ret = (float64x2_t) __builtin_neon_vld1q_v(__p0, 42); \
	52216	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vld1q_v(__p0, 42)); \
51123	52217	__ret; \
51124	52218	})
51125	52219	#else
51126	52220	#define vld1q_f64(__p0) __extension__ ({ \
51127	52221	float64x2_t __ret; \
51128		__ret = (float64x2_t) __builtin_neon_vld1q_v(__p0, 42); \
51129		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	52222	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vld1q_v(__p0, 42)); \
	52223	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
	52224	__ret; \
	52225	})
	52226	#endif
	52227
	52228	#ifdef __LITTLE_ENDIAN__
	52229	#define vld1q_mf8(__p0) __extension__ ({ \
	52230	mfloat8x16_t __ret; \
	52231	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vld1q_v(__p0, 44)); \
	52232	__ret; \
	52233	})
	52234	#else
	52235	#define vld1q_mf8(__p0) __extension__ ({ \
	52236	mfloat8x16_t __ret; \
	52237	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vld1q_v(__p0, 44)); \
	52238	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
51130	52239	__ret; \
51131	52240	})
51132	52241	#endif
51133	52242
51134	52243	#define vld1_f64(__p0) __extension__ ({ \
51135	52244	float64x1_t __ret; \
51136		__ret = (float64x1_t) __builtin_neon_vld1_v(__p0, 10); \
	52245	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vld1_v(__p0, 10)); \
	52246	__ret; \
	52247	})
	52248	#ifdef __LITTLE_ENDIAN__
	52249	#define vld1_mf8(__p0) __extension__ ({ \
	52250	mfloat8x8_t __ret; \
	52251	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vld1_v(__p0, 12)); \
	52252	__ret; \
	52253	})
	52254	#else
	52255	#define vld1_mf8(__p0) __extension__ ({ \
	52256	mfloat8x8_t __ret; \
	52257	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vld1_v(__p0, 12)); \
	52258	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
51137	52259	__ret; \
51138	52260	})
	52261	#endif
	52262
51139	52263	#define vld1_dup_p64(__p0) __extension__ ({ \
51140	52264	poly64x1_t __ret; \
51141		__ret = (poly64x1_t) __builtin_neon_vld1_dup_v(__p0, 6); \
	52265	__ret = __builtin_bit_cast(poly64x1_t, __builtin_neon_vld1_dup_v(__p0, 6)); \
51142	52266	__ret; \
51143	52267	})
51144	52268	#ifdef __LITTLE_ENDIAN__
51145	52269	#define vld1q_dup_p64(__p0) __extension__ ({ \
51146	52270	poly64x2_t __ret; \
51147		__ret = (poly64x2_t) __builtin_neon_vld1q_dup_v(__p0, 38); \
	52271	__ret = __builtin_bit_cast(poly64x2_t, __builtin_neon_vld1q_dup_v(__p0, 38)); \
51148	52272	__ret; \
51149	52273	})
51150	52274	#else
51151	52275	#define vld1q_dup_p64(__p0) __extension__ ({ \
51152	52276	poly64x2_t __ret; \
51153		__ret = (poly64x2_t) __builtin_neon_vld1q_dup_v(__p0, 38); \
51154		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	52277	__ret = __builtin_bit_cast(poly64x2_t, __builtin_neon_vld1q_dup_v(__p0, 38)); \
	52278	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
51155	52279	__ret; \
51156	52280	})
51157	52281	#endif
...	...	@@ -51159,43 +52283,73 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51159	52283	#ifdef __LITTLE_ENDIAN__
51160	52284	#define vld1q_dup_f64(__p0) __extension__ ({ \
51161	52285	float64x2_t __ret; \
51162		__ret = (float64x2_t) __builtin_neon_vld1q_dup_v(__p0, 42); \
	52286	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vld1q_dup_v(__p0, 42)); \
51163	52287	__ret; \
51164	52288	})
51165	52289	#else
51166	52290	#define vld1q_dup_f64(__p0) __extension__ ({ \
51167	52291	float64x2_t __ret; \
51168		__ret = (float64x2_t) __builtin_neon_vld1q_dup_v(__p0, 42); \
51169		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	52292	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vld1q_dup_v(__p0, 42)); \
	52293	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
	52294	__ret; \
	52295	})
	52296	#endif
	52297
	52298	#ifdef __LITTLE_ENDIAN__
	52299	#define vld1q_dup_mf8(__p0) __extension__ ({ \
	52300	mfloat8x16_t __ret; \
	52301	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vld1q_dup_v(__p0, 44)); \
	52302	__ret; \
	52303	})
	52304	#else
	52305	#define vld1q_dup_mf8(__p0) __extension__ ({ \
	52306	mfloat8x16_t __ret; \
	52307	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vld1q_dup_v(__p0, 44)); \
	52308	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
51170	52309	__ret; \
51171	52310	})
51172	52311	#endif
51173	52312
51174	52313	#define vld1_dup_f64(__p0) __extension__ ({ \
51175	52314	float64x1_t __ret; \
51176		__ret = (float64x1_t) __builtin_neon_vld1_dup_v(__p0, 10); \
	52315	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vld1_dup_v(__p0, 10)); \
	52316	__ret; \
	52317	})
	52318	#ifdef __LITTLE_ENDIAN__
	52319	#define vld1_dup_mf8(__p0) __extension__ ({ \
	52320	mfloat8x8_t __ret; \
	52321	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vld1_dup_v(__p0, 12)); \
	52322	__ret; \
	52323	})
	52324	#else
	52325	#define vld1_dup_mf8(__p0) __extension__ ({ \
	52326	mfloat8x8_t __ret; \
	52327	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vld1_dup_v(__p0, 12)); \
	52328	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
51177	52329	__ret; \
51178	52330	})
	52331	#endif
	52332
51179	52333	#define vld1_lane_p64(__p0, __p1, __p2) __extension__ ({ \
51180	52334	poly64x1_t __ret; \
51181	52335	poly64x1_t __s1 = __p1; \
51182		__ret = (poly64x1_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__s1, __p2, 6); \
	52336	__ret = __builtin_bit_cast(poly64x1_t, __builtin_neon_vld1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 6)); \
51183	52337	__ret; \
51184	52338	})
51185	52339	#ifdef __LITTLE_ENDIAN__
51186	52340	#define vld1q_lane_p64(__p0, __p1, __p2) __extension__ ({ \
51187	52341	poly64x2_t __ret; \
51188	52342	poly64x2_t __s1 = __p1; \
51189		__ret = (poly64x2_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__s1, __p2, 38); \
	52343	__ret = __builtin_bit_cast(poly64x2_t, __builtin_neon_vld1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 38)); \
51190	52344	__ret; \
51191	52345	})
51192	52346	#else
51193	52347	#define vld1q_lane_p64(__p0, __p1, __p2) __extension__ ({ \
51194	52348	poly64x2_t __ret; \
51195	52349	poly64x2_t __s1 = __p1; \
51196		poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
51197		__ret = (poly64x2_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 38); \
51198		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	52350	poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	52351	__ret = __builtin_bit_cast(poly64x2_t, __builtin_neon_vld1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 38)); \
	52352	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
51199	52353	__ret; \
51200	52354	})
51201	52355	#endif
...	...	@@ -51204,16 +52358,34 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51204	52358	#define vld1q_lane_f64(__p0, __p1, __p2) __extension__ ({ \
51205	52359	float64x2_t __ret; \
51206	52360	float64x2_t __s1 = __p1; \
51207		__ret = (float64x2_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__s1, __p2, 42); \
	52361	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vld1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 42)); \
51208	52362	__ret; \
51209	52363	})
51210	52364	#else
51211	52365	#define vld1q_lane_f64(__p0, __p1, __p2) __extension__ ({ \
51212	52366	float64x2_t __ret; \
51213	52367	float64x2_t __s1 = __p1; \
51214		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
51215		__ret = (float64x2_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 42); \
51216		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	52368	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	52369	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vld1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 42)); \
	52370	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
	52371	__ret; \
	52372	})
	52373	#endif
	52374
	52375	#ifdef __LITTLE_ENDIAN__
	52376	#define vld1q_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	52377	mfloat8x16_t __ret; \
	52378	mfloat8x16_t __s1 = __p1; \
	52379	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vld1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 44)); \
	52380	__ret; \
	52381	})
	52382	#else
	52383	#define vld1q_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	52384	mfloat8x16_t __ret; \
	52385	mfloat8x16_t __s1 = __p1; \
	52386	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	52387	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vld1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 44)); \
	52388	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8); \
51217	52389	__ret; \
51218	52390	})
51219	52391	#endif
...	...	@@ -51221,9 +52393,27 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51221	52393	#define vld1_lane_f64(__p0, __p1, __p2) __extension__ ({ \
51222	52394	float64x1_t __ret; \
51223	52395	float64x1_t __s1 = __p1; \
51224		__ret = (float64x1_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__s1, __p2, 10); \
	52396	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vld1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 10)); \
	52397	__ret; \
	52398	})
	52399	#ifdef __LITTLE_ENDIAN__
	52400	#define vld1_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	52401	mfloat8x8_t __ret; \
	52402	mfloat8x8_t __s1 = __p1; \
	52403	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vld1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 12)); \
51225	52404	__ret; \
51226	52405	})
	52406	#else
	52407	#define vld1_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	52408	mfloat8x8_t __ret; \
	52409	mfloat8x8_t __s1 = __p1; \
	52410	mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	52411	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vld1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), __p2, 12)); \
	52412	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8); \
	52413	__ret; \
	52414	})
	52415	#endif
	52416
51227	52417	#define vld1_p64_x2(__p0) __extension__ ({ \
51228	52418	poly64x1x2_t __ret; \
51229	52419	__builtin_neon_vld1_x2_v(&__ret, __p0, 6); \
...	...	@@ -51240,8 +52430,8 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51240	52430	poly64x2x2_t __ret; \
51241	52431	__builtin_neon_vld1q_x2_v(&__ret, __p0, 38); \
51242	52432	\
51243		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
51244		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
	52433	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	52434	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
51245	52435	__ret; \
51246	52436	})
51247	52437	#endif
...	...	@@ -51257,8 +52447,25 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51257	52447	float64x2x2_t __ret; \
51258	52448	__builtin_neon_vld1q_x2_v(&__ret, __p0, 42); \
51259	52449	\
51260		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
51261		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
	52450	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	52451	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	52452	__ret; \
	52453	})
	52454	#endif
	52455
	52456	#ifdef __LITTLE_ENDIAN__
	52457	#define vld1q_mf8_x2(__p0) __extension__ ({ \
	52458	mfloat8x16x2_t __ret; \
	52459	__builtin_neon_vld1q_x2_v(&__ret, __p0, 44); \
	52460	__ret; \
	52461	})
	52462	#else
	52463	#define vld1q_mf8_x2(__p0) __extension__ ({ \
	52464	mfloat8x16x2_t __ret; \
	52465	__builtin_neon_vld1q_x2_v(&__ret, __p0, 44); \
	52466	\
	52467	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	52468	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
51262	52469	__ret; \
51263	52470	})
51264	52471	#endif
...	...	@@ -51268,6 +52475,23 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51268	52475	__builtin_neon_vld1_x2_v(&__ret, __p0, 10); \
51269	52476	__ret; \
51270	52477	})
	52478	#ifdef __LITTLE_ENDIAN__
	52479	#define vld1_mf8_x2(__p0) __extension__ ({ \
	52480	mfloat8x8x2_t __ret; \
	52481	__builtin_neon_vld1_x2_v(&__ret, __p0, 12); \
	52482	__ret; \
	52483	})
	52484	#else
	52485	#define vld1_mf8_x2(__p0) __extension__ ({ \
	52486	mfloat8x8x2_t __ret; \
	52487	__builtin_neon_vld1_x2_v(&__ret, __p0, 12); \
	52488	\
	52489	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	52490	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	52491	__ret; \
	52492	})
	52493	#endif
	52494
51271	52495	#define vld1_p64_x3(__p0) __extension__ ({ \
51272	52496	poly64x1x3_t __ret; \
51273	52497	__builtin_neon_vld1_x3_v(&__ret, __p0, 6); \
...	...	@@ -51284,9 +52508,9 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51284	52508	poly64x2x3_t __ret; \
51285	52509	__builtin_neon_vld1q_x3_v(&__ret, __p0, 38); \
51286	52510	\
51287		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
51288		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
51289		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
	52511	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	52512	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	52513	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
51290	52514	__ret; \
51291	52515	})
51292	52516	#endif
...	...	@@ -51302,9 +52526,27 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51302	52526	float64x2x3_t __ret; \
51303	52527	__builtin_neon_vld1q_x3_v(&__ret, __p0, 42); \
51304	52528	\
51305		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
51306		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
51307		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
	52529	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	52530	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	52531	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
	52532	__ret; \
	52533	})
	52534	#endif
	52535
	52536	#ifdef __LITTLE_ENDIAN__
	52537	#define vld1q_mf8_x3(__p0) __extension__ ({ \
	52538	mfloat8x16x3_t __ret; \
	52539	__builtin_neon_vld1q_x3_v(&__ret, __p0, 44); \
	52540	__ret; \
	52541	})
	52542	#else
	52543	#define vld1q_mf8_x3(__p0) __extension__ ({ \
	52544	mfloat8x16x3_t __ret; \
	52545	__builtin_neon_vld1q_x3_v(&__ret, __p0, 44); \
	52546	\
	52547	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	52548	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	52549	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
51308	52550	__ret; \
51309	52551	})
51310	52552	#endif
...	...	@@ -51314,6 +52556,24 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51314	52556	__builtin_neon_vld1_x3_v(&__ret, __p0, 10); \
51315	52557	__ret; \
51316	52558	})
	52559	#ifdef __LITTLE_ENDIAN__
	52560	#define vld1_mf8_x3(__p0) __extension__ ({ \
	52561	mfloat8x8x3_t __ret; \
	52562	__builtin_neon_vld1_x3_v(&__ret, __p0, 12); \
	52563	__ret; \
	52564	})
	52565	#else
	52566	#define vld1_mf8_x3(__p0) __extension__ ({ \
	52567	mfloat8x8x3_t __ret; \
	52568	__builtin_neon_vld1_x3_v(&__ret, __p0, 12); \
	52569	\
	52570	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	52571	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	52572	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
	52573	__ret; \
	52574	})
	52575	#endif
	52576
51317	52577	#define vld1_p64_x4(__p0) __extension__ ({ \
51318	52578	poly64x1x4_t __ret; \
51319	52579	__builtin_neon_vld1_x4_v(&__ret, __p0, 6); \
...	...	@@ -51330,10 +52590,10 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51330	52590	poly64x2x4_t __ret; \
51331	52591	__builtin_neon_vld1q_x4_v(&__ret, __p0, 38); \
51332	52592	\
51333		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
51334		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
51335		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
51336		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
	52593	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	52594	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	52595	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
	52596	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_64); \
51337	52597	__ret; \
51338	52598	})
51339	52599	#endif
...	...	@@ -51349,10 +52609,29 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51349	52609	float64x2x4_t __ret; \
51350	52610	__builtin_neon_vld1q_x4_v(&__ret, __p0, 42); \
51351	52611	\
51352		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
51353		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
51354		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
51355		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
	52612	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	52613	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	52614	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
	52615	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_64); \
	52616	__ret; \
	52617	})
	52618	#endif
	52619
	52620	#ifdef __LITTLE_ENDIAN__
	52621	#define vld1q_mf8_x4(__p0) __extension__ ({ \
	52622	mfloat8x16x4_t __ret; \
	52623	__builtin_neon_vld1q_x4_v(&__ret, __p0, 44); \
	52624	__ret; \
	52625	})
	52626	#else
	52627	#define vld1q_mf8_x4(__p0) __extension__ ({ \
	52628	mfloat8x16x4_t __ret; \
	52629	__builtin_neon_vld1q_x4_v(&__ret, __p0, 44); \
	52630	\
	52631	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	52632	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	52633	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
	52634	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_8); \
51356	52635	__ret; \
51357	52636	})
51358	52637	#endif
...	...	@@ -51362,6 +52641,25 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51362	52641	__builtin_neon_vld1_x4_v(&__ret, __p0, 10); \
51363	52642	__ret; \
51364	52643	})
	52644	#ifdef __LITTLE_ENDIAN__
	52645	#define vld1_mf8_x4(__p0) __extension__ ({ \
	52646	mfloat8x8x4_t __ret; \
	52647	__builtin_neon_vld1_x4_v(&__ret, __p0, 12); \
	52648	__ret; \
	52649	})
	52650	#else
	52651	#define vld1_mf8_x4(__p0) __extension__ ({ \
	52652	mfloat8x8x4_t __ret; \
	52653	__builtin_neon_vld1_x4_v(&__ret, __p0, 12); \
	52654	\
	52655	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	52656	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	52657	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
	52658	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_8); \
	52659	__ret; \
	52660	})
	52661	#endif
	52662
51365	52663	#define vld2_p64(__p0) __extension__ ({ \
51366	52664	poly64x1x2_t __ret; \
51367	52665	__builtin_neon_vld2_v(&__ret, __p0, 6); \
...	...	@@ -51378,8 +52676,8 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51378	52676	poly64x2x2_t __ret; \
51379	52677	__builtin_neon_vld2q_v(&__ret, __p0, 38); \
51380	52678	\
51381		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
51382		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
	52679	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	52680	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
51383	52681	__ret; \
51384	52682	})
51385	52683	#endif
...	...	@@ -51395,8 +52693,8 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51395	52693	uint64x2x2_t __ret; \
51396	52694	__builtin_neon_vld2q_v(&__ret, __p0, 51); \
51397	52695	\
51398		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
51399		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
	52696	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	52697	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
51400	52698	__ret; \
51401	52699	})
51402	52700	#endif
...	...	@@ -51412,8 +52710,8 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51412	52710	float64x2x2_t __ret; \
51413	52711	__builtin_neon_vld2q_v(&__ret, __p0, 42); \
51414	52712	\
51415		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
51416		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
	52713	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	52714	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
51417	52715	__ret; \
51418	52716	})
51419	52717	#endif
...	...	@@ -51429,8 +52727,25 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51429	52727	int64x2x2_t __ret; \
51430	52728	__builtin_neon_vld2q_v(&__ret, __p0, 35); \
51431	52729	\
51432		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
51433		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
	52730	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	52731	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	52732	__ret; \
	52733	})
	52734	#endif
	52735
	52736	#ifdef __LITTLE_ENDIAN__
	52737	#define vld2q_mf8(__p0) __extension__ ({ \
	52738	mfloat8x16x2_t __ret; \
	52739	__builtin_neon_vld2q_v(&__ret, __p0, 44); \
	52740	__ret; \
	52741	})
	52742	#else
	52743	#define vld2q_mf8(__p0) __extension__ ({ \
	52744	mfloat8x16x2_t __ret; \
	52745	__builtin_neon_vld2q_v(&__ret, __p0, 44); \
	52746	\
	52747	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	52748	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
51434	52749	__ret; \
51435	52750	})
51436	52751	#endif
...	...	@@ -51440,6 +52755,23 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51440	52755	__builtin_neon_vld2_v(&__ret, __p0, 10); \
51441	52756	__ret; \
51442	52757	})
	52758	#ifdef __LITTLE_ENDIAN__
	52759	#define vld2_mf8(__p0) __extension__ ({ \
	52760	mfloat8x8x2_t __ret; \
	52761	__builtin_neon_vld2_v(&__ret, __p0, 12); \
	52762	__ret; \
	52763	})
	52764	#else
	52765	#define vld2_mf8(__p0) __extension__ ({ \
	52766	mfloat8x8x2_t __ret; \
	52767	__builtin_neon_vld2_v(&__ret, __p0, 12); \
	52768	\
	52769	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	52770	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	52771	__ret; \
	52772	})
	52773	#endif
	52774
51443	52775	#define vld2_dup_p64(__p0) __extension__ ({ \
51444	52776	poly64x1x2_t __ret; \
51445	52777	__builtin_neon_vld2_dup_v(&__ret, __p0, 6); \
...	...	@@ -51456,8 +52788,8 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51456	52788	poly64x2x2_t __ret; \
51457	52789	__builtin_neon_vld2q_dup_v(&__ret, __p0, 38); \
51458	52790	\
51459		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
51460		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
	52791	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	52792	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
51461	52793	__ret; \
51462	52794	})
51463	52795	#endif
...	...	@@ -51473,8 +52805,25 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51473	52805	float64x2x2_t __ret; \
51474	52806	__builtin_neon_vld2q_dup_v(&__ret, __p0, 42); \
51475	52807	\
51476		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
51477		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
	52808	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	52809	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	52810	__ret; \
	52811	})
	52812	#endif
	52813
	52814	#ifdef __LITTLE_ENDIAN__
	52815	#define vld2q_dup_mf8(__p0) __extension__ ({ \
	52816	mfloat8x16x2_t __ret; \
	52817	__builtin_neon_vld2q_dup_v(&__ret, __p0, 44); \
	52818	__ret; \
	52819	})
	52820	#else
	52821	#define vld2q_dup_mf8(__p0) __extension__ ({ \
	52822	mfloat8x16x2_t __ret; \
	52823	__builtin_neon_vld2q_dup_v(&__ret, __p0, 44); \
	52824	\
	52825	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	52826	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
51478	52827	__ret; \
51479	52828	})
51480	52829	#endif
...	...	@@ -51484,17 +52833,34 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51484	52833	__builtin_neon_vld2_dup_v(&__ret, __p0, 10); \
51485	52834	__ret; \
51486	52835	})
	52836	#ifdef __LITTLE_ENDIAN__
	52837	#define vld2_dup_mf8(__p0) __extension__ ({ \
	52838	mfloat8x8x2_t __ret; \
	52839	__builtin_neon_vld2_dup_v(&__ret, __p0, 12); \
	52840	__ret; \
	52841	})
	52842	#else
	52843	#define vld2_dup_mf8(__p0) __extension__ ({ \
	52844	mfloat8x8x2_t __ret; \
	52845	__builtin_neon_vld2_dup_v(&__ret, __p0, 12); \
	52846	\
	52847	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	52848	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	52849	__ret; \
	52850	})
	52851	#endif
	52852
51487	52853	#define vld2_lane_p64(__p0, __p1, __p2) __extension__ ({ \
51488	52854	poly64x1x2_t __ret; \
51489	52855	poly64x1x2_t __s1 = __p1; \
51490		__builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 6); \
	52856	__builtin_neon_vld2_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 6); \
51491	52857	__ret; \
51492	52858	})
51493	52859	#ifdef __LITTLE_ENDIAN__
51494	52860	#define vld2q_lane_p8(__p0, __p1, __p2) __extension__ ({ \
51495	52861	poly8x16x2_t __ret; \
51496	52862	poly8x16x2_t __s1 = __p1; \
51497		__builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 36); \
	52863	__builtin_neon_vld2q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 36); \
51498	52864	__ret; \
51499	52865	})
51500	52866	#else
...	...	@@ -51502,12 +52868,12 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51502	52868	poly8x16x2_t __ret; \
51503	52869	poly8x16x2_t __s1 = __p1; \
51504	52870	poly8x16x2_t __rev1; \
51505		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
51506		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
51507		__builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 36); \
	52871	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	52872	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	52873	__builtin_neon_vld2q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 36); \
51508	52874	\
51509		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
51510		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	52875	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	52876	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
51511	52877	__ret; \
51512	52878	})
51513	52879	#endif
...	...	@@ -51516,7 +52882,7 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51516	52882	#define vld2q_lane_p64(__p0, __p1, __p2) __extension__ ({ \
51517	52883	poly64x2x2_t __ret; \
51518	52884	poly64x2x2_t __s1 = __p1; \
51519		__builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 38); \
	52885	__builtin_neon_vld2q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 38); \
51520	52886	__ret; \
51521	52887	})
51522	52888	#else
...	...	@@ -51524,12 +52890,12 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51524	52890	poly64x2x2_t __ret; \
51525	52891	poly64x2x2_t __s1 = __p1; \
51526	52892	poly64x2x2_t __rev1; \
51527		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
51528		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
51529		__builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 38); \
	52893	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	52894	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	52895	__builtin_neon_vld2q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 38); \
51530	52896	\
51531		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
51532		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
	52897	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	52898	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
51533	52899	__ret; \
51534	52900	})
51535	52901	#endif
...	...	@@ -51538,7 +52904,7 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51538	52904	#define vld2q_lane_u8(__p0, __p1, __p2) __extension__ ({ \
51539	52905	uint8x16x2_t __ret; \
51540	52906	uint8x16x2_t __s1 = __p1; \
51541		__builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 48); \
	52907	__builtin_neon_vld2q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 48); \
51542	52908	__ret; \
51543	52909	})
51544	52910	#else
...	...	@@ -51546,12 +52912,12 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51546	52912	uint8x16x2_t __ret; \
51547	52913	uint8x16x2_t __s1 = __p1; \
51548	52914	uint8x16x2_t __rev1; \
51549		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
51550		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
51551		__builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 48); \
	52915	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	52916	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	52917	__builtin_neon_vld2q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 48); \
51552	52918	\
51553		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
51554		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	52919	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	52920	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
51555	52921	__ret; \
51556	52922	})
51557	52923	#endif
...	...	@@ -51560,7 +52926,7 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51560	52926	#define vld2q_lane_u64(__p0, __p1, __p2) __extension__ ({ \
51561	52927	uint64x2x2_t __ret; \
51562	52928	uint64x2x2_t __s1 = __p1; \
51563		__builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 51); \
	52929	__builtin_neon_vld2q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 51); \
51564	52930	__ret; \
51565	52931	})
51566	52932	#else
...	...	@@ -51568,12 +52934,12 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51568	52934	uint64x2x2_t __ret; \
51569	52935	uint64x2x2_t __s1 = __p1; \
51570	52936	uint64x2x2_t __rev1; \
51571		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
51572		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
51573		__builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 51); \
	52937	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	52938	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	52939	__builtin_neon_vld2q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 51); \
51574	52940	\
51575		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
51576		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
	52941	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	52942	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
51577	52943	__ret; \
51578	52944	})
51579	52945	#endif
...	...	@@ -51582,7 +52948,7 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51582	52948	#define vld2q_lane_s8(__p0, __p1, __p2) __extension__ ({ \
51583	52949	int8x16x2_t __ret; \
51584	52950	int8x16x2_t __s1 = __p1; \
51585		__builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 32); \
	52951	__builtin_neon_vld2q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 32); \
51586	52952	__ret; \
51587	52953	})
51588	52954	#else
...	...	@@ -51590,12 +52956,12 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51590	52956	int8x16x2_t __ret; \
51591	52957	int8x16x2_t __s1 = __p1; \
51592	52958	int8x16x2_t __rev1; \
51593		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
51594		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
51595		__builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 32); \
	52959	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	52960	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	52961	__builtin_neon_vld2q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 32); \
51596	52962	\
51597		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
51598		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	52963	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	52964	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
51599	52965	__ret; \
51600	52966	})
51601	52967	#endif
...	...	@@ -51604,7 +52970,7 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51604	52970	#define vld2q_lane_f64(__p0, __p1, __p2) __extension__ ({ \
51605	52971	float64x2x2_t __ret; \
51606	52972	float64x2x2_t __s1 = __p1; \
51607		__builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 42); \
	52973	__builtin_neon_vld2q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 42); \
51608	52974	__ret; \
51609	52975	})
51610	52976	#else
...	...	@@ -51612,12 +52978,12 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51612	52978	float64x2x2_t __ret; \
51613	52979	float64x2x2_t __s1 = __p1; \
51614	52980	float64x2x2_t __rev1; \
51615		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
51616		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
51617		__builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 42); \
	52981	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	52982	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	52983	__builtin_neon_vld2q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 42); \
51618	52984	\
51619		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
51620		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
	52985	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	52986	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
51621	52987	__ret; \
51622	52988	})
51623	52989	#endif
...	...	@@ -51626,7 +52992,7 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51626	52992	#define vld2q_lane_s64(__p0, __p1, __p2) __extension__ ({ \
51627	52993	int64x2x2_t __ret; \
51628	52994	int64x2x2_t __s1 = __p1; \
51629		__builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 35); \
	52995	__builtin_neon_vld2q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 35); \
51630	52996	__ret; \
51631	52997	})
51632	52998	#else
...	...	@@ -51634,12 +53000,34 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51634	53000	int64x2x2_t __ret; \
51635	53001	int64x2x2_t __s1 = __p1; \
51636	53002	int64x2x2_t __rev1; \
51637		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
51638		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
51639		__builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 35); \
	53003	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	53004	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	53005	__builtin_neon_vld2q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 35); \
	53006	\
	53007	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	53008	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	53009	__ret; \
	53010	})
	53011	#endif
	53012
	53013	#ifdef __LITTLE_ENDIAN__
	53014	#define vld2q_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	53015	mfloat8x16x2_t __ret; \
	53016	mfloat8x16x2_t __s1 = __p1; \
	53017	__builtin_neon_vld2q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 44); \
	53018	__ret; \
	53019	})
	53020	#else
	53021	#define vld2q_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	53022	mfloat8x16x2_t __ret; \
	53023	mfloat8x16x2_t __s1 = __p1; \
	53024	mfloat8x16x2_t __rev1; \
	53025	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	53026	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	53027	__builtin_neon_vld2q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 44); \
51640	53028	\
51641		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
51642		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
	53029	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	53030	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
51643	53031	__ret; \
51644	53032	})
51645	53033	#endif
...	...	@@ -51647,21 +53035,43 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51647	53035	#define vld2_lane_u64(__p0, __p1, __p2) __extension__ ({ \
51648	53036	uint64x1x2_t __ret; \
51649	53037	uint64x1x2_t __s1 = __p1; \
51650		__builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 19); \
	53038	__builtin_neon_vld2_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 19); \
51651	53039	__ret; \
51652	53040	})
51653	53041	#define vld2_lane_f64(__p0, __p1, __p2) __extension__ ({ \
51654	53042	float64x1x2_t __ret; \
51655	53043	float64x1x2_t __s1 = __p1; \
51656		__builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 10); \
	53044	__builtin_neon_vld2_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 10); \
51657	53045	__ret; \
51658	53046	})
51659	53047	#define vld2_lane_s64(__p0, __p1, __p2) __extension__ ({ \
51660	53048	int64x1x2_t __ret; \
51661	53049	int64x1x2_t __s1 = __p1; \
51662		__builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 3); \
	53050	__builtin_neon_vld2_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 3); \
51663	53051	__ret; \
51664	53052	})
	53053	#ifdef __LITTLE_ENDIAN__
	53054	#define vld2_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	53055	mfloat8x8x2_t __ret; \
	53056	mfloat8x8x2_t __s1 = __p1; \
	53057	__builtin_neon_vld2_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 12); \
	53058	__ret; \
	53059	})
	53060	#else
	53061	#define vld2_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	53062	mfloat8x8x2_t __ret; \
	53063	mfloat8x8x2_t __s1 = __p1; \
	53064	mfloat8x8x2_t __rev1; \
	53065	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	53066	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	53067	__builtin_neon_vld2_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __p2, 12); \
	53068	\
	53069	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	53070	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	53071	__ret; \
	53072	})
	53073	#endif
	53074
51665	53075	#define vld3_p64(__p0) __extension__ ({ \
51666	53076	poly64x1x3_t __ret; \
51667	53077	__builtin_neon_vld3_v(&__ret, __p0, 6); \
...	...	@@ -51678,9 +53088,9 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51678	53088	poly64x2x3_t __ret; \
51679	53089	__builtin_neon_vld3q_v(&__ret, __p0, 38); \
51680	53090	\
51681		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
51682		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
51683		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
	53091	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	53092	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	53093	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
51684	53094	__ret; \
51685	53095	})
51686	53096	#endif
...	...	@@ -51696,9 +53106,9 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51696	53106	uint64x2x3_t __ret; \
51697	53107	__builtin_neon_vld3q_v(&__ret, __p0, 51); \
51698	53108	\
51699		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
51700		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
51701		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
	53109	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	53110	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	53111	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
51702	53112	__ret; \
51703	53113	})
51704	53114	#endif
...	...	@@ -51714,9 +53124,9 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51714	53124	float64x2x3_t __ret; \
51715	53125	__builtin_neon_vld3q_v(&__ret, __p0, 42); \
51716	53126	\
51717		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
51718		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
51719		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
	53127	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	53128	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	53129	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
51720	53130	__ret; \
51721	53131	})
51722	53132	#endif
...	...	@@ -51732,9 +53142,27 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51732	53142	int64x2x3_t __ret; \
51733	53143	__builtin_neon_vld3q_v(&__ret, __p0, 35); \
51734	53144	\
51735		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
51736		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
51737		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
	53145	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	53146	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	53147	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
	53148	__ret; \
	53149	})
	53150	#endif
	53151
	53152	#ifdef __LITTLE_ENDIAN__
	53153	#define vld3q_mf8(__p0) __extension__ ({ \
	53154	mfloat8x16x3_t __ret; \
	53155	__builtin_neon_vld3q_v(&__ret, __p0, 44); \
	53156	__ret; \
	53157	})
	53158	#else
	53159	#define vld3q_mf8(__p0) __extension__ ({ \
	53160	mfloat8x16x3_t __ret; \
	53161	__builtin_neon_vld3q_v(&__ret, __p0, 44); \
	53162	\
	53163	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	53164	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	53165	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
51738	53166	__ret; \
51739	53167	})
51740	53168	#endif
...	...	@@ -51744,6 +53172,24 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51744	53172	__builtin_neon_vld3_v(&__ret, __p0, 10); \
51745	53173	__ret; \
51746	53174	})
	53175	#ifdef __LITTLE_ENDIAN__
	53176	#define vld3_mf8(__p0) __extension__ ({ \
	53177	mfloat8x8x3_t __ret; \
	53178	__builtin_neon_vld3_v(&__ret, __p0, 12); \
	53179	__ret; \
	53180	})
	53181	#else
	53182	#define vld3_mf8(__p0) __extension__ ({ \
	53183	mfloat8x8x3_t __ret; \
	53184	__builtin_neon_vld3_v(&__ret, __p0, 12); \
	53185	\
	53186	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	53187	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	53188	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
	53189	__ret; \
	53190	})
	53191	#endif
	53192
51747	53193	#define vld3_dup_p64(__p0) __extension__ ({ \
51748	53194	poly64x1x3_t __ret; \
51749	53195	__builtin_neon_vld3_dup_v(&__ret, __p0, 6); \
...	...	@@ -51760,9 +53206,9 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51760	53206	poly64x2x3_t __ret; \
51761	53207	__builtin_neon_vld3q_dup_v(&__ret, __p0, 38); \
51762	53208	\
51763		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
51764		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
51765		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
	53209	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	53210	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	53211	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
51766	53212	__ret; \
51767	53213	})
51768	53214	#endif
...	...	@@ -51778,9 +53224,27 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51778	53224	float64x2x3_t __ret; \
51779	53225	__builtin_neon_vld3q_dup_v(&__ret, __p0, 42); \
51780	53226	\
51781		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
51782		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
51783		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
	53227	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	53228	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	53229	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
	53230	__ret; \
	53231	})
	53232	#endif
	53233
	53234	#ifdef __LITTLE_ENDIAN__
	53235	#define vld3q_dup_mf8(__p0) __extension__ ({ \
	53236	mfloat8x16x3_t __ret; \
	53237	__builtin_neon_vld3q_dup_v(&__ret, __p0, 44); \
	53238	__ret; \
	53239	})
	53240	#else
	53241	#define vld3q_dup_mf8(__p0) __extension__ ({ \
	53242	mfloat8x16x3_t __ret; \
	53243	__builtin_neon_vld3q_dup_v(&__ret, __p0, 44); \
	53244	\
	53245	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	53246	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	53247	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
51784	53248	__ret; \
51785	53249	})
51786	53250	#endif
...	...	@@ -51790,17 +53254,35 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51790	53254	__builtin_neon_vld3_dup_v(&__ret, __p0, 10); \
51791	53255	__ret; \
51792	53256	})
	53257	#ifdef __LITTLE_ENDIAN__
	53258	#define vld3_dup_mf8(__p0) __extension__ ({ \
	53259	mfloat8x8x3_t __ret; \
	53260	__builtin_neon_vld3_dup_v(&__ret, __p0, 12); \
	53261	__ret; \
	53262	})
	53263	#else
	53264	#define vld3_dup_mf8(__p0) __extension__ ({ \
	53265	mfloat8x8x3_t __ret; \
	53266	__builtin_neon_vld3_dup_v(&__ret, __p0, 12); \
	53267	\
	53268	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	53269	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	53270	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
	53271	__ret; \
	53272	})
	53273	#endif
	53274
51793	53275	#define vld3_lane_p64(__p0, __p1, __p2) __extension__ ({ \
51794	53276	poly64x1x3_t __ret; \
51795	53277	poly64x1x3_t __s1 = __p1; \
51796		__builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 6); \
	53278	__builtin_neon_vld3_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 6); \
51797	53279	__ret; \
51798	53280	})
51799	53281	#ifdef __LITTLE_ENDIAN__
51800	53282	#define vld3q_lane_p8(__p0, __p1, __p2) __extension__ ({ \
51801	53283	poly8x16x3_t __ret; \
51802	53284	poly8x16x3_t __s1 = __p1; \
51803		__builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 36); \
	53285	__builtin_neon_vld3q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 36); \
51804	53286	__ret; \
51805	53287	})
51806	53288	#else
...	...	@@ -51808,14 +53290,14 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51808	53290	poly8x16x3_t __ret; \
51809	53291	poly8x16x3_t __s1 = __p1; \
51810	53292	poly8x16x3_t __rev1; \
51811		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
51812		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
51813		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
51814		__builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 36); \
	53293	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	53294	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	53295	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	53296	__builtin_neon_vld3q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 36); \
51815	53297	\
51816		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
51817		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
51818		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	53298	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	53299	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	53300	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
51819	53301	__ret; \
51820	53302	})
51821	53303	#endif
...	...	@@ -51824,7 +53306,7 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51824	53306	#define vld3q_lane_p64(__p0, __p1, __p2) __extension__ ({ \
51825	53307	poly64x2x3_t __ret; \
51826	53308	poly64x2x3_t __s1 = __p1; \
51827		__builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 38); \
	53309	__builtin_neon_vld3q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 38); \
51828	53310	__ret; \
51829	53311	})
51830	53312	#else
...	...	@@ -51832,14 +53314,14 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51832	53314	poly64x2x3_t __ret; \
51833	53315	poly64x2x3_t __s1 = __p1; \
51834	53316	poly64x2x3_t __rev1; \
51835		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
51836		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
51837		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
51838		__builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 38); \
	53317	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	53318	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	53319	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	53320	__builtin_neon_vld3q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 38); \
51839	53321	\
51840		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
51841		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
51842		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
	53322	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	53323	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	53324	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
51843	53325	__ret; \
51844	53326	})
51845	53327	#endif
...	...	@@ -51848,7 +53330,7 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51848	53330	#define vld3q_lane_u8(__p0, __p1, __p2) __extension__ ({ \
51849	53331	uint8x16x3_t __ret; \
51850	53332	uint8x16x3_t __s1 = __p1; \
51851		__builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 48); \
	53333	__builtin_neon_vld3q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 48); \
51852	53334	__ret; \
51853	53335	})
51854	53336	#else
...	...	@@ -51856,14 +53338,14 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51856	53338	uint8x16x3_t __ret; \
51857	53339	uint8x16x3_t __s1 = __p1; \
51858	53340	uint8x16x3_t __rev1; \
51859		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
51860		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
51861		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
51862		__builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 48); \
	53341	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	53342	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	53343	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	53344	__builtin_neon_vld3q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 48); \
51863	53345	\
51864		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
51865		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
51866		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	53346	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	53347	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	53348	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
51867	53349	__ret; \
51868	53350	})
51869	53351	#endif
...	...	@@ -51872,7 +53354,7 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51872	53354	#define vld3q_lane_u64(__p0, __p1, __p2) __extension__ ({ \
51873	53355	uint64x2x3_t __ret; \
51874	53356	uint64x2x3_t __s1 = __p1; \
51875		__builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 51); \
	53357	__builtin_neon_vld3q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 51); \
51876	53358	__ret; \
51877	53359	})
51878	53360	#else
...	...	@@ -51880,14 +53362,14 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51880	53362	uint64x2x3_t __ret; \
51881	53363	uint64x2x3_t __s1 = __p1; \
51882	53364	uint64x2x3_t __rev1; \
51883		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
51884		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
51885		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
51886		__builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 51); \
	53365	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	53366	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	53367	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	53368	__builtin_neon_vld3q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 51); \
51887	53369	\
51888		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
51889		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
51890		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
	53370	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	53371	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	53372	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
51891	53373	__ret; \
51892	53374	})
51893	53375	#endif
...	...	@@ -51896,7 +53378,7 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51896	53378	#define vld3q_lane_s8(__p0, __p1, __p2) __extension__ ({ \
51897	53379	int8x16x3_t __ret; \
51898	53380	int8x16x3_t __s1 = __p1; \
51899		__builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 32); \
	53381	__builtin_neon_vld3q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 32); \
51900	53382	__ret; \
51901	53383	})
51902	53384	#else
...	...	@@ -51904,14 +53386,14 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51904	53386	int8x16x3_t __ret; \
51905	53387	int8x16x3_t __s1 = __p1; \
51906	53388	int8x16x3_t __rev1; \
51907		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
51908		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
51909		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
51910		__builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 32); \
	53389	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	53390	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	53391	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	53392	__builtin_neon_vld3q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 32); \
51911	53393	\
51912		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
51913		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
51914		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	53394	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	53395	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	53396	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
51915	53397	__ret; \
51916	53398	})
51917	53399	#endif
...	...	@@ -51920,7 +53402,7 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51920	53402	#define vld3q_lane_f64(__p0, __p1, __p2) __extension__ ({ \
51921	53403	float64x2x3_t __ret; \
51922	53404	float64x2x3_t __s1 = __p1; \
51923		__builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 42); \
	53405	__builtin_neon_vld3q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 42); \
51924	53406	__ret; \
51925	53407	})
51926	53408	#else
...	...	@@ -51928,14 +53410,14 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51928	53410	float64x2x3_t __ret; \
51929	53411	float64x2x3_t __s1 = __p1; \
51930	53412	float64x2x3_t __rev1; \
51931		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
51932		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
51933		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
51934		__builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 42); \
	53413	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	53414	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	53415	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	53416	__builtin_neon_vld3q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 42); \
51935	53417	\
51936		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
51937		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
51938		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
	53418	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	53419	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	53420	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
51939	53421	__ret; \
51940	53422	})
51941	53423	#endif
...	...	@@ -51944,7 +53426,7 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51944	53426	#define vld3q_lane_s64(__p0, __p1, __p2) __extension__ ({ \
51945	53427	int64x2x3_t __ret; \
51946	53428	int64x2x3_t __s1 = __p1; \
51947		__builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 35); \
	53429	__builtin_neon_vld3q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 35); \
51948	53430	__ret; \
51949	53431	})
51950	53432	#else
...	...	@@ -51952,14 +53434,38 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51952	53434	int64x2x3_t __ret; \
51953	53435	int64x2x3_t __s1 = __p1; \
51954	53436	int64x2x3_t __rev1; \
51955		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
51956		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
51957		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
51958		__builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 35); \
	53437	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	53438	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	53439	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	53440	__builtin_neon_vld3q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 35); \
51959	53441	\
51960		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
51961		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
51962		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
	53442	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	53443	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	53444	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
	53445	__ret; \
	53446	})
	53447	#endif
	53448
	53449	#ifdef __LITTLE_ENDIAN__
	53450	#define vld3q_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	53451	mfloat8x16x3_t __ret; \
	53452	mfloat8x16x3_t __s1 = __p1; \
	53453	__builtin_neon_vld3q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 44); \
	53454	__ret; \
	53455	})
	53456	#else
	53457	#define vld3q_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	53458	mfloat8x16x3_t __ret; \
	53459	mfloat8x16x3_t __s1 = __p1; \
	53460	mfloat8x16x3_t __rev1; \
	53461	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	53462	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	53463	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	53464	__builtin_neon_vld3q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 44); \
	53465	\
	53466	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	53467	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	53468	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
51963	53469	__ret; \
51964	53470	})
51965	53471	#endif
...	...	@@ -51967,21 +53473,45 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51967	53473	#define vld3_lane_u64(__p0, __p1, __p2) __extension__ ({ \
51968	53474	uint64x1x3_t __ret; \
51969	53475	uint64x1x3_t __s1 = __p1; \
51970		__builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 19); \
	53476	__builtin_neon_vld3_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 19); \
51971	53477	__ret; \
51972	53478	})
51973	53479	#define vld3_lane_f64(__p0, __p1, __p2) __extension__ ({ \
51974	53480	float64x1x3_t __ret; \
51975	53481	float64x1x3_t __s1 = __p1; \
51976		__builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 10); \
	53482	__builtin_neon_vld3_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 10); \
51977	53483	__ret; \
51978	53484	})
51979	53485	#define vld3_lane_s64(__p0, __p1, __p2) __extension__ ({ \
51980	53486	int64x1x3_t __ret; \
51981	53487	int64x1x3_t __s1 = __p1; \
51982		__builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 3); \
	53488	__builtin_neon_vld3_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 3); \
	53489	__ret; \
	53490	})
	53491	#ifdef __LITTLE_ENDIAN__
	53492	#define vld3_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	53493	mfloat8x8x3_t __ret; \
	53494	mfloat8x8x3_t __s1 = __p1; \
	53495	__builtin_neon_vld3_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 12); \
	53496	__ret; \
	53497	})
	53498	#else
	53499	#define vld3_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	53500	mfloat8x8x3_t __ret; \
	53501	mfloat8x8x3_t __s1 = __p1; \
	53502	mfloat8x8x3_t __rev1; \
	53503	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	53504	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	53505	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	53506	__builtin_neon_vld3_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __p2, 12); \
	53507	\
	53508	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	53509	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	53510	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
51983	53511	__ret; \
51984	53512	})
	53513	#endif
	53514
51985	53515	#define vld4_p64(__p0) __extension__ ({ \
51986	53516	poly64x1x4_t __ret; \
51987	53517	__builtin_neon_vld4_v(&__ret, __p0, 6); \
...	...	@@ -51998,10 +53528,10 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
51998	53528	poly64x2x4_t __ret; \
51999	53529	__builtin_neon_vld4q_v(&__ret, __p0, 38); \
52000	53530	\
52001		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
52002		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
52003		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
52004		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
	53531	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	53532	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	53533	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
	53534	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_64); \
52005	53535	__ret; \
52006	53536	})
52007	53537	#endif
...	...	@@ -52017,10 +53547,10 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
52017	53547	uint64x2x4_t __ret; \
52018	53548	__builtin_neon_vld4q_v(&__ret, __p0, 51); \
52019	53549	\
52020		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
52021		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
52022		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
52023		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
	53550	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	53551	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	53552	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
	53553	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_64); \
52024	53554	__ret; \
52025	53555	})
52026	53556	#endif
...	...	@@ -52036,10 +53566,10 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
52036	53566	float64x2x4_t __ret; \
52037	53567	__builtin_neon_vld4q_v(&__ret, __p0, 42); \
52038	53568	\
52039		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
52040		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
52041		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
52042		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
	53569	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	53570	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	53571	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
	53572	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_64); \
52043	53573	__ret; \
52044	53574	})
52045	53575	#endif
...	...	@@ -52055,10 +53585,29 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
52055	53585	int64x2x4_t __ret; \
52056	53586	__builtin_neon_vld4q_v(&__ret, __p0, 35); \
52057	53587	\
52058		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
52059		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
52060		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
52061		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
	53588	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	53589	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	53590	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
	53591	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_64); \
	53592	__ret; \
	53593	})
	53594	#endif
	53595
	53596	#ifdef __LITTLE_ENDIAN__
	53597	#define vld4q_mf8(__p0) __extension__ ({ \
	53598	mfloat8x16x4_t __ret; \
	53599	__builtin_neon_vld4q_v(&__ret, __p0, 44); \
	53600	__ret; \
	53601	})
	53602	#else
	53603	#define vld4q_mf8(__p0) __extension__ ({ \
	53604	mfloat8x16x4_t __ret; \
	53605	__builtin_neon_vld4q_v(&__ret, __p0, 44); \
	53606	\
	53607	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	53608	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	53609	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
	53610	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_8); \
52062	53611	__ret; \
52063	53612	})
52064	53613	#endif
...	...	@@ -52068,6 +53617,25 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
52068	53617	__builtin_neon_vld4_v(&__ret, __p0, 10); \
52069	53618	__ret; \
52070	53619	})
	53620	#ifdef __LITTLE_ENDIAN__
	53621	#define vld4_mf8(__p0) __extension__ ({ \
	53622	mfloat8x8x4_t __ret; \
	53623	__builtin_neon_vld4_v(&__ret, __p0, 12); \
	53624	__ret; \
	53625	})
	53626	#else
	53627	#define vld4_mf8(__p0) __extension__ ({ \
	53628	mfloat8x8x4_t __ret; \
	53629	__builtin_neon_vld4_v(&__ret, __p0, 12); \
	53630	\
	53631	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	53632	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	53633	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
	53634	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_8); \
	53635	__ret; \
	53636	})
	53637	#endif
	53638
52071	53639	#define vld4_dup_p64(__p0) __extension__ ({ \
52072	53640	poly64x1x4_t __ret; \
52073	53641	__builtin_neon_vld4_dup_v(&__ret, __p0, 6); \
...	...	@@ -52084,10 +53652,10 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
52084	53652	poly64x2x4_t __ret; \
52085	53653	__builtin_neon_vld4q_dup_v(&__ret, __p0, 38); \
52086	53654	\
52087		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
52088		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
52089		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
52090		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
	53655	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	53656	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	53657	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
	53658	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_64); \
52091	53659	__ret; \
52092	53660	})
52093	53661	#endif
...	...	@@ -52103,10 +53671,29 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
52103	53671	float64x2x4_t __ret; \
52104	53672	__builtin_neon_vld4q_dup_v(&__ret, __p0, 42); \
52105	53673	\
52106		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
52107		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
52108		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
52109		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
	53674	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	53675	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	53676	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
	53677	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_64); \
	53678	__ret; \
	53679	})
	53680	#endif
	53681
	53682	#ifdef __LITTLE_ENDIAN__
	53683	#define vld4q_dup_mf8(__p0) __extension__ ({ \
	53684	mfloat8x16x4_t __ret; \
	53685	__builtin_neon_vld4q_dup_v(&__ret, __p0, 44); \
	53686	__ret; \
	53687	})
	53688	#else
	53689	#define vld4q_dup_mf8(__p0) __extension__ ({ \
	53690	mfloat8x16x4_t __ret; \
	53691	__builtin_neon_vld4q_dup_v(&__ret, __p0, 44); \
	53692	\
	53693	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	53694	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	53695	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
	53696	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_8); \
52110	53697	__ret; \
52111	53698	})
52112	53699	#endif
...	...	@@ -52116,17 +53703,36 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
52116	53703	__builtin_neon_vld4_dup_v(&__ret, __p0, 10); \
52117	53704	__ret; \
52118	53705	})
	53706	#ifdef __LITTLE_ENDIAN__
	53707	#define vld4_dup_mf8(__p0) __extension__ ({ \
	53708	mfloat8x8x4_t __ret; \
	53709	__builtin_neon_vld4_dup_v(&__ret, __p0, 12); \
	53710	__ret; \
	53711	})
	53712	#else
	53713	#define vld4_dup_mf8(__p0) __extension__ ({ \
	53714	mfloat8x8x4_t __ret; \
	53715	__builtin_neon_vld4_dup_v(&__ret, __p0, 12); \
	53716	\
	53717	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	53718	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	53719	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
	53720	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_8); \
	53721	__ret; \
	53722	})
	53723	#endif
	53724
52119	53725	#define vld4_lane_p64(__p0, __p1, __p2) __extension__ ({ \
52120	53726	poly64x1x4_t __ret; \
52121	53727	poly64x1x4_t __s1 = __p1; \
52122		__builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 6); \
	53728	__builtin_neon_vld4_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 6); \
52123	53729	__ret; \
52124	53730	})
52125	53731	#ifdef __LITTLE_ENDIAN__
52126	53732	#define vld4q_lane_p8(__p0, __p1, __p2) __extension__ ({ \
52127	53733	poly8x16x4_t __ret; \
52128	53734	poly8x16x4_t __s1 = __p1; \
52129		__builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 36); \
	53735	__builtin_neon_vld4q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 36); \
52130	53736	__ret; \
52131	53737	})
52132	53738	#else
...	...	@@ -52134,16 +53740,16 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
52134	53740	poly8x16x4_t __ret; \
52135	53741	poly8x16x4_t __s1 = __p1; \
52136	53742	poly8x16x4_t __rev1; \
52137		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
52138		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
52139		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
52140		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
52141		__builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 36); \
	53743	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	53744	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	53745	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	53746	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_8); \
	53747	__builtin_neon_vld4q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 36); \
52142	53748	\
52143		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
52144		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
52145		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
52146		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	53749	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	53750	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	53751	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
	53752	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_8); \
52147	53753	__ret; \
52148	53754	})
52149	53755	#endif
...	...	@@ -52152,7 +53758,7 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
52152	53758	#define vld4q_lane_p64(__p0, __p1, __p2) __extension__ ({ \
52153	53759	poly64x2x4_t __ret; \
52154	53760	poly64x2x4_t __s1 = __p1; \
52155		__builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 38); \
	53761	__builtin_neon_vld4q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 38); \
52156	53762	__ret; \
52157	53763	})
52158	53764	#else
...	...	@@ -52160,16 +53766,16 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
52160	53766	poly64x2x4_t __ret; \
52161	53767	poly64x2x4_t __s1 = __p1; \
52162	53768	poly64x2x4_t __rev1; \
52163		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
52164		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
52165		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
52166		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
52167		__builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 38); \
	53769	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	53770	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	53771	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	53772	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_64); \
	53773	__builtin_neon_vld4q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 38); \
52168	53774	\
52169		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
52170		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
52171		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
52172		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
	53775	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	53776	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	53777	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
	53778	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_64); \
52173	53779	__ret; \
52174	53780	})
52175	53781	#endif
...	...	@@ -52178,7 +53784,7 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
52178	53784	#define vld4q_lane_u8(__p0, __p1, __p2) __extension__ ({ \
52179	53785	uint8x16x4_t __ret; \
52180	53786	uint8x16x4_t __s1 = __p1; \
52181		__builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 48); \
	53787	__builtin_neon_vld4q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 48); \
52182	53788	__ret; \
52183	53789	})
52184	53790	#else
...	...	@@ -52186,16 +53792,16 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
52186	53792	uint8x16x4_t __ret; \
52187	53793	uint8x16x4_t __s1 = __p1; \
52188	53794	uint8x16x4_t __rev1; \
52189		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
52190		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
52191		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
52192		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
52193		__builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 48); \
	53795	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	53796	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	53797	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	53798	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_8); \
	53799	__builtin_neon_vld4q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 48); \
52194	53800	\
52195		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
52196		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
52197		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
52198		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	53801	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	53802	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	53803	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
	53804	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_8); \
52199	53805	__ret; \
52200	53806	})
52201	53807	#endif
...	...	@@ -52204,7 +53810,7 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
52204	53810	#define vld4q_lane_u64(__p0, __p1, __p2) __extension__ ({ \
52205	53811	uint64x2x4_t __ret; \
52206	53812	uint64x2x4_t __s1 = __p1; \
52207		__builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 51); \
	53813	__builtin_neon_vld4q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 51); \
52208	53814	__ret; \
52209	53815	})
52210	53816	#else
...	...	@@ -52212,16 +53818,16 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
52212	53818	uint64x2x4_t __ret; \
52213	53819	uint64x2x4_t __s1 = __p1; \
52214	53820	uint64x2x4_t __rev1; \
52215		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
52216		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
52217		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
52218		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
52219		__builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 51); \
	53821	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	53822	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	53823	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	53824	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_64); \
	53825	__builtin_neon_vld4q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 51); \
52220	53826	\
52221		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
52222		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
52223		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
52224		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
	53827	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	53828	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	53829	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
	53830	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_64); \
52225	53831	__ret; \
52226	53832	})
52227	53833	#endif
...	...	@@ -52230,7 +53836,7 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
52230	53836	#define vld4q_lane_s8(__p0, __p1, __p2) __extension__ ({ \
52231	53837	int8x16x4_t __ret; \
52232	53838	int8x16x4_t __s1 = __p1; \
52233		__builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 32); \
	53839	__builtin_neon_vld4q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 32); \
52234	53840	__ret; \
52235	53841	})
52236	53842	#else
...	...	@@ -52238,16 +53844,16 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
52238	53844	int8x16x4_t __ret; \
52239	53845	int8x16x4_t __s1 = __p1; \
52240	53846	int8x16x4_t __rev1; \
52241		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
52242		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
52243		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
52244		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
52245		__builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 32); \
	53847	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	53848	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	53849	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	53850	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_8); \
	53851	__builtin_neon_vld4q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 32); \
52246	53852	\
52247		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
52248		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
52249		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
52250		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
	53853	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	53854	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	53855	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
	53856	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_8); \
52251	53857	__ret; \
52252	53858	})
52253	53859	#endif
...	...	@@ -52256,7 +53862,7 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
52256	53862	#define vld4q_lane_f64(__p0, __p1, __p2) __extension__ ({ \
52257	53863	float64x2x4_t __ret; \
52258	53864	float64x2x4_t __s1 = __p1; \
52259		__builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 42); \
	53865	__builtin_neon_vld4q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 42); \
52260	53866	__ret; \
52261	53867	})
52262	53868	#else
...	...	@@ -52264,16 +53870,16 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
52264	53870	float64x2x4_t __ret; \
52265	53871	float64x2x4_t __s1 = __p1; \
52266	53872	float64x2x4_t __rev1; \
52267		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
52268		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
52269		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
52270		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
52271		__builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 42); \
	53873	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	53874	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	53875	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	53876	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_64); \
	53877	__builtin_neon_vld4q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 42); \
52272	53878	\
52273		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
52274		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
52275		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
52276		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
	53879	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	53880	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	53881	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
	53882	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_64); \
52277	53883	__ret; \
52278	53884	})
52279	53885	#endif
...	...	@@ -52282,7 +53888,7 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
52282	53888	#define vld4q_lane_s64(__p0, __p1, __p2) __extension__ ({ \
52283	53889	int64x2x4_t __ret; \
52284	53890	int64x2x4_t __s1 = __p1; \
52285		__builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 35); \
	53891	__builtin_neon_vld4q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 35); \
52286	53892	__ret; \
52287	53893	})
52288	53894	#else
...	...	@@ -52290,16 +53896,42 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
52290	53896	int64x2x4_t __ret; \
52291	53897	int64x2x4_t __s1 = __p1; \
52292	53898	int64x2x4_t __rev1; \
52293		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
52294		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
52295		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
52296		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
52297		__builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 35); \
	53899	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	53900	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	53901	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	53902	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_64); \
	53903	__builtin_neon_vld4q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 35); \
52298	53904	\
52299		__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
52300		__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
52301		__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
52302		__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
	53905	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_64); \
	53906	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_64); \
	53907	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_64); \
	53908	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_64); \
	53909	__ret; \
	53910	})
	53911	#endif
	53912
	53913	#ifdef __LITTLE_ENDIAN__
	53914	#define vld4q_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	53915	mfloat8x16x4_t __ret; \
	53916	mfloat8x16x4_t __s1 = __p1; \
	53917	__builtin_neon_vld4q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 44); \
	53918	__ret; \
	53919	})
	53920	#else
	53921	#define vld4q_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	53922	mfloat8x16x4_t __ret; \
	53923	mfloat8x16x4_t __s1 = __p1; \
	53924	mfloat8x16x4_t __rev1; \
	53925	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	53926	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	53927	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	53928	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_8); \
	53929	__builtin_neon_vld4q_lane_v(&__ret, __p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 44); \
	53930	\
	53931	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_128_8); \
	53932	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_128_8); \
	53933	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_128_8); \
	53934	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_128_8); \
52303	53935	__ret; \
52304	53936	})
52305	53937	#endif
...	...	@@ -52307,59 +53939,85 @@ __ai __attribute__((target("neon"))) float64x1_t vget_low_f64(float64x2_t __p0)
52307	53939	#define vld4_lane_u64(__p0, __p1, __p2) __extension__ ({ \
52308	53940	uint64x1x4_t __ret; \
52309	53941	uint64x1x4_t __s1 = __p1; \
52310		__builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 19); \
	53942	__builtin_neon_vld4_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 19); \
52311	53943	__ret; \
52312	53944	})
52313	53945	#define vld4_lane_f64(__p0, __p1, __p2) __extension__ ({ \
52314	53946	float64x1x4_t __ret; \
52315	53947	float64x1x4_t __s1 = __p1; \
52316		__builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 10); \
	53948	__builtin_neon_vld4_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 10); \
52317	53949	__ret; \
52318	53950	})
52319	53951	#define vld4_lane_s64(__p0, __p1, __p2) __extension__ ({ \
52320	53952	int64x1x4_t __ret; \
52321	53953	int64x1x4_t __s1 = __p1; \
52322		__builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 3); \
	53954	__builtin_neon_vld4_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 3); \
	53955	__ret; \
	53956	})
	53957	#ifdef __LITTLE_ENDIAN__
	53958	#define vld4_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	53959	mfloat8x8x4_t __ret; \
	53960	mfloat8x8x4_t __s1 = __p1; \
	53961	__builtin_neon_vld4_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 12); \
52323	53962	__ret; \
52324	53963	})
	53964	#else
	53965	#define vld4_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	53966	mfloat8x8x4_t __ret; \
	53967	mfloat8x8x4_t __s1 = __p1; \
	53968	mfloat8x8x4_t __rev1; \
	53969	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	53970	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	53971	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	53972	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_8); \
	53973	__builtin_neon_vld4_lane_v(&__ret, __p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), __p2, 12); \
	53974	\
	53975	__ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], __lane_reverse_64_8); \
	53976	__ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], __lane_reverse_64_8); \
	53977	__ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], __lane_reverse_64_8); \
	53978	__ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], __lane_reverse_64_8); \
	53979	__ret; \
	53980	})
	53981	#endif
	53982
52325	53983	#define vldrq_p128(__p0) __extension__ ({ \
52326	53984	poly128_t __ret; \
52327		__ret = (poly128_t) __builtin_neon_vldrq_p128(__p0); \
	53985	__ret = __builtin_bit_cast(poly128_t, __builtin_neon_vldrq_p128(__p0)); \
52328	53986	__ret; \
52329	53987	})
52330	53988	#ifdef __LITTLE_ENDIAN__
52331	53989	__ai __attribute__((target("neon"))) float64x2_t vmaxq_f64(float64x2_t __p0, float64x2_t __p1) {
52332	53990	float64x2_t __ret;
52333		__ret = (float64x2_t) __builtin_neon_vmaxq_v((int8x16_t)__p0, (int8x16_t)__p1, 42);
	53991	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vmaxq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 42));
52334	53992	return __ret;
52335	53993	}
52336	53994	#else
52337	53995	__ai __attribute__((target("neon"))) float64x2_t vmaxq_f64(float64x2_t __p0, float64x2_t __p1) {
52338	53996	float64x2_t __ret;
52339		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
52340		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
52341		__ret = (float64x2_t) __builtin_neon_vmaxq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 42);
52342		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	53997	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	53998	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	53999	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vmaxq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 42));
	54000	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
52343	54001	return __ret;
52344	54002	}
52345	54003	#endif
52346	54004
52347	54005	__ai __attribute__((target("neon"))) float64x1_t vmax_f64(float64x1_t __p0, float64x1_t __p1) {
52348	54006	float64x1_t __ret;
52349		__ret = (float64x1_t) __builtin_neon_vmax_v((int8x8_t)__p0, (int8x8_t)__p1, 10);
	54007	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vmax_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 10));
52350	54008	return __ret;
52351	54009	}
52352	54010	#ifdef __LITTLE_ENDIAN__
52353	54011	__ai __attribute__((target("neon"))) float64_t vmaxnmvq_f64(float64x2_t __p0) {
52354	54012	float64_t __ret;
52355		__ret = (float64_t) __builtin_neon_vmaxnmvq_f64(__p0);
	54013	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vmaxnmvq_f64(__p0));
52356	54014	return __ret;
52357	54015	}
52358	54016	#else
52359	54017	__ai __attribute__((target("neon"))) float64_t vmaxnmvq_f64(float64x2_t __p0) {
52360	54018	float64_t __ret;
52361		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
52362		__ret = (float64_t) __builtin_neon_vmaxnmvq_f64(__rev0);
	54019	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	54020	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vmaxnmvq_f64(__rev0));
52363	54021	return __ret;
52364	54022	}
52365	54023	#endif
...	...	@@ -52367,14 +54025,14 @@ __ai __attribute__((target("neon"))) float64_t vmaxnmvq_f64(float64x2_t __p0) {
52367	54025	#ifdef __LITTLE_ENDIAN__
52368	54026	__ai __attribute__((target("neon"))) float32_t vmaxnmvq_f32(float32x4_t __p0) {
52369	54027	float32_t __ret;
52370		__ret = (float32_t) __builtin_neon_vmaxnmvq_f32(__p0);
	54028	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vmaxnmvq_f32(__p0));
52371	54029	return __ret;
52372	54030	}
52373	54031	#else
52374	54032	__ai __attribute__((target("neon"))) float32_t vmaxnmvq_f32(float32x4_t __p0) {
52375	54033	float32_t __ret;
52376		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
52377		__ret = (float32_t) __builtin_neon_vmaxnmvq_f32(__rev0);
	54034	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	54035	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vmaxnmvq_f32(__rev0));
52378	54036	return __ret;
52379	54037	}
52380	54038	#endif
...	...	@@ -52382,14 +54040,14 @@ __ai __attribute__((target("neon"))) float32_t vmaxnmvq_f32(float32x4_t __p0) {
52382	54040	#ifdef __LITTLE_ENDIAN__
52383	54041	__ai __attribute__((target("neon"))) float32_t vmaxnmv_f32(float32x2_t __p0) {
52384	54042	float32_t __ret;
52385		__ret = (float32_t) __builtin_neon_vmaxnmv_f32(__p0);
	54043	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vmaxnmv_f32(__p0));
52386	54044	return __ret;
52387	54045	}
52388	54046	#else
52389	54047	__ai __attribute__((target("neon"))) float32_t vmaxnmv_f32(float32x2_t __p0) {
52390	54048	float32_t __ret;
52391		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
52392		__ret = (float32_t) __builtin_neon_vmaxnmv_f32(__rev0);
	54049	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	54050	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vmaxnmv_f32(__rev0));
52393	54051	return __ret;
52394	54052	}
52395	54053	#endif
...	...	@@ -52397,14 +54055,14 @@ __ai __attribute__((target("neon"))) float32_t vmaxnmv_f32(float32x2_t __p0) {
52397	54055	#ifdef __LITTLE_ENDIAN__
52398	54056	__ai __attribute__((target("neon"))) uint8_t vmaxvq_u8(uint8x16_t __p0) {
52399	54057	uint8_t __ret;
52400		__ret = (uint8_t) __builtin_neon_vmaxvq_u8(__p0);
	54058	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vmaxvq_u8(__p0));
52401	54059	return __ret;
52402	54060	}
52403	54061	#else
52404	54062	__ai __attribute__((target("neon"))) uint8_t vmaxvq_u8(uint8x16_t __p0) {
52405	54063	uint8_t __ret;
52406		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
52407		__ret = (uint8_t) __builtin_neon_vmaxvq_u8(__rev0);
	54064	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	54065	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vmaxvq_u8(__rev0));
52408	54066	return __ret;
52409	54067	}
52410	54068	#endif
...	...	@@ -52412,14 +54070,14 @@ __ai __attribute__((target("neon"))) uint8_t vmaxvq_u8(uint8x16_t __p0) {
52412	54070	#ifdef __LITTLE_ENDIAN__
52413	54071	__ai __attribute__((target("neon"))) uint32_t vmaxvq_u32(uint32x4_t __p0) {
52414	54072	uint32_t __ret;
52415		__ret = (uint32_t) __builtin_neon_vmaxvq_u32(__p0);
	54073	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vmaxvq_u32(__p0));
52416	54074	return __ret;
52417	54075	}
52418	54076	#else
52419	54077	__ai __attribute__((target("neon"))) uint32_t vmaxvq_u32(uint32x4_t __p0) {
52420	54078	uint32_t __ret;
52421		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
52422		__ret = (uint32_t) __builtin_neon_vmaxvq_u32(__rev0);
	54079	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	54080	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vmaxvq_u32(__rev0));
52423	54081	return __ret;
52424	54082	}
52425	54083	#endif
...	...	@@ -52427,14 +54085,14 @@ __ai __attribute__((target("neon"))) uint32_t vmaxvq_u32(uint32x4_t __p0) {
52427	54085	#ifdef __LITTLE_ENDIAN__
52428	54086	__ai __attribute__((target("neon"))) uint16_t vmaxvq_u16(uint16x8_t __p0) {
52429	54087	uint16_t __ret;
52430		__ret = (uint16_t) __builtin_neon_vmaxvq_u16(__p0);
	54088	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vmaxvq_u16(__p0));
52431	54089	return __ret;
52432	54090	}
52433	54091	#else
52434	54092	__ai __attribute__((target("neon"))) uint16_t vmaxvq_u16(uint16x8_t __p0) {
52435	54093	uint16_t __ret;
52436		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
52437		__ret = (uint16_t) __builtin_neon_vmaxvq_u16(__rev0);
	54094	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	54095	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vmaxvq_u16(__rev0));
52438	54096	return __ret;
52439	54097	}
52440	54098	#endif
...	...	@@ -52442,14 +54100,14 @@ __ai __attribute__((target("neon"))) uint16_t vmaxvq_u16(uint16x8_t __p0) {
52442	54100	#ifdef __LITTLE_ENDIAN__
52443	54101	__ai __attribute__((target("neon"))) int8_t vmaxvq_s8(int8x16_t __p0) {
52444	54102	int8_t __ret;
52445		__ret = (int8_t) __builtin_neon_vmaxvq_s8(__p0);
	54103	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vmaxvq_s8(__p0));
52446	54104	return __ret;
52447	54105	}
52448	54106	#else
52449	54107	__ai __attribute__((target("neon"))) int8_t vmaxvq_s8(int8x16_t __p0) {
52450	54108	int8_t __ret;
52451		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
52452		__ret = (int8_t) __builtin_neon_vmaxvq_s8(__rev0);
	54109	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	54110	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vmaxvq_s8(__rev0));
52453	54111	return __ret;
52454	54112	}
52455	54113	#endif
...	...	@@ -52457,14 +54115,14 @@ __ai __attribute__((target("neon"))) int8_t vmaxvq_s8(int8x16_t __p0) {
52457	54115	#ifdef __LITTLE_ENDIAN__
52458	54116	__ai __attribute__((target("neon"))) float64_t vmaxvq_f64(float64x2_t __p0) {
52459	54117	float64_t __ret;
52460		__ret = (float64_t) __builtin_neon_vmaxvq_f64(__p0);
	54118	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vmaxvq_f64(__p0));
52461	54119	return __ret;
52462	54120	}
52463	54121	#else
52464	54122	__ai __attribute__((target("neon"))) float64_t vmaxvq_f64(float64x2_t __p0) {
52465	54123	float64_t __ret;
52466		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
52467		__ret = (float64_t) __builtin_neon_vmaxvq_f64(__rev0);
	54124	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	54125	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vmaxvq_f64(__rev0));
52468	54126	return __ret;
52469	54127	}
52470	54128	#endif
...	...	@@ -52472,14 +54130,14 @@ __ai __attribute__((target("neon"))) float64_t vmaxvq_f64(float64x2_t __p0) {
52472	54130	#ifdef __LITTLE_ENDIAN__
52473	54131	__ai __attribute__((target("neon"))) float32_t vmaxvq_f32(float32x4_t __p0) {
52474	54132	float32_t __ret;
52475		__ret = (float32_t) __builtin_neon_vmaxvq_f32(__p0);
	54133	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vmaxvq_f32(__p0));
52476	54134	return __ret;
52477	54135	}
52478	54136	#else
52479	54137	__ai __attribute__((target("neon"))) float32_t vmaxvq_f32(float32x4_t __p0) {
52480	54138	float32_t __ret;
52481		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
52482		__ret = (float32_t) __builtin_neon_vmaxvq_f32(__rev0);
	54139	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	54140	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vmaxvq_f32(__rev0));
52483	54141	return __ret;
52484	54142	}
52485	54143	#endif
...	...	@@ -52487,14 +54145,14 @@ __ai __attribute__((target("neon"))) float32_t vmaxvq_f32(float32x4_t __p0) {
52487	54145	#ifdef __LITTLE_ENDIAN__
52488	54146	__ai __attribute__((target("neon"))) int32_t vmaxvq_s32(int32x4_t __p0) {
52489	54147	int32_t __ret;
52490		__ret = (int32_t) __builtin_neon_vmaxvq_s32(__p0);
	54148	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vmaxvq_s32(__p0));
52491	54149	return __ret;
52492	54150	}
52493	54151	#else
52494	54152	__ai __attribute__((target("neon"))) int32_t vmaxvq_s32(int32x4_t __p0) {
52495	54153	int32_t __ret;
52496		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
52497		__ret = (int32_t) __builtin_neon_vmaxvq_s32(__rev0);
	54154	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	54155	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vmaxvq_s32(__rev0));
52498	54156	return __ret;
52499	54157	}
52500	54158	#endif
...	...	@@ -52502,14 +54160,14 @@ __ai __attribute__((target("neon"))) int32_t vmaxvq_s32(int32x4_t __p0) {
52502	54160	#ifdef __LITTLE_ENDIAN__
52503	54161	__ai __attribute__((target("neon"))) int16_t vmaxvq_s16(int16x8_t __p0) {
52504	54162	int16_t __ret;
52505		__ret = (int16_t) __builtin_neon_vmaxvq_s16(__p0);
	54163	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vmaxvq_s16(__p0));
52506	54164	return __ret;
52507	54165	}
52508	54166	#else
52509	54167	__ai __attribute__((target("neon"))) int16_t vmaxvq_s16(int16x8_t __p0) {
52510	54168	int16_t __ret;
52511		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
52512		__ret = (int16_t) __builtin_neon_vmaxvq_s16(__rev0);
	54169	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	54170	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vmaxvq_s16(__rev0));
52513	54171	return __ret;
52514	54172	}
52515	54173	#endif
...	...	@@ -52517,14 +54175,14 @@ __ai __attribute__((target("neon"))) int16_t vmaxvq_s16(int16x8_t __p0) {
52517	54175	#ifdef __LITTLE_ENDIAN__
52518	54176	__ai __attribute__((target("neon"))) uint8_t vmaxv_u8(uint8x8_t __p0) {
52519	54177	uint8_t __ret;
52520		__ret = (uint8_t) __builtin_neon_vmaxv_u8(__p0);
	54178	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vmaxv_u8(__p0));
52521	54179	return __ret;
52522	54180	}
52523	54181	#else
52524	54182	__ai __attribute__((target("neon"))) uint8_t vmaxv_u8(uint8x8_t __p0) {
52525	54183	uint8_t __ret;
52526		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
52527		__ret = (uint8_t) __builtin_neon_vmaxv_u8(__rev0);
	54184	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	54185	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vmaxv_u8(__rev0));
52528	54186	return __ret;
52529	54187	}
52530	54188	#endif
...	...	@@ -52532,14 +54190,14 @@ __ai __attribute__((target("neon"))) uint8_t vmaxv_u8(uint8x8_t __p0) {
52532	54190	#ifdef __LITTLE_ENDIAN__
52533	54191	__ai __attribute__((target("neon"))) uint32_t vmaxv_u32(uint32x2_t __p0) {
52534	54192	uint32_t __ret;
52535		__ret = (uint32_t) __builtin_neon_vmaxv_u32(__p0);
	54193	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vmaxv_u32(__p0));
52536	54194	return __ret;
52537	54195	}
52538	54196	#else
52539	54197	__ai __attribute__((target("neon"))) uint32_t vmaxv_u32(uint32x2_t __p0) {
52540	54198	uint32_t __ret;
52541		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
52542		__ret = (uint32_t) __builtin_neon_vmaxv_u32(__rev0);
	54199	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	54200	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vmaxv_u32(__rev0));
52543	54201	return __ret;
52544	54202	}
52545	54203	#endif
...	...	@@ -52547,14 +54205,14 @@ __ai __attribute__((target("neon"))) uint32_t vmaxv_u32(uint32x2_t __p0) {
52547	54205	#ifdef __LITTLE_ENDIAN__
52548	54206	__ai __attribute__((target("neon"))) uint16_t vmaxv_u16(uint16x4_t __p0) {
52549	54207	uint16_t __ret;
52550		__ret = (uint16_t) __builtin_neon_vmaxv_u16(__p0);
	54208	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vmaxv_u16(__p0));
52551	54209	return __ret;
52552	54210	}
52553	54211	#else
52554	54212	__ai __attribute__((target("neon"))) uint16_t vmaxv_u16(uint16x4_t __p0) {
52555	54213	uint16_t __ret;
52556		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
52557		__ret = (uint16_t) __builtin_neon_vmaxv_u16(__rev0);
	54214	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	54215	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vmaxv_u16(__rev0));
52558	54216	return __ret;
52559	54217	}
52560	54218	#endif
...	...	@@ -52562,14 +54220,14 @@ __ai __attribute__((target("neon"))) uint16_t vmaxv_u16(uint16x4_t __p0) {
52562	54220	#ifdef __LITTLE_ENDIAN__
52563	54221	__ai __attribute__((target("neon"))) int8_t vmaxv_s8(int8x8_t __p0) {
52564	54222	int8_t __ret;
52565		__ret = (int8_t) __builtin_neon_vmaxv_s8(__p0);
	54223	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vmaxv_s8(__p0));
52566	54224	return __ret;
52567	54225	}
52568	54226	#else
52569	54227	__ai __attribute__((target("neon"))) int8_t vmaxv_s8(int8x8_t __p0) {
52570	54228	int8_t __ret;
52571		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
52572		__ret = (int8_t) __builtin_neon_vmaxv_s8(__rev0);
	54229	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	54230	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vmaxv_s8(__rev0));
52573	54231	return __ret;
52574	54232	}
52575	54233	#endif
...	...	@@ -52577,14 +54235,14 @@ __ai __attribute__((target("neon"))) int8_t vmaxv_s8(int8x8_t __p0) {
52577	54235	#ifdef __LITTLE_ENDIAN__
52578	54236	__ai __attribute__((target("neon"))) float32_t vmaxv_f32(float32x2_t __p0) {
52579	54237	float32_t __ret;
52580		__ret = (float32_t) __builtin_neon_vmaxv_f32(__p0);
	54238	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vmaxv_f32(__p0));
52581	54239	return __ret;
52582	54240	}
52583	54241	#else
52584	54242	__ai __attribute__((target("neon"))) float32_t vmaxv_f32(float32x2_t __p0) {
52585	54243	float32_t __ret;
52586		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
52587		__ret = (float32_t) __builtin_neon_vmaxv_f32(__rev0);
	54244	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	54245	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vmaxv_f32(__rev0));
52588	54246	return __ret;
52589	54247	}
52590	54248	#endif
...	...	@@ -52592,14 +54250,14 @@ __ai __attribute__((target("neon"))) float32_t vmaxv_f32(float32x2_t __p0) {
52592	54250	#ifdef __LITTLE_ENDIAN__
52593	54251	__ai __attribute__((target("neon"))) int32_t vmaxv_s32(int32x2_t __p0) {
52594	54252	int32_t __ret;
52595		__ret = (int32_t) __builtin_neon_vmaxv_s32(__p0);
	54253	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vmaxv_s32(__p0));
52596	54254	return __ret;
52597	54255	}
52598	54256	#else
52599	54257	__ai __attribute__((target("neon"))) int32_t vmaxv_s32(int32x2_t __p0) {
52600	54258	int32_t __ret;
52601		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
52602		__ret = (int32_t) __builtin_neon_vmaxv_s32(__rev0);
	54259	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	54260	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vmaxv_s32(__rev0));
52603	54261	return __ret;
52604	54262	}
52605	54263	#endif
...	...	@@ -52607,14 +54265,14 @@ __ai __attribute__((target("neon"))) int32_t vmaxv_s32(int32x2_t __p0) {
52607	54265	#ifdef __LITTLE_ENDIAN__
52608	54266	__ai __attribute__((target("neon"))) int16_t vmaxv_s16(int16x4_t __p0) {
52609	54267	int16_t __ret;
52610		__ret = (int16_t) __builtin_neon_vmaxv_s16(__p0);
	54268	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vmaxv_s16(__p0));
52611	54269	return __ret;
52612	54270	}
52613	54271	#else
52614	54272	__ai __attribute__((target("neon"))) int16_t vmaxv_s16(int16x4_t __p0) {
52615	54273	int16_t __ret;
52616		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
52617		__ret = (int16_t) __builtin_neon_vmaxv_s16(__rev0);
	54274	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	54275	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vmaxv_s16(__rev0));
52618	54276	return __ret;
52619	54277	}
52620	54278	#endif
...	...	@@ -52622,36 +54280,36 @@ __ai __attribute__((target("neon"))) int16_t vmaxv_s16(int16x4_t __p0) {
52622	54280	#ifdef __LITTLE_ENDIAN__
52623	54281	__ai __attribute__((target("neon"))) float64x2_t vminq_f64(float64x2_t __p0, float64x2_t __p1) {
52624	54282	float64x2_t __ret;
52625		__ret = (float64x2_t) __builtin_neon_vminq_v((int8x16_t)__p0, (int8x16_t)__p1, 42);
	54283	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vminq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 42));
52626	54284	return __ret;
52627	54285	}
52628	54286	#else
52629	54287	__ai __attribute__((target("neon"))) float64x2_t vminq_f64(float64x2_t __p0, float64x2_t __p1) {
52630	54288	float64x2_t __ret;
52631		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
52632		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
52633		__ret = (float64x2_t) __builtin_neon_vminq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 42);
52634		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	54289	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	54290	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	54291	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vminq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 42));
	54292	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
52635	54293	return __ret;
52636	54294	}
52637	54295	#endif
52638	54296
52639	54297	__ai __attribute__((target("neon"))) float64x1_t vmin_f64(float64x1_t __p0, float64x1_t __p1) {
52640	54298	float64x1_t __ret;
52641		__ret = (float64x1_t) __builtin_neon_vmin_v((int8x8_t)__p0, (int8x8_t)__p1, 10);
	54299	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vmin_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 10));
52642	54300	return __ret;
52643	54301	}
52644	54302	#ifdef __LITTLE_ENDIAN__
52645	54303	__ai __attribute__((target("neon"))) float64_t vminnmvq_f64(float64x2_t __p0) {
52646	54304	float64_t __ret;
52647		__ret = (float64_t) __builtin_neon_vminnmvq_f64(__p0);
	54305	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vminnmvq_f64(__p0));
52648	54306	return __ret;
52649	54307	}
52650	54308	#else
52651	54309	__ai __attribute__((target("neon"))) float64_t vminnmvq_f64(float64x2_t __p0) {
52652	54310	float64_t __ret;
52653		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
52654		__ret = (float64_t) __builtin_neon_vminnmvq_f64(__rev0);
	54311	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	54312	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vminnmvq_f64(__rev0));
52655	54313	return __ret;
52656	54314	}
52657	54315	#endif
...	...	@@ -52659,14 +54317,14 @@ __ai __attribute__((target("neon"))) float64_t vminnmvq_f64(float64x2_t __p0) {
52659	54317	#ifdef __LITTLE_ENDIAN__
52660	54318	__ai __attribute__((target("neon"))) float32_t vminnmvq_f32(float32x4_t __p0) {
52661	54319	float32_t __ret;
52662		__ret = (float32_t) __builtin_neon_vminnmvq_f32(__p0);
	54320	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vminnmvq_f32(__p0));
52663	54321	return __ret;
52664	54322	}
52665	54323	#else
52666	54324	__ai __attribute__((target("neon"))) float32_t vminnmvq_f32(float32x4_t __p0) {
52667	54325	float32_t __ret;
52668		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
52669		__ret = (float32_t) __builtin_neon_vminnmvq_f32(__rev0);
	54326	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	54327	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vminnmvq_f32(__rev0));
52670	54328	return __ret;
52671	54329	}
52672	54330	#endif
...	...	@@ -52674,14 +54332,14 @@ __ai __attribute__((target("neon"))) float32_t vminnmvq_f32(float32x4_t __p0) {
52674	54332	#ifdef __LITTLE_ENDIAN__
52675	54333	__ai __attribute__((target("neon"))) float32_t vminnmv_f32(float32x2_t __p0) {
52676	54334	float32_t __ret;
52677		__ret = (float32_t) __builtin_neon_vminnmv_f32(__p0);
	54335	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vminnmv_f32(__p0));
52678	54336	return __ret;
52679	54337	}
52680	54338	#else
52681	54339	__ai __attribute__((target("neon"))) float32_t vminnmv_f32(float32x2_t __p0) {
52682	54340	float32_t __ret;
52683		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
52684		__ret = (float32_t) __builtin_neon_vminnmv_f32(__rev0);
	54341	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	54342	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vminnmv_f32(__rev0));
52685	54343	return __ret;
52686	54344	}
52687	54345	#endif
...	...	@@ -52689,14 +54347,14 @@ __ai __attribute__((target("neon"))) float32_t vminnmv_f32(float32x2_t __p0) {
52689	54347	#ifdef __LITTLE_ENDIAN__
52690	54348	__ai __attribute__((target("neon"))) uint8_t vminvq_u8(uint8x16_t __p0) {
52691	54349	uint8_t __ret;
52692		__ret = (uint8_t) __builtin_neon_vminvq_u8(__p0);
	54350	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vminvq_u8(__p0));
52693	54351	return __ret;
52694	54352	}
52695	54353	#else
52696	54354	__ai __attribute__((target("neon"))) uint8_t vminvq_u8(uint8x16_t __p0) {
52697	54355	uint8_t __ret;
52698		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
52699		__ret = (uint8_t) __builtin_neon_vminvq_u8(__rev0);
	54356	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	54357	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vminvq_u8(__rev0));
52700	54358	return __ret;
52701	54359	}
52702	54360	#endif
...	...	@@ -52704,14 +54362,14 @@ __ai __attribute__((target("neon"))) uint8_t vminvq_u8(uint8x16_t __p0) {
52704	54362	#ifdef __LITTLE_ENDIAN__
52705	54363	__ai __attribute__((target("neon"))) uint32_t vminvq_u32(uint32x4_t __p0) {
52706	54364	uint32_t __ret;
52707		__ret = (uint32_t) __builtin_neon_vminvq_u32(__p0);
	54365	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vminvq_u32(__p0));
52708	54366	return __ret;
52709	54367	}
52710	54368	#else
52711	54369	__ai __attribute__((target("neon"))) uint32_t vminvq_u32(uint32x4_t __p0) {
52712	54370	uint32_t __ret;
52713		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
52714		__ret = (uint32_t) __builtin_neon_vminvq_u32(__rev0);
	54371	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	54372	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vminvq_u32(__rev0));
52715	54373	return __ret;
52716	54374	}
52717	54375	#endif
...	...	@@ -52719,14 +54377,14 @@ __ai __attribute__((target("neon"))) uint32_t vminvq_u32(uint32x4_t __p0) {
52719	54377	#ifdef __LITTLE_ENDIAN__
52720	54378	__ai __attribute__((target("neon"))) uint16_t vminvq_u16(uint16x8_t __p0) {
52721	54379	uint16_t __ret;
52722		__ret = (uint16_t) __builtin_neon_vminvq_u16(__p0);
	54380	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vminvq_u16(__p0));
52723	54381	return __ret;
52724	54382	}
52725	54383	#else
52726	54384	__ai __attribute__((target("neon"))) uint16_t vminvq_u16(uint16x8_t __p0) {
52727	54385	uint16_t __ret;
52728		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
52729		__ret = (uint16_t) __builtin_neon_vminvq_u16(__rev0);
	54386	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	54387	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vminvq_u16(__rev0));
52730	54388	return __ret;
52731	54389	}
52732	54390	#endif
...	...	@@ -52734,14 +54392,14 @@ __ai __attribute__((target("neon"))) uint16_t vminvq_u16(uint16x8_t __p0) {
52734	54392	#ifdef __LITTLE_ENDIAN__
52735	54393	__ai __attribute__((target("neon"))) int8_t vminvq_s8(int8x16_t __p0) {
52736	54394	int8_t __ret;
52737		__ret = (int8_t) __builtin_neon_vminvq_s8(__p0);
	54395	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vminvq_s8(__p0));
52738	54396	return __ret;
52739	54397	}
52740	54398	#else
52741	54399	__ai __attribute__((target("neon"))) int8_t vminvq_s8(int8x16_t __p0) {
52742	54400	int8_t __ret;
52743		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
52744		__ret = (int8_t) __builtin_neon_vminvq_s8(__rev0);
	54401	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	54402	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vminvq_s8(__rev0));
52745	54403	return __ret;
52746	54404	}
52747	54405	#endif
...	...	@@ -52749,14 +54407,14 @@ __ai __attribute__((target("neon"))) int8_t vminvq_s8(int8x16_t __p0) {
52749	54407	#ifdef __LITTLE_ENDIAN__
52750	54408	__ai __attribute__((target("neon"))) float64_t vminvq_f64(float64x2_t __p0) {
52751	54409	float64_t __ret;
52752		__ret = (float64_t) __builtin_neon_vminvq_f64(__p0);
	54410	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vminvq_f64(__p0));
52753	54411	return __ret;
52754	54412	}
52755	54413	#else
52756	54414	__ai __attribute__((target("neon"))) float64_t vminvq_f64(float64x2_t __p0) {
52757	54415	float64_t __ret;
52758		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
52759		__ret = (float64_t) __builtin_neon_vminvq_f64(__rev0);
	54416	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	54417	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vminvq_f64(__rev0));
52760	54418	return __ret;
52761	54419	}
52762	54420	#endif
...	...	@@ -52764,14 +54422,14 @@ __ai __attribute__((target("neon"))) float64_t vminvq_f64(float64x2_t __p0) {
52764	54422	#ifdef __LITTLE_ENDIAN__
52765	54423	__ai __attribute__((target("neon"))) float32_t vminvq_f32(float32x4_t __p0) {
52766	54424	float32_t __ret;
52767		__ret = (float32_t) __builtin_neon_vminvq_f32(__p0);
	54425	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vminvq_f32(__p0));
52768	54426	return __ret;
52769	54427	}
52770	54428	#else
52771	54429	__ai __attribute__((target("neon"))) float32_t vminvq_f32(float32x4_t __p0) {
52772	54430	float32_t __ret;
52773		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
52774		__ret = (float32_t) __builtin_neon_vminvq_f32(__rev0);
	54431	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	54432	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vminvq_f32(__rev0));
52775	54433	return __ret;
52776	54434	}
52777	54435	#endif
...	...	@@ -52779,14 +54437,14 @@ __ai __attribute__((target("neon"))) float32_t vminvq_f32(float32x4_t __p0) {
52779	54437	#ifdef __LITTLE_ENDIAN__
52780	54438	__ai __attribute__((target("neon"))) int32_t vminvq_s32(int32x4_t __p0) {
52781	54439	int32_t __ret;
52782		__ret = (int32_t) __builtin_neon_vminvq_s32(__p0);
	54440	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vminvq_s32(__p0));
52783	54441	return __ret;
52784	54442	}
52785	54443	#else
52786	54444	__ai __attribute__((target("neon"))) int32_t vminvq_s32(int32x4_t __p0) {
52787	54445	int32_t __ret;
52788		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
52789		__ret = (int32_t) __builtin_neon_vminvq_s32(__rev0);
	54446	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	54447	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vminvq_s32(__rev0));
52790	54448	return __ret;
52791	54449	}
52792	54450	#endif
...	...	@@ -52794,14 +54452,14 @@ __ai __attribute__((target("neon"))) int32_t vminvq_s32(int32x4_t __p0) {
52794	54452	#ifdef __LITTLE_ENDIAN__
52795	54453	__ai __attribute__((target("neon"))) int16_t vminvq_s16(int16x8_t __p0) {
52796	54454	int16_t __ret;
52797		__ret = (int16_t) __builtin_neon_vminvq_s16(__p0);
	54455	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vminvq_s16(__p0));
52798	54456	return __ret;
52799	54457	}
52800	54458	#else
52801	54459	__ai __attribute__((target("neon"))) int16_t vminvq_s16(int16x8_t __p0) {
52802	54460	int16_t __ret;
52803		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
52804		__ret = (int16_t) __builtin_neon_vminvq_s16(__rev0);
	54461	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	54462	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vminvq_s16(__rev0));
52805	54463	return __ret;
52806	54464	}
52807	54465	#endif
...	...	@@ -52809,14 +54467,14 @@ __ai __attribute__((target("neon"))) int16_t vminvq_s16(int16x8_t __p0) {
52809	54467	#ifdef __LITTLE_ENDIAN__
52810	54468	__ai __attribute__((target("neon"))) uint8_t vminv_u8(uint8x8_t __p0) {
52811	54469	uint8_t __ret;
52812		__ret = (uint8_t) __builtin_neon_vminv_u8(__p0);
	54470	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vminv_u8(__p0));
52813	54471	return __ret;
52814	54472	}
52815	54473	#else
52816	54474	__ai __attribute__((target("neon"))) uint8_t vminv_u8(uint8x8_t __p0) {
52817	54475	uint8_t __ret;
52818		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
52819		__ret = (uint8_t) __builtin_neon_vminv_u8(__rev0);
	54476	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	54477	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vminv_u8(__rev0));
52820	54478	return __ret;
52821	54479	}
52822	54480	#endif
...	...	@@ -52824,14 +54482,14 @@ __ai __attribute__((target("neon"))) uint8_t vminv_u8(uint8x8_t __p0) {
52824	54482	#ifdef __LITTLE_ENDIAN__
52825	54483	__ai __attribute__((target("neon"))) uint32_t vminv_u32(uint32x2_t __p0) {
52826	54484	uint32_t __ret;
52827		__ret = (uint32_t) __builtin_neon_vminv_u32(__p0);
	54485	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vminv_u32(__p0));
52828	54486	return __ret;
52829	54487	}
52830	54488	#else
52831	54489	__ai __attribute__((target("neon"))) uint32_t vminv_u32(uint32x2_t __p0) {
52832	54490	uint32_t __ret;
52833		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
52834		__ret = (uint32_t) __builtin_neon_vminv_u32(__rev0);
	54491	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	54492	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vminv_u32(__rev0));
52835	54493	return __ret;
52836	54494	}
52837	54495	#endif
...	...	@@ -52839,14 +54497,14 @@ __ai __attribute__((target("neon"))) uint32_t vminv_u32(uint32x2_t __p0) {
52839	54497	#ifdef __LITTLE_ENDIAN__
52840	54498	__ai __attribute__((target("neon"))) uint16_t vminv_u16(uint16x4_t __p0) {
52841	54499	uint16_t __ret;
52842		__ret = (uint16_t) __builtin_neon_vminv_u16(__p0);
	54500	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vminv_u16(__p0));
52843	54501	return __ret;
52844	54502	}
52845	54503	#else
52846	54504	__ai __attribute__((target("neon"))) uint16_t vminv_u16(uint16x4_t __p0) {
52847	54505	uint16_t __ret;
52848		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
52849		__ret = (uint16_t) __builtin_neon_vminv_u16(__rev0);
	54506	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	54507	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vminv_u16(__rev0));
52850	54508	return __ret;
52851	54509	}
52852	54510	#endif
...	...	@@ -52854,14 +54512,14 @@ __ai __attribute__((target("neon"))) uint16_t vminv_u16(uint16x4_t __p0) {
52854	54512	#ifdef __LITTLE_ENDIAN__
52855	54513	__ai __attribute__((target("neon"))) int8_t vminv_s8(int8x8_t __p0) {
52856	54514	int8_t __ret;
52857		__ret = (int8_t) __builtin_neon_vminv_s8(__p0);
	54515	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vminv_s8(__p0));
52858	54516	return __ret;
52859	54517	}
52860	54518	#else
52861	54519	__ai __attribute__((target("neon"))) int8_t vminv_s8(int8x8_t __p0) {
52862	54520	int8_t __ret;
52863		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
52864		__ret = (int8_t) __builtin_neon_vminv_s8(__rev0);
	54521	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	54522	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vminv_s8(__rev0));
52865	54523	return __ret;
52866	54524	}
52867	54525	#endif
...	...	@@ -52869,14 +54527,14 @@ __ai __attribute__((target("neon"))) int8_t vminv_s8(int8x8_t __p0) {
52869	54527	#ifdef __LITTLE_ENDIAN__
52870	54528	__ai __attribute__((target("neon"))) float32_t vminv_f32(float32x2_t __p0) {
52871	54529	float32_t __ret;
52872		__ret = (float32_t) __builtin_neon_vminv_f32(__p0);
	54530	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vminv_f32(__p0));
52873	54531	return __ret;
52874	54532	}
52875	54533	#else
52876	54534	__ai __attribute__((target("neon"))) float32_t vminv_f32(float32x2_t __p0) {
52877	54535	float32_t __ret;
52878		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
52879		__ret = (float32_t) __builtin_neon_vminv_f32(__rev0);
	54536	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	54537	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vminv_f32(__rev0));
52880	54538	return __ret;
52881	54539	}
52882	54540	#endif
...	...	@@ -52884,14 +54542,14 @@ __ai __attribute__((target("neon"))) float32_t vminv_f32(float32x2_t __p0) {
52884	54542	#ifdef __LITTLE_ENDIAN__
52885	54543	__ai __attribute__((target("neon"))) int32_t vminv_s32(int32x2_t __p0) {
52886	54544	int32_t __ret;
52887		__ret = (int32_t) __builtin_neon_vminv_s32(__p0);
	54545	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vminv_s32(__p0));
52888	54546	return __ret;
52889	54547	}
52890	54548	#else
52891	54549	__ai __attribute__((target("neon"))) int32_t vminv_s32(int32x2_t __p0) {
52892	54550	int32_t __ret;
52893		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
52894		__ret = (int32_t) __builtin_neon_vminv_s32(__rev0);
	54551	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	54552	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vminv_s32(__rev0));
52895	54553	return __ret;
52896	54554	}
52897	54555	#endif
...	...	@@ -52899,14 +54557,14 @@ __ai __attribute__((target("neon"))) int32_t vminv_s32(int32x2_t __p0) {
52899	54557	#ifdef __LITTLE_ENDIAN__
52900	54558	__ai __attribute__((target("neon"))) int16_t vminv_s16(int16x4_t __p0) {
52901	54559	int16_t __ret;
52902		__ret = (int16_t) __builtin_neon_vminv_s16(__p0);
	54560	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vminv_s16(__p0));
52903	54561	return __ret;
52904	54562	}
52905	54563	#else
52906	54564	__ai __attribute__((target("neon"))) int16_t vminv_s16(int16x4_t __p0) {
52907	54565	int16_t __ret;
52908		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
52909		__ret = (int16_t) __builtin_neon_vminv_s16(__rev0);
	54566	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	54567	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vminv_s16(__rev0));
52910	54568	return __ret;
52911	54569	}
52912	54570	#endif
...	...	@@ -52920,11 +54578,11 @@ __ai __attribute__((target("neon"))) float64x2_t vmlaq_f64(float64x2_t __p0, flo
52920	54578	#else
52921	54579	__ai __attribute__((target("neon"))) float64x2_t vmlaq_f64(float64x2_t __p0, float64x2_t __p1, float64x2_t __p2) {
52922	54580	float64x2_t __ret;
52923		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
52924		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
52925		float64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
	54581	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	54582	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	54583	float64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_64);
52926	54584	__ret = __rev0 + __rev1 * __rev2;
52927		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	54585	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
52928	54586	return __ret;
52929	54587	}
52930	54588	#endif
...	...	@@ -52935,1084 +54593,1084 @@ __ai __attribute__((target("neon"))) float64x1_t vmla_f64(float64x1_t __p0, floa
52935	54593	return __ret;
52936	54594	}
52937	54595	#ifdef __LITTLE_ENDIAN__
52938		#define vmlaq_laneq_u32(__p0_448, __p1_448, __p2_448, __p3_448) __extension__ ({ \
52939		uint32x4_t __ret_448; \
52940		uint32x4_t __s0_448 = __p0_448; \
52941		uint32x4_t __s1_448 = __p1_448; \
52942		uint32x4_t __s2_448 = __p2_448; \
52943		__ret_448 = __s0_448 + __s1_448 * splatq_laneq_u32(__s2_448, __p3_448); \
52944		__ret_448; \
52945		})
52946		#else
52947		#define vmlaq_laneq_u32(__p0_449, __p1_449, __p2_449, __p3_449) __extension__ ({ \
52948		uint32x4_t __ret_449; \
52949		uint32x4_t __s0_449 = __p0_449; \
52950		uint32x4_t __s1_449 = __p1_449; \
52951		uint32x4_t __s2_449 = __p2_449; \
52952		uint32x4_t __rev0_449; __rev0_449 = __builtin_shufflevector(__s0_449, __s0_449, 3, 2, 1, 0); \
52953		uint32x4_t __rev1_449; __rev1_449 = __builtin_shufflevector(__s1_449, __s1_449, 3, 2, 1, 0); \
52954		uint32x4_t __rev2_449; __rev2_449 = __builtin_shufflevector(__s2_449, __s2_449, 3, 2, 1, 0); \
52955		__ret_449 = __rev0_449 + __rev1_449 * __noswap_splatq_laneq_u32(__rev2_449, __p3_449); \
52956		__ret_449 = __builtin_shufflevector(__ret_449, __ret_449, 3, 2, 1, 0); \
52957		__ret_449; \
52958		})
52959		#endif
52960
52961		#ifdef __LITTLE_ENDIAN__
52962		#define vmlaq_laneq_u16(__p0_450, __p1_450, __p2_450, __p3_450) __extension__ ({ \
52963		uint16x8_t __ret_450; \
52964		uint16x8_t __s0_450 = __p0_450; \
52965		uint16x8_t __s1_450 = __p1_450; \
52966		uint16x8_t __s2_450 = __p2_450; \
52967		__ret_450 = __s0_450 + __s1_450 * splatq_laneq_u16(__s2_450, __p3_450); \
52968		__ret_450; \
52969		})
52970		#else
52971		#define vmlaq_laneq_u16(__p0_451, __p1_451, __p2_451, __p3_451) __extension__ ({ \
52972		uint16x8_t __ret_451; \
52973		uint16x8_t __s0_451 = __p0_451; \
52974		uint16x8_t __s1_451 = __p1_451; \
52975		uint16x8_t __s2_451 = __p2_451; \
52976		uint16x8_t __rev0_451; __rev0_451 = __builtin_shufflevector(__s0_451, __s0_451, 7, 6, 5, 4, 3, 2, 1, 0); \
52977		uint16x8_t __rev1_451; __rev1_451 = __builtin_shufflevector(__s1_451, __s1_451, 7, 6, 5, 4, 3, 2, 1, 0); \
52978		uint16x8_t __rev2_451; __rev2_451 = __builtin_shufflevector(__s2_451, __s2_451, 7, 6, 5, 4, 3, 2, 1, 0); \
52979		__ret_451 = __rev0_451 + __rev1_451 * __noswap_splatq_laneq_u16(__rev2_451, __p3_451); \
52980		__ret_451 = __builtin_shufflevector(__ret_451, __ret_451, 7, 6, 5, 4, 3, 2, 1, 0); \
52981		__ret_451; \
52982		})
52983		#endif
52984
52985		#ifdef __LITTLE_ENDIAN__
52986		#define vmlaq_laneq_f32(__p0_452, __p1_452, __p2_452, __p3_452) __extension__ ({ \
52987		float32x4_t __ret_452; \
52988		float32x4_t __s0_452 = __p0_452; \
52989		float32x4_t __s1_452 = __p1_452; \
52990		float32x4_t __s2_452 = __p2_452; \
52991		__ret_452 = __s0_452 + __s1_452 * splatq_laneq_f32(__s2_452, __p3_452); \
52992		__ret_452; \
52993		})
52994		#else
52995		#define vmlaq_laneq_f32(__p0_453, __p1_453, __p2_453, __p3_453) __extension__ ({ \
52996		float32x4_t __ret_453; \
52997		float32x4_t __s0_453 = __p0_453; \
52998		float32x4_t __s1_453 = __p1_453; \
52999		float32x4_t __s2_453 = __p2_453; \
53000		float32x4_t __rev0_453; __rev0_453 = __builtin_shufflevector(__s0_453, __s0_453, 3, 2, 1, 0); \
53001		float32x4_t __rev1_453; __rev1_453 = __builtin_shufflevector(__s1_453, __s1_453, 3, 2, 1, 0); \
53002		float32x4_t __rev2_453; __rev2_453 = __builtin_shufflevector(__s2_453, __s2_453, 3, 2, 1, 0); \
53003		__ret_453 = __rev0_453 + __rev1_453 * __noswap_splatq_laneq_f32(__rev2_453, __p3_453); \
53004		__ret_453 = __builtin_shufflevector(__ret_453, __ret_453, 3, 2, 1, 0); \
53005		__ret_453; \
53006		})
53007		#endif
53008
53009		#ifdef __LITTLE_ENDIAN__
53010		#define vmlaq_laneq_s32(__p0_454, __p1_454, __p2_454, __p3_454) __extension__ ({ \
53011		int32x4_t __ret_454; \
53012		int32x4_t __s0_454 = __p0_454; \
53013		int32x4_t __s1_454 = __p1_454; \
53014		int32x4_t __s2_454 = __p2_454; \
53015		__ret_454 = __s0_454 + __s1_454 * splatq_laneq_s32(__s2_454, __p3_454); \
53016		__ret_454; \
53017		})
53018		#else
53019		#define vmlaq_laneq_s32(__p0_455, __p1_455, __p2_455, __p3_455) __extension__ ({ \
53020		int32x4_t __ret_455; \
53021		int32x4_t __s0_455 = __p0_455; \
53022		int32x4_t __s1_455 = __p1_455; \
53023		int32x4_t __s2_455 = __p2_455; \
53024		int32x4_t __rev0_455; __rev0_455 = __builtin_shufflevector(__s0_455, __s0_455, 3, 2, 1, 0); \
53025		int32x4_t __rev1_455; __rev1_455 = __builtin_shufflevector(__s1_455, __s1_455, 3, 2, 1, 0); \
53026		int32x4_t __rev2_455; __rev2_455 = __builtin_shufflevector(__s2_455, __s2_455, 3, 2, 1, 0); \
53027		__ret_455 = __rev0_455 + __rev1_455 * __noswap_splatq_laneq_s32(__rev2_455, __p3_455); \
53028		__ret_455 = __builtin_shufflevector(__ret_455, __ret_455, 3, 2, 1, 0); \
53029		__ret_455; \
53030		})
53031		#endif
53032
53033		#ifdef __LITTLE_ENDIAN__
53034		#define vmlaq_laneq_s16(__p0_456, __p1_456, __p2_456, __p3_456) __extension__ ({ \
53035		int16x8_t __ret_456; \
53036		int16x8_t __s0_456 = __p0_456; \
53037		int16x8_t __s1_456 = __p1_456; \
53038		int16x8_t __s2_456 = __p2_456; \
53039		__ret_456 = __s0_456 + __s1_456 * splatq_laneq_s16(__s2_456, __p3_456); \
53040		__ret_456; \
53041		})
53042		#else
53043		#define vmlaq_laneq_s16(__p0_457, __p1_457, __p2_457, __p3_457) __extension__ ({ \
53044		int16x8_t __ret_457; \
53045		int16x8_t __s0_457 = __p0_457; \
53046		int16x8_t __s1_457 = __p1_457; \
53047		int16x8_t __s2_457 = __p2_457; \
53048		int16x8_t __rev0_457; __rev0_457 = __builtin_shufflevector(__s0_457, __s0_457, 7, 6, 5, 4, 3, 2, 1, 0); \
53049		int16x8_t __rev1_457; __rev1_457 = __builtin_shufflevector(__s1_457, __s1_457, 7, 6, 5, 4, 3, 2, 1, 0); \
53050		int16x8_t __rev2_457; __rev2_457 = __builtin_shufflevector(__s2_457, __s2_457, 7, 6, 5, 4, 3, 2, 1, 0); \
53051		__ret_457 = __rev0_457 + __rev1_457 * __noswap_splatq_laneq_s16(__rev2_457, __p3_457); \
53052		__ret_457 = __builtin_shufflevector(__ret_457, __ret_457, 7, 6, 5, 4, 3, 2, 1, 0); \
53053		__ret_457; \
53054		})
53055		#endif
53056
53057		#ifdef __LITTLE_ENDIAN__
53058		#define vmla_laneq_u32(__p0_458, __p1_458, __p2_458, __p3_458) __extension__ ({ \
53059		uint32x2_t __ret_458; \
53060		uint32x2_t __s0_458 = __p0_458; \
53061		uint32x2_t __s1_458 = __p1_458; \
53062		uint32x4_t __s2_458 = __p2_458; \
53063		__ret_458 = __s0_458 + __s1_458 * splat_laneq_u32(__s2_458, __p3_458); \
53064		__ret_458; \
53065		})
53066		#else
53067		#define vmla_laneq_u32(__p0_459, __p1_459, __p2_459, __p3_459) __extension__ ({ \
53068		uint32x2_t __ret_459; \
53069		uint32x2_t __s0_459 = __p0_459; \
53070		uint32x2_t __s1_459 = __p1_459; \
53071		uint32x4_t __s2_459 = __p2_459; \
53072		uint32x2_t __rev0_459; __rev0_459 = __builtin_shufflevector(__s0_459, __s0_459, 1, 0); \
53073		uint32x2_t __rev1_459; __rev1_459 = __builtin_shufflevector(__s1_459, __s1_459, 1, 0); \
53074		uint32x4_t __rev2_459; __rev2_459 = __builtin_shufflevector(__s2_459, __s2_459, 3, 2, 1, 0); \
53075		__ret_459 = __rev0_459 + __rev1_459 * __noswap_splat_laneq_u32(__rev2_459, __p3_459); \
53076		__ret_459 = __builtin_shufflevector(__ret_459, __ret_459, 1, 0); \
53077		__ret_459; \
53078		})
53079		#endif
53080
53081		#ifdef __LITTLE_ENDIAN__
53082		#define vmla_laneq_u16(__p0_460, __p1_460, __p2_460, __p3_460) __extension__ ({ \
53083		uint16x4_t __ret_460; \
53084		uint16x4_t __s0_460 = __p0_460; \
53085		uint16x4_t __s1_460 = __p1_460; \
53086		uint16x8_t __s2_460 = __p2_460; \
53087		__ret_460 = __s0_460 + __s1_460 * splat_laneq_u16(__s2_460, __p3_460); \
53088		__ret_460; \
53089		})
53090		#else
53091		#define vmla_laneq_u16(__p0_461, __p1_461, __p2_461, __p3_461) __extension__ ({ \
53092		uint16x4_t __ret_461; \
53093		uint16x4_t __s0_461 = __p0_461; \
53094		uint16x4_t __s1_461 = __p1_461; \
53095		uint16x8_t __s2_461 = __p2_461; \
53096		uint16x4_t __rev0_461; __rev0_461 = __builtin_shufflevector(__s0_461, __s0_461, 3, 2, 1, 0); \
53097		uint16x4_t __rev1_461; __rev1_461 = __builtin_shufflevector(__s1_461, __s1_461, 3, 2, 1, 0); \
53098		uint16x8_t __rev2_461; __rev2_461 = __builtin_shufflevector(__s2_461, __s2_461, 7, 6, 5, 4, 3, 2, 1, 0); \
53099		__ret_461 = __rev0_461 + __rev1_461 * __noswap_splat_laneq_u16(__rev2_461, __p3_461); \
53100		__ret_461 = __builtin_shufflevector(__ret_461, __ret_461, 3, 2, 1, 0); \
53101		__ret_461; \
53102		})
53103		#endif
53104
53105		#ifdef __LITTLE_ENDIAN__
53106		#define vmla_laneq_f32(__p0_462, __p1_462, __p2_462, __p3_462) __extension__ ({ \
53107		float32x2_t __ret_462; \
53108		float32x2_t __s0_462 = __p0_462; \
53109		float32x2_t __s1_462 = __p1_462; \
53110		float32x4_t __s2_462 = __p2_462; \
53111		__ret_462 = __s0_462 + __s1_462 * splat_laneq_f32(__s2_462, __p3_462); \
53112		__ret_462; \
53113		})
53114		#else
53115		#define vmla_laneq_f32(__p0_463, __p1_463, __p2_463, __p3_463) __extension__ ({ \
53116		float32x2_t __ret_463; \
53117		float32x2_t __s0_463 = __p0_463; \
53118		float32x2_t __s1_463 = __p1_463; \
53119		float32x4_t __s2_463 = __p2_463; \
53120		float32x2_t __rev0_463; __rev0_463 = __builtin_shufflevector(__s0_463, __s0_463, 1, 0); \
53121		float32x2_t __rev1_463; __rev1_463 = __builtin_shufflevector(__s1_463, __s1_463, 1, 0); \
53122		float32x4_t __rev2_463; __rev2_463 = __builtin_shufflevector(__s2_463, __s2_463, 3, 2, 1, 0); \
53123		__ret_463 = __rev0_463 + __rev1_463 * __noswap_splat_laneq_f32(__rev2_463, __p3_463); \
53124		__ret_463 = __builtin_shufflevector(__ret_463, __ret_463, 1, 0); \
53125		__ret_463; \
53126		})
53127		#endif
53128
53129		#ifdef __LITTLE_ENDIAN__
53130		#define vmla_laneq_s32(__p0_464, __p1_464, __p2_464, __p3_464) __extension__ ({ \
53131		int32x2_t __ret_464; \
53132		int32x2_t __s0_464 = __p0_464; \
53133		int32x2_t __s1_464 = __p1_464; \
53134		int32x4_t __s2_464 = __p2_464; \
53135		__ret_464 = __s0_464 + __s1_464 * splat_laneq_s32(__s2_464, __p3_464); \
	54596	#define vmlaq_laneq_u32(__p0_464, __p1_464, __p2_464, __p3_464) __extension__ ({ \
	54597	uint32x4_t __ret_464; \
	54598	uint32x4_t __s0_464 = __p0_464; \
	54599	uint32x4_t __s1_464 = __p1_464; \
	54600	uint32x4_t __s2_464 = __p2_464; \
	54601	__ret_464 = __s0_464 + __s1_464 * splatq_laneq_u32(__s2_464, __p3_464); \
53136	54602	__ret_464; \
53137	54603	})
53138	54604	#else
53139		#define vmla_laneq_s32(__p0_465, __p1_465, __p2_465, __p3_465) __extension__ ({ \
53140		int32x2_t __ret_465; \
53141		int32x2_t __s0_465 = __p0_465; \
53142		int32x2_t __s1_465 = __p1_465; \
53143		int32x4_t __s2_465 = __p2_465; \
53144		int32x2_t __rev0_465; __rev0_465 = __builtin_shufflevector(__s0_465, __s0_465, 1, 0); \
53145		int32x2_t __rev1_465; __rev1_465 = __builtin_shufflevector(__s1_465, __s1_465, 1, 0); \
53146		int32x4_t __rev2_465; __rev2_465 = __builtin_shufflevector(__s2_465, __s2_465, 3, 2, 1, 0); \
53147		__ret_465 = __rev0_465 + __rev1_465 * __noswap_splat_laneq_s32(__rev2_465, __p3_465); \
53148		__ret_465 = __builtin_shufflevector(__ret_465, __ret_465, 1, 0); \
	54605	#define vmlaq_laneq_u32(__p0_465, __p1_465, __p2_465, __p3_465) __extension__ ({ \
	54606	uint32x4_t __ret_465; \
	54607	uint32x4_t __s0_465 = __p0_465; \
	54608	uint32x4_t __s1_465 = __p1_465; \
	54609	uint32x4_t __s2_465 = __p2_465; \
	54610	uint32x4_t __rev0_465; __rev0_465 = __builtin_shufflevector(__s0_465, __s0_465, __lane_reverse_128_32); \
	54611	uint32x4_t __rev1_465; __rev1_465 = __builtin_shufflevector(__s1_465, __s1_465, __lane_reverse_128_32); \
	54612	uint32x4_t __rev2_465; __rev2_465 = __builtin_shufflevector(__s2_465, __s2_465, __lane_reverse_128_32); \
	54613	__ret_465 = __rev0_465 + __rev1_465 * __noswap_splatq_laneq_u32(__rev2_465, __p3_465); \
	54614	__ret_465 = __builtin_shufflevector(__ret_465, __ret_465, __lane_reverse_128_32); \
53149	54615	__ret_465; \
53150	54616	})
53151	54617	#endif
53152	54618
53153	54619	#ifdef __LITTLE_ENDIAN__
53154		#define vmla_laneq_s16(__p0_466, __p1_466, __p2_466, __p3_466) __extension__ ({ \
53155		int16x4_t __ret_466; \
53156		int16x4_t __s0_466 = __p0_466; \
53157		int16x4_t __s1_466 = __p1_466; \
53158		int16x8_t __s2_466 = __p2_466; \
53159		__ret_466 = __s0_466 + __s1_466 * splat_laneq_s16(__s2_466, __p3_466); \
	54620	#define vmlaq_laneq_u16(__p0_466, __p1_466, __p2_466, __p3_466) __extension__ ({ \
	54621	uint16x8_t __ret_466; \
	54622	uint16x8_t __s0_466 = __p0_466; \
	54623	uint16x8_t __s1_466 = __p1_466; \
	54624	uint16x8_t __s2_466 = __p2_466; \
	54625	__ret_466 = __s0_466 + __s1_466 * splatq_laneq_u16(__s2_466, __p3_466); \
53160	54626	__ret_466; \
53161	54627	})
53162	54628	#else
53163		#define vmla_laneq_s16(__p0_467, __p1_467, __p2_467, __p3_467) __extension__ ({ \
53164		int16x4_t __ret_467; \
53165		int16x4_t __s0_467 = __p0_467; \
53166		int16x4_t __s1_467 = __p1_467; \
53167		int16x8_t __s2_467 = __p2_467; \
53168		int16x4_t __rev0_467; __rev0_467 = __builtin_shufflevector(__s0_467, __s0_467, 3, 2, 1, 0); \
53169		int16x4_t __rev1_467; __rev1_467 = __builtin_shufflevector(__s1_467, __s1_467, 3, 2, 1, 0); \
53170		int16x8_t __rev2_467; __rev2_467 = __builtin_shufflevector(__s2_467, __s2_467, 7, 6, 5, 4, 3, 2, 1, 0); \
53171		__ret_467 = __rev0_467 + __rev1_467 * __noswap_splat_laneq_s16(__rev2_467, __p3_467); \
53172		__ret_467 = __builtin_shufflevector(__ret_467, __ret_467, 3, 2, 1, 0); \
	54629	#define vmlaq_laneq_u16(__p0_467, __p1_467, __p2_467, __p3_467) __extension__ ({ \
	54630	uint16x8_t __ret_467; \
	54631	uint16x8_t __s0_467 = __p0_467; \
	54632	uint16x8_t __s1_467 = __p1_467; \
	54633	uint16x8_t __s2_467 = __p2_467; \
	54634	uint16x8_t __rev0_467; __rev0_467 = __builtin_shufflevector(__s0_467, __s0_467, __lane_reverse_128_16); \
	54635	uint16x8_t __rev1_467; __rev1_467 = __builtin_shufflevector(__s1_467, __s1_467, __lane_reverse_128_16); \
	54636	uint16x8_t __rev2_467; __rev2_467 = __builtin_shufflevector(__s2_467, __s2_467, __lane_reverse_128_16); \
	54637	__ret_467 = __rev0_467 + __rev1_467 * __noswap_splatq_laneq_u16(__rev2_467, __p3_467); \
	54638	__ret_467 = __builtin_shufflevector(__ret_467, __ret_467, __lane_reverse_128_16); \
53173	54639	__ret_467; \
53174	54640	})
53175	54641	#endif
53176	54642
53177	54643	#ifdef __LITTLE_ENDIAN__
53178		#define vmlal_high_lane_u32(__p0_468, __p1_468, __p2_468, __p3_468) __extension__ ({ \
53179		uint64x2_t __ret_468; \
53180		uint64x2_t __s0_468 = __p0_468; \
53181		uint32x4_t __s1_468 = __p1_468; \
53182		uint32x2_t __s2_468 = __p2_468; \
53183		__ret_468 = __s0_468 + vmull_u32(vget_high_u32(__s1_468), splat_lane_u32(__s2_468, __p3_468)); \
	54644	#define vmlaq_laneq_f32(__p0_468, __p1_468, __p2_468, __p3_468) __extension__ ({ \
	54645	float32x4_t __ret_468; \
	54646	float32x4_t __s0_468 = __p0_468; \
	54647	float32x4_t __s1_468 = __p1_468; \
	54648	float32x4_t __s2_468 = __p2_468; \
	54649	__ret_468 = __s0_468 + __s1_468 * splatq_laneq_f32(__s2_468, __p3_468); \
53184	54650	__ret_468; \
53185	54651	})
53186	54652	#else
53187		#define vmlal_high_lane_u32(__p0_469, __p1_469, __p2_469, __p3_469) __extension__ ({ \
53188		uint64x2_t __ret_469; \
53189		uint64x2_t __s0_469 = __p0_469; \
53190		uint32x4_t __s1_469 = __p1_469; \
53191		uint32x2_t __s2_469 = __p2_469; \
53192		uint64x2_t __rev0_469; __rev0_469 = __builtin_shufflevector(__s0_469, __s0_469, 1, 0); \
53193		uint32x4_t __rev1_469; __rev1_469 = __builtin_shufflevector(__s1_469, __s1_469, 3, 2, 1, 0); \
53194		uint32x2_t __rev2_469; __rev2_469 = __builtin_shufflevector(__s2_469, __s2_469, 1, 0); \
53195		__ret_469 = __rev0_469 + __noswap_vmull_u32(__noswap_vget_high_u32(__rev1_469), __noswap_splat_lane_u32(__rev2_469, __p3_469)); \
53196		__ret_469 = __builtin_shufflevector(__ret_469, __ret_469, 1, 0); \
	54653	#define vmlaq_laneq_f32(__p0_469, __p1_469, __p2_469, __p3_469) __extension__ ({ \
	54654	float32x4_t __ret_469; \
	54655	float32x4_t __s0_469 = __p0_469; \
	54656	float32x4_t __s1_469 = __p1_469; \
	54657	float32x4_t __s2_469 = __p2_469; \
	54658	float32x4_t __rev0_469; __rev0_469 = __builtin_shufflevector(__s0_469, __s0_469, __lane_reverse_128_32); \
	54659	float32x4_t __rev1_469; __rev1_469 = __builtin_shufflevector(__s1_469, __s1_469, __lane_reverse_128_32); \
	54660	float32x4_t __rev2_469; __rev2_469 = __builtin_shufflevector(__s2_469, __s2_469, __lane_reverse_128_32); \
	54661	__ret_469 = __rev0_469 + __rev1_469 * __noswap_splatq_laneq_f32(__rev2_469, __p3_469); \
	54662	__ret_469 = __builtin_shufflevector(__ret_469, __ret_469, __lane_reverse_128_32); \
53197	54663	__ret_469; \
53198	54664	})
53199	54665	#endif
53200	54666
53201	54667	#ifdef __LITTLE_ENDIAN__
53202		#define vmlal_high_lane_u16(__p0_470, __p1_470, __p2_470, __p3_470) __extension__ ({ \
53203		uint32x4_t __ret_470; \
53204		uint32x4_t __s0_470 = __p0_470; \
53205		uint16x8_t __s1_470 = __p1_470; \
53206		uint16x4_t __s2_470 = __p2_470; \
53207		__ret_470 = __s0_470 + vmull_u16(vget_high_u16(__s1_470), splat_lane_u16(__s2_470, __p3_470)); \
	54668	#define vmlaq_laneq_s32(__p0_470, __p1_470, __p2_470, __p3_470) __extension__ ({ \
	54669	int32x4_t __ret_470; \
	54670	int32x4_t __s0_470 = __p0_470; \
	54671	int32x4_t __s1_470 = __p1_470; \
	54672	int32x4_t __s2_470 = __p2_470; \
	54673	__ret_470 = __s0_470 + __s1_470 * splatq_laneq_s32(__s2_470, __p3_470); \
53208	54674	__ret_470; \
53209	54675	})
53210	54676	#else
53211		#define vmlal_high_lane_u16(__p0_471, __p1_471, __p2_471, __p3_471) __extension__ ({ \
53212		uint32x4_t __ret_471; \
53213		uint32x4_t __s0_471 = __p0_471; \
53214		uint16x8_t __s1_471 = __p1_471; \
53215		uint16x4_t __s2_471 = __p2_471; \
53216		uint32x4_t __rev0_471; __rev0_471 = __builtin_shufflevector(__s0_471, __s0_471, 3, 2, 1, 0); \
53217		uint16x8_t __rev1_471; __rev1_471 = __builtin_shufflevector(__s1_471, __s1_471, 7, 6, 5, 4, 3, 2, 1, 0); \
53218		uint16x4_t __rev2_471; __rev2_471 = __builtin_shufflevector(__s2_471, __s2_471, 3, 2, 1, 0); \
53219		__ret_471 = __rev0_471 + __noswap_vmull_u16(__noswap_vget_high_u16(__rev1_471), __noswap_splat_lane_u16(__rev2_471, __p3_471)); \
53220		__ret_471 = __builtin_shufflevector(__ret_471, __ret_471, 3, 2, 1, 0); \
	54677	#define vmlaq_laneq_s32(__p0_471, __p1_471, __p2_471, __p3_471) __extension__ ({ \
	54678	int32x4_t __ret_471; \
	54679	int32x4_t __s0_471 = __p0_471; \
	54680	int32x4_t __s1_471 = __p1_471; \
	54681	int32x4_t __s2_471 = __p2_471; \
	54682	int32x4_t __rev0_471; __rev0_471 = __builtin_shufflevector(__s0_471, __s0_471, __lane_reverse_128_32); \
	54683	int32x4_t __rev1_471; __rev1_471 = __builtin_shufflevector(__s1_471, __s1_471, __lane_reverse_128_32); \
	54684	int32x4_t __rev2_471; __rev2_471 = __builtin_shufflevector(__s2_471, __s2_471, __lane_reverse_128_32); \
	54685	__ret_471 = __rev0_471 + __rev1_471 * __noswap_splatq_laneq_s32(__rev2_471, __p3_471); \
	54686	__ret_471 = __builtin_shufflevector(__ret_471, __ret_471, __lane_reverse_128_32); \
53221	54687	__ret_471; \
53222	54688	})
53223	54689	#endif
53224	54690
53225	54691	#ifdef __LITTLE_ENDIAN__
53226		#define vmlal_high_lane_s32(__p0_472, __p1_472, __p2_472, __p3_472) __extension__ ({ \
53227		int64x2_t __ret_472; \
53228		int64x2_t __s0_472 = __p0_472; \
53229		int32x4_t __s1_472 = __p1_472; \
53230		int32x2_t __s2_472 = __p2_472; \
53231		__ret_472 = __s0_472 + vmull_s32(vget_high_s32(__s1_472), splat_lane_s32(__s2_472, __p3_472)); \
	54692	#define vmlaq_laneq_s16(__p0_472, __p1_472, __p2_472, __p3_472) __extension__ ({ \
	54693	int16x8_t __ret_472; \
	54694	int16x8_t __s0_472 = __p0_472; \
	54695	int16x8_t __s1_472 = __p1_472; \
	54696	int16x8_t __s2_472 = __p2_472; \
	54697	__ret_472 = __s0_472 + __s1_472 * splatq_laneq_s16(__s2_472, __p3_472); \
53232	54698	__ret_472; \
53233	54699	})
53234	54700	#else
53235		#define vmlal_high_lane_s32(__p0_473, __p1_473, __p2_473, __p3_473) __extension__ ({ \
53236		int64x2_t __ret_473; \
53237		int64x2_t __s0_473 = __p0_473; \
53238		int32x4_t __s1_473 = __p1_473; \
53239		int32x2_t __s2_473 = __p2_473; \
53240		int64x2_t __rev0_473; __rev0_473 = __builtin_shufflevector(__s0_473, __s0_473, 1, 0); \
53241		int32x4_t __rev1_473; __rev1_473 = __builtin_shufflevector(__s1_473, __s1_473, 3, 2, 1, 0); \
53242		int32x2_t __rev2_473; __rev2_473 = __builtin_shufflevector(__s2_473, __s2_473, 1, 0); \
53243		__ret_473 = __rev0_473 + __noswap_vmull_s32(__noswap_vget_high_s32(__rev1_473), __noswap_splat_lane_s32(__rev2_473, __p3_473)); \
53244		__ret_473 = __builtin_shufflevector(__ret_473, __ret_473, 1, 0); \
	54701	#define vmlaq_laneq_s16(__p0_473, __p1_473, __p2_473, __p3_473) __extension__ ({ \
	54702	int16x8_t __ret_473; \
	54703	int16x8_t __s0_473 = __p0_473; \
	54704	int16x8_t __s1_473 = __p1_473; \
	54705	int16x8_t __s2_473 = __p2_473; \
	54706	int16x8_t __rev0_473; __rev0_473 = __builtin_shufflevector(__s0_473, __s0_473, __lane_reverse_128_16); \
	54707	int16x8_t __rev1_473; __rev1_473 = __builtin_shufflevector(__s1_473, __s1_473, __lane_reverse_128_16); \
	54708	int16x8_t __rev2_473; __rev2_473 = __builtin_shufflevector(__s2_473, __s2_473, __lane_reverse_128_16); \
	54709	__ret_473 = __rev0_473 + __rev1_473 * __noswap_splatq_laneq_s16(__rev2_473, __p3_473); \
	54710	__ret_473 = __builtin_shufflevector(__ret_473, __ret_473, __lane_reverse_128_16); \
53245	54711	__ret_473; \
53246	54712	})
53247	54713	#endif
53248	54714
53249	54715	#ifdef __LITTLE_ENDIAN__
53250		#define vmlal_high_lane_s16(__p0_474, __p1_474, __p2_474, __p3_474) __extension__ ({ \
53251		int32x4_t __ret_474; \
53252		int32x4_t __s0_474 = __p0_474; \
53253		int16x8_t __s1_474 = __p1_474; \
53254		int16x4_t __s2_474 = __p2_474; \
53255		__ret_474 = __s0_474 + vmull_s16(vget_high_s16(__s1_474), splat_lane_s16(__s2_474, __p3_474)); \
	54716	#define vmla_laneq_u32(__p0_474, __p1_474, __p2_474, __p3_474) __extension__ ({ \
	54717	uint32x2_t __ret_474; \
	54718	uint32x2_t __s0_474 = __p0_474; \
	54719	uint32x2_t __s1_474 = __p1_474; \
	54720	uint32x4_t __s2_474 = __p2_474; \
	54721	__ret_474 = __s0_474 + __s1_474 * splat_laneq_u32(__s2_474, __p3_474); \
53256	54722	__ret_474; \
53257	54723	})
53258	54724	#else
53259		#define vmlal_high_lane_s16(__p0_475, __p1_475, __p2_475, __p3_475) __extension__ ({ \
53260		int32x4_t __ret_475; \
53261		int32x4_t __s0_475 = __p0_475; \
53262		int16x8_t __s1_475 = __p1_475; \
53263		int16x4_t __s2_475 = __p2_475; \
53264		int32x4_t __rev0_475; __rev0_475 = __builtin_shufflevector(__s0_475, __s0_475, 3, 2, 1, 0); \
53265		int16x8_t __rev1_475; __rev1_475 = __builtin_shufflevector(__s1_475, __s1_475, 7, 6, 5, 4, 3, 2, 1, 0); \
53266		int16x4_t __rev2_475; __rev2_475 = __builtin_shufflevector(__s2_475, __s2_475, 3, 2, 1, 0); \
53267		__ret_475 = __rev0_475 + __noswap_vmull_s16(__noswap_vget_high_s16(__rev1_475), __noswap_splat_lane_s16(__rev2_475, __p3_475)); \
53268		__ret_475 = __builtin_shufflevector(__ret_475, __ret_475, 3, 2, 1, 0); \
	54725	#define vmla_laneq_u32(__p0_475, __p1_475, __p2_475, __p3_475) __extension__ ({ \
	54726	uint32x2_t __ret_475; \
	54727	uint32x2_t __s0_475 = __p0_475; \
	54728	uint32x2_t __s1_475 = __p1_475; \
	54729	uint32x4_t __s2_475 = __p2_475; \
	54730	uint32x2_t __rev0_475; __rev0_475 = __builtin_shufflevector(__s0_475, __s0_475, __lane_reverse_64_32); \
	54731	uint32x2_t __rev1_475; __rev1_475 = __builtin_shufflevector(__s1_475, __s1_475, __lane_reverse_64_32); \
	54732	uint32x4_t __rev2_475; __rev2_475 = __builtin_shufflevector(__s2_475, __s2_475, __lane_reverse_128_32); \
	54733	__ret_475 = __rev0_475 + __rev1_475 * __noswap_splat_laneq_u32(__rev2_475, __p3_475); \
	54734	__ret_475 = __builtin_shufflevector(__ret_475, __ret_475, __lane_reverse_64_32); \
53269	54735	__ret_475; \
53270	54736	})
53271	54737	#endif
53272	54738
53273	54739	#ifdef __LITTLE_ENDIAN__
53274		#define vmlal_high_laneq_u32(__p0_476, __p1_476, __p2_476, __p3_476) __extension__ ({ \
53275		uint64x2_t __ret_476; \
53276		uint64x2_t __s0_476 = __p0_476; \
53277		uint32x4_t __s1_476 = __p1_476; \
53278		uint32x4_t __s2_476 = __p2_476; \
53279		__ret_476 = __s0_476 + vmull_u32(vget_high_u32(__s1_476), splat_laneq_u32(__s2_476, __p3_476)); \
	54740	#define vmla_laneq_u16(__p0_476, __p1_476, __p2_476, __p3_476) __extension__ ({ \
	54741	uint16x4_t __ret_476; \
	54742	uint16x4_t __s0_476 = __p0_476; \
	54743	uint16x4_t __s1_476 = __p1_476; \
	54744	uint16x8_t __s2_476 = __p2_476; \
	54745	__ret_476 = __s0_476 + __s1_476 * splat_laneq_u16(__s2_476, __p3_476); \
53280	54746	__ret_476; \
53281	54747	})
53282	54748	#else
53283		#define vmlal_high_laneq_u32(__p0_477, __p1_477, __p2_477, __p3_477) __extension__ ({ \
53284		uint64x2_t __ret_477; \
53285		uint64x2_t __s0_477 = __p0_477; \
53286		uint32x4_t __s1_477 = __p1_477; \
53287		uint32x4_t __s2_477 = __p2_477; \
53288		uint64x2_t __rev0_477; __rev0_477 = __builtin_shufflevector(__s0_477, __s0_477, 1, 0); \
53289		uint32x4_t __rev1_477; __rev1_477 = __builtin_shufflevector(__s1_477, __s1_477, 3, 2, 1, 0); \
53290		uint32x4_t __rev2_477; __rev2_477 = __builtin_shufflevector(__s2_477, __s2_477, 3, 2, 1, 0); \
53291		__ret_477 = __rev0_477 + __noswap_vmull_u32(__noswap_vget_high_u32(__rev1_477), __noswap_splat_laneq_u32(__rev2_477, __p3_477)); \
53292		__ret_477 = __builtin_shufflevector(__ret_477, __ret_477, 1, 0); \
	54749	#define vmla_laneq_u16(__p0_477, __p1_477, __p2_477, __p3_477) __extension__ ({ \
	54750	uint16x4_t __ret_477; \
	54751	uint16x4_t __s0_477 = __p0_477; \
	54752	uint16x4_t __s1_477 = __p1_477; \
	54753	uint16x8_t __s2_477 = __p2_477; \
	54754	uint16x4_t __rev0_477; __rev0_477 = __builtin_shufflevector(__s0_477, __s0_477, __lane_reverse_64_16); \
	54755	uint16x4_t __rev1_477; __rev1_477 = __builtin_shufflevector(__s1_477, __s1_477, __lane_reverse_64_16); \
	54756	uint16x8_t __rev2_477; __rev2_477 = __builtin_shufflevector(__s2_477, __s2_477, __lane_reverse_128_16); \
	54757	__ret_477 = __rev0_477 + __rev1_477 * __noswap_splat_laneq_u16(__rev2_477, __p3_477); \
	54758	__ret_477 = __builtin_shufflevector(__ret_477, __ret_477, __lane_reverse_64_16); \
53293	54759	__ret_477; \
53294	54760	})
53295	54761	#endif
53296	54762
53297	54763	#ifdef __LITTLE_ENDIAN__
53298		#define vmlal_high_laneq_u16(__p0_478, __p1_478, __p2_478, __p3_478) __extension__ ({ \
53299		uint32x4_t __ret_478; \
53300		uint32x4_t __s0_478 = __p0_478; \
53301		uint16x8_t __s1_478 = __p1_478; \
53302		uint16x8_t __s2_478 = __p2_478; \
53303		__ret_478 = __s0_478 + vmull_u16(vget_high_u16(__s1_478), splat_laneq_u16(__s2_478, __p3_478)); \
	54764	#define vmla_laneq_f32(__p0_478, __p1_478, __p2_478, __p3_478) __extension__ ({ \
	54765	float32x2_t __ret_478; \
	54766	float32x2_t __s0_478 = __p0_478; \
	54767	float32x2_t __s1_478 = __p1_478; \
	54768	float32x4_t __s2_478 = __p2_478; \
	54769	__ret_478 = __s0_478 + __s1_478 * splat_laneq_f32(__s2_478, __p3_478); \
53304	54770	__ret_478; \
53305	54771	})
53306	54772	#else
53307		#define vmlal_high_laneq_u16(__p0_479, __p1_479, __p2_479, __p3_479) __extension__ ({ \
53308		uint32x4_t __ret_479; \
53309		uint32x4_t __s0_479 = __p0_479; \
53310		uint16x8_t __s1_479 = __p1_479; \
53311		uint16x8_t __s2_479 = __p2_479; \
53312		uint32x4_t __rev0_479; __rev0_479 = __builtin_shufflevector(__s0_479, __s0_479, 3, 2, 1, 0); \
53313		uint16x8_t __rev1_479; __rev1_479 = __builtin_shufflevector(__s1_479, __s1_479, 7, 6, 5, 4, 3, 2, 1, 0); \
53314		uint16x8_t __rev2_479; __rev2_479 = __builtin_shufflevector(__s2_479, __s2_479, 7, 6, 5, 4, 3, 2, 1, 0); \
53315		__ret_479 = __rev0_479 + __noswap_vmull_u16(__noswap_vget_high_u16(__rev1_479), __noswap_splat_laneq_u16(__rev2_479, __p3_479)); \
53316		__ret_479 = __builtin_shufflevector(__ret_479, __ret_479, 3, 2, 1, 0); \
	54773	#define vmla_laneq_f32(__p0_479, __p1_479, __p2_479, __p3_479) __extension__ ({ \
	54774	float32x2_t __ret_479; \
	54775	float32x2_t __s0_479 = __p0_479; \
	54776	float32x2_t __s1_479 = __p1_479; \
	54777	float32x4_t __s2_479 = __p2_479; \
	54778	float32x2_t __rev0_479; __rev0_479 = __builtin_shufflevector(__s0_479, __s0_479, __lane_reverse_64_32); \
	54779	float32x2_t __rev1_479; __rev1_479 = __builtin_shufflevector(__s1_479, __s1_479, __lane_reverse_64_32); \
	54780	float32x4_t __rev2_479; __rev2_479 = __builtin_shufflevector(__s2_479, __s2_479, __lane_reverse_128_32); \
	54781	__ret_479 = __rev0_479 + __rev1_479 * __noswap_splat_laneq_f32(__rev2_479, __p3_479); \
	54782	__ret_479 = __builtin_shufflevector(__ret_479, __ret_479, __lane_reverse_64_32); \
53317	54783	__ret_479; \
53318	54784	})
53319	54785	#endif
53320	54786
53321	54787	#ifdef __LITTLE_ENDIAN__
53322		#define vmlal_high_laneq_s32(__p0_480, __p1_480, __p2_480, __p3_480) __extension__ ({ \
53323		int64x2_t __ret_480; \
53324		int64x2_t __s0_480 = __p0_480; \
53325		int32x4_t __s1_480 = __p1_480; \
	54788	#define vmla_laneq_s32(__p0_480, __p1_480, __p2_480, __p3_480) __extension__ ({ \
	54789	int32x2_t __ret_480; \
	54790	int32x2_t __s0_480 = __p0_480; \
	54791	int32x2_t __s1_480 = __p1_480; \
53326	54792	int32x4_t __s2_480 = __p2_480; \
53327		__ret_480 = __s0_480 + vmull_s32(vget_high_s32(__s1_480), splat_laneq_s32(__s2_480, __p3_480)); \
	54793	__ret_480 = __s0_480 + __s1_480 * splat_laneq_s32(__s2_480, __p3_480); \
53328	54794	__ret_480; \
53329	54795	})
53330	54796	#else
53331		#define vmlal_high_laneq_s32(__p0_481, __p1_481, __p2_481, __p3_481) __extension__ ({ \
53332		int64x2_t __ret_481; \
53333		int64x2_t __s0_481 = __p0_481; \
53334		int32x4_t __s1_481 = __p1_481; \
	54797	#define vmla_laneq_s32(__p0_481, __p1_481, __p2_481, __p3_481) __extension__ ({ \
	54798	int32x2_t __ret_481; \
	54799	int32x2_t __s0_481 = __p0_481; \
	54800	int32x2_t __s1_481 = __p1_481; \
53335	54801	int32x4_t __s2_481 = __p2_481; \
53336		int64x2_t __rev0_481; __rev0_481 = __builtin_shufflevector(__s0_481, __s0_481, 1, 0); \
53337		int32x4_t __rev1_481; __rev1_481 = __builtin_shufflevector(__s1_481, __s1_481, 3, 2, 1, 0); \
53338		int32x4_t __rev2_481; __rev2_481 = __builtin_shufflevector(__s2_481, __s2_481, 3, 2, 1, 0); \
53339		__ret_481 = __rev0_481 + __noswap_vmull_s32(__noswap_vget_high_s32(__rev1_481), __noswap_splat_laneq_s32(__rev2_481, __p3_481)); \
53340		__ret_481 = __builtin_shufflevector(__ret_481, __ret_481, 1, 0); \
	54802	int32x2_t __rev0_481; __rev0_481 = __builtin_shufflevector(__s0_481, __s0_481, __lane_reverse_64_32); \
	54803	int32x2_t __rev1_481; __rev1_481 = __builtin_shufflevector(__s1_481, __s1_481, __lane_reverse_64_32); \
	54804	int32x4_t __rev2_481; __rev2_481 = __builtin_shufflevector(__s2_481, __s2_481, __lane_reverse_128_32); \
	54805	__ret_481 = __rev0_481 + __rev1_481 * __noswap_splat_laneq_s32(__rev2_481, __p3_481); \
	54806	__ret_481 = __builtin_shufflevector(__ret_481, __ret_481, __lane_reverse_64_32); \
53341	54807	__ret_481; \
53342	54808	})
53343	54809	#endif
53344	54810
53345	54811	#ifdef __LITTLE_ENDIAN__
53346		#define vmlal_high_laneq_s16(__p0_482, __p1_482, __p2_482, __p3_482) __extension__ ({ \
53347		int32x4_t __ret_482; \
53348		int32x4_t __s0_482 = __p0_482; \
53349		int16x8_t __s1_482 = __p1_482; \
	54812	#define vmla_laneq_s16(__p0_482, __p1_482, __p2_482, __p3_482) __extension__ ({ \
	54813	int16x4_t __ret_482; \
	54814	int16x4_t __s0_482 = __p0_482; \
	54815	int16x4_t __s1_482 = __p1_482; \
53350	54816	int16x8_t __s2_482 = __p2_482; \
53351		__ret_482 = __s0_482 + vmull_s16(vget_high_s16(__s1_482), splat_laneq_s16(__s2_482, __p3_482)); \
	54817	__ret_482 = __s0_482 + __s1_482 * splat_laneq_s16(__s2_482, __p3_482); \
53352	54818	__ret_482; \
53353	54819	})
53354	54820	#else
53355		#define vmlal_high_laneq_s16(__p0_483, __p1_483, __p2_483, __p3_483) __extension__ ({ \
53356		int32x4_t __ret_483; \
53357		int32x4_t __s0_483 = __p0_483; \
53358		int16x8_t __s1_483 = __p1_483; \
	54821	#define vmla_laneq_s16(__p0_483, __p1_483, __p2_483, __p3_483) __extension__ ({ \
	54822	int16x4_t __ret_483; \
	54823	int16x4_t __s0_483 = __p0_483; \
	54824	int16x4_t __s1_483 = __p1_483; \
53359	54825	int16x8_t __s2_483 = __p2_483; \
53360		int32x4_t __rev0_483; __rev0_483 = __builtin_shufflevector(__s0_483, __s0_483, 3, 2, 1, 0); \
53361		int16x8_t __rev1_483; __rev1_483 = __builtin_shufflevector(__s1_483, __s1_483, 7, 6, 5, 4, 3, 2, 1, 0); \
53362		int16x8_t __rev2_483; __rev2_483 = __builtin_shufflevector(__s2_483, __s2_483, 7, 6, 5, 4, 3, 2, 1, 0); \
53363		__ret_483 = __rev0_483 + __noswap_vmull_s16(__noswap_vget_high_s16(__rev1_483), __noswap_splat_laneq_s16(__rev2_483, __p3_483)); \
53364		__ret_483 = __builtin_shufflevector(__ret_483, __ret_483, 3, 2, 1, 0); \
	54826	int16x4_t __rev0_483; __rev0_483 = __builtin_shufflevector(__s0_483, __s0_483, __lane_reverse_64_16); \
	54827	int16x4_t __rev1_483; __rev1_483 = __builtin_shufflevector(__s1_483, __s1_483, __lane_reverse_64_16); \
	54828	int16x8_t __rev2_483; __rev2_483 = __builtin_shufflevector(__s2_483, __s2_483, __lane_reverse_128_16); \
	54829	__ret_483 = __rev0_483 + __rev1_483 * __noswap_splat_laneq_s16(__rev2_483, __p3_483); \
	54830	__ret_483 = __builtin_shufflevector(__ret_483, __ret_483, __lane_reverse_64_16); \
53365	54831	__ret_483; \
53366	54832	})
53367	54833	#endif
53368	54834
53369	54835	#ifdef __LITTLE_ENDIAN__
53370		#define vmlal_laneq_u32(__p0_484, __p1_484, __p2_484, __p3_484) __extension__ ({ \
	54836	#define vmlal_high_lane_u32(__p0_484, __p1_484, __p2_484, __p3_484) __extension__ ({ \
53371	54837	uint64x2_t __ret_484; \
53372	54838	uint64x2_t __s0_484 = __p0_484; \
53373		uint32x2_t __s1_484 = __p1_484; \
53374		uint32x4_t __s2_484 = __p2_484; \
53375		__ret_484 = __s0_484 + vmull_u32(__s1_484, splat_laneq_u32(__s2_484, __p3_484)); \
	54839	uint32x4_t __s1_484 = __p1_484; \
	54840	uint32x2_t __s2_484 = __p2_484; \
	54841	__ret_484 = __s0_484 + vmull_u32(vget_high_u32(__s1_484), splat_lane_u32(__s2_484, __p3_484)); \
53376	54842	__ret_484; \
53377	54843	})
53378	54844	#else
53379		#define vmlal_laneq_u32(__p0_485, __p1_485, __p2_485, __p3_485) __extension__ ({ \
	54845	#define vmlal_high_lane_u32(__p0_485, __p1_485, __p2_485, __p3_485) __extension__ ({ \
53380	54846	uint64x2_t __ret_485; \
53381	54847	uint64x2_t __s0_485 = __p0_485; \
53382		uint32x2_t __s1_485 = __p1_485; \
53383		uint32x4_t __s2_485 = __p2_485; \
53384		uint64x2_t __rev0_485; __rev0_485 = __builtin_shufflevector(__s0_485, __s0_485, 1, 0); \
53385		uint32x2_t __rev1_485; __rev1_485 = __builtin_shufflevector(__s1_485, __s1_485, 1, 0); \
53386		uint32x4_t __rev2_485; __rev2_485 = __builtin_shufflevector(__s2_485, __s2_485, 3, 2, 1, 0); \
53387		__ret_485 = __rev0_485 + __noswap_vmull_u32(__rev1_485, __noswap_splat_laneq_u32(__rev2_485, __p3_485)); \
53388		__ret_485 = __builtin_shufflevector(__ret_485, __ret_485, 1, 0); \
	54848	uint32x4_t __s1_485 = __p1_485; \
	54849	uint32x2_t __s2_485 = __p2_485; \
	54850	uint64x2_t __rev0_485; __rev0_485 = __builtin_shufflevector(__s0_485, __s0_485, __lane_reverse_128_64); \
	54851	uint32x4_t __rev1_485; __rev1_485 = __builtin_shufflevector(__s1_485, __s1_485, __lane_reverse_128_32); \
	54852	uint32x2_t __rev2_485; __rev2_485 = __builtin_shufflevector(__s2_485, __s2_485, __lane_reverse_64_32); \
	54853	__ret_485 = __rev0_485 + __noswap_vmull_u32(__noswap_vget_high_u32(__rev1_485), __noswap_splat_lane_u32(__rev2_485, __p3_485)); \
	54854	__ret_485 = __builtin_shufflevector(__ret_485, __ret_485, __lane_reverse_128_64); \
53389	54855	__ret_485; \
53390	54856	})
53391	54857	#endif
53392	54858
53393	54859	#ifdef __LITTLE_ENDIAN__
53394		#define vmlal_laneq_u16(__p0_486, __p1_486, __p2_486, __p3_486) __extension__ ({ \
	54860	#define vmlal_high_lane_u16(__p0_486, __p1_486, __p2_486, __p3_486) __extension__ ({ \
53395	54861	uint32x4_t __ret_486; \
53396	54862	uint32x4_t __s0_486 = __p0_486; \
53397		uint16x4_t __s1_486 = __p1_486; \
53398		uint16x8_t __s2_486 = __p2_486; \
53399		__ret_486 = __s0_486 + vmull_u16(__s1_486, splat_laneq_u16(__s2_486, __p3_486)); \
	54863	uint16x8_t __s1_486 = __p1_486; \
	54864	uint16x4_t __s2_486 = __p2_486; \
	54865	__ret_486 = __s0_486 + vmull_u16(vget_high_u16(__s1_486), splat_lane_u16(__s2_486, __p3_486)); \
53400	54866	__ret_486; \
53401	54867	})
53402	54868	#else
53403		#define vmlal_laneq_u16(__p0_487, __p1_487, __p2_487, __p3_487) __extension__ ({ \
	54869	#define vmlal_high_lane_u16(__p0_487, __p1_487, __p2_487, __p3_487) __extension__ ({ \
53404	54870	uint32x4_t __ret_487; \
53405	54871	uint32x4_t __s0_487 = __p0_487; \
53406		uint16x4_t __s1_487 = __p1_487; \
53407		uint16x8_t __s2_487 = __p2_487; \
53408		uint32x4_t __rev0_487; __rev0_487 = __builtin_shufflevector(__s0_487, __s0_487, 3, 2, 1, 0); \
53409		uint16x4_t __rev1_487; __rev1_487 = __builtin_shufflevector(__s1_487, __s1_487, 3, 2, 1, 0); \
53410		uint16x8_t __rev2_487; __rev2_487 = __builtin_shufflevector(__s2_487, __s2_487, 7, 6, 5, 4, 3, 2, 1, 0); \
53411		__ret_487 = __rev0_487 + __noswap_vmull_u16(__rev1_487, __noswap_splat_laneq_u16(__rev2_487, __p3_487)); \
53412		__ret_487 = __builtin_shufflevector(__ret_487, __ret_487, 3, 2, 1, 0); \
	54872	uint16x8_t __s1_487 = __p1_487; \
	54873	uint16x4_t __s2_487 = __p2_487; \
	54874	uint32x4_t __rev0_487; __rev0_487 = __builtin_shufflevector(__s0_487, __s0_487, __lane_reverse_128_32); \
	54875	uint16x8_t __rev1_487; __rev1_487 = __builtin_shufflevector(__s1_487, __s1_487, __lane_reverse_128_16); \
	54876	uint16x4_t __rev2_487; __rev2_487 = __builtin_shufflevector(__s2_487, __s2_487, __lane_reverse_64_16); \
	54877	__ret_487 = __rev0_487 + __noswap_vmull_u16(__noswap_vget_high_u16(__rev1_487), __noswap_splat_lane_u16(__rev2_487, __p3_487)); \
	54878	__ret_487 = __builtin_shufflevector(__ret_487, __ret_487, __lane_reverse_128_32); \
53413	54879	__ret_487; \
53414	54880	})
53415	54881	#endif
53416	54882
53417	54883	#ifdef __LITTLE_ENDIAN__
53418		#define vmlal_laneq_s32(__p0_488, __p1_488, __p2_488, __p3_488) __extension__ ({ \
	54884	#define vmlal_high_lane_s32(__p0_488, __p1_488, __p2_488, __p3_488) __extension__ ({ \
53419	54885	int64x2_t __ret_488; \
53420	54886	int64x2_t __s0_488 = __p0_488; \
53421		int32x2_t __s1_488 = __p1_488; \
53422		int32x4_t __s2_488 = __p2_488; \
53423		__ret_488 = __s0_488 + vmull_s32(__s1_488, splat_laneq_s32(__s2_488, __p3_488)); \
	54887	int32x4_t __s1_488 = __p1_488; \
	54888	int32x2_t __s2_488 = __p2_488; \
	54889	__ret_488 = __s0_488 + vmull_s32(vget_high_s32(__s1_488), splat_lane_s32(__s2_488, __p3_488)); \
53424	54890	__ret_488; \
53425	54891	})
53426	54892	#else
53427		#define vmlal_laneq_s32(__p0_489, __p1_489, __p2_489, __p3_489) __extension__ ({ \
	54893	#define vmlal_high_lane_s32(__p0_489, __p1_489, __p2_489, __p3_489) __extension__ ({ \
53428	54894	int64x2_t __ret_489; \
53429	54895	int64x2_t __s0_489 = __p0_489; \
53430		int32x2_t __s1_489 = __p1_489; \
53431		int32x4_t __s2_489 = __p2_489; \
53432		int64x2_t __rev0_489; __rev0_489 = __builtin_shufflevector(__s0_489, __s0_489, 1, 0); \
53433		int32x2_t __rev1_489; __rev1_489 = __builtin_shufflevector(__s1_489, __s1_489, 1, 0); \
53434		int32x4_t __rev2_489; __rev2_489 = __builtin_shufflevector(__s2_489, __s2_489, 3, 2, 1, 0); \
53435		__ret_489 = __rev0_489 + __noswap_vmull_s32(__rev1_489, __noswap_splat_laneq_s32(__rev2_489, __p3_489)); \
53436		__ret_489 = __builtin_shufflevector(__ret_489, __ret_489, 1, 0); \
	54896	int32x4_t __s1_489 = __p1_489; \
	54897	int32x2_t __s2_489 = __p2_489; \
	54898	int64x2_t __rev0_489; __rev0_489 = __builtin_shufflevector(__s0_489, __s0_489, __lane_reverse_128_64); \
	54899	int32x4_t __rev1_489; __rev1_489 = __builtin_shufflevector(__s1_489, __s1_489, __lane_reverse_128_32); \
	54900	int32x2_t __rev2_489; __rev2_489 = __builtin_shufflevector(__s2_489, __s2_489, __lane_reverse_64_32); \
	54901	__ret_489 = __rev0_489 + __noswap_vmull_s32(__noswap_vget_high_s32(__rev1_489), __noswap_splat_lane_s32(__rev2_489, __p3_489)); \
	54902	__ret_489 = __builtin_shufflevector(__ret_489, __ret_489, __lane_reverse_128_64); \
53437	54903	__ret_489; \
53438	54904	})
53439	54905	#endif
53440	54906
53441	54907	#ifdef __LITTLE_ENDIAN__
53442		#define vmlal_laneq_s16(__p0_490, __p1_490, __p2_490, __p3_490) __extension__ ({ \
	54908	#define vmlal_high_lane_s16(__p0_490, __p1_490, __p2_490, __p3_490) __extension__ ({ \
53443	54909	int32x4_t __ret_490; \
53444	54910	int32x4_t __s0_490 = __p0_490; \
53445		int16x4_t __s1_490 = __p1_490; \
53446		int16x8_t __s2_490 = __p2_490; \
53447		__ret_490 = __s0_490 + vmull_s16(__s1_490, splat_laneq_s16(__s2_490, __p3_490)); \
	54911	int16x8_t __s1_490 = __p1_490; \
	54912	int16x4_t __s2_490 = __p2_490; \
	54913	__ret_490 = __s0_490 + vmull_s16(vget_high_s16(__s1_490), splat_lane_s16(__s2_490, __p3_490)); \
53448	54914	__ret_490; \
53449	54915	})
53450	54916	#else
53451		#define vmlal_laneq_s16(__p0_491, __p1_491, __p2_491, __p3_491) __extension__ ({ \
	54917	#define vmlal_high_lane_s16(__p0_491, __p1_491, __p2_491, __p3_491) __extension__ ({ \
53452	54918	int32x4_t __ret_491; \
53453	54919	int32x4_t __s0_491 = __p0_491; \
53454		int16x4_t __s1_491 = __p1_491; \
53455		int16x8_t __s2_491 = __p2_491; \
53456		int32x4_t __rev0_491; __rev0_491 = __builtin_shufflevector(__s0_491, __s0_491, 3, 2, 1, 0); \
53457		int16x4_t __rev1_491; __rev1_491 = __builtin_shufflevector(__s1_491, __s1_491, 3, 2, 1, 0); \
53458		int16x8_t __rev2_491; __rev2_491 = __builtin_shufflevector(__s2_491, __s2_491, 7, 6, 5, 4, 3, 2, 1, 0); \
53459		__ret_491 = __rev0_491 + __noswap_vmull_s16(__rev1_491, __noswap_splat_laneq_s16(__rev2_491, __p3_491)); \
53460		__ret_491 = __builtin_shufflevector(__ret_491, __ret_491, 3, 2, 1, 0); \
	54920	int16x8_t __s1_491 = __p1_491; \
	54921	int16x4_t __s2_491 = __p2_491; \
	54922	int32x4_t __rev0_491; __rev0_491 = __builtin_shufflevector(__s0_491, __s0_491, __lane_reverse_128_32); \
	54923	int16x8_t __rev1_491; __rev1_491 = __builtin_shufflevector(__s1_491, __s1_491, __lane_reverse_128_16); \
	54924	int16x4_t __rev2_491; __rev2_491 = __builtin_shufflevector(__s2_491, __s2_491, __lane_reverse_64_16); \
	54925	__ret_491 = __rev0_491 + __noswap_vmull_s16(__noswap_vget_high_s16(__rev1_491), __noswap_splat_lane_s16(__rev2_491, __p3_491)); \
	54926	__ret_491 = __builtin_shufflevector(__ret_491, __ret_491, __lane_reverse_128_32); \
53461	54927	__ret_491; \
53462	54928	})
53463	54929	#endif
53464	54930
53465	54931	#ifdef __LITTLE_ENDIAN__
53466		__ai __attribute__((target("neon"))) float64x2_t vmlsq_f64(float64x2_t __p0, float64x2_t __p1, float64x2_t __p2) {
53467		float64x2_t __ret;
53468		__ret = __p0 - __p1 * __p2;
53469		return __ret;
53470		}
53471		#else
53472		__ai __attribute__((target("neon"))) float64x2_t vmlsq_f64(float64x2_t __p0, float64x2_t __p1, float64x2_t __p2) {
53473		float64x2_t __ret;
53474		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
53475		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
53476		float64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
53477		__ret = __rev0 - __rev1 * __rev2;
53478		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
53479		return __ret;
53480		}
53481		#endif
53482
53483		__ai __attribute__((target("neon"))) float64x1_t vmls_f64(float64x1_t __p0, float64x1_t __p1, float64x1_t __p2) {
53484		float64x1_t __ret;
53485		__ret = __p0 - __p1 * __p2;
53486		return __ret;
53487		}
53488		#ifdef __LITTLE_ENDIAN__
53489		#define vmlsq_laneq_u32(__p0_492, __p1_492, __p2_492, __p3_492) __extension__ ({ \
53490		uint32x4_t __ret_492; \
53491		uint32x4_t __s0_492 = __p0_492; \
	54932	#define vmlal_high_laneq_u32(__p0_492, __p1_492, __p2_492, __p3_492) __extension__ ({ \
	54933	uint64x2_t __ret_492; \
	54934	uint64x2_t __s0_492 = __p0_492; \
53492	54935	uint32x4_t __s1_492 = __p1_492; \
53493	54936	uint32x4_t __s2_492 = __p2_492; \
53494		__ret_492 = __s0_492 - __s1_492 * splatq_laneq_u32(__s2_492, __p3_492); \
	54937	__ret_492 = __s0_492 + vmull_u32(vget_high_u32(__s1_492), splat_laneq_u32(__s2_492, __p3_492)); \
53495	54938	__ret_492; \
53496	54939	})
53497	54940	#else
53498		#define vmlsq_laneq_u32(__p0_493, __p1_493, __p2_493, __p3_493) __extension__ ({ \
53499		uint32x4_t __ret_493; \
53500		uint32x4_t __s0_493 = __p0_493; \
	54941	#define vmlal_high_laneq_u32(__p0_493, __p1_493, __p2_493, __p3_493) __extension__ ({ \
	54942	uint64x2_t __ret_493; \
	54943	uint64x2_t __s0_493 = __p0_493; \
53501	54944	uint32x4_t __s1_493 = __p1_493; \
53502	54945	uint32x4_t __s2_493 = __p2_493; \
53503		uint32x4_t __rev0_493; __rev0_493 = __builtin_shufflevector(__s0_493, __s0_493, 3, 2, 1, 0); \
53504		uint32x4_t __rev1_493; __rev1_493 = __builtin_shufflevector(__s1_493, __s1_493, 3, 2, 1, 0); \
53505		uint32x4_t __rev2_493; __rev2_493 = __builtin_shufflevector(__s2_493, __s2_493, 3, 2, 1, 0); \
53506		__ret_493 = __rev0_493 - __rev1_493 * __noswap_splatq_laneq_u32(__rev2_493, __p3_493); \
53507		__ret_493 = __builtin_shufflevector(__ret_493, __ret_493, 3, 2, 1, 0); \
	54946	uint64x2_t __rev0_493; __rev0_493 = __builtin_shufflevector(__s0_493, __s0_493, __lane_reverse_128_64); \
	54947	uint32x4_t __rev1_493; __rev1_493 = __builtin_shufflevector(__s1_493, __s1_493, __lane_reverse_128_32); \
	54948	uint32x4_t __rev2_493; __rev2_493 = __builtin_shufflevector(__s2_493, __s2_493, __lane_reverse_128_32); \
	54949	__ret_493 = __rev0_493 + __noswap_vmull_u32(__noswap_vget_high_u32(__rev1_493), __noswap_splat_laneq_u32(__rev2_493, __p3_493)); \
	54950	__ret_493 = __builtin_shufflevector(__ret_493, __ret_493, __lane_reverse_128_64); \
53508	54951	__ret_493; \
53509	54952	})
53510	54953	#endif
53511	54954
53512	54955	#ifdef __LITTLE_ENDIAN__
53513		#define vmlsq_laneq_u16(__p0_494, __p1_494, __p2_494, __p3_494) __extension__ ({ \
53514		uint16x8_t __ret_494; \
53515		uint16x8_t __s0_494 = __p0_494; \
	54956	#define vmlal_high_laneq_u16(__p0_494, __p1_494, __p2_494, __p3_494) __extension__ ({ \
	54957	uint32x4_t __ret_494; \
	54958	uint32x4_t __s0_494 = __p0_494; \
53516	54959	uint16x8_t __s1_494 = __p1_494; \
53517	54960	uint16x8_t __s2_494 = __p2_494; \
53518		__ret_494 = __s0_494 - __s1_494 * splatq_laneq_u16(__s2_494, __p3_494); \
	54961	__ret_494 = __s0_494 + vmull_u16(vget_high_u16(__s1_494), splat_laneq_u16(__s2_494, __p3_494)); \
53519	54962	__ret_494; \
53520	54963	})
53521	54964	#else
53522		#define vmlsq_laneq_u16(__p0_495, __p1_495, __p2_495, __p3_495) __extension__ ({ \
53523		uint16x8_t __ret_495; \
53524		uint16x8_t __s0_495 = __p0_495; \
	54965	#define vmlal_high_laneq_u16(__p0_495, __p1_495, __p2_495, __p3_495) __extension__ ({ \
	54966	uint32x4_t __ret_495; \
	54967	uint32x4_t __s0_495 = __p0_495; \
53525	54968	uint16x8_t __s1_495 = __p1_495; \
53526	54969	uint16x8_t __s2_495 = __p2_495; \
53527		uint16x8_t __rev0_495; __rev0_495 = __builtin_shufflevector(__s0_495, __s0_495, 7, 6, 5, 4, 3, 2, 1, 0); \
53528		uint16x8_t __rev1_495; __rev1_495 = __builtin_shufflevector(__s1_495, __s1_495, 7, 6, 5, 4, 3, 2, 1, 0); \
53529		uint16x8_t __rev2_495; __rev2_495 = __builtin_shufflevector(__s2_495, __s2_495, 7, 6, 5, 4, 3, 2, 1, 0); \
53530		__ret_495 = __rev0_495 - __rev1_495 * __noswap_splatq_laneq_u16(__rev2_495, __p3_495); \
53531		__ret_495 = __builtin_shufflevector(__ret_495, __ret_495, 7, 6, 5, 4, 3, 2, 1, 0); \
	54970	uint32x4_t __rev0_495; __rev0_495 = __builtin_shufflevector(__s0_495, __s0_495, __lane_reverse_128_32); \
	54971	uint16x8_t __rev1_495; __rev1_495 = __builtin_shufflevector(__s1_495, __s1_495, __lane_reverse_128_16); \
	54972	uint16x8_t __rev2_495; __rev2_495 = __builtin_shufflevector(__s2_495, __s2_495, __lane_reverse_128_16); \
	54973	__ret_495 = __rev0_495 + __noswap_vmull_u16(__noswap_vget_high_u16(__rev1_495), __noswap_splat_laneq_u16(__rev2_495, __p3_495)); \
	54974	__ret_495 = __builtin_shufflevector(__ret_495, __ret_495, __lane_reverse_128_32); \
53532	54975	__ret_495; \
53533	54976	})
53534	54977	#endif
53535	54978
53536	54979	#ifdef __LITTLE_ENDIAN__
53537		#define vmlsq_laneq_f32(__p0_496, __p1_496, __p2_496, __p3_496) __extension__ ({ \
53538		float32x4_t __ret_496; \
53539		float32x4_t __s0_496 = __p0_496; \
53540		float32x4_t __s1_496 = __p1_496; \
53541		float32x4_t __s2_496 = __p2_496; \
53542		__ret_496 = __s0_496 - __s1_496 * splatq_laneq_f32(__s2_496, __p3_496); \
	54980	#define vmlal_high_laneq_s32(__p0_496, __p1_496, __p2_496, __p3_496) __extension__ ({ \
	54981	int64x2_t __ret_496; \
	54982	int64x2_t __s0_496 = __p0_496; \
	54983	int32x4_t __s1_496 = __p1_496; \
	54984	int32x4_t __s2_496 = __p2_496; \
	54985	__ret_496 = __s0_496 + vmull_s32(vget_high_s32(__s1_496), splat_laneq_s32(__s2_496, __p3_496)); \
53543	54986	__ret_496; \
53544	54987	})
53545	54988	#else
53546		#define vmlsq_laneq_f32(__p0_497, __p1_497, __p2_497, __p3_497) __extension__ ({ \
53547		float32x4_t __ret_497; \
53548		float32x4_t __s0_497 = __p0_497; \
53549		float32x4_t __s1_497 = __p1_497; \
53550		float32x4_t __s2_497 = __p2_497; \
53551		float32x4_t __rev0_497; __rev0_497 = __builtin_shufflevector(__s0_497, __s0_497, 3, 2, 1, 0); \
53552		float32x4_t __rev1_497; __rev1_497 = __builtin_shufflevector(__s1_497, __s1_497, 3, 2, 1, 0); \
53553		float32x4_t __rev2_497; __rev2_497 = __builtin_shufflevector(__s2_497, __s2_497, 3, 2, 1, 0); \
53554		__ret_497 = __rev0_497 - __rev1_497 * __noswap_splatq_laneq_f32(__rev2_497, __p3_497); \
53555		__ret_497 = __builtin_shufflevector(__ret_497, __ret_497, 3, 2, 1, 0); \
	54989	#define vmlal_high_laneq_s32(__p0_497, __p1_497, __p2_497, __p3_497) __extension__ ({ \
	54990	int64x2_t __ret_497; \
	54991	int64x2_t __s0_497 = __p0_497; \
	54992	int32x4_t __s1_497 = __p1_497; \
	54993	int32x4_t __s2_497 = __p2_497; \
	54994	int64x2_t __rev0_497; __rev0_497 = __builtin_shufflevector(__s0_497, __s0_497, __lane_reverse_128_64); \
	54995	int32x4_t __rev1_497; __rev1_497 = __builtin_shufflevector(__s1_497, __s1_497, __lane_reverse_128_32); \
	54996	int32x4_t __rev2_497; __rev2_497 = __builtin_shufflevector(__s2_497, __s2_497, __lane_reverse_128_32); \
	54997	__ret_497 = __rev0_497 + __noswap_vmull_s32(__noswap_vget_high_s32(__rev1_497), __noswap_splat_laneq_s32(__rev2_497, __p3_497)); \
	54998	__ret_497 = __builtin_shufflevector(__ret_497, __ret_497, __lane_reverse_128_64); \
53556	54999	__ret_497; \
53557	55000	})
53558	55001	#endif
53559	55002
53560	55003	#ifdef __LITTLE_ENDIAN__
53561		#define vmlsq_laneq_s32(__p0_498, __p1_498, __p2_498, __p3_498) __extension__ ({ \
	55004	#define vmlal_high_laneq_s16(__p0_498, __p1_498, __p2_498, __p3_498) __extension__ ({ \
53562	55005	int32x4_t __ret_498; \
53563	55006	int32x4_t __s0_498 = __p0_498; \
53564		int32x4_t __s1_498 = __p1_498; \
53565		int32x4_t __s2_498 = __p2_498; \
53566		__ret_498 = __s0_498 - __s1_498 * splatq_laneq_s32(__s2_498, __p3_498); \
	55007	int16x8_t __s1_498 = __p1_498; \
	55008	int16x8_t __s2_498 = __p2_498; \
	55009	__ret_498 = __s0_498 + vmull_s16(vget_high_s16(__s1_498), splat_laneq_s16(__s2_498, __p3_498)); \
53567	55010	__ret_498; \
53568	55011	})
53569	55012	#else
53570		#define vmlsq_laneq_s32(__p0_499, __p1_499, __p2_499, __p3_499) __extension__ ({ \
	55013	#define vmlal_high_laneq_s16(__p0_499, __p1_499, __p2_499, __p3_499) __extension__ ({ \
53571	55014	int32x4_t __ret_499; \
53572	55015	int32x4_t __s0_499 = __p0_499; \
53573		int32x4_t __s1_499 = __p1_499; \
53574		int32x4_t __s2_499 = __p2_499; \
53575		int32x4_t __rev0_499; __rev0_499 = __builtin_shufflevector(__s0_499, __s0_499, 3, 2, 1, 0); \
53576		int32x4_t __rev1_499; __rev1_499 = __builtin_shufflevector(__s1_499, __s1_499, 3, 2, 1, 0); \
53577		int32x4_t __rev2_499; __rev2_499 = __builtin_shufflevector(__s2_499, __s2_499, 3, 2, 1, 0); \
53578		__ret_499 = __rev0_499 - __rev1_499 * __noswap_splatq_laneq_s32(__rev2_499, __p3_499); \
53579		__ret_499 = __builtin_shufflevector(__ret_499, __ret_499, 3, 2, 1, 0); \
	55016	int16x8_t __s1_499 = __p1_499; \
	55017	int16x8_t __s2_499 = __p2_499; \
	55018	int32x4_t __rev0_499; __rev0_499 = __builtin_shufflevector(__s0_499, __s0_499, __lane_reverse_128_32); \
	55019	int16x8_t __rev1_499; __rev1_499 = __builtin_shufflevector(__s1_499, __s1_499, __lane_reverse_128_16); \
	55020	int16x8_t __rev2_499; __rev2_499 = __builtin_shufflevector(__s2_499, __s2_499, __lane_reverse_128_16); \
	55021	__ret_499 = __rev0_499 + __noswap_vmull_s16(__noswap_vget_high_s16(__rev1_499), __noswap_splat_laneq_s16(__rev2_499, __p3_499)); \
	55022	__ret_499 = __builtin_shufflevector(__ret_499, __ret_499, __lane_reverse_128_32); \
53580	55023	__ret_499; \
53581	55024	})
53582	55025	#endif
53583	55026
53584	55027	#ifdef __LITTLE_ENDIAN__
53585		#define vmlsq_laneq_s16(__p0_500, __p1_500, __p2_500, __p3_500) __extension__ ({ \
53586		int16x8_t __ret_500; \
53587		int16x8_t __s0_500 = __p0_500; \
53588		int16x8_t __s1_500 = __p1_500; \
53589		int16x8_t __s2_500 = __p2_500; \
53590		__ret_500 = __s0_500 - __s1_500 * splatq_laneq_s16(__s2_500, __p3_500); \
	55028	#define vmlal_laneq_u32(__p0_500, __p1_500, __p2_500, __p3_500) __extension__ ({ \
	55029	uint64x2_t __ret_500; \
	55030	uint64x2_t __s0_500 = __p0_500; \
	55031	uint32x2_t __s1_500 = __p1_500; \
	55032	uint32x4_t __s2_500 = __p2_500; \
	55033	__ret_500 = __s0_500 + vmull_u32(__s1_500, splat_laneq_u32(__s2_500, __p3_500)); \
53591	55034	__ret_500; \
53592	55035	})
53593	55036	#else
53594		#define vmlsq_laneq_s16(__p0_501, __p1_501, __p2_501, __p3_501) __extension__ ({ \
53595		int16x8_t __ret_501; \
53596		int16x8_t __s0_501 = __p0_501; \
53597		int16x8_t __s1_501 = __p1_501; \
53598		int16x8_t __s2_501 = __p2_501; \
53599		int16x8_t __rev0_501; __rev0_501 = __builtin_shufflevector(__s0_501, __s0_501, 7, 6, 5, 4, 3, 2, 1, 0); \
53600		int16x8_t __rev1_501; __rev1_501 = __builtin_shufflevector(__s1_501, __s1_501, 7, 6, 5, 4, 3, 2, 1, 0); \
53601		int16x8_t __rev2_501; __rev2_501 = __builtin_shufflevector(__s2_501, __s2_501, 7, 6, 5, 4, 3, 2, 1, 0); \
53602		__ret_501 = __rev0_501 - __rev1_501 * __noswap_splatq_laneq_s16(__rev2_501, __p3_501); \
53603		__ret_501 = __builtin_shufflevector(__ret_501, __ret_501, 7, 6, 5, 4, 3, 2, 1, 0); \
	55037	#define vmlal_laneq_u32(__p0_501, __p1_501, __p2_501, __p3_501) __extension__ ({ \
	55038	uint64x2_t __ret_501; \
	55039	uint64x2_t __s0_501 = __p0_501; \
	55040	uint32x2_t __s1_501 = __p1_501; \
	55041	uint32x4_t __s2_501 = __p2_501; \
	55042	uint64x2_t __rev0_501; __rev0_501 = __builtin_shufflevector(__s0_501, __s0_501, __lane_reverse_128_64); \
	55043	uint32x2_t __rev1_501; __rev1_501 = __builtin_shufflevector(__s1_501, __s1_501, __lane_reverse_64_32); \
	55044	uint32x4_t __rev2_501; __rev2_501 = __builtin_shufflevector(__s2_501, __s2_501, __lane_reverse_128_32); \
	55045	__ret_501 = __rev0_501 + __noswap_vmull_u32(__rev1_501, __noswap_splat_laneq_u32(__rev2_501, __p3_501)); \
	55046	__ret_501 = __builtin_shufflevector(__ret_501, __ret_501, __lane_reverse_128_64); \
53604	55047	__ret_501; \
53605	55048	})
53606	55049	#endif
53607	55050
53608	55051	#ifdef __LITTLE_ENDIAN__
53609		#define vmls_laneq_u32(__p0_502, __p1_502, __p2_502, __p3_502) __extension__ ({ \
53610		uint32x2_t __ret_502; \
53611		uint32x2_t __s0_502 = __p0_502; \
53612		uint32x2_t __s1_502 = __p1_502; \
53613		uint32x4_t __s2_502 = __p2_502; \
53614		__ret_502 = __s0_502 - __s1_502 * splat_laneq_u32(__s2_502, __p3_502); \
	55052	#define vmlal_laneq_u16(__p0_502, __p1_502, __p2_502, __p3_502) __extension__ ({ \
	55053	uint32x4_t __ret_502; \
	55054	uint32x4_t __s0_502 = __p0_502; \
	55055	uint16x4_t __s1_502 = __p1_502; \
	55056	uint16x8_t __s2_502 = __p2_502; \
	55057	__ret_502 = __s0_502 + vmull_u16(__s1_502, splat_laneq_u16(__s2_502, __p3_502)); \
53615	55058	__ret_502; \
53616	55059	})
53617	55060	#else
53618		#define vmls_laneq_u32(__p0_503, __p1_503, __p2_503, __p3_503) __extension__ ({ \
53619		uint32x2_t __ret_503; \
53620		uint32x2_t __s0_503 = __p0_503; \
53621		uint32x2_t __s1_503 = __p1_503; \
53622		uint32x4_t __s2_503 = __p2_503; \
53623		uint32x2_t __rev0_503; __rev0_503 = __builtin_shufflevector(__s0_503, __s0_503, 1, 0); \
53624		uint32x2_t __rev1_503; __rev1_503 = __builtin_shufflevector(__s1_503, __s1_503, 1, 0); \
53625		uint32x4_t __rev2_503; __rev2_503 = __builtin_shufflevector(__s2_503, __s2_503, 3, 2, 1, 0); \
53626		__ret_503 = __rev0_503 - __rev1_503 * __noswap_splat_laneq_u32(__rev2_503, __p3_503); \
53627		__ret_503 = __builtin_shufflevector(__ret_503, __ret_503, 1, 0); \
	55061	#define vmlal_laneq_u16(__p0_503, __p1_503, __p2_503, __p3_503) __extension__ ({ \
	55062	uint32x4_t __ret_503; \
	55063	uint32x4_t __s0_503 = __p0_503; \
	55064	uint16x4_t __s1_503 = __p1_503; \
	55065	uint16x8_t __s2_503 = __p2_503; \
	55066	uint32x4_t __rev0_503; __rev0_503 = __builtin_shufflevector(__s0_503, __s0_503, __lane_reverse_128_32); \
	55067	uint16x4_t __rev1_503; __rev1_503 = __builtin_shufflevector(__s1_503, __s1_503, __lane_reverse_64_16); \
	55068	uint16x8_t __rev2_503; __rev2_503 = __builtin_shufflevector(__s2_503, __s2_503, __lane_reverse_128_16); \
	55069	__ret_503 = __rev0_503 + __noswap_vmull_u16(__rev1_503, __noswap_splat_laneq_u16(__rev2_503, __p3_503)); \
	55070	__ret_503 = __builtin_shufflevector(__ret_503, __ret_503, __lane_reverse_128_32); \
53628	55071	__ret_503; \
53629	55072	})
53630	55073	#endif
53631	55074
53632	55075	#ifdef __LITTLE_ENDIAN__
53633		#define vmls_laneq_u16(__p0_504, __p1_504, __p2_504, __p3_504) __extension__ ({ \
53634		uint16x4_t __ret_504; \
53635		uint16x4_t __s0_504 = __p0_504; \
53636		uint16x4_t __s1_504 = __p1_504; \
53637		uint16x8_t __s2_504 = __p2_504; \
53638		__ret_504 = __s0_504 - __s1_504 * splat_laneq_u16(__s2_504, __p3_504); \
	55076	#define vmlal_laneq_s32(__p0_504, __p1_504, __p2_504, __p3_504) __extension__ ({ \
	55077	int64x2_t __ret_504; \
	55078	int64x2_t __s0_504 = __p0_504; \
	55079	int32x2_t __s1_504 = __p1_504; \
	55080	int32x4_t __s2_504 = __p2_504; \
	55081	__ret_504 = __s0_504 + vmull_s32(__s1_504, splat_laneq_s32(__s2_504, __p3_504)); \
53639	55082	__ret_504; \
53640	55083	})
53641	55084	#else
53642		#define vmls_laneq_u16(__p0_505, __p1_505, __p2_505, __p3_505) __extension__ ({ \
53643		uint16x4_t __ret_505; \
53644		uint16x4_t __s0_505 = __p0_505; \
53645		uint16x4_t __s1_505 = __p1_505; \
53646		uint16x8_t __s2_505 = __p2_505; \
53647		uint16x4_t __rev0_505; __rev0_505 = __builtin_shufflevector(__s0_505, __s0_505, 3, 2, 1, 0); \
53648		uint16x4_t __rev1_505; __rev1_505 = __builtin_shufflevector(__s1_505, __s1_505, 3, 2, 1, 0); \
53649		uint16x8_t __rev2_505; __rev2_505 = __builtin_shufflevector(__s2_505, __s2_505, 7, 6, 5, 4, 3, 2, 1, 0); \
53650		__ret_505 = __rev0_505 - __rev1_505 * __noswap_splat_laneq_u16(__rev2_505, __p3_505); \
53651		__ret_505 = __builtin_shufflevector(__ret_505, __ret_505, 3, 2, 1, 0); \
	55085	#define vmlal_laneq_s32(__p0_505, __p1_505, __p2_505, __p3_505) __extension__ ({ \
	55086	int64x2_t __ret_505; \
	55087	int64x2_t __s0_505 = __p0_505; \
	55088	int32x2_t __s1_505 = __p1_505; \
	55089	int32x4_t __s2_505 = __p2_505; \
	55090	int64x2_t __rev0_505; __rev0_505 = __builtin_shufflevector(__s0_505, __s0_505, __lane_reverse_128_64); \
	55091	int32x2_t __rev1_505; __rev1_505 = __builtin_shufflevector(__s1_505, __s1_505, __lane_reverse_64_32); \
	55092	int32x4_t __rev2_505; __rev2_505 = __builtin_shufflevector(__s2_505, __s2_505, __lane_reverse_128_32); \
	55093	__ret_505 = __rev0_505 + __noswap_vmull_s32(__rev1_505, __noswap_splat_laneq_s32(__rev2_505, __p3_505)); \
	55094	__ret_505 = __builtin_shufflevector(__ret_505, __ret_505, __lane_reverse_128_64); \
53652	55095	__ret_505; \
53653	55096	})
53654	55097	#endif
53655	55098
53656	55099	#ifdef __LITTLE_ENDIAN__
53657		#define vmls_laneq_f32(__p0_506, __p1_506, __p2_506, __p3_506) __extension__ ({ \
53658		float32x2_t __ret_506; \
53659		float32x2_t __s0_506 = __p0_506; \
53660		float32x2_t __s1_506 = __p1_506; \
53661		float32x4_t __s2_506 = __p2_506; \
53662		__ret_506 = __s0_506 - __s1_506 * splat_laneq_f32(__s2_506, __p3_506); \
	55100	#define vmlal_laneq_s16(__p0_506, __p1_506, __p2_506, __p3_506) __extension__ ({ \
	55101	int32x4_t __ret_506; \
	55102	int32x4_t __s0_506 = __p0_506; \
	55103	int16x4_t __s1_506 = __p1_506; \
	55104	int16x8_t __s2_506 = __p2_506; \
	55105	__ret_506 = __s0_506 + vmull_s16(__s1_506, splat_laneq_s16(__s2_506, __p3_506)); \
53663	55106	__ret_506; \
53664	55107	})
53665	55108	#else
53666		#define vmls_laneq_f32(__p0_507, __p1_507, __p2_507, __p3_507) __extension__ ({ \
53667		float32x2_t __ret_507; \
53668		float32x2_t __s0_507 = __p0_507; \
53669		float32x2_t __s1_507 = __p1_507; \
53670		float32x4_t __s2_507 = __p2_507; \
53671		float32x2_t __rev0_507; __rev0_507 = __builtin_shufflevector(__s0_507, __s0_507, 1, 0); \
53672		float32x2_t __rev1_507; __rev1_507 = __builtin_shufflevector(__s1_507, __s1_507, 1, 0); \
53673		float32x4_t __rev2_507; __rev2_507 = __builtin_shufflevector(__s2_507, __s2_507, 3, 2, 1, 0); \
53674		__ret_507 = __rev0_507 - __rev1_507 * __noswap_splat_laneq_f32(__rev2_507, __p3_507); \
53675		__ret_507 = __builtin_shufflevector(__ret_507, __ret_507, 1, 0); \
	55109	#define vmlal_laneq_s16(__p0_507, __p1_507, __p2_507, __p3_507) __extension__ ({ \
	55110	int32x4_t __ret_507; \
	55111	int32x4_t __s0_507 = __p0_507; \
	55112	int16x4_t __s1_507 = __p1_507; \
	55113	int16x8_t __s2_507 = __p2_507; \
	55114	int32x4_t __rev0_507; __rev0_507 = __builtin_shufflevector(__s0_507, __s0_507, __lane_reverse_128_32); \
	55115	int16x4_t __rev1_507; __rev1_507 = __builtin_shufflevector(__s1_507, __s1_507, __lane_reverse_64_16); \
	55116	int16x8_t __rev2_507; __rev2_507 = __builtin_shufflevector(__s2_507, __s2_507, __lane_reverse_128_16); \
	55117	__ret_507 = __rev0_507 + __noswap_vmull_s16(__rev1_507, __noswap_splat_laneq_s16(__rev2_507, __p3_507)); \
	55118	__ret_507 = __builtin_shufflevector(__ret_507, __ret_507, __lane_reverse_128_32); \
53676	55119	__ret_507; \
53677	55120	})
53678	55121	#endif
53679	55122
53680	55123	#ifdef __LITTLE_ENDIAN__
53681		#define vmls_laneq_s32(__p0_508, __p1_508, __p2_508, __p3_508) __extension__ ({ \
53682		int32x2_t __ret_508; \
53683		int32x2_t __s0_508 = __p0_508; \
53684		int32x2_t __s1_508 = __p1_508; \
53685		int32x4_t __s2_508 = __p2_508; \
53686		__ret_508 = __s0_508 - __s1_508 * splat_laneq_s32(__s2_508, __p3_508); \
	55124	__ai __attribute__((target("neon"))) float64x2_t vmlsq_f64(float64x2_t __p0, float64x2_t __p1, float64x2_t __p2) {
	55125	float64x2_t __ret;
	55126	__ret = __p0 - __p1 * __p2;
	55127	return __ret;
	55128	}
	55129	#else
	55130	__ai __attribute__((target("neon"))) float64x2_t vmlsq_f64(float64x2_t __p0, float64x2_t __p1, float64x2_t __p2) {
	55131	float64x2_t __ret;
	55132	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	55133	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	55134	float64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_64);
	55135	__ret = __rev0 - __rev1 * __rev2;
	55136	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
	55137	return __ret;
	55138	}
	55139	#endif
	55140
	55141	__ai __attribute__((target("neon"))) float64x1_t vmls_f64(float64x1_t __p0, float64x1_t __p1, float64x1_t __p2) {
	55142	float64x1_t __ret;
	55143	__ret = __p0 - __p1 * __p2;
	55144	return __ret;
	55145	}
	55146	#ifdef __LITTLE_ENDIAN__
	55147	#define vmlsq_laneq_u32(__p0_508, __p1_508, __p2_508, __p3_508) __extension__ ({ \
	55148	uint32x4_t __ret_508; \
	55149	uint32x4_t __s0_508 = __p0_508; \
	55150	uint32x4_t __s1_508 = __p1_508; \
	55151	uint32x4_t __s2_508 = __p2_508; \
	55152	__ret_508 = __s0_508 - __s1_508 * splatq_laneq_u32(__s2_508, __p3_508); \
53687	55153	__ret_508; \
53688	55154	})
53689	55155	#else
53690		#define vmls_laneq_s32(__p0_509, __p1_509, __p2_509, __p3_509) __extension__ ({ \
53691		int32x2_t __ret_509; \
53692		int32x2_t __s0_509 = __p0_509; \
53693		int32x2_t __s1_509 = __p1_509; \
53694		int32x4_t __s2_509 = __p2_509; \
53695		int32x2_t __rev0_509; __rev0_509 = __builtin_shufflevector(__s0_509, __s0_509, 1, 0); \
53696		int32x2_t __rev1_509; __rev1_509 = __builtin_shufflevector(__s1_509, __s1_509, 1, 0); \
53697		int32x4_t __rev2_509; __rev2_509 = __builtin_shufflevector(__s2_509, __s2_509, 3, 2, 1, 0); \
53698		__ret_509 = __rev0_509 - __rev1_509 * __noswap_splat_laneq_s32(__rev2_509, __p3_509); \
53699		__ret_509 = __builtin_shufflevector(__ret_509, __ret_509, 1, 0); \
	55156	#define vmlsq_laneq_u32(__p0_509, __p1_509, __p2_509, __p3_509) __extension__ ({ \
	55157	uint32x4_t __ret_509; \
	55158	uint32x4_t __s0_509 = __p0_509; \
	55159	uint32x4_t __s1_509 = __p1_509; \
	55160	uint32x4_t __s2_509 = __p2_509; \
	55161	uint32x4_t __rev0_509; __rev0_509 = __builtin_shufflevector(__s0_509, __s0_509, __lane_reverse_128_32); \
	55162	uint32x4_t __rev1_509; __rev1_509 = __builtin_shufflevector(__s1_509, __s1_509, __lane_reverse_128_32); \
	55163	uint32x4_t __rev2_509; __rev2_509 = __builtin_shufflevector(__s2_509, __s2_509, __lane_reverse_128_32); \
	55164	__ret_509 = __rev0_509 - __rev1_509 * __noswap_splatq_laneq_u32(__rev2_509, __p3_509); \
	55165	__ret_509 = __builtin_shufflevector(__ret_509, __ret_509, __lane_reverse_128_32); \
53700	55166	__ret_509; \
53701	55167	})
53702	55168	#endif
53703	55169
53704	55170	#ifdef __LITTLE_ENDIAN__
53705		#define vmls_laneq_s16(__p0_510, __p1_510, __p2_510, __p3_510) __extension__ ({ \
53706		int16x4_t __ret_510; \
53707		int16x4_t __s0_510 = __p0_510; \
53708		int16x4_t __s1_510 = __p1_510; \
53709		int16x8_t __s2_510 = __p2_510; \
53710		__ret_510 = __s0_510 - __s1_510 * splat_laneq_s16(__s2_510, __p3_510); \
	55171	#define vmlsq_laneq_u16(__p0_510, __p1_510, __p2_510, __p3_510) __extension__ ({ \
	55172	uint16x8_t __ret_510; \
	55173	uint16x8_t __s0_510 = __p0_510; \
	55174	uint16x8_t __s1_510 = __p1_510; \
	55175	uint16x8_t __s2_510 = __p2_510; \
	55176	__ret_510 = __s0_510 - __s1_510 * splatq_laneq_u16(__s2_510, __p3_510); \
53711	55177	__ret_510; \
53712	55178	})
53713	55179	#else
53714		#define vmls_laneq_s16(__p0_511, __p1_511, __p2_511, __p3_511) __extension__ ({ \
53715		int16x4_t __ret_511; \
53716		int16x4_t __s0_511 = __p0_511; \
53717		int16x4_t __s1_511 = __p1_511; \
53718		int16x8_t __s2_511 = __p2_511; \
53719		int16x4_t __rev0_511; __rev0_511 = __builtin_shufflevector(__s0_511, __s0_511, 3, 2, 1, 0); \
53720		int16x4_t __rev1_511; __rev1_511 = __builtin_shufflevector(__s1_511, __s1_511, 3, 2, 1, 0); \
53721		int16x8_t __rev2_511; __rev2_511 = __builtin_shufflevector(__s2_511, __s2_511, 7, 6, 5, 4, 3, 2, 1, 0); \
53722		__ret_511 = __rev0_511 - __rev1_511 * __noswap_splat_laneq_s16(__rev2_511, __p3_511); \
53723		__ret_511 = __builtin_shufflevector(__ret_511, __ret_511, 3, 2, 1, 0); \
	55180	#define vmlsq_laneq_u16(__p0_511, __p1_511, __p2_511, __p3_511) __extension__ ({ \
	55181	uint16x8_t __ret_511; \
	55182	uint16x8_t __s0_511 = __p0_511; \
	55183	uint16x8_t __s1_511 = __p1_511; \
	55184	uint16x8_t __s2_511 = __p2_511; \
	55185	uint16x8_t __rev0_511; __rev0_511 = __builtin_shufflevector(__s0_511, __s0_511, __lane_reverse_128_16); \
	55186	uint16x8_t __rev1_511; __rev1_511 = __builtin_shufflevector(__s1_511, __s1_511, __lane_reverse_128_16); \
	55187	uint16x8_t __rev2_511; __rev2_511 = __builtin_shufflevector(__s2_511, __s2_511, __lane_reverse_128_16); \
	55188	__ret_511 = __rev0_511 - __rev1_511 * __noswap_splatq_laneq_u16(__rev2_511, __p3_511); \
	55189	__ret_511 = __builtin_shufflevector(__ret_511, __ret_511, __lane_reverse_128_16); \
53724	55190	__ret_511; \
53725	55191	})
53726	55192	#endif
53727	55193
53728	55194	#ifdef __LITTLE_ENDIAN__
53729		#define vmlsl_high_lane_u32(__p0_512, __p1_512, __p2_512, __p3_512) __extension__ ({ \
53730		uint64x2_t __ret_512; \
53731		uint64x2_t __s0_512 = __p0_512; \
53732		uint32x4_t __s1_512 = __p1_512; \
53733		uint32x2_t __s2_512 = __p2_512; \
53734		__ret_512 = __s0_512 - vmull_u32(vget_high_u32(__s1_512), splat_lane_u32(__s2_512, __p3_512)); \
	55195	#define vmlsq_laneq_f32(__p0_512, __p1_512, __p2_512, __p3_512) __extension__ ({ \
	55196	float32x4_t __ret_512; \
	55197	float32x4_t __s0_512 = __p0_512; \
	55198	float32x4_t __s1_512 = __p1_512; \
	55199	float32x4_t __s2_512 = __p2_512; \
	55200	__ret_512 = __s0_512 - __s1_512 * splatq_laneq_f32(__s2_512, __p3_512); \
53735	55201	__ret_512; \
53736	55202	})
53737	55203	#else
53738		#define vmlsl_high_lane_u32(__p0_513, __p1_513, __p2_513, __p3_513) __extension__ ({ \
53739		uint64x2_t __ret_513; \
53740		uint64x2_t __s0_513 = __p0_513; \
53741		uint32x4_t __s1_513 = __p1_513; \
53742		uint32x2_t __s2_513 = __p2_513; \
53743		uint64x2_t __rev0_513; __rev0_513 = __builtin_shufflevector(__s0_513, __s0_513, 1, 0); \
53744		uint32x4_t __rev1_513; __rev1_513 = __builtin_shufflevector(__s1_513, __s1_513, 3, 2, 1, 0); \
53745		uint32x2_t __rev2_513; __rev2_513 = __builtin_shufflevector(__s2_513, __s2_513, 1, 0); \
53746		__ret_513 = __rev0_513 - __noswap_vmull_u32(__noswap_vget_high_u32(__rev1_513), __noswap_splat_lane_u32(__rev2_513, __p3_513)); \
53747		__ret_513 = __builtin_shufflevector(__ret_513, __ret_513, 1, 0); \
	55204	#define vmlsq_laneq_f32(__p0_513, __p1_513, __p2_513, __p3_513) __extension__ ({ \
	55205	float32x4_t __ret_513; \
	55206	float32x4_t __s0_513 = __p0_513; \
	55207	float32x4_t __s1_513 = __p1_513; \
	55208	float32x4_t __s2_513 = __p2_513; \
	55209	float32x4_t __rev0_513; __rev0_513 = __builtin_shufflevector(__s0_513, __s0_513, __lane_reverse_128_32); \
	55210	float32x4_t __rev1_513; __rev1_513 = __builtin_shufflevector(__s1_513, __s1_513, __lane_reverse_128_32); \
	55211	float32x4_t __rev2_513; __rev2_513 = __builtin_shufflevector(__s2_513, __s2_513, __lane_reverse_128_32); \
	55212	__ret_513 = __rev0_513 - __rev1_513 * __noswap_splatq_laneq_f32(__rev2_513, __p3_513); \
	55213	__ret_513 = __builtin_shufflevector(__ret_513, __ret_513, __lane_reverse_128_32); \
53748	55214	__ret_513; \
53749	55215	})
53750	55216	#endif
53751	55217
53752	55218	#ifdef __LITTLE_ENDIAN__
53753		#define vmlsl_high_lane_u16(__p0_514, __p1_514, __p2_514, __p3_514) __extension__ ({ \
53754		uint32x4_t __ret_514; \
53755		uint32x4_t __s0_514 = __p0_514; \
53756		uint16x8_t __s1_514 = __p1_514; \
53757		uint16x4_t __s2_514 = __p2_514; \
53758		__ret_514 = __s0_514 - vmull_u16(vget_high_u16(__s1_514), splat_lane_u16(__s2_514, __p3_514)); \
	55219	#define vmlsq_laneq_s32(__p0_514, __p1_514, __p2_514, __p3_514) __extension__ ({ \
	55220	int32x4_t __ret_514; \
	55221	int32x4_t __s0_514 = __p0_514; \
	55222	int32x4_t __s1_514 = __p1_514; \
	55223	int32x4_t __s2_514 = __p2_514; \
	55224	__ret_514 = __s0_514 - __s1_514 * splatq_laneq_s32(__s2_514, __p3_514); \
53759	55225	__ret_514; \
53760	55226	})
53761	55227	#else
53762		#define vmlsl_high_lane_u16(__p0_515, __p1_515, __p2_515, __p3_515) __extension__ ({ \
53763		uint32x4_t __ret_515; \
53764		uint32x4_t __s0_515 = __p0_515; \
53765		uint16x8_t __s1_515 = __p1_515; \
53766		uint16x4_t __s2_515 = __p2_515; \
53767		uint32x4_t __rev0_515; __rev0_515 = __builtin_shufflevector(__s0_515, __s0_515, 3, 2, 1, 0); \
53768		uint16x8_t __rev1_515; __rev1_515 = __builtin_shufflevector(__s1_515, __s1_515, 7, 6, 5, 4, 3, 2, 1, 0); \
53769		uint16x4_t __rev2_515; __rev2_515 = __builtin_shufflevector(__s2_515, __s2_515, 3, 2, 1, 0); \
53770		__ret_515 = __rev0_515 - __noswap_vmull_u16(__noswap_vget_high_u16(__rev1_515), __noswap_splat_lane_u16(__rev2_515, __p3_515)); \
53771		__ret_515 = __builtin_shufflevector(__ret_515, __ret_515, 3, 2, 1, 0); \
	55228	#define vmlsq_laneq_s32(__p0_515, __p1_515, __p2_515, __p3_515) __extension__ ({ \
	55229	int32x4_t __ret_515; \
	55230	int32x4_t __s0_515 = __p0_515; \
	55231	int32x4_t __s1_515 = __p1_515; \
	55232	int32x4_t __s2_515 = __p2_515; \
	55233	int32x4_t __rev0_515; __rev0_515 = __builtin_shufflevector(__s0_515, __s0_515, __lane_reverse_128_32); \
	55234	int32x4_t __rev1_515; __rev1_515 = __builtin_shufflevector(__s1_515, __s1_515, __lane_reverse_128_32); \
	55235	int32x4_t __rev2_515; __rev2_515 = __builtin_shufflevector(__s2_515, __s2_515, __lane_reverse_128_32); \
	55236	__ret_515 = __rev0_515 - __rev1_515 * __noswap_splatq_laneq_s32(__rev2_515, __p3_515); \
	55237	__ret_515 = __builtin_shufflevector(__ret_515, __ret_515, __lane_reverse_128_32); \
53772	55238	__ret_515; \
53773	55239	})
53774	55240	#endif
53775	55241
53776	55242	#ifdef __LITTLE_ENDIAN__
53777		#define vmlsl_high_lane_s32(__p0_516, __p1_516, __p2_516, __p3_516) __extension__ ({ \
53778		int64x2_t __ret_516; \
53779		int64x2_t __s0_516 = __p0_516; \
53780		int32x4_t __s1_516 = __p1_516; \
53781		int32x2_t __s2_516 = __p2_516; \
53782		__ret_516 = __s0_516 - vmull_s32(vget_high_s32(__s1_516), splat_lane_s32(__s2_516, __p3_516)); \
	55243	#define vmlsq_laneq_s16(__p0_516, __p1_516, __p2_516, __p3_516) __extension__ ({ \
	55244	int16x8_t __ret_516; \
	55245	int16x8_t __s0_516 = __p0_516; \
	55246	int16x8_t __s1_516 = __p1_516; \
	55247	int16x8_t __s2_516 = __p2_516; \
	55248	__ret_516 = __s0_516 - __s1_516 * splatq_laneq_s16(__s2_516, __p3_516); \
53783	55249	__ret_516; \
53784	55250	})
53785	55251	#else
53786		#define vmlsl_high_lane_s32(__p0_517, __p1_517, __p2_517, __p3_517) __extension__ ({ \
53787		int64x2_t __ret_517; \
53788		int64x2_t __s0_517 = __p0_517; \
53789		int32x4_t __s1_517 = __p1_517; \
53790		int32x2_t __s2_517 = __p2_517; \
53791		int64x2_t __rev0_517; __rev0_517 = __builtin_shufflevector(__s0_517, __s0_517, 1, 0); \
53792		int32x4_t __rev1_517; __rev1_517 = __builtin_shufflevector(__s1_517, __s1_517, 3, 2, 1, 0); \
53793		int32x2_t __rev2_517; __rev2_517 = __builtin_shufflevector(__s2_517, __s2_517, 1, 0); \
53794		__ret_517 = __rev0_517 - __noswap_vmull_s32(__noswap_vget_high_s32(__rev1_517), __noswap_splat_lane_s32(__rev2_517, __p3_517)); \
53795		__ret_517 = __builtin_shufflevector(__ret_517, __ret_517, 1, 0); \
	55252	#define vmlsq_laneq_s16(__p0_517, __p1_517, __p2_517, __p3_517) __extension__ ({ \
	55253	int16x8_t __ret_517; \
	55254	int16x8_t __s0_517 = __p0_517; \
	55255	int16x8_t __s1_517 = __p1_517; \
	55256	int16x8_t __s2_517 = __p2_517; \
	55257	int16x8_t __rev0_517; __rev0_517 = __builtin_shufflevector(__s0_517, __s0_517, __lane_reverse_128_16); \
	55258	int16x8_t __rev1_517; __rev1_517 = __builtin_shufflevector(__s1_517, __s1_517, __lane_reverse_128_16); \
	55259	int16x8_t __rev2_517; __rev2_517 = __builtin_shufflevector(__s2_517, __s2_517, __lane_reverse_128_16); \
	55260	__ret_517 = __rev0_517 - __rev1_517 * __noswap_splatq_laneq_s16(__rev2_517, __p3_517); \
	55261	__ret_517 = __builtin_shufflevector(__ret_517, __ret_517, __lane_reverse_128_16); \
53796	55262	__ret_517; \
53797	55263	})
53798	55264	#endif
53799	55265
53800	55266	#ifdef __LITTLE_ENDIAN__
53801		#define vmlsl_high_lane_s16(__p0_518, __p1_518, __p2_518, __p3_518) __extension__ ({ \
53802		int32x4_t __ret_518; \
53803		int32x4_t __s0_518 = __p0_518; \
53804		int16x8_t __s1_518 = __p1_518; \
53805		int16x4_t __s2_518 = __p2_518; \
53806		__ret_518 = __s0_518 - vmull_s16(vget_high_s16(__s1_518), splat_lane_s16(__s2_518, __p3_518)); \
	55267	#define vmls_laneq_u32(__p0_518, __p1_518, __p2_518, __p3_518) __extension__ ({ \
	55268	uint32x2_t __ret_518; \
	55269	uint32x2_t __s0_518 = __p0_518; \
	55270	uint32x2_t __s1_518 = __p1_518; \
	55271	uint32x4_t __s2_518 = __p2_518; \
	55272	__ret_518 = __s0_518 - __s1_518 * splat_laneq_u32(__s2_518, __p3_518); \
53807	55273	__ret_518; \
53808	55274	})
53809	55275	#else
53810		#define vmlsl_high_lane_s16(__p0_519, __p1_519, __p2_519, __p3_519) __extension__ ({ \
53811		int32x4_t __ret_519; \
53812		int32x4_t __s0_519 = __p0_519; \
53813		int16x8_t __s1_519 = __p1_519; \
53814		int16x4_t __s2_519 = __p2_519; \
53815		int32x4_t __rev0_519; __rev0_519 = __builtin_shufflevector(__s0_519, __s0_519, 3, 2, 1, 0); \
53816		int16x8_t __rev1_519; __rev1_519 = __builtin_shufflevector(__s1_519, __s1_519, 7, 6, 5, 4, 3, 2, 1, 0); \
53817		int16x4_t __rev2_519; __rev2_519 = __builtin_shufflevector(__s2_519, __s2_519, 3, 2, 1, 0); \
53818		__ret_519 = __rev0_519 - __noswap_vmull_s16(__noswap_vget_high_s16(__rev1_519), __noswap_splat_lane_s16(__rev2_519, __p3_519)); \
53819		__ret_519 = __builtin_shufflevector(__ret_519, __ret_519, 3, 2, 1, 0); \
	55276	#define vmls_laneq_u32(__p0_519, __p1_519, __p2_519, __p3_519) __extension__ ({ \
	55277	uint32x2_t __ret_519; \
	55278	uint32x2_t __s0_519 = __p0_519; \
	55279	uint32x2_t __s1_519 = __p1_519; \
	55280	uint32x4_t __s2_519 = __p2_519; \
	55281	uint32x2_t __rev0_519; __rev0_519 = __builtin_shufflevector(__s0_519, __s0_519, __lane_reverse_64_32); \
	55282	uint32x2_t __rev1_519; __rev1_519 = __builtin_shufflevector(__s1_519, __s1_519, __lane_reverse_64_32); \
	55283	uint32x4_t __rev2_519; __rev2_519 = __builtin_shufflevector(__s2_519, __s2_519, __lane_reverse_128_32); \
	55284	__ret_519 = __rev0_519 - __rev1_519 * __noswap_splat_laneq_u32(__rev2_519, __p3_519); \
	55285	__ret_519 = __builtin_shufflevector(__ret_519, __ret_519, __lane_reverse_64_32); \
53820	55286	__ret_519; \
53821	55287	})
53822	55288	#endif
53823	55289
53824	55290	#ifdef __LITTLE_ENDIAN__
53825		#define vmlsl_high_laneq_u32(__p0_520, __p1_520, __p2_520, __p3_520) __extension__ ({ \
53826		uint64x2_t __ret_520; \
53827		uint64x2_t __s0_520 = __p0_520; \
53828		uint32x4_t __s1_520 = __p1_520; \
53829		uint32x4_t __s2_520 = __p2_520; \
53830		__ret_520 = __s0_520 - vmull_u32(vget_high_u32(__s1_520), splat_laneq_u32(__s2_520, __p3_520)); \
	55291	#define vmls_laneq_u16(__p0_520, __p1_520, __p2_520, __p3_520) __extension__ ({ \
	55292	uint16x4_t __ret_520; \
	55293	uint16x4_t __s0_520 = __p0_520; \
	55294	uint16x4_t __s1_520 = __p1_520; \
	55295	uint16x8_t __s2_520 = __p2_520; \
	55296	__ret_520 = __s0_520 - __s1_520 * splat_laneq_u16(__s2_520, __p3_520); \
53831	55297	__ret_520; \
53832	55298	})
53833	55299	#else
53834		#define vmlsl_high_laneq_u32(__p0_521, __p1_521, __p2_521, __p3_521) __extension__ ({ \
53835		uint64x2_t __ret_521; \
53836		uint64x2_t __s0_521 = __p0_521; \
53837		uint32x4_t __s1_521 = __p1_521; \
53838		uint32x4_t __s2_521 = __p2_521; \
53839		uint64x2_t __rev0_521; __rev0_521 = __builtin_shufflevector(__s0_521, __s0_521, 1, 0); \
53840		uint32x4_t __rev1_521; __rev1_521 = __builtin_shufflevector(__s1_521, __s1_521, 3, 2, 1, 0); \
53841		uint32x4_t __rev2_521; __rev2_521 = __builtin_shufflevector(__s2_521, __s2_521, 3, 2, 1, 0); \
53842		__ret_521 = __rev0_521 - __noswap_vmull_u32(__noswap_vget_high_u32(__rev1_521), __noswap_splat_laneq_u32(__rev2_521, __p3_521)); \
53843		__ret_521 = __builtin_shufflevector(__ret_521, __ret_521, 1, 0); \
	55300	#define vmls_laneq_u16(__p0_521, __p1_521, __p2_521, __p3_521) __extension__ ({ \
	55301	uint16x4_t __ret_521; \
	55302	uint16x4_t __s0_521 = __p0_521; \
	55303	uint16x4_t __s1_521 = __p1_521; \
	55304	uint16x8_t __s2_521 = __p2_521; \
	55305	uint16x4_t __rev0_521; __rev0_521 = __builtin_shufflevector(__s0_521, __s0_521, __lane_reverse_64_16); \
	55306	uint16x4_t __rev1_521; __rev1_521 = __builtin_shufflevector(__s1_521, __s1_521, __lane_reverse_64_16); \
	55307	uint16x8_t __rev2_521; __rev2_521 = __builtin_shufflevector(__s2_521, __s2_521, __lane_reverse_128_16); \
	55308	__ret_521 = __rev0_521 - __rev1_521 * __noswap_splat_laneq_u16(__rev2_521, __p3_521); \
	55309	__ret_521 = __builtin_shufflevector(__ret_521, __ret_521, __lane_reverse_64_16); \
53844	55310	__ret_521; \
53845	55311	})
53846	55312	#endif
53847	55313
53848	55314	#ifdef __LITTLE_ENDIAN__
53849		#define vmlsl_high_laneq_u16(__p0_522, __p1_522, __p2_522, __p3_522) __extension__ ({ \
53850		uint32x4_t __ret_522; \
53851		uint32x4_t __s0_522 = __p0_522; \
53852		uint16x8_t __s1_522 = __p1_522; \
53853		uint16x8_t __s2_522 = __p2_522; \
53854		__ret_522 = __s0_522 - vmull_u16(vget_high_u16(__s1_522), splat_laneq_u16(__s2_522, __p3_522)); \
	55315	#define vmls_laneq_f32(__p0_522, __p1_522, __p2_522, __p3_522) __extension__ ({ \
	55316	float32x2_t __ret_522; \
	55317	float32x2_t __s0_522 = __p0_522; \
	55318	float32x2_t __s1_522 = __p1_522; \
	55319	float32x4_t __s2_522 = __p2_522; \
	55320	__ret_522 = __s0_522 - __s1_522 * splat_laneq_f32(__s2_522, __p3_522); \
53855	55321	__ret_522; \
53856	55322	})
53857	55323	#else
53858		#define vmlsl_high_laneq_u16(__p0_523, __p1_523, __p2_523, __p3_523) __extension__ ({ \
53859		uint32x4_t __ret_523; \
53860		uint32x4_t __s0_523 = __p0_523; \
53861		uint16x8_t __s1_523 = __p1_523; \
53862		uint16x8_t __s2_523 = __p2_523; \
53863		uint32x4_t __rev0_523; __rev0_523 = __builtin_shufflevector(__s0_523, __s0_523, 3, 2, 1, 0); \
53864		uint16x8_t __rev1_523; __rev1_523 = __builtin_shufflevector(__s1_523, __s1_523, 7, 6, 5, 4, 3, 2, 1, 0); \
53865		uint16x8_t __rev2_523; __rev2_523 = __builtin_shufflevector(__s2_523, __s2_523, 7, 6, 5, 4, 3, 2, 1, 0); \
53866		__ret_523 = __rev0_523 - __noswap_vmull_u16(__noswap_vget_high_u16(__rev1_523), __noswap_splat_laneq_u16(__rev2_523, __p3_523)); \
53867		__ret_523 = __builtin_shufflevector(__ret_523, __ret_523, 3, 2, 1, 0); \
	55324	#define vmls_laneq_f32(__p0_523, __p1_523, __p2_523, __p3_523) __extension__ ({ \
	55325	float32x2_t __ret_523; \
	55326	float32x2_t __s0_523 = __p0_523; \
	55327	float32x2_t __s1_523 = __p1_523; \
	55328	float32x4_t __s2_523 = __p2_523; \
	55329	float32x2_t __rev0_523; __rev0_523 = __builtin_shufflevector(__s0_523, __s0_523, __lane_reverse_64_32); \
	55330	float32x2_t __rev1_523; __rev1_523 = __builtin_shufflevector(__s1_523, __s1_523, __lane_reverse_64_32); \
	55331	float32x4_t __rev2_523; __rev2_523 = __builtin_shufflevector(__s2_523, __s2_523, __lane_reverse_128_32); \
	55332	__ret_523 = __rev0_523 - __rev1_523 * __noswap_splat_laneq_f32(__rev2_523, __p3_523); \
	55333	__ret_523 = __builtin_shufflevector(__ret_523, __ret_523, __lane_reverse_64_32); \
53868	55334	__ret_523; \
53869	55335	})
53870	55336	#endif
53871	55337
53872	55338	#ifdef __LITTLE_ENDIAN__
53873		#define vmlsl_high_laneq_s32(__p0_524, __p1_524, __p2_524, __p3_524) __extension__ ({ \
53874		int64x2_t __ret_524; \
53875		int64x2_t __s0_524 = __p0_524; \
53876		int32x4_t __s1_524 = __p1_524; \
	55339	#define vmls_laneq_s32(__p0_524, __p1_524, __p2_524, __p3_524) __extension__ ({ \
	55340	int32x2_t __ret_524; \
	55341	int32x2_t __s0_524 = __p0_524; \
	55342	int32x2_t __s1_524 = __p1_524; \
53877	55343	int32x4_t __s2_524 = __p2_524; \
53878		__ret_524 = __s0_524 - vmull_s32(vget_high_s32(__s1_524), splat_laneq_s32(__s2_524, __p3_524)); \
	55344	__ret_524 = __s0_524 - __s1_524 * splat_laneq_s32(__s2_524, __p3_524); \
53879	55345	__ret_524; \
53880	55346	})
53881	55347	#else
53882		#define vmlsl_high_laneq_s32(__p0_525, __p1_525, __p2_525, __p3_525) __extension__ ({ \
53883		int64x2_t __ret_525; \
53884		int64x2_t __s0_525 = __p0_525; \
53885		int32x4_t __s1_525 = __p1_525; \
	55348	#define vmls_laneq_s32(__p0_525, __p1_525, __p2_525, __p3_525) __extension__ ({ \
	55349	int32x2_t __ret_525; \
	55350	int32x2_t __s0_525 = __p0_525; \
	55351	int32x2_t __s1_525 = __p1_525; \
53886	55352	int32x4_t __s2_525 = __p2_525; \
53887		int64x2_t __rev0_525; __rev0_525 = __builtin_shufflevector(__s0_525, __s0_525, 1, 0); \
53888		int32x4_t __rev1_525; __rev1_525 = __builtin_shufflevector(__s1_525, __s1_525, 3, 2, 1, 0); \
53889		int32x4_t __rev2_525; __rev2_525 = __builtin_shufflevector(__s2_525, __s2_525, 3, 2, 1, 0); \
53890		__ret_525 = __rev0_525 - __noswap_vmull_s32(__noswap_vget_high_s32(__rev1_525), __noswap_splat_laneq_s32(__rev2_525, __p3_525)); \
53891		__ret_525 = __builtin_shufflevector(__ret_525, __ret_525, 1, 0); \
	55353	int32x2_t __rev0_525; __rev0_525 = __builtin_shufflevector(__s0_525, __s0_525, __lane_reverse_64_32); \
	55354	int32x2_t __rev1_525; __rev1_525 = __builtin_shufflevector(__s1_525, __s1_525, __lane_reverse_64_32); \
	55355	int32x4_t __rev2_525; __rev2_525 = __builtin_shufflevector(__s2_525, __s2_525, __lane_reverse_128_32); \
	55356	__ret_525 = __rev0_525 - __rev1_525 * __noswap_splat_laneq_s32(__rev2_525, __p3_525); \
	55357	__ret_525 = __builtin_shufflevector(__ret_525, __ret_525, __lane_reverse_64_32); \
53892	55358	__ret_525; \
53893	55359	})
53894	55360	#endif
53895	55361
53896	55362	#ifdef __LITTLE_ENDIAN__
53897		#define vmlsl_high_laneq_s16(__p0_526, __p1_526, __p2_526, __p3_526) __extension__ ({ \
53898		int32x4_t __ret_526; \
53899		int32x4_t __s0_526 = __p0_526; \
53900		int16x8_t __s1_526 = __p1_526; \
	55363	#define vmls_laneq_s16(__p0_526, __p1_526, __p2_526, __p3_526) __extension__ ({ \
	55364	int16x4_t __ret_526; \
	55365	int16x4_t __s0_526 = __p0_526; \
	55366	int16x4_t __s1_526 = __p1_526; \
53901	55367	int16x8_t __s2_526 = __p2_526; \
53902		__ret_526 = __s0_526 - vmull_s16(vget_high_s16(__s1_526), splat_laneq_s16(__s2_526, __p3_526)); \
	55368	__ret_526 = __s0_526 - __s1_526 * splat_laneq_s16(__s2_526, __p3_526); \
53903	55369	__ret_526; \
53904	55370	})
53905	55371	#else
53906		#define vmlsl_high_laneq_s16(__p0_527, __p1_527, __p2_527, __p3_527) __extension__ ({ \
53907		int32x4_t __ret_527; \
53908		int32x4_t __s0_527 = __p0_527; \
53909		int16x8_t __s1_527 = __p1_527; \
	55372	#define vmls_laneq_s16(__p0_527, __p1_527, __p2_527, __p3_527) __extension__ ({ \
	55373	int16x4_t __ret_527; \
	55374	int16x4_t __s0_527 = __p0_527; \
	55375	int16x4_t __s1_527 = __p1_527; \
53910	55376	int16x8_t __s2_527 = __p2_527; \
53911		int32x4_t __rev0_527; __rev0_527 = __builtin_shufflevector(__s0_527, __s0_527, 3, 2, 1, 0); \
53912		int16x8_t __rev1_527; __rev1_527 = __builtin_shufflevector(__s1_527, __s1_527, 7, 6, 5, 4, 3, 2, 1, 0); \
53913		int16x8_t __rev2_527; __rev2_527 = __builtin_shufflevector(__s2_527, __s2_527, 7, 6, 5, 4, 3, 2, 1, 0); \
53914		__ret_527 = __rev0_527 - __noswap_vmull_s16(__noswap_vget_high_s16(__rev1_527), __noswap_splat_laneq_s16(__rev2_527, __p3_527)); \
53915		__ret_527 = __builtin_shufflevector(__ret_527, __ret_527, 3, 2, 1, 0); \
	55377	int16x4_t __rev0_527; __rev0_527 = __builtin_shufflevector(__s0_527, __s0_527, __lane_reverse_64_16); \
	55378	int16x4_t __rev1_527; __rev1_527 = __builtin_shufflevector(__s1_527, __s1_527, __lane_reverse_64_16); \
	55379	int16x8_t __rev2_527; __rev2_527 = __builtin_shufflevector(__s2_527, __s2_527, __lane_reverse_128_16); \
	55380	__ret_527 = __rev0_527 - __rev1_527 * __noswap_splat_laneq_s16(__rev2_527, __p3_527); \
	55381	__ret_527 = __builtin_shufflevector(__ret_527, __ret_527, __lane_reverse_64_16); \
53916	55382	__ret_527; \
53917	55383	})
53918	55384	#endif
53919	55385
53920	55386	#ifdef __LITTLE_ENDIAN__
53921		#define vmlsl_laneq_u32(__p0_528, __p1_528, __p2_528, __p3_528) __extension__ ({ \
	55387	#define vmlsl_high_lane_u32(__p0_528, __p1_528, __p2_528, __p3_528) __extension__ ({ \
53922	55388	uint64x2_t __ret_528; \
53923	55389	uint64x2_t __s0_528 = __p0_528; \
53924		uint32x2_t __s1_528 = __p1_528; \
53925		uint32x4_t __s2_528 = __p2_528; \
53926		__ret_528 = __s0_528 - vmull_u32(__s1_528, splat_laneq_u32(__s2_528, __p3_528)); \
	55390	uint32x4_t __s1_528 = __p1_528; \
	55391	uint32x2_t __s2_528 = __p2_528; \
	55392	__ret_528 = __s0_528 - vmull_u32(vget_high_u32(__s1_528), splat_lane_u32(__s2_528, __p3_528)); \
53927	55393	__ret_528; \
53928	55394	})
53929	55395	#else
53930		#define vmlsl_laneq_u32(__p0_529, __p1_529, __p2_529, __p3_529) __extension__ ({ \
	55396	#define vmlsl_high_lane_u32(__p0_529, __p1_529, __p2_529, __p3_529) __extension__ ({ \
53931	55397	uint64x2_t __ret_529; \
53932	55398	uint64x2_t __s0_529 = __p0_529; \
53933		uint32x2_t __s1_529 = __p1_529; \
53934		uint32x4_t __s2_529 = __p2_529; \
53935		uint64x2_t __rev0_529; __rev0_529 = __builtin_shufflevector(__s0_529, __s0_529, 1, 0); \
53936		uint32x2_t __rev1_529; __rev1_529 = __builtin_shufflevector(__s1_529, __s1_529, 1, 0); \
53937		uint32x4_t __rev2_529; __rev2_529 = __builtin_shufflevector(__s2_529, __s2_529, 3, 2, 1, 0); \
53938		__ret_529 = __rev0_529 - __noswap_vmull_u32(__rev1_529, __noswap_splat_laneq_u32(__rev2_529, __p3_529)); \
53939		__ret_529 = __builtin_shufflevector(__ret_529, __ret_529, 1, 0); \
	55399	uint32x4_t __s1_529 = __p1_529; \
	55400	uint32x2_t __s2_529 = __p2_529; \
	55401	uint64x2_t __rev0_529; __rev0_529 = __builtin_shufflevector(__s0_529, __s0_529, __lane_reverse_128_64); \
	55402	uint32x4_t __rev1_529; __rev1_529 = __builtin_shufflevector(__s1_529, __s1_529, __lane_reverse_128_32); \
	55403	uint32x2_t __rev2_529; __rev2_529 = __builtin_shufflevector(__s2_529, __s2_529, __lane_reverse_64_32); \
	55404	__ret_529 = __rev0_529 - __noswap_vmull_u32(__noswap_vget_high_u32(__rev1_529), __noswap_splat_lane_u32(__rev2_529, __p3_529)); \
	55405	__ret_529 = __builtin_shufflevector(__ret_529, __ret_529, __lane_reverse_128_64); \
53940	55406	__ret_529; \
53941	55407	})
53942	55408	#endif
53943	55409
53944	55410	#ifdef __LITTLE_ENDIAN__
53945		#define vmlsl_laneq_u16(__p0_530, __p1_530, __p2_530, __p3_530) __extension__ ({ \
	55411	#define vmlsl_high_lane_u16(__p0_530, __p1_530, __p2_530, __p3_530) __extension__ ({ \
53946	55412	uint32x4_t __ret_530; \
53947	55413	uint32x4_t __s0_530 = __p0_530; \
53948		uint16x4_t __s1_530 = __p1_530; \
53949		uint16x8_t __s2_530 = __p2_530; \
53950		__ret_530 = __s0_530 - vmull_u16(__s1_530, splat_laneq_u16(__s2_530, __p3_530)); \
	55414	uint16x8_t __s1_530 = __p1_530; \
	55415	uint16x4_t __s2_530 = __p2_530; \
	55416	__ret_530 = __s0_530 - vmull_u16(vget_high_u16(__s1_530), splat_lane_u16(__s2_530, __p3_530)); \
53951	55417	__ret_530; \
53952	55418	})
53953	55419	#else
53954		#define vmlsl_laneq_u16(__p0_531, __p1_531, __p2_531, __p3_531) __extension__ ({ \
	55420	#define vmlsl_high_lane_u16(__p0_531, __p1_531, __p2_531, __p3_531) __extension__ ({ \
53955	55421	uint32x4_t __ret_531; \
53956	55422	uint32x4_t __s0_531 = __p0_531; \
53957		uint16x4_t __s1_531 = __p1_531; \
53958		uint16x8_t __s2_531 = __p2_531; \
53959		uint32x4_t __rev0_531; __rev0_531 = __builtin_shufflevector(__s0_531, __s0_531, 3, 2, 1, 0); \
53960		uint16x4_t __rev1_531; __rev1_531 = __builtin_shufflevector(__s1_531, __s1_531, 3, 2, 1, 0); \
53961		uint16x8_t __rev2_531; __rev2_531 = __builtin_shufflevector(__s2_531, __s2_531, 7, 6, 5, 4, 3, 2, 1, 0); \
53962		__ret_531 = __rev0_531 - __noswap_vmull_u16(__rev1_531, __noswap_splat_laneq_u16(__rev2_531, __p3_531)); \
53963		__ret_531 = __builtin_shufflevector(__ret_531, __ret_531, 3, 2, 1, 0); \
	55423	uint16x8_t __s1_531 = __p1_531; \
	55424	uint16x4_t __s2_531 = __p2_531; \
	55425	uint32x4_t __rev0_531; __rev0_531 = __builtin_shufflevector(__s0_531, __s0_531, __lane_reverse_128_32); \
	55426	uint16x8_t __rev1_531; __rev1_531 = __builtin_shufflevector(__s1_531, __s1_531, __lane_reverse_128_16); \
	55427	uint16x4_t __rev2_531; __rev2_531 = __builtin_shufflevector(__s2_531, __s2_531, __lane_reverse_64_16); \
	55428	__ret_531 = __rev0_531 - __noswap_vmull_u16(__noswap_vget_high_u16(__rev1_531), __noswap_splat_lane_u16(__rev2_531, __p3_531)); \
	55429	__ret_531 = __builtin_shufflevector(__ret_531, __ret_531, __lane_reverse_128_32); \
53964	55430	__ret_531; \
53965	55431	})
53966	55432	#endif
53967	55433
53968	55434	#ifdef __LITTLE_ENDIAN__
53969		#define vmlsl_laneq_s32(__p0_532, __p1_532, __p2_532, __p3_532) __extension__ ({ \
	55435	#define vmlsl_high_lane_s32(__p0_532, __p1_532, __p2_532, __p3_532) __extension__ ({ \
53970	55436	int64x2_t __ret_532; \
53971	55437	int64x2_t __s0_532 = __p0_532; \
53972		int32x2_t __s1_532 = __p1_532; \
53973		int32x4_t __s2_532 = __p2_532; \
53974		__ret_532 = __s0_532 - vmull_s32(__s1_532, splat_laneq_s32(__s2_532, __p3_532)); \
	55438	int32x4_t __s1_532 = __p1_532; \
	55439	int32x2_t __s2_532 = __p2_532; \
	55440	__ret_532 = __s0_532 - vmull_s32(vget_high_s32(__s1_532), splat_lane_s32(__s2_532, __p3_532)); \
53975	55441	__ret_532; \
53976	55442	})
53977	55443	#else
53978		#define vmlsl_laneq_s32(__p0_533, __p1_533, __p2_533, __p3_533) __extension__ ({ \
	55444	#define vmlsl_high_lane_s32(__p0_533, __p1_533, __p2_533, __p3_533) __extension__ ({ \
53979	55445	int64x2_t __ret_533; \
53980	55446	int64x2_t __s0_533 = __p0_533; \
53981		int32x2_t __s1_533 = __p1_533; \
53982		int32x4_t __s2_533 = __p2_533; \
53983		int64x2_t __rev0_533; __rev0_533 = __builtin_shufflevector(__s0_533, __s0_533, 1, 0); \
53984		int32x2_t __rev1_533; __rev1_533 = __builtin_shufflevector(__s1_533, __s1_533, 1, 0); \
53985		int32x4_t __rev2_533; __rev2_533 = __builtin_shufflevector(__s2_533, __s2_533, 3, 2, 1, 0); \
53986		__ret_533 = __rev0_533 - __noswap_vmull_s32(__rev1_533, __noswap_splat_laneq_s32(__rev2_533, __p3_533)); \
53987		__ret_533 = __builtin_shufflevector(__ret_533, __ret_533, 1, 0); \
	55447	int32x4_t __s1_533 = __p1_533; \
	55448	int32x2_t __s2_533 = __p2_533; \
	55449	int64x2_t __rev0_533; __rev0_533 = __builtin_shufflevector(__s0_533, __s0_533, __lane_reverse_128_64); \
	55450	int32x4_t __rev1_533; __rev1_533 = __builtin_shufflevector(__s1_533, __s1_533, __lane_reverse_128_32); \
	55451	int32x2_t __rev2_533; __rev2_533 = __builtin_shufflevector(__s2_533, __s2_533, __lane_reverse_64_32); \
	55452	__ret_533 = __rev0_533 - __noswap_vmull_s32(__noswap_vget_high_s32(__rev1_533), __noswap_splat_lane_s32(__rev2_533, __p3_533)); \
	55453	__ret_533 = __builtin_shufflevector(__ret_533, __ret_533, __lane_reverse_128_64); \
53988	55454	__ret_533; \
53989	55455	})
53990	55456	#endif
53991	55457
53992	55458	#ifdef __LITTLE_ENDIAN__
53993		#define vmlsl_laneq_s16(__p0_534, __p1_534, __p2_534, __p3_534) __extension__ ({ \
	55459	#define vmlsl_high_lane_s16(__p0_534, __p1_534, __p2_534, __p3_534) __extension__ ({ \
53994	55460	int32x4_t __ret_534; \
53995	55461	int32x4_t __s0_534 = __p0_534; \
53996		int16x4_t __s1_534 = __p1_534; \
53997		int16x8_t __s2_534 = __p2_534; \
53998		__ret_534 = __s0_534 - vmull_s16(__s1_534, splat_laneq_s16(__s2_534, __p3_534)); \
	55462	int16x8_t __s1_534 = __p1_534; \
	55463	int16x4_t __s2_534 = __p2_534; \
	55464	__ret_534 = __s0_534 - vmull_s16(vget_high_s16(__s1_534), splat_lane_s16(__s2_534, __p3_534)); \
53999	55465	__ret_534; \
54000	55466	})
54001	55467	#else
54002		#define vmlsl_laneq_s16(__p0_535, __p1_535, __p2_535, __p3_535) __extension__ ({ \
	55468	#define vmlsl_high_lane_s16(__p0_535, __p1_535, __p2_535, __p3_535) __extension__ ({ \
54003	55469	int32x4_t __ret_535; \
54004	55470	int32x4_t __s0_535 = __p0_535; \
54005		int16x4_t __s1_535 = __p1_535; \
54006		int16x8_t __s2_535 = __p2_535; \
54007		int32x4_t __rev0_535; __rev0_535 = __builtin_shufflevector(__s0_535, __s0_535, 3, 2, 1, 0); \
54008		int16x4_t __rev1_535; __rev1_535 = __builtin_shufflevector(__s1_535, __s1_535, 3, 2, 1, 0); \
54009		int16x8_t __rev2_535; __rev2_535 = __builtin_shufflevector(__s2_535, __s2_535, 7, 6, 5, 4, 3, 2, 1, 0); \
54010		__ret_535 = __rev0_535 - __noswap_vmull_s16(__rev1_535, __noswap_splat_laneq_s16(__rev2_535, __p3_535)); \
54011		__ret_535 = __builtin_shufflevector(__ret_535, __ret_535, 3, 2, 1, 0); \
	55471	int16x8_t __s1_535 = __p1_535; \
	55472	int16x4_t __s2_535 = __p2_535; \
	55473	int32x4_t __rev0_535; __rev0_535 = __builtin_shufflevector(__s0_535, __s0_535, __lane_reverse_128_32); \
	55474	int16x8_t __rev1_535; __rev1_535 = __builtin_shufflevector(__s1_535, __s1_535, __lane_reverse_128_16); \
	55475	int16x4_t __rev2_535; __rev2_535 = __builtin_shufflevector(__s2_535, __s2_535, __lane_reverse_64_16); \
	55476	__ret_535 = __rev0_535 - __noswap_vmull_s16(__noswap_vget_high_s16(__rev1_535), __noswap_splat_lane_s16(__rev2_535, __p3_535)); \
	55477	__ret_535 = __builtin_shufflevector(__ret_535, __ret_535, __lane_reverse_128_32); \
54012	55478	__ret_535; \
54013	55479	})
54014	55480	#endif
54015	55481
	55482	#ifdef __LITTLE_ENDIAN__
	55483	#define vmlsl_high_laneq_u32(__p0_536, __p1_536, __p2_536, __p3_536) __extension__ ({ \
	55484	uint64x2_t __ret_536; \
	55485	uint64x2_t __s0_536 = __p0_536; \
	55486	uint32x4_t __s1_536 = __p1_536; \
	55487	uint32x4_t __s2_536 = __p2_536; \
	55488	__ret_536 = __s0_536 - vmull_u32(vget_high_u32(__s1_536), splat_laneq_u32(__s2_536, __p3_536)); \
	55489	__ret_536; \
	55490	})
	55491	#else
	55492	#define vmlsl_high_laneq_u32(__p0_537, __p1_537, __p2_537, __p3_537) __extension__ ({ \
	55493	uint64x2_t __ret_537; \
	55494	uint64x2_t __s0_537 = __p0_537; \
	55495	uint32x4_t __s1_537 = __p1_537; \
	55496	uint32x4_t __s2_537 = __p2_537; \
	55497	uint64x2_t __rev0_537; __rev0_537 = __builtin_shufflevector(__s0_537, __s0_537, __lane_reverse_128_64); \
	55498	uint32x4_t __rev1_537; __rev1_537 = __builtin_shufflevector(__s1_537, __s1_537, __lane_reverse_128_32); \
	55499	uint32x4_t __rev2_537; __rev2_537 = __builtin_shufflevector(__s2_537, __s2_537, __lane_reverse_128_32); \
	55500	__ret_537 = __rev0_537 - __noswap_vmull_u32(__noswap_vget_high_u32(__rev1_537), __noswap_splat_laneq_u32(__rev2_537, __p3_537)); \
	55501	__ret_537 = __builtin_shufflevector(__ret_537, __ret_537, __lane_reverse_128_64); \
	55502	__ret_537; \
	55503	})
	55504	#endif
	55505
	55506	#ifdef __LITTLE_ENDIAN__
	55507	#define vmlsl_high_laneq_u16(__p0_538, __p1_538, __p2_538, __p3_538) __extension__ ({ \
	55508	uint32x4_t __ret_538; \
	55509	uint32x4_t __s0_538 = __p0_538; \
	55510	uint16x8_t __s1_538 = __p1_538; \
	55511	uint16x8_t __s2_538 = __p2_538; \
	55512	__ret_538 = __s0_538 - vmull_u16(vget_high_u16(__s1_538), splat_laneq_u16(__s2_538, __p3_538)); \
	55513	__ret_538; \
	55514	})
	55515	#else
	55516	#define vmlsl_high_laneq_u16(__p0_539, __p1_539, __p2_539, __p3_539) __extension__ ({ \
	55517	uint32x4_t __ret_539; \
	55518	uint32x4_t __s0_539 = __p0_539; \
	55519	uint16x8_t __s1_539 = __p1_539; \
	55520	uint16x8_t __s2_539 = __p2_539; \
	55521	uint32x4_t __rev0_539; __rev0_539 = __builtin_shufflevector(__s0_539, __s0_539, __lane_reverse_128_32); \
	55522	uint16x8_t __rev1_539; __rev1_539 = __builtin_shufflevector(__s1_539, __s1_539, __lane_reverse_128_16); \
	55523	uint16x8_t __rev2_539; __rev2_539 = __builtin_shufflevector(__s2_539, __s2_539, __lane_reverse_128_16); \
	55524	__ret_539 = __rev0_539 - __noswap_vmull_u16(__noswap_vget_high_u16(__rev1_539), __noswap_splat_laneq_u16(__rev2_539, __p3_539)); \
	55525	__ret_539 = __builtin_shufflevector(__ret_539, __ret_539, __lane_reverse_128_32); \
	55526	__ret_539; \
	55527	})
	55528	#endif
	55529
	55530	#ifdef __LITTLE_ENDIAN__
	55531	#define vmlsl_high_laneq_s32(__p0_540, __p1_540, __p2_540, __p3_540) __extension__ ({ \
	55532	int64x2_t __ret_540; \
	55533	int64x2_t __s0_540 = __p0_540; \
	55534	int32x4_t __s1_540 = __p1_540; \
	55535	int32x4_t __s2_540 = __p2_540; \
	55536	__ret_540 = __s0_540 - vmull_s32(vget_high_s32(__s1_540), splat_laneq_s32(__s2_540, __p3_540)); \
	55537	__ret_540; \
	55538	})
	55539	#else
	55540	#define vmlsl_high_laneq_s32(__p0_541, __p1_541, __p2_541, __p3_541) __extension__ ({ \
	55541	int64x2_t __ret_541; \
	55542	int64x2_t __s0_541 = __p0_541; \
	55543	int32x4_t __s1_541 = __p1_541; \
	55544	int32x4_t __s2_541 = __p2_541; \
	55545	int64x2_t __rev0_541; __rev0_541 = __builtin_shufflevector(__s0_541, __s0_541, __lane_reverse_128_64); \
	55546	int32x4_t __rev1_541; __rev1_541 = __builtin_shufflevector(__s1_541, __s1_541, __lane_reverse_128_32); \
	55547	int32x4_t __rev2_541; __rev2_541 = __builtin_shufflevector(__s2_541, __s2_541, __lane_reverse_128_32); \
	55548	__ret_541 = __rev0_541 - __noswap_vmull_s32(__noswap_vget_high_s32(__rev1_541), __noswap_splat_laneq_s32(__rev2_541, __p3_541)); \
	55549	__ret_541 = __builtin_shufflevector(__ret_541, __ret_541, __lane_reverse_128_64); \
	55550	__ret_541; \
	55551	})
	55552	#endif
	55553
	55554	#ifdef __LITTLE_ENDIAN__
	55555	#define vmlsl_high_laneq_s16(__p0_542, __p1_542, __p2_542, __p3_542) __extension__ ({ \
	55556	int32x4_t __ret_542; \
	55557	int32x4_t __s0_542 = __p0_542; \
	55558	int16x8_t __s1_542 = __p1_542; \
	55559	int16x8_t __s2_542 = __p2_542; \
	55560	__ret_542 = __s0_542 - vmull_s16(vget_high_s16(__s1_542), splat_laneq_s16(__s2_542, __p3_542)); \
	55561	__ret_542; \
	55562	})
	55563	#else
	55564	#define vmlsl_high_laneq_s16(__p0_543, __p1_543, __p2_543, __p3_543) __extension__ ({ \
	55565	int32x4_t __ret_543; \
	55566	int32x4_t __s0_543 = __p0_543; \
	55567	int16x8_t __s1_543 = __p1_543; \
	55568	int16x8_t __s2_543 = __p2_543; \
	55569	int32x4_t __rev0_543; __rev0_543 = __builtin_shufflevector(__s0_543, __s0_543, __lane_reverse_128_32); \
	55570	int16x8_t __rev1_543; __rev1_543 = __builtin_shufflevector(__s1_543, __s1_543, __lane_reverse_128_16); \
	55571	int16x8_t __rev2_543; __rev2_543 = __builtin_shufflevector(__s2_543, __s2_543, __lane_reverse_128_16); \
	55572	__ret_543 = __rev0_543 - __noswap_vmull_s16(__noswap_vget_high_s16(__rev1_543), __noswap_splat_laneq_s16(__rev2_543, __p3_543)); \
	55573	__ret_543 = __builtin_shufflevector(__ret_543, __ret_543, __lane_reverse_128_32); \
	55574	__ret_543; \
	55575	})
	55576	#endif
	55577
	55578	#ifdef __LITTLE_ENDIAN__
	55579	#define vmlsl_laneq_u32(__p0_544, __p1_544, __p2_544, __p3_544) __extension__ ({ \
	55580	uint64x2_t __ret_544; \
	55581	uint64x2_t __s0_544 = __p0_544; \
	55582	uint32x2_t __s1_544 = __p1_544; \
	55583	uint32x4_t __s2_544 = __p2_544; \
	55584	__ret_544 = __s0_544 - vmull_u32(__s1_544, splat_laneq_u32(__s2_544, __p3_544)); \
	55585	__ret_544; \
	55586	})
	55587	#else
	55588	#define vmlsl_laneq_u32(__p0_545, __p1_545, __p2_545, __p3_545) __extension__ ({ \
	55589	uint64x2_t __ret_545; \
	55590	uint64x2_t __s0_545 = __p0_545; \
	55591	uint32x2_t __s1_545 = __p1_545; \
	55592	uint32x4_t __s2_545 = __p2_545; \
	55593	uint64x2_t __rev0_545; __rev0_545 = __builtin_shufflevector(__s0_545, __s0_545, __lane_reverse_128_64); \
	55594	uint32x2_t __rev1_545; __rev1_545 = __builtin_shufflevector(__s1_545, __s1_545, __lane_reverse_64_32); \
	55595	uint32x4_t __rev2_545; __rev2_545 = __builtin_shufflevector(__s2_545, __s2_545, __lane_reverse_128_32); \
	55596	__ret_545 = __rev0_545 - __noswap_vmull_u32(__rev1_545, __noswap_splat_laneq_u32(__rev2_545, __p3_545)); \
	55597	__ret_545 = __builtin_shufflevector(__ret_545, __ret_545, __lane_reverse_128_64); \
	55598	__ret_545; \
	55599	})
	55600	#endif
	55601
	55602	#ifdef __LITTLE_ENDIAN__
	55603	#define vmlsl_laneq_u16(__p0_546, __p1_546, __p2_546, __p3_546) __extension__ ({ \
	55604	uint32x4_t __ret_546; \
	55605	uint32x4_t __s0_546 = __p0_546; \
	55606	uint16x4_t __s1_546 = __p1_546; \
	55607	uint16x8_t __s2_546 = __p2_546; \
	55608	__ret_546 = __s0_546 - vmull_u16(__s1_546, splat_laneq_u16(__s2_546, __p3_546)); \
	55609	__ret_546; \
	55610	})
	55611	#else
	55612	#define vmlsl_laneq_u16(__p0_547, __p1_547, __p2_547, __p3_547) __extension__ ({ \
	55613	uint32x4_t __ret_547; \
	55614	uint32x4_t __s0_547 = __p0_547; \
	55615	uint16x4_t __s1_547 = __p1_547; \
	55616	uint16x8_t __s2_547 = __p2_547; \
	55617	uint32x4_t __rev0_547; __rev0_547 = __builtin_shufflevector(__s0_547, __s0_547, __lane_reverse_128_32); \
	55618	uint16x4_t __rev1_547; __rev1_547 = __builtin_shufflevector(__s1_547, __s1_547, __lane_reverse_64_16); \
	55619	uint16x8_t __rev2_547; __rev2_547 = __builtin_shufflevector(__s2_547, __s2_547, __lane_reverse_128_16); \
	55620	__ret_547 = __rev0_547 - __noswap_vmull_u16(__rev1_547, __noswap_splat_laneq_u16(__rev2_547, __p3_547)); \
	55621	__ret_547 = __builtin_shufflevector(__ret_547, __ret_547, __lane_reverse_128_32); \
	55622	__ret_547; \
	55623	})
	55624	#endif
	55625
	55626	#ifdef __LITTLE_ENDIAN__
	55627	#define vmlsl_laneq_s32(__p0_548, __p1_548, __p2_548, __p3_548) __extension__ ({ \
	55628	int64x2_t __ret_548; \
	55629	int64x2_t __s0_548 = __p0_548; \
	55630	int32x2_t __s1_548 = __p1_548; \
	55631	int32x4_t __s2_548 = __p2_548; \
	55632	__ret_548 = __s0_548 - vmull_s32(__s1_548, splat_laneq_s32(__s2_548, __p3_548)); \
	55633	__ret_548; \
	55634	})
	55635	#else
	55636	#define vmlsl_laneq_s32(__p0_549, __p1_549, __p2_549, __p3_549) __extension__ ({ \
	55637	int64x2_t __ret_549; \
	55638	int64x2_t __s0_549 = __p0_549; \
	55639	int32x2_t __s1_549 = __p1_549; \
	55640	int32x4_t __s2_549 = __p2_549; \
	55641	int64x2_t __rev0_549; __rev0_549 = __builtin_shufflevector(__s0_549, __s0_549, __lane_reverse_128_64); \
	55642	int32x2_t __rev1_549; __rev1_549 = __builtin_shufflevector(__s1_549, __s1_549, __lane_reverse_64_32); \
	55643	int32x4_t __rev2_549; __rev2_549 = __builtin_shufflevector(__s2_549, __s2_549, __lane_reverse_128_32); \
	55644	__ret_549 = __rev0_549 - __noswap_vmull_s32(__rev1_549, __noswap_splat_laneq_s32(__rev2_549, __p3_549)); \
	55645	__ret_549 = __builtin_shufflevector(__ret_549, __ret_549, __lane_reverse_128_64); \
	55646	__ret_549; \
	55647	})
	55648	#endif
	55649
	55650	#ifdef __LITTLE_ENDIAN__
	55651	#define vmlsl_laneq_s16(__p0_550, __p1_550, __p2_550, __p3_550) __extension__ ({ \
	55652	int32x4_t __ret_550; \
	55653	int32x4_t __s0_550 = __p0_550; \
	55654	int16x4_t __s1_550 = __p1_550; \
	55655	int16x8_t __s2_550 = __p2_550; \
	55656	__ret_550 = __s0_550 - vmull_s16(__s1_550, splat_laneq_s16(__s2_550, __p3_550)); \
	55657	__ret_550; \
	55658	})
	55659	#else
	55660	#define vmlsl_laneq_s16(__p0_551, __p1_551, __p2_551, __p3_551) __extension__ ({ \
	55661	int32x4_t __ret_551; \
	55662	int32x4_t __s0_551 = __p0_551; \
	55663	int16x4_t __s1_551 = __p1_551; \
	55664	int16x8_t __s2_551 = __p2_551; \
	55665	int32x4_t __rev0_551; __rev0_551 = __builtin_shufflevector(__s0_551, __s0_551, __lane_reverse_128_32); \
	55666	int16x4_t __rev1_551; __rev1_551 = __builtin_shufflevector(__s1_551, __s1_551, __lane_reverse_64_16); \
	55667	int16x8_t __rev2_551; __rev2_551 = __builtin_shufflevector(__s2_551, __s2_551, __lane_reverse_128_16); \
	55668	__ret_551 = __rev0_551 - __noswap_vmull_s16(__rev1_551, __noswap_splat_laneq_s16(__rev2_551, __p3_551)); \
	55669	__ret_551 = __builtin_shufflevector(__ret_551, __ret_551, __lane_reverse_128_32); \
	55670	__ret_551; \
	55671	})
	55672	#endif
	55673
54016	55674	__ai __attribute__((target("neon"))) poly64x1_t vmov_n_p64(poly64_t __p0) {
54017	55675	poly64x1_t __ret;
54018	55676	__ret = (poly64x1_t) {__p0};
...	...	@@ -54028,7 +55686,7 @@ __ai __attribute__((target("neon"))) poly64x2_t vmovq_n_p64(poly64_t __p0) {
54028	55686	__ai __attribute__((target("neon"))) poly64x2_t vmovq_n_p64(poly64_t __p0) {
54029	55687	poly64x2_t __ret;
54030	55688	__ret = (poly64x2_t) {__p0, __p0};
54031		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	55689	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
54032	55690	return __ret;
54033	55691	}
54034	55692	#endif
...	...	@@ -54043,7 +55701,7 @@ __ai __attribute__((target("neon"))) float64x2_t vmovq_n_f64(float64_t __p0) {
54043	55701	__ai __attribute__((target("neon"))) float64x2_t vmovq_n_f64(float64_t __p0) {
54044	55702	float64x2_t __ret;
54045	55703	__ret = (float64x2_t) {__p0, __p0};
54046		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	55704	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
54047	55705	return __ret;
54048	55706	}
54049	55707	#endif
...	...	@@ -54054,146 +55712,146 @@ __ai __attribute__((target("neon"))) float64x1_t vmov_n_f64(float64_t __p0) {
54054	55712	return __ret;
54055	55713	}
54056	55714	#ifdef __LITTLE_ENDIAN__
54057		__ai __attribute__((target("neon"))) uint16x8_t vmovl_high_u8(uint8x16_t __p0_536) {
54058		uint16x8_t __ret_536;
54059		uint8x8_t __a1_536 = vget_high_u8(__p0_536);
54060		__ret_536 = (uint16x8_t)(vshll_n_u8(__a1_536, 0));
54061		return __ret_536;
	55715	__ai __attribute__((target("neon"))) uint16x8_t vmovl_high_u8(uint8x16_t __p0_552) {
	55716	uint16x8_t __ret_552;
	55717	uint8x8_t __a1_552 = vget_high_u8(__p0_552);
	55718	__ret_552 = __builtin_bit_cast(uint16x8_t, vshll_n_u8(__a1_552, 0));
	55719	return __ret_552;
54062	55720	}
54063	55721	#else
54064		__ai __attribute__((target("neon"))) uint16x8_t vmovl_high_u8(uint8x16_t __p0_537) {
54065		uint16x8_t __ret_537;
54066		uint8x16_t __rev0_537; __rev0_537 = __builtin_shufflevector(__p0_537, __p0_537, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
54067		uint8x8_t __a1_537 = __noswap_vget_high_u8(__rev0_537);
54068		__ret_537 = (uint16x8_t)(__noswap_vshll_n_u8(__a1_537, 0));
54069		__ret_537 = __builtin_shufflevector(__ret_537, __ret_537, 7, 6, 5, 4, 3, 2, 1, 0);
54070		return __ret_537;
	55722	__ai __attribute__((target("neon"))) uint16x8_t vmovl_high_u8(uint8x16_t __p0_553) {
	55723	uint16x8_t __ret_553;
	55724	uint8x16_t __rev0_553; __rev0_553 = __builtin_shufflevector(__p0_553, __p0_553, __lane_reverse_128_8);
	55725	uint8x8_t __a1_553 = __noswap_vget_high_u8(__rev0_553);
	55726	__ret_553 = __builtin_bit_cast(uint16x8_t, __noswap_vshll_n_u8(__a1_553, 0));
	55727	__ret_553 = __builtin_shufflevector(__ret_553, __ret_553, __lane_reverse_128_16);
	55728	return __ret_553;
54071	55729	}
54072		__ai __attribute__((target("neon"))) uint16x8_t __noswap_vmovl_high_u8(uint8x16_t __p0_538) {
54073		uint16x8_t __ret_538;
54074		uint8x8_t __a1_538 = __noswap_vget_high_u8(__p0_538);
54075		__ret_538 = (uint16x8_t)(__noswap_vshll_n_u8(__a1_538, 0));
54076		return __ret_538;
	55730	__ai __attribute__((target("neon"))) uint16x8_t __noswap_vmovl_high_u8(uint8x16_t __p0_554) {
	55731	uint16x8_t __ret_554;
	55732	uint8x8_t __a1_554 = __noswap_vget_high_u8(__p0_554);
	55733	__ret_554 = __builtin_bit_cast(uint16x8_t, __noswap_vshll_n_u8(__a1_554, 0));
	55734	return __ret_554;
54077	55735	}
54078	55736	#endif
54079	55737
54080	55738	#ifdef __LITTLE_ENDIAN__
54081		__ai __attribute__((target("neon"))) uint64x2_t vmovl_high_u32(uint32x4_t __p0_539) {
54082		uint64x2_t __ret_539;
54083		uint32x2_t __a1_539 = vget_high_u32(__p0_539);
54084		__ret_539 = (uint64x2_t)(vshll_n_u32(__a1_539, 0));
54085		return __ret_539;
	55739	__ai __attribute__((target("neon"))) uint64x2_t vmovl_high_u32(uint32x4_t __p0_555) {
	55740	uint64x2_t __ret_555;
	55741	uint32x2_t __a1_555 = vget_high_u32(__p0_555);
	55742	__ret_555 = __builtin_bit_cast(uint64x2_t, vshll_n_u32(__a1_555, 0));
	55743	return __ret_555;
54086	55744	}
54087	55745	#else
54088		__ai __attribute__((target("neon"))) uint64x2_t vmovl_high_u32(uint32x4_t __p0_540) {
54089		uint64x2_t __ret_540;
54090		uint32x4_t __rev0_540; __rev0_540 = __builtin_shufflevector(__p0_540, __p0_540, 3, 2, 1, 0);
54091		uint32x2_t __a1_540 = __noswap_vget_high_u32(__rev0_540);
54092		__ret_540 = (uint64x2_t)(__noswap_vshll_n_u32(__a1_540, 0));
54093		__ret_540 = __builtin_shufflevector(__ret_540, __ret_540, 1, 0);
54094		return __ret_540;
	55746	__ai __attribute__((target("neon"))) uint64x2_t vmovl_high_u32(uint32x4_t __p0_556) {
	55747	uint64x2_t __ret_556;
	55748	uint32x4_t __rev0_556; __rev0_556 = __builtin_shufflevector(__p0_556, __p0_556, __lane_reverse_128_32);
	55749	uint32x2_t __a1_556 = __noswap_vget_high_u32(__rev0_556);
	55750	__ret_556 = __builtin_bit_cast(uint64x2_t, __noswap_vshll_n_u32(__a1_556, 0));
	55751	__ret_556 = __builtin_shufflevector(__ret_556, __ret_556, __lane_reverse_128_64);
	55752	return __ret_556;
54095	55753	}
54096		__ai __attribute__((target("neon"))) uint64x2_t __noswap_vmovl_high_u32(uint32x4_t __p0_541) {
54097		uint64x2_t __ret_541;
54098		uint32x2_t __a1_541 = __noswap_vget_high_u32(__p0_541);
54099		__ret_541 = (uint64x2_t)(__noswap_vshll_n_u32(__a1_541, 0));
54100		return __ret_541;
	55754	__ai __attribute__((target("neon"))) uint64x2_t __noswap_vmovl_high_u32(uint32x4_t __p0_557) {
	55755	uint64x2_t __ret_557;
	55756	uint32x2_t __a1_557 = __noswap_vget_high_u32(__p0_557);
	55757	__ret_557 = __builtin_bit_cast(uint64x2_t, __noswap_vshll_n_u32(__a1_557, 0));
	55758	return __ret_557;
54101	55759	}
54102	55760	#endif
54103	55761
54104	55762	#ifdef __LITTLE_ENDIAN__
54105		__ai __attribute__((target("neon"))) uint32x4_t vmovl_high_u16(uint16x8_t __p0_542) {
54106		uint32x4_t __ret_542;
54107		uint16x4_t __a1_542 = vget_high_u16(__p0_542);
54108		__ret_542 = (uint32x4_t)(vshll_n_u16(__a1_542, 0));
54109		return __ret_542;
	55763	__ai __attribute__((target("neon"))) uint32x4_t vmovl_high_u16(uint16x8_t __p0_558) {
	55764	uint32x4_t __ret_558;
	55765	uint16x4_t __a1_558 = vget_high_u16(__p0_558);
	55766	__ret_558 = __builtin_bit_cast(uint32x4_t, vshll_n_u16(__a1_558, 0));
	55767	return __ret_558;
54110	55768	}
54111	55769	#else
54112		__ai __attribute__((target("neon"))) uint32x4_t vmovl_high_u16(uint16x8_t __p0_543) {
54113		uint32x4_t __ret_543;
54114		uint16x8_t __rev0_543; __rev0_543 = __builtin_shufflevector(__p0_543, __p0_543, 7, 6, 5, 4, 3, 2, 1, 0);
54115		uint16x4_t __a1_543 = __noswap_vget_high_u16(__rev0_543);
54116		__ret_543 = (uint32x4_t)(__noswap_vshll_n_u16(__a1_543, 0));
54117		__ret_543 = __builtin_shufflevector(__ret_543, __ret_543, 3, 2, 1, 0);
54118		return __ret_543;
	55770	__ai __attribute__((target("neon"))) uint32x4_t vmovl_high_u16(uint16x8_t __p0_559) {
	55771	uint32x4_t __ret_559;
	55772	uint16x8_t __rev0_559; __rev0_559 = __builtin_shufflevector(__p0_559, __p0_559, __lane_reverse_128_16);
	55773	uint16x4_t __a1_559 = __noswap_vget_high_u16(__rev0_559);
	55774	__ret_559 = __builtin_bit_cast(uint32x4_t, __noswap_vshll_n_u16(__a1_559, 0));
	55775	__ret_559 = __builtin_shufflevector(__ret_559, __ret_559, __lane_reverse_128_32);
	55776	return __ret_559;
54119	55777	}
54120		__ai __attribute__((target("neon"))) uint32x4_t __noswap_vmovl_high_u16(uint16x8_t __p0_544) {
54121		uint32x4_t __ret_544;
54122		uint16x4_t __a1_544 = __noswap_vget_high_u16(__p0_544);
54123		__ret_544 = (uint32x4_t)(__noswap_vshll_n_u16(__a1_544, 0));
54124		return __ret_544;
	55778	__ai __attribute__((target("neon"))) uint32x4_t __noswap_vmovl_high_u16(uint16x8_t __p0_560) {
	55779	uint32x4_t __ret_560;
	55780	uint16x4_t __a1_560 = __noswap_vget_high_u16(__p0_560);
	55781	__ret_560 = __builtin_bit_cast(uint32x4_t, __noswap_vshll_n_u16(__a1_560, 0));
	55782	return __ret_560;
54125	55783	}
54126	55784	#endif
54127	55785
54128	55786	#ifdef __LITTLE_ENDIAN__
54129		__ai __attribute__((target("neon"))) int16x8_t vmovl_high_s8(int8x16_t __p0_545) {
54130		int16x8_t __ret_545;
54131		int8x8_t __a1_545 = vget_high_s8(__p0_545);
54132		__ret_545 = (int16x8_t)(vshll_n_s8(__a1_545, 0));
54133		return __ret_545;
	55787	__ai __attribute__((target("neon"))) int16x8_t vmovl_high_s8(int8x16_t __p0_561) {
	55788	int16x8_t __ret_561;
	55789	int8x8_t __a1_561 = vget_high_s8(__p0_561);
	55790	__ret_561 = __builtin_bit_cast(int16x8_t, vshll_n_s8(__a1_561, 0));
	55791	return __ret_561;
54134	55792	}
54135	55793	#else
54136		__ai __attribute__((target("neon"))) int16x8_t vmovl_high_s8(int8x16_t __p0_546) {
54137		int16x8_t __ret_546;
54138		int8x16_t __rev0_546; __rev0_546 = __builtin_shufflevector(__p0_546, __p0_546, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
54139		int8x8_t __a1_546 = __noswap_vget_high_s8(__rev0_546);
54140		__ret_546 = (int16x8_t)(__noswap_vshll_n_s8(__a1_546, 0));
54141		__ret_546 = __builtin_shufflevector(__ret_546, __ret_546, 7, 6, 5, 4, 3, 2, 1, 0);
54142		return __ret_546;
	55794	__ai __attribute__((target("neon"))) int16x8_t vmovl_high_s8(int8x16_t __p0_562) {
	55795	int16x8_t __ret_562;
	55796	int8x16_t __rev0_562; __rev0_562 = __builtin_shufflevector(__p0_562, __p0_562, __lane_reverse_128_8);
	55797	int8x8_t __a1_562 = __noswap_vget_high_s8(__rev0_562);
	55798	__ret_562 = __builtin_bit_cast(int16x8_t, __noswap_vshll_n_s8(__a1_562, 0));
	55799	__ret_562 = __builtin_shufflevector(__ret_562, __ret_562, __lane_reverse_128_16);
	55800	return __ret_562;
54143	55801	}
54144		__ai __attribute__((target("neon"))) int16x8_t __noswap_vmovl_high_s8(int8x16_t __p0_547) {
54145		int16x8_t __ret_547;
54146		int8x8_t __a1_547 = __noswap_vget_high_s8(__p0_547);
54147		__ret_547 = (int16x8_t)(__noswap_vshll_n_s8(__a1_547, 0));
54148		return __ret_547;
	55802	__ai __attribute__((target("neon"))) int16x8_t __noswap_vmovl_high_s8(int8x16_t __p0_563) {
	55803	int16x8_t __ret_563;
	55804	int8x8_t __a1_563 = __noswap_vget_high_s8(__p0_563);
	55805	__ret_563 = __builtin_bit_cast(int16x8_t, __noswap_vshll_n_s8(__a1_563, 0));
	55806	return __ret_563;
54149	55807	}
54150	55808	#endif
54151	55809
54152	55810	#ifdef __LITTLE_ENDIAN__
54153		__ai __attribute__((target("neon"))) int64x2_t vmovl_high_s32(int32x4_t __p0_548) {
54154		int64x2_t __ret_548;
54155		int32x2_t __a1_548 = vget_high_s32(__p0_548);
54156		__ret_548 = (int64x2_t)(vshll_n_s32(__a1_548, 0));
54157		return __ret_548;
	55811	__ai __attribute__((target("neon"))) int64x2_t vmovl_high_s32(int32x4_t __p0_564) {
	55812	int64x2_t __ret_564;
	55813	int32x2_t __a1_564 = vget_high_s32(__p0_564);
	55814	__ret_564 = __builtin_bit_cast(int64x2_t, vshll_n_s32(__a1_564, 0));
	55815	return __ret_564;
54158	55816	}
54159	55817	#else
54160		__ai __attribute__((target("neon"))) int64x2_t vmovl_high_s32(int32x4_t __p0_549) {
54161		int64x2_t __ret_549;
54162		int32x4_t __rev0_549; __rev0_549 = __builtin_shufflevector(__p0_549, __p0_549, 3, 2, 1, 0);
54163		int32x2_t __a1_549 = __noswap_vget_high_s32(__rev0_549);
54164		__ret_549 = (int64x2_t)(__noswap_vshll_n_s32(__a1_549, 0));
54165		__ret_549 = __builtin_shufflevector(__ret_549, __ret_549, 1, 0);
54166		return __ret_549;
	55818	__ai __attribute__((target("neon"))) int64x2_t vmovl_high_s32(int32x4_t __p0_565) {
	55819	int64x2_t __ret_565;
	55820	int32x4_t __rev0_565; __rev0_565 = __builtin_shufflevector(__p0_565, __p0_565, __lane_reverse_128_32);
	55821	int32x2_t __a1_565 = __noswap_vget_high_s32(__rev0_565);
	55822	__ret_565 = __builtin_bit_cast(int64x2_t, __noswap_vshll_n_s32(__a1_565, 0));
	55823	__ret_565 = __builtin_shufflevector(__ret_565, __ret_565, __lane_reverse_128_64);
	55824	return __ret_565;
54167	55825	}
54168		__ai __attribute__((target("neon"))) int64x2_t __noswap_vmovl_high_s32(int32x4_t __p0_550) {
54169		int64x2_t __ret_550;
54170		int32x2_t __a1_550 = __noswap_vget_high_s32(__p0_550);
54171		__ret_550 = (int64x2_t)(__noswap_vshll_n_s32(__a1_550, 0));
54172		return __ret_550;
	55826	__ai __attribute__((target("neon"))) int64x2_t __noswap_vmovl_high_s32(int32x4_t __p0_566) {
	55827	int64x2_t __ret_566;
	55828	int32x2_t __a1_566 = __noswap_vget_high_s32(__p0_566);
	55829	__ret_566 = __builtin_bit_cast(int64x2_t, __noswap_vshll_n_s32(__a1_566, 0));
	55830	return __ret_566;
54173	55831	}
54174	55832	#endif
54175	55833
54176	55834	#ifdef __LITTLE_ENDIAN__
54177		__ai __attribute__((target("neon"))) int32x4_t vmovl_high_s16(int16x8_t __p0_551) {
54178		int32x4_t __ret_551;
54179		int16x4_t __a1_551 = vget_high_s16(__p0_551);
54180		__ret_551 = (int32x4_t)(vshll_n_s16(__a1_551, 0));
54181		return __ret_551;
	55835	__ai __attribute__((target("neon"))) int32x4_t vmovl_high_s16(int16x8_t __p0_567) {
	55836	int32x4_t __ret_567;
	55837	int16x4_t __a1_567 = vget_high_s16(__p0_567);
	55838	__ret_567 = __builtin_bit_cast(int32x4_t, vshll_n_s16(__a1_567, 0));
	55839	return __ret_567;
54182	55840	}
54183	55841	#else
54184		__ai __attribute__((target("neon"))) int32x4_t vmovl_high_s16(int16x8_t __p0_552) {
54185		int32x4_t __ret_552;
54186		int16x8_t __rev0_552; __rev0_552 = __builtin_shufflevector(__p0_552, __p0_552, 7, 6, 5, 4, 3, 2, 1, 0);
54187		int16x4_t __a1_552 = __noswap_vget_high_s16(__rev0_552);
54188		__ret_552 = (int32x4_t)(__noswap_vshll_n_s16(__a1_552, 0));
54189		__ret_552 = __builtin_shufflevector(__ret_552, __ret_552, 3, 2, 1, 0);
54190		return __ret_552;
	55842	__ai __attribute__((target("neon"))) int32x4_t vmovl_high_s16(int16x8_t __p0_568) {
	55843	int32x4_t __ret_568;
	55844	int16x8_t __rev0_568; __rev0_568 = __builtin_shufflevector(__p0_568, __p0_568, __lane_reverse_128_16);
	55845	int16x4_t __a1_568 = __noswap_vget_high_s16(__rev0_568);
	55846	__ret_568 = __builtin_bit_cast(int32x4_t, __noswap_vshll_n_s16(__a1_568, 0));
	55847	__ret_568 = __builtin_shufflevector(__ret_568, __ret_568, __lane_reverse_128_32);
	55848	return __ret_568;
54191	55849	}
54192		__ai __attribute__((target("neon"))) int32x4_t __noswap_vmovl_high_s16(int16x8_t __p0_553) {
54193		int32x4_t __ret_553;
54194		int16x4_t __a1_553 = __noswap_vget_high_s16(__p0_553);
54195		__ret_553 = (int32x4_t)(__noswap_vshll_n_s16(__a1_553, 0));
54196		return __ret_553;
	55850	__ai __attribute__((target("neon"))) int32x4_t __noswap_vmovl_high_s16(int16x8_t __p0_569) {
	55851	int32x4_t __ret_569;
	55852	int16x4_t __a1_569 = __noswap_vget_high_s16(__p0_569);
	55853	__ret_569 = __builtin_bit_cast(int32x4_t, __noswap_vshll_n_s16(__a1_569, 0));
	55854	return __ret_569;
54197	55855	}
54198	55856	#endif
54199	55857
...	...	@@ -54206,10 +55864,10 @@ __ai __attribute__((target("neon"))) uint16x8_t vmovn_high_u32(uint16x4_t __p0,
54206	55864	#else
54207	55865	__ai __attribute__((target("neon"))) uint16x8_t vmovn_high_u32(uint16x4_t __p0, uint32x4_t __p1) {
54208	55866	uint16x8_t __ret;
54209		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
54210		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	55867	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	55868	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
54211	55869	__ret = __noswap_vcombine_u16(__rev0, __noswap_vmovn_u32(__rev1));
54212		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	55870	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
54213	55871	return __ret;
54214	55872	}
54215	55873	#endif
...	...	@@ -54223,10 +55881,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vmovn_high_u64(uint32x2_t __p0,
54223	55881	#else
54224	55882	__ai __attribute__((target("neon"))) uint32x4_t vmovn_high_u64(uint32x2_t __p0, uint64x2_t __p1) {
54225	55883	uint32x4_t __ret;
54226		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
54227		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	55884	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	55885	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
54228	55886	__ret = __noswap_vcombine_u32(__rev0, __noswap_vmovn_u64(__rev1));
54229		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	55887	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
54230	55888	return __ret;
54231	55889	}
54232	55890	#endif
...	...	@@ -54240,10 +55898,10 @@ __ai __attribute__((target("neon"))) uint8x16_t vmovn_high_u16(uint8x8_t __p0, u
54240	55898	#else
54241	55899	__ai __attribute__((target("neon"))) uint8x16_t vmovn_high_u16(uint8x8_t __p0, uint16x8_t __p1) {
54242	55900	uint8x16_t __ret;
54243		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
54244		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	55901	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	55902	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
54245	55903	__ret = __noswap_vcombine_u8(__rev0, __noswap_vmovn_u16(__rev1));
54246		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	55904	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
54247	55905	return __ret;
54248	55906	}
54249	55907	#endif
...	...	@@ -54257,10 +55915,10 @@ __ai __attribute__((target("neon"))) int16x8_t vmovn_high_s32(int16x4_t __p0, in
54257	55915	#else
54258	55916	__ai __attribute__((target("neon"))) int16x8_t vmovn_high_s32(int16x4_t __p0, int32x4_t __p1) {
54259	55917	int16x8_t __ret;
54260		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
54261		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	55918	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	55919	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
54262	55920	__ret = __noswap_vcombine_s16(__rev0, __noswap_vmovn_s32(__rev1));
54263		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	55921	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
54264	55922	return __ret;
54265	55923	}
54266	55924	#endif
...	...	@@ -54274,10 +55932,10 @@ __ai __attribute__((target("neon"))) int32x4_t vmovn_high_s64(int32x2_t __p0, in
54274	55932	#else
54275	55933	__ai __attribute__((target("neon"))) int32x4_t vmovn_high_s64(int32x2_t __p0, int64x2_t __p1) {
54276	55934	int32x4_t __ret;
54277		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
54278		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	55935	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	55936	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
54279	55937	__ret = __noswap_vcombine_s32(__rev0, __noswap_vmovn_s64(__rev1));
54280		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	55938	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
54281	55939	return __ret;
54282	55940	}
54283	55941	#endif
...	...	@@ -54291,10 +55949,10 @@ __ai __attribute__((target("neon"))) int8x16_t vmovn_high_s16(int8x8_t __p0, int
54291	55949	#else
54292	55950	__ai __attribute__((target("neon"))) int8x16_t vmovn_high_s16(int8x8_t __p0, int16x8_t __p1) {
54293	55951	int8x16_t __ret;
54294		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
54295		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	55952	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	55953	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
54296	55954	__ret = __noswap_vcombine_s8(__rev0, __noswap_vmovn_s16(__rev1));
54297		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	55955	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
54298	55956	return __ret;
54299	55957	}
54300	55958	#endif
...	...	@@ -54308,10 +55966,10 @@ __ai __attribute__((target("neon"))) float64x2_t vmulq_f64(float64x2_t __p0, flo
54308	55966	#else
54309	55967	__ai __attribute__((target("neon"))) float64x2_t vmulq_f64(float64x2_t __p0, float64x2_t __p1) {
54310	55968	float64x2_t __ret;
54311		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
54312		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	55969	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	55970	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
54313	55971	__ret = __rev0 * __rev1;
54314		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	55972	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
54315	55973	return __ret;
54316	55974	}
54317	55975	#endif
...	...	@@ -54321,29 +55979,29 @@ __ai __attribute__((target("neon"))) float64x1_t vmul_f64(float64x1_t __p0, floa
54321	55979	__ret = __p0 * __p1;
54322	55980	return __ret;
54323	55981	}
54324		#define vmuld_lane_f64(__p0_554, __p1_554, __p2_554) __extension__ ({ \
54325		float64_t __ret_554; \
54326		float64_t __s0_554 = __p0_554; \
54327		float64x1_t __s1_554 = __p1_554; \
54328		__ret_554 = __s0_554 * vget_lane_f64(__s1_554, __p2_554); \
54329		__ret_554; \
	55982	#define vmuld_lane_f64(__p0_570, __p1_570, __p2_570) __extension__ ({ \
	55983	float64_t __ret_570; \
	55984	float64_t __s0_570 = __p0_570; \
	55985	float64x1_t __s1_570 = __p1_570; \
	55986	__ret_570 = __s0_570 * vget_lane_f64(__s1_570, __p2_570); \
	55987	__ret_570; \
54330	55988	})
54331	55989	#ifdef __LITTLE_ENDIAN__
54332		#define vmuls_lane_f32(__p0_555, __p1_555, __p2_555) __extension__ ({ \
54333		float32_t __ret_555; \
54334		float32_t __s0_555 = __p0_555; \
54335		float32x2_t __s1_555 = __p1_555; \
54336		__ret_555 = __s0_555 * vget_lane_f32(__s1_555, __p2_555); \
54337		__ret_555; \
	55990	#define vmuls_lane_f32(__p0_571, __p1_571, __p2_571) __extension__ ({ \
	55991	float32_t __ret_571; \
	55992	float32_t __s0_571 = __p0_571; \
	55993	float32x2_t __s1_571 = __p1_571; \
	55994	__ret_571 = __s0_571 * vget_lane_f32(__s1_571, __p2_571); \
	55995	__ret_571; \
54338	55996	})
54339	55997	#else
54340		#define vmuls_lane_f32(__p0_556, __p1_556, __p2_556) __extension__ ({ \
54341		float32_t __ret_556; \
54342		float32_t __s0_556 = __p0_556; \
54343		float32x2_t __s1_556 = __p1_556; \
54344		float32x2_t __rev1_556; __rev1_556 = __builtin_shufflevector(__s1_556, __s1_556, 1, 0); \
54345		__ret_556 = __s0_556 * __noswap_vget_lane_f32(__rev1_556, __p2_556); \
54346		__ret_556; \
	55998	#define vmuls_lane_f32(__p0_572, __p1_572, __p2_572) __extension__ ({ \
	55999	float32_t __ret_572; \
	56000	float32_t __s0_572 = __p0_572; \
	56001	float32x2_t __s1_572 = __p1_572; \
	56002	float32x2_t __rev1_572; __rev1_572 = __builtin_shufflevector(__s1_572, __s1_572, __lane_reverse_64_32); \
	56003	__ret_572 = __s0_572 * __noswap_vget_lane_f32(__rev1_572, __p2_572); \
	56004	__ret_572; \
54347	56005	})
54348	56006	#endif
54349	56007
...	...	@@ -54351,64 +56009,64 @@ __ai __attribute__((target("neon"))) float64x1_t vmul_f64(float64x1_t __p0, floa
54351	56009	float64x1_t __ret; \
54352	56010	float64x1_t __s0 = __p0; \
54353	56011	float64x1_t __s1 = __p1; \
54354		__ret = (float64x1_t) __builtin_neon_vmul_lane_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 10); \
	56012	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vmul_lane_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 10)); \
54355	56013	__ret; \
54356	56014	})
54357	56015	#ifdef __LITTLE_ENDIAN__
54358		#define vmulq_lane_f64(__p0_557, __p1_557, __p2_557) __extension__ ({ \
54359		float64x2_t __ret_557; \
54360		float64x2_t __s0_557 = __p0_557; \
54361		float64x1_t __s1_557 = __p1_557; \
54362		__ret_557 = __s0_557 * splatq_lane_f64(__s1_557, __p2_557); \
54363		__ret_557; \
	56016	#define vmulq_lane_f64(__p0_573, __p1_573, __p2_573) __extension__ ({ \
	56017	float64x2_t __ret_573; \
	56018	float64x2_t __s0_573 = __p0_573; \
	56019	float64x1_t __s1_573 = __p1_573; \
	56020	__ret_573 = __s0_573 * splatq_lane_f64(__s1_573, __p2_573); \
	56021	__ret_573; \
54364	56022	})
54365	56023	#else
54366		#define vmulq_lane_f64(__p0_558, __p1_558, __p2_558) __extension__ ({ \
54367		float64x2_t __ret_558; \
54368		float64x2_t __s0_558 = __p0_558; \
54369		float64x1_t __s1_558 = __p1_558; \
54370		float64x2_t __rev0_558; __rev0_558 = __builtin_shufflevector(__s0_558, __s0_558, 1, 0); \
54371		__ret_558 = __rev0_558 * __noswap_splatq_lane_f64(__s1_558, __p2_558); \
54372		__ret_558 = __builtin_shufflevector(__ret_558, __ret_558, 1, 0); \
54373		__ret_558; \
	56024	#define vmulq_lane_f64(__p0_574, __p1_574, __p2_574) __extension__ ({ \
	56025	float64x2_t __ret_574; \
	56026	float64x2_t __s0_574 = __p0_574; \
	56027	float64x1_t __s1_574 = __p1_574; \
	56028	float64x2_t __rev0_574; __rev0_574 = __builtin_shufflevector(__s0_574, __s0_574, __lane_reverse_128_64); \
	56029	__ret_574 = __rev0_574 * __noswap_splatq_lane_f64(__s1_574, __p2_574); \
	56030	__ret_574 = __builtin_shufflevector(__ret_574, __ret_574, __lane_reverse_128_64); \
	56031	__ret_574; \
54374	56032	})
54375	56033	#endif
54376	56034
54377	56035	#ifdef __LITTLE_ENDIAN__
54378		#define vmuld_laneq_f64(__p0_559, __p1_559, __p2_559) __extension__ ({ \
54379		float64_t __ret_559; \
54380		float64_t __s0_559 = __p0_559; \
54381		float64x2_t __s1_559 = __p1_559; \
54382		__ret_559 = __s0_559 * vgetq_lane_f64(__s1_559, __p2_559); \
54383		__ret_559; \
	56036	#define vmuld_laneq_f64(__p0_575, __p1_575, __p2_575) __extension__ ({ \
	56037	float64_t __ret_575; \
	56038	float64_t __s0_575 = __p0_575; \
	56039	float64x2_t __s1_575 = __p1_575; \
	56040	__ret_575 = __s0_575 * vgetq_lane_f64(__s1_575, __p2_575); \
	56041	__ret_575; \
54384	56042	})
54385	56043	#else
54386		#define vmuld_laneq_f64(__p0_560, __p1_560, __p2_560) __extension__ ({ \
54387		float64_t __ret_560; \
54388		float64_t __s0_560 = __p0_560; \
54389		float64x2_t __s1_560 = __p1_560; \
54390		float64x2_t __rev1_560; __rev1_560 = __builtin_shufflevector(__s1_560, __s1_560, 1, 0); \
54391		__ret_560 = __s0_560 * __noswap_vgetq_lane_f64(__rev1_560, __p2_560); \
54392		__ret_560; \
	56044	#define vmuld_laneq_f64(__p0_576, __p1_576, __p2_576) __extension__ ({ \
	56045	float64_t __ret_576; \
	56046	float64_t __s0_576 = __p0_576; \
	56047	float64x2_t __s1_576 = __p1_576; \
	56048	float64x2_t __rev1_576; __rev1_576 = __builtin_shufflevector(__s1_576, __s1_576, __lane_reverse_128_64); \
	56049	__ret_576 = __s0_576 * __noswap_vgetq_lane_f64(__rev1_576, __p2_576); \
	56050	__ret_576; \
54393	56051	})
54394	56052	#endif
54395	56053
54396	56054	#ifdef __LITTLE_ENDIAN__
54397		#define vmuls_laneq_f32(__p0_561, __p1_561, __p2_561) __extension__ ({ \
54398		float32_t __ret_561; \
54399		float32_t __s0_561 = __p0_561; \
54400		float32x4_t __s1_561 = __p1_561; \
54401		__ret_561 = __s0_561 * vgetq_lane_f32(__s1_561, __p2_561); \
54402		__ret_561; \
	56055	#define vmuls_laneq_f32(__p0_577, __p1_577, __p2_577) __extension__ ({ \
	56056	float32_t __ret_577; \
	56057	float32_t __s0_577 = __p0_577; \
	56058	float32x4_t __s1_577 = __p1_577; \
	56059	__ret_577 = __s0_577 * vgetq_lane_f32(__s1_577, __p2_577); \
	56060	__ret_577; \
54403	56061	})
54404	56062	#else
54405		#define vmuls_laneq_f32(__p0_562, __p1_562, __p2_562) __extension__ ({ \
54406		float32_t __ret_562; \
54407		float32_t __s0_562 = __p0_562; \
54408		float32x4_t __s1_562 = __p1_562; \
54409		float32x4_t __rev1_562; __rev1_562 = __builtin_shufflevector(__s1_562, __s1_562, 3, 2, 1, 0); \
54410		__ret_562 = __s0_562 * __noswap_vgetq_lane_f32(__rev1_562, __p2_562); \
54411		__ret_562; \
	56063	#define vmuls_laneq_f32(__p0_578, __p1_578, __p2_578) __extension__ ({ \
	56064	float32_t __ret_578; \
	56065	float32_t __s0_578 = __p0_578; \
	56066	float32x4_t __s1_578 = __p1_578; \
	56067	float32x4_t __rev1_578; __rev1_578 = __builtin_shufflevector(__s1_578, __s1_578, __lane_reverse_128_32); \
	56068	__ret_578 = __s0_578 * __noswap_vgetq_lane_f32(__rev1_578, __p2_578); \
	56069	__ret_578; \
54412	56070	})
54413	56071	#endif
54414	56072
...	...	@@ -54417,7 +56075,7 @@ __ai __attribute__((target("neon"))) float64x1_t vmul_f64(float64x1_t __p0, floa
54417	56075	float64x1_t __ret; \
54418	56076	float64x1_t __s0 = __p0; \
54419	56077	float64x2_t __s1 = __p1; \
54420		__ret = (float64x1_t) __builtin_neon_vmul_laneq_v((int8x8_t)__s0, (int8x16_t)__s1, __p2, 10); \
	56078	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vmul_laneq_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 10)); \
54421	56079	__ret; \
54422	56080	})
54423	56081	#else
...	...	@@ -54425,246 +56083,246 @@ __ai __attribute__((target("neon"))) float64x1_t vmul_f64(float64x1_t __p0, floa
54425	56083	float64x1_t __ret; \
54426	56084	float64x1_t __s0 = __p0; \
54427	56085	float64x2_t __s1 = __p1; \
54428		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
54429		__ret = (float64x1_t) __builtin_neon_vmul_laneq_v((int8x8_t)__s0, (int8x16_t)__rev1, __p2, 10); \
	56086	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	56087	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vmul_laneq_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 10)); \
54430	56088	__ret; \
54431	56089	})
54432	56090	#endif
54433	56091
54434	56092	#ifdef __LITTLE_ENDIAN__
54435		#define vmulq_laneq_u32(__p0_563, __p1_563, __p2_563) __extension__ ({ \
54436		uint32x4_t __ret_563; \
54437		uint32x4_t __s0_563 = __p0_563; \
54438		uint32x4_t __s1_563 = __p1_563; \
54439		__ret_563 = __s0_563 * splatq_laneq_u32(__s1_563, __p2_563); \
54440		__ret_563; \
	56093	#define vmulq_laneq_u32(__p0_579, __p1_579, __p2_579) __extension__ ({ \
	56094	uint32x4_t __ret_579; \
	56095	uint32x4_t __s0_579 = __p0_579; \
	56096	uint32x4_t __s1_579 = __p1_579; \
	56097	__ret_579 = __s0_579 * splatq_laneq_u32(__s1_579, __p2_579); \
	56098	__ret_579; \
54441	56099	})
54442	56100	#else
54443		#define vmulq_laneq_u32(__p0_564, __p1_564, __p2_564) __extension__ ({ \
54444		uint32x4_t __ret_564; \
54445		uint32x4_t __s0_564 = __p0_564; \
54446		uint32x4_t __s1_564 = __p1_564; \
54447		uint32x4_t __rev0_564; __rev0_564 = __builtin_shufflevector(__s0_564, __s0_564, 3, 2, 1, 0); \
54448		uint32x4_t __rev1_564; __rev1_564 = __builtin_shufflevector(__s1_564, __s1_564, 3, 2, 1, 0); \
54449		__ret_564 = __rev0_564 * __noswap_splatq_laneq_u32(__rev1_564, __p2_564); \
54450		__ret_564 = __builtin_shufflevector(__ret_564, __ret_564, 3, 2, 1, 0); \
54451		__ret_564; \
	56101	#define vmulq_laneq_u32(__p0_580, __p1_580, __p2_580) __extension__ ({ \
	56102	uint32x4_t __ret_580; \
	56103	uint32x4_t __s0_580 = __p0_580; \
	56104	uint32x4_t __s1_580 = __p1_580; \
	56105	uint32x4_t __rev0_580; __rev0_580 = __builtin_shufflevector(__s0_580, __s0_580, __lane_reverse_128_32); \
	56106	uint32x4_t __rev1_580; __rev1_580 = __builtin_shufflevector(__s1_580, __s1_580, __lane_reverse_128_32); \
	56107	__ret_580 = __rev0_580 * __noswap_splatq_laneq_u32(__rev1_580, __p2_580); \
	56108	__ret_580 = __builtin_shufflevector(__ret_580, __ret_580, __lane_reverse_128_32); \
	56109	__ret_580; \
54452	56110	})
54453	56111	#endif
54454	56112
54455	56113	#ifdef __LITTLE_ENDIAN__
54456		#define vmulq_laneq_u16(__p0_565, __p1_565, __p2_565) __extension__ ({ \
54457		uint16x8_t __ret_565; \
54458		uint16x8_t __s0_565 = __p0_565; \
54459		uint16x8_t __s1_565 = __p1_565; \
54460		__ret_565 = __s0_565 * splatq_laneq_u16(__s1_565, __p2_565); \
54461		__ret_565; \
	56114	#define vmulq_laneq_u16(__p0_581, __p1_581, __p2_581) __extension__ ({ \
	56115	uint16x8_t __ret_581; \
	56116	uint16x8_t __s0_581 = __p0_581; \
	56117	uint16x8_t __s1_581 = __p1_581; \
	56118	__ret_581 = __s0_581 * splatq_laneq_u16(__s1_581, __p2_581); \
	56119	__ret_581; \
54462	56120	})
54463	56121	#else
54464		#define vmulq_laneq_u16(__p0_566, __p1_566, __p2_566) __extension__ ({ \
54465		uint16x8_t __ret_566; \
54466		uint16x8_t __s0_566 = __p0_566; \
54467		uint16x8_t __s1_566 = __p1_566; \
54468		uint16x8_t __rev0_566; __rev0_566 = __builtin_shufflevector(__s0_566, __s0_566, 7, 6, 5, 4, 3, 2, 1, 0); \
54469		uint16x8_t __rev1_566; __rev1_566 = __builtin_shufflevector(__s1_566, __s1_566, 7, 6, 5, 4, 3, 2, 1, 0); \
54470		__ret_566 = __rev0_566 * __noswap_splatq_laneq_u16(__rev1_566, __p2_566); \
54471		__ret_566 = __builtin_shufflevector(__ret_566, __ret_566, 7, 6, 5, 4, 3, 2, 1, 0); \
54472		__ret_566; \
	56122	#define vmulq_laneq_u16(__p0_582, __p1_582, __p2_582) __extension__ ({ \
	56123	uint16x8_t __ret_582; \
	56124	uint16x8_t __s0_582 = __p0_582; \
	56125	uint16x8_t __s1_582 = __p1_582; \
	56126	uint16x8_t __rev0_582; __rev0_582 = __builtin_shufflevector(__s0_582, __s0_582, __lane_reverse_128_16); \
	56127	uint16x8_t __rev1_582; __rev1_582 = __builtin_shufflevector(__s1_582, __s1_582, __lane_reverse_128_16); \
	56128	__ret_582 = __rev0_582 * __noswap_splatq_laneq_u16(__rev1_582, __p2_582); \
	56129	__ret_582 = __builtin_shufflevector(__ret_582, __ret_582, __lane_reverse_128_16); \
	56130	__ret_582; \
54473	56131	})
54474	56132	#endif
54475	56133
54476	56134	#ifdef __LITTLE_ENDIAN__
54477		#define vmulq_laneq_f64(__p0_567, __p1_567, __p2_567) __extension__ ({ \
54478		float64x2_t __ret_567; \
54479		float64x2_t __s0_567 = __p0_567; \
54480		float64x2_t __s1_567 = __p1_567; \
54481		__ret_567 = __s0_567 * splatq_laneq_f64(__s1_567, __p2_567); \
54482		__ret_567; \
	56135	#define vmulq_laneq_f64(__p0_583, __p1_583, __p2_583) __extension__ ({ \
	56136	float64x2_t __ret_583; \
	56137	float64x2_t __s0_583 = __p0_583; \
	56138	float64x2_t __s1_583 = __p1_583; \
	56139	__ret_583 = __s0_583 * splatq_laneq_f64(__s1_583, __p2_583); \
	56140	__ret_583; \
54483	56141	})
54484	56142	#else
54485		#define vmulq_laneq_f64(__p0_568, __p1_568, __p2_568) __extension__ ({ \
54486		float64x2_t __ret_568; \
54487		float64x2_t __s0_568 = __p0_568; \
54488		float64x2_t __s1_568 = __p1_568; \
54489		float64x2_t __rev0_568; __rev0_568 = __builtin_shufflevector(__s0_568, __s0_568, 1, 0); \
54490		float64x2_t __rev1_568; __rev1_568 = __builtin_shufflevector(__s1_568, __s1_568, 1, 0); \
54491		__ret_568 = __rev0_568 * __noswap_splatq_laneq_f64(__rev1_568, __p2_568); \
54492		__ret_568 = __builtin_shufflevector(__ret_568, __ret_568, 1, 0); \
54493		__ret_568; \
	56143	#define vmulq_laneq_f64(__p0_584, __p1_584, __p2_584) __extension__ ({ \
	56144	float64x2_t __ret_584; \
	56145	float64x2_t __s0_584 = __p0_584; \
	56146	float64x2_t __s1_584 = __p1_584; \
	56147	float64x2_t __rev0_584; __rev0_584 = __builtin_shufflevector(__s0_584, __s0_584, __lane_reverse_128_64); \
	56148	float64x2_t __rev1_584; __rev1_584 = __builtin_shufflevector(__s1_584, __s1_584, __lane_reverse_128_64); \
	56149	__ret_584 = __rev0_584 * __noswap_splatq_laneq_f64(__rev1_584, __p2_584); \
	56150	__ret_584 = __builtin_shufflevector(__ret_584, __ret_584, __lane_reverse_128_64); \
	56151	__ret_584; \
54494	56152	})
54495	56153	#endif
54496	56154
54497	56155	#ifdef __LITTLE_ENDIAN__
54498		#define vmulq_laneq_f32(__p0_569, __p1_569, __p2_569) __extension__ ({ \
54499		float32x4_t __ret_569; \
54500		float32x4_t __s0_569 = __p0_569; \
54501		float32x4_t __s1_569 = __p1_569; \
54502		__ret_569 = __s0_569 * splatq_laneq_f32(__s1_569, __p2_569); \
54503		__ret_569; \
	56156	#define vmulq_laneq_f32(__p0_585, __p1_585, __p2_585) __extension__ ({ \
	56157	float32x4_t __ret_585; \
	56158	float32x4_t __s0_585 = __p0_585; \
	56159	float32x4_t __s1_585 = __p1_585; \
	56160	__ret_585 = __s0_585 * splatq_laneq_f32(__s1_585, __p2_585); \
	56161	__ret_585; \
54504	56162	})
54505	56163	#else
54506		#define vmulq_laneq_f32(__p0_570, __p1_570, __p2_570) __extension__ ({ \
54507		float32x4_t __ret_570; \
54508		float32x4_t __s0_570 = __p0_570; \
54509		float32x4_t __s1_570 = __p1_570; \
54510		float32x4_t __rev0_570; __rev0_570 = __builtin_shufflevector(__s0_570, __s0_570, 3, 2, 1, 0); \
54511		float32x4_t __rev1_570; __rev1_570 = __builtin_shufflevector(__s1_570, __s1_570, 3, 2, 1, 0); \
54512		__ret_570 = __rev0_570 * __noswap_splatq_laneq_f32(__rev1_570, __p2_570); \
54513		__ret_570 = __builtin_shufflevector(__ret_570, __ret_570, 3, 2, 1, 0); \
54514		__ret_570; \
	56164	#define vmulq_laneq_f32(__p0_586, __p1_586, __p2_586) __extension__ ({ \
	56165	float32x4_t __ret_586; \
	56166	float32x4_t __s0_586 = __p0_586; \
	56167	float32x4_t __s1_586 = __p1_586; \
	56168	float32x4_t __rev0_586; __rev0_586 = __builtin_shufflevector(__s0_586, __s0_586, __lane_reverse_128_32); \
	56169	float32x4_t __rev1_586; __rev1_586 = __builtin_shufflevector(__s1_586, __s1_586, __lane_reverse_128_32); \
	56170	__ret_586 = __rev0_586 * __noswap_splatq_laneq_f32(__rev1_586, __p2_586); \
	56171	__ret_586 = __builtin_shufflevector(__ret_586, __ret_586, __lane_reverse_128_32); \
	56172	__ret_586; \
54515	56173	})
54516	56174	#endif
54517	56175
54518	56176	#ifdef __LITTLE_ENDIAN__
54519		#define vmulq_laneq_s32(__p0_571, __p1_571, __p2_571) __extension__ ({ \
54520		int32x4_t __ret_571; \
54521		int32x4_t __s0_571 = __p0_571; \
54522		int32x4_t __s1_571 = __p1_571; \
54523		__ret_571 = __s0_571 * splatq_laneq_s32(__s1_571, __p2_571); \
54524		__ret_571; \
	56177	#define vmulq_laneq_s32(__p0_587, __p1_587, __p2_587) __extension__ ({ \
	56178	int32x4_t __ret_587; \
	56179	int32x4_t __s0_587 = __p0_587; \
	56180	int32x4_t __s1_587 = __p1_587; \
	56181	__ret_587 = __s0_587 * splatq_laneq_s32(__s1_587, __p2_587); \
	56182	__ret_587; \
54525	56183	})
54526	56184	#else
54527		#define vmulq_laneq_s32(__p0_572, __p1_572, __p2_572) __extension__ ({ \
54528		int32x4_t __ret_572; \
54529		int32x4_t __s0_572 = __p0_572; \
54530		int32x4_t __s1_572 = __p1_572; \
54531		int32x4_t __rev0_572; __rev0_572 = __builtin_shufflevector(__s0_572, __s0_572, 3, 2, 1, 0); \
54532		int32x4_t __rev1_572; __rev1_572 = __builtin_shufflevector(__s1_572, __s1_572, 3, 2, 1, 0); \
54533		__ret_572 = __rev0_572 * __noswap_splatq_laneq_s32(__rev1_572, __p2_572); \
54534		__ret_572 = __builtin_shufflevector(__ret_572, __ret_572, 3, 2, 1, 0); \
54535		__ret_572; \
	56185	#define vmulq_laneq_s32(__p0_588, __p1_588, __p2_588) __extension__ ({ \
	56186	int32x4_t __ret_588; \
	56187	int32x4_t __s0_588 = __p0_588; \
	56188	int32x4_t __s1_588 = __p1_588; \
	56189	int32x4_t __rev0_588; __rev0_588 = __builtin_shufflevector(__s0_588, __s0_588, __lane_reverse_128_32); \
	56190	int32x4_t __rev1_588; __rev1_588 = __builtin_shufflevector(__s1_588, __s1_588, __lane_reverse_128_32); \
	56191	__ret_588 = __rev0_588 * __noswap_splatq_laneq_s32(__rev1_588, __p2_588); \
	56192	__ret_588 = __builtin_shufflevector(__ret_588, __ret_588, __lane_reverse_128_32); \
	56193	__ret_588; \
54536	56194	})
54537	56195	#endif
54538	56196
54539	56197	#ifdef __LITTLE_ENDIAN__
54540		#define vmulq_laneq_s16(__p0_573, __p1_573, __p2_573) __extension__ ({ \
54541		int16x8_t __ret_573; \
54542		int16x8_t __s0_573 = __p0_573; \
54543		int16x8_t __s1_573 = __p1_573; \
54544		__ret_573 = __s0_573 * splatq_laneq_s16(__s1_573, __p2_573); \
54545		__ret_573; \
	56198	#define vmulq_laneq_s16(__p0_589, __p1_589, __p2_589) __extension__ ({ \
	56199	int16x8_t __ret_589; \
	56200	int16x8_t __s0_589 = __p0_589; \
	56201	int16x8_t __s1_589 = __p1_589; \
	56202	__ret_589 = __s0_589 * splatq_laneq_s16(__s1_589, __p2_589); \
	56203	__ret_589; \
54546	56204	})
54547	56205	#else
54548		#define vmulq_laneq_s16(__p0_574, __p1_574, __p2_574) __extension__ ({ \
54549		int16x8_t __ret_574; \
54550		int16x8_t __s0_574 = __p0_574; \
54551		int16x8_t __s1_574 = __p1_574; \
54552		int16x8_t __rev0_574; __rev0_574 = __builtin_shufflevector(__s0_574, __s0_574, 7, 6, 5, 4, 3, 2, 1, 0); \
54553		int16x8_t __rev1_574; __rev1_574 = __builtin_shufflevector(__s1_574, __s1_574, 7, 6, 5, 4, 3, 2, 1, 0); \
54554		__ret_574 = __rev0_574 * __noswap_splatq_laneq_s16(__rev1_574, __p2_574); \
54555		__ret_574 = __builtin_shufflevector(__ret_574, __ret_574, 7, 6, 5, 4, 3, 2, 1, 0); \
54556		__ret_574; \
	56206	#define vmulq_laneq_s16(__p0_590, __p1_590, __p2_590) __extension__ ({ \
	56207	int16x8_t __ret_590; \
	56208	int16x8_t __s0_590 = __p0_590; \
	56209	int16x8_t __s1_590 = __p1_590; \
	56210	int16x8_t __rev0_590; __rev0_590 = __builtin_shufflevector(__s0_590, __s0_590, __lane_reverse_128_16); \
	56211	int16x8_t __rev1_590; __rev1_590 = __builtin_shufflevector(__s1_590, __s1_590, __lane_reverse_128_16); \
	56212	__ret_590 = __rev0_590 * __noswap_splatq_laneq_s16(__rev1_590, __p2_590); \
	56213	__ret_590 = __builtin_shufflevector(__ret_590, __ret_590, __lane_reverse_128_16); \
	56214	__ret_590; \
54557	56215	})
54558	56216	#endif
54559	56217
54560	56218	#ifdef __LITTLE_ENDIAN__
54561		#define vmul_laneq_u32(__p0_575, __p1_575, __p2_575) __extension__ ({ \
54562		uint32x2_t __ret_575; \
54563		uint32x2_t __s0_575 = __p0_575; \
54564		uint32x4_t __s1_575 = __p1_575; \
54565		__ret_575 = __s0_575 * splat_laneq_u32(__s1_575, __p2_575); \
54566		__ret_575; \
	56219	#define vmul_laneq_u32(__p0_591, __p1_591, __p2_591) __extension__ ({ \
	56220	uint32x2_t __ret_591; \
	56221	uint32x2_t __s0_591 = __p0_591; \
	56222	uint32x4_t __s1_591 = __p1_591; \
	56223	__ret_591 = __s0_591 * splat_laneq_u32(__s1_591, __p2_591); \
	56224	__ret_591; \
54567	56225	})
54568	56226	#else
54569		#define vmul_laneq_u32(__p0_576, __p1_576, __p2_576) __extension__ ({ \
54570		uint32x2_t __ret_576; \
54571		uint32x2_t __s0_576 = __p0_576; \
54572		uint32x4_t __s1_576 = __p1_576; \
54573		uint32x2_t __rev0_576; __rev0_576 = __builtin_shufflevector(__s0_576, __s0_576, 1, 0); \
54574		uint32x4_t __rev1_576; __rev1_576 = __builtin_shufflevector(__s1_576, __s1_576, 3, 2, 1, 0); \
54575		__ret_576 = __rev0_576 * __noswap_splat_laneq_u32(__rev1_576, __p2_576); \
54576		__ret_576 = __builtin_shufflevector(__ret_576, __ret_576, 1, 0); \
54577		__ret_576; \
	56227	#define vmul_laneq_u32(__p0_592, __p1_592, __p2_592) __extension__ ({ \
	56228	uint32x2_t __ret_592; \
	56229	uint32x2_t __s0_592 = __p0_592; \
	56230	uint32x4_t __s1_592 = __p1_592; \
	56231	uint32x2_t __rev0_592; __rev0_592 = __builtin_shufflevector(__s0_592, __s0_592, __lane_reverse_64_32); \
	56232	uint32x4_t __rev1_592; __rev1_592 = __builtin_shufflevector(__s1_592, __s1_592, __lane_reverse_128_32); \
	56233	__ret_592 = __rev0_592 * __noswap_splat_laneq_u32(__rev1_592, __p2_592); \
	56234	__ret_592 = __builtin_shufflevector(__ret_592, __ret_592, __lane_reverse_64_32); \
	56235	__ret_592; \
54578	56236	})
54579	56237	#endif
54580	56238
54581	56239	#ifdef __LITTLE_ENDIAN__
54582		#define vmul_laneq_u16(__p0_577, __p1_577, __p2_577) __extension__ ({ \
54583		uint16x4_t __ret_577; \
54584		uint16x4_t __s0_577 = __p0_577; \
54585		uint16x8_t __s1_577 = __p1_577; \
54586		__ret_577 = __s0_577 * splat_laneq_u16(__s1_577, __p2_577); \
54587		__ret_577; \
	56240	#define vmul_laneq_u16(__p0_593, __p1_593, __p2_593) __extension__ ({ \
	56241	uint16x4_t __ret_593; \
	56242	uint16x4_t __s0_593 = __p0_593; \
	56243	uint16x8_t __s1_593 = __p1_593; \
	56244	__ret_593 = __s0_593 * splat_laneq_u16(__s1_593, __p2_593); \
	56245	__ret_593; \
54588	56246	})
54589	56247	#else
54590		#define vmul_laneq_u16(__p0_578, __p1_578, __p2_578) __extension__ ({ \
54591		uint16x4_t __ret_578; \
54592		uint16x4_t __s0_578 = __p0_578; \
54593		uint16x8_t __s1_578 = __p1_578; \
54594		uint16x4_t __rev0_578; __rev0_578 = __builtin_shufflevector(__s0_578, __s0_578, 3, 2, 1, 0); \
54595		uint16x8_t __rev1_578; __rev1_578 = __builtin_shufflevector(__s1_578, __s1_578, 7, 6, 5, 4, 3, 2, 1, 0); \
54596		__ret_578 = __rev0_578 * __noswap_splat_laneq_u16(__rev1_578, __p2_578); \
54597		__ret_578 = __builtin_shufflevector(__ret_578, __ret_578, 3, 2, 1, 0); \
54598		__ret_578; \
	56248	#define vmul_laneq_u16(__p0_594, __p1_594, __p2_594) __extension__ ({ \
	56249	uint16x4_t __ret_594; \
	56250	uint16x4_t __s0_594 = __p0_594; \
	56251	uint16x8_t __s1_594 = __p1_594; \
	56252	uint16x4_t __rev0_594; __rev0_594 = __builtin_shufflevector(__s0_594, __s0_594, __lane_reverse_64_16); \
	56253	uint16x8_t __rev1_594; __rev1_594 = __builtin_shufflevector(__s1_594, __s1_594, __lane_reverse_128_16); \
	56254	__ret_594 = __rev0_594 * __noswap_splat_laneq_u16(__rev1_594, __p2_594); \
	56255	__ret_594 = __builtin_shufflevector(__ret_594, __ret_594, __lane_reverse_64_16); \
	56256	__ret_594; \
54599	56257	})
54600	56258	#endif
54601	56259
54602	56260	#ifdef __LITTLE_ENDIAN__
54603		#define vmul_laneq_f32(__p0_579, __p1_579, __p2_579) __extension__ ({ \
54604		float32x2_t __ret_579; \
54605		float32x2_t __s0_579 = __p0_579; \
54606		float32x4_t __s1_579 = __p1_579; \
54607		__ret_579 = __s0_579 * splat_laneq_f32(__s1_579, __p2_579); \
54608		__ret_579; \
	56261	#define vmul_laneq_f32(__p0_595, __p1_595, __p2_595) __extension__ ({ \
	56262	float32x2_t __ret_595; \
	56263	float32x2_t __s0_595 = __p0_595; \
	56264	float32x4_t __s1_595 = __p1_595; \
	56265	__ret_595 = __s0_595 * splat_laneq_f32(__s1_595, __p2_595); \
	56266	__ret_595; \
54609	56267	})
54610	56268	#else
54611		#define vmul_laneq_f32(__p0_580, __p1_580, __p2_580) __extension__ ({ \
54612		float32x2_t __ret_580; \
54613		float32x2_t __s0_580 = __p0_580; \
54614		float32x4_t __s1_580 = __p1_580; \
54615		float32x2_t __rev0_580; __rev0_580 = __builtin_shufflevector(__s0_580, __s0_580, 1, 0); \
54616		float32x4_t __rev1_580; __rev1_580 = __builtin_shufflevector(__s1_580, __s1_580, 3, 2, 1, 0); \
54617		__ret_580 = __rev0_580 * __noswap_splat_laneq_f32(__rev1_580, __p2_580); \
54618		__ret_580 = __builtin_shufflevector(__ret_580, __ret_580, 1, 0); \
54619		__ret_580; \
	56269	#define vmul_laneq_f32(__p0_596, __p1_596, __p2_596) __extension__ ({ \
	56270	float32x2_t __ret_596; \
	56271	float32x2_t __s0_596 = __p0_596; \
	56272	float32x4_t __s1_596 = __p1_596; \
	56273	float32x2_t __rev0_596; __rev0_596 = __builtin_shufflevector(__s0_596, __s0_596, __lane_reverse_64_32); \
	56274	float32x4_t __rev1_596; __rev1_596 = __builtin_shufflevector(__s1_596, __s1_596, __lane_reverse_128_32); \
	56275	__ret_596 = __rev0_596 * __noswap_splat_laneq_f32(__rev1_596, __p2_596); \
	56276	__ret_596 = __builtin_shufflevector(__ret_596, __ret_596, __lane_reverse_64_32); \
	56277	__ret_596; \
54620	56278	})
54621	56279	#endif
54622	56280
54623	56281	#ifdef __LITTLE_ENDIAN__
54624		#define vmul_laneq_s32(__p0_581, __p1_581, __p2_581) __extension__ ({ \
54625		int32x2_t __ret_581; \
54626		int32x2_t __s0_581 = __p0_581; \
54627		int32x4_t __s1_581 = __p1_581; \
54628		__ret_581 = __s0_581 * splat_laneq_s32(__s1_581, __p2_581); \
54629		__ret_581; \
	56282	#define vmul_laneq_s32(__p0_597, __p1_597, __p2_597) __extension__ ({ \
	56283	int32x2_t __ret_597; \
	56284	int32x2_t __s0_597 = __p0_597; \
	56285	int32x4_t __s1_597 = __p1_597; \
	56286	__ret_597 = __s0_597 * splat_laneq_s32(__s1_597, __p2_597); \
	56287	__ret_597; \
54630	56288	})
54631	56289	#else
54632		#define vmul_laneq_s32(__p0_582, __p1_582, __p2_582) __extension__ ({ \
54633		int32x2_t __ret_582; \
54634		int32x2_t __s0_582 = __p0_582; \
54635		int32x4_t __s1_582 = __p1_582; \
54636		int32x2_t __rev0_582; __rev0_582 = __builtin_shufflevector(__s0_582, __s0_582, 1, 0); \
54637		int32x4_t __rev1_582; __rev1_582 = __builtin_shufflevector(__s1_582, __s1_582, 3, 2, 1, 0); \
54638		__ret_582 = __rev0_582 * __noswap_splat_laneq_s32(__rev1_582, __p2_582); \
54639		__ret_582 = __builtin_shufflevector(__ret_582, __ret_582, 1, 0); \
54640		__ret_582; \
	56290	#define vmul_laneq_s32(__p0_598, __p1_598, __p2_598) __extension__ ({ \
	56291	int32x2_t __ret_598; \
	56292	int32x2_t __s0_598 = __p0_598; \
	56293	int32x4_t __s1_598 = __p1_598; \
	56294	int32x2_t __rev0_598; __rev0_598 = __builtin_shufflevector(__s0_598, __s0_598, __lane_reverse_64_32); \
	56295	int32x4_t __rev1_598; __rev1_598 = __builtin_shufflevector(__s1_598, __s1_598, __lane_reverse_128_32); \
	56296	__ret_598 = __rev0_598 * __noswap_splat_laneq_s32(__rev1_598, __p2_598); \
	56297	__ret_598 = __builtin_shufflevector(__ret_598, __ret_598, __lane_reverse_64_32); \
	56298	__ret_598; \
54641	56299	})
54642	56300	#endif
54643	56301
54644	56302	#ifdef __LITTLE_ENDIAN__
54645		#define vmul_laneq_s16(__p0_583, __p1_583, __p2_583) __extension__ ({ \
54646		int16x4_t __ret_583; \
54647		int16x4_t __s0_583 = __p0_583; \
54648		int16x8_t __s1_583 = __p1_583; \
54649		__ret_583 = __s0_583 * splat_laneq_s16(__s1_583, __p2_583); \
54650		__ret_583; \
	56303	#define vmul_laneq_s16(__p0_599, __p1_599, __p2_599) __extension__ ({ \
	56304	int16x4_t __ret_599; \
	56305	int16x4_t __s0_599 = __p0_599; \
	56306	int16x8_t __s1_599 = __p1_599; \
	56307	__ret_599 = __s0_599 * splat_laneq_s16(__s1_599, __p2_599); \
	56308	__ret_599; \
54651	56309	})
54652	56310	#else
54653		#define vmul_laneq_s16(__p0_584, __p1_584, __p2_584) __extension__ ({ \
54654		int16x4_t __ret_584; \
54655		int16x4_t __s0_584 = __p0_584; \
54656		int16x8_t __s1_584 = __p1_584; \
54657		int16x4_t __rev0_584; __rev0_584 = __builtin_shufflevector(__s0_584, __s0_584, 3, 2, 1, 0); \
54658		int16x8_t __rev1_584; __rev1_584 = __builtin_shufflevector(__s1_584, __s1_584, 7, 6, 5, 4, 3, 2, 1, 0); \
54659		__ret_584 = __rev0_584 * __noswap_splat_laneq_s16(__rev1_584, __p2_584); \
54660		__ret_584 = __builtin_shufflevector(__ret_584, __ret_584, 3, 2, 1, 0); \
54661		__ret_584; \
	56311	#define vmul_laneq_s16(__p0_600, __p1_600, __p2_600) __extension__ ({ \
	56312	int16x4_t __ret_600; \
	56313	int16x4_t __s0_600 = __p0_600; \
	56314	int16x8_t __s1_600 = __p1_600; \
	56315	int16x4_t __rev0_600; __rev0_600 = __builtin_shufflevector(__s0_600, __s0_600, __lane_reverse_64_16); \
	56316	int16x8_t __rev1_600; __rev1_600 = __builtin_shufflevector(__s1_600, __s1_600, __lane_reverse_128_16); \
	56317	__ret_600 = __rev0_600 * __noswap_splat_laneq_s16(__rev1_600, __p2_600); \
	56318	__ret_600 = __builtin_shufflevector(__ret_600, __ret_600, __lane_reverse_64_16); \
	56319	__ret_600; \
54662	56320	})
54663	56321	#endif
54664	56322
54665	56323	__ai __attribute__((target("neon"))) float64x1_t vmul_n_f64(float64x1_t __p0, float64_t __p1) {
54666	56324	float64x1_t __ret;
54667		__ret = (float64x1_t) __builtin_neon_vmul_n_f64((float64x1_t)__p0, __p1);
	56325	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vmul_n_f64(__p0, __p1));
54668	56326	return __ret;
54669	56327	}
54670	56328	#ifdef __LITTLE_ENDIAN__
...	...	@@ -54676,9 +56334,9 @@ __ai __attribute__((target("neon"))) float64x2_t vmulq_n_f64(float64x2_t __p0, f
54676	56334	#else
54677	56335	__ai __attribute__((target("neon"))) float64x2_t vmulq_n_f64(float64x2_t __p0, float64_t __p1) {
54678	56336	float64x2_t __ret;
54679		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
	56337	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
54680	56338	__ret = __rev0 * (float64x2_t) {__p1, __p1};
54681		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	56339	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
54682	56340	return __ret;
54683	56341	}
54684	56342	#endif
...	...	@@ -54692,10 +56350,10 @@ __ai __attribute__((target("neon"))) poly16x8_t vmull_high_p8(poly8x16_t __p0, p
54692	56350	#else
54693	56351	__ai __attribute__((target("neon"))) poly16x8_t vmull_high_p8(poly8x16_t __p0, poly8x16_t __p1) {
54694	56352	poly16x8_t __ret;
54695		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
54696		poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	56353	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	56354	poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
54697	56355	__ret = __noswap_vmull_p8(__noswap_vget_high_p8(__rev0), __noswap_vget_high_p8(__rev1));
54698		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	56356	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
54699	56357	return __ret;
54700	56358	}
54701	56359	#endif
...	...	@@ -54709,10 +56367,10 @@ __ai __attribute__((target("neon"))) uint16x8_t vmull_high_u8(uint8x16_t __p0, u
54709	56367	#else
54710	56368	__ai __attribute__((target("neon"))) uint16x8_t vmull_high_u8(uint8x16_t __p0, uint8x16_t __p1) {
54711	56369	uint16x8_t __ret;
54712		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
54713		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	56370	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	56371	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
54714	56372	__ret = __noswap_vmull_u8(__noswap_vget_high_u8(__rev0), __noswap_vget_high_u8(__rev1));
54715		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	56373	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
54716	56374	return __ret;
54717	56375	}
54718	56376	#endif
...	...	@@ -54726,10 +56384,10 @@ __ai __attribute__((target("neon"))) uint64x2_t vmull_high_u32(uint32x4_t __p0,
54726	56384	#else
54727	56385	__ai __attribute__((target("neon"))) uint64x2_t vmull_high_u32(uint32x4_t __p0, uint32x4_t __p1) {
54728	56386	uint64x2_t __ret;
54729		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
54730		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	56387	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	56388	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
54731	56389	__ret = __noswap_vmull_u32(__noswap_vget_high_u32(__rev0), __noswap_vget_high_u32(__rev1));
54732		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	56390	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
54733	56391	return __ret;
54734	56392	}
54735	56393	#endif
...	...	@@ -54743,10 +56401,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vmull_high_u16(uint16x8_t __p0,
54743	56401	#else
54744	56402	__ai __attribute__((target("neon"))) uint32x4_t vmull_high_u16(uint16x8_t __p0, uint16x8_t __p1) {
54745	56403	uint32x4_t __ret;
54746		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
54747		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	56404	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	56405	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
54748	56406	__ret = __noswap_vmull_u16(__noswap_vget_high_u16(__rev0), __noswap_vget_high_u16(__rev1));
54749		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	56407	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
54750	56408	return __ret;
54751	56409	}
54752	56410	#endif
...	...	@@ -54760,10 +56418,10 @@ __ai __attribute__((target("neon"))) int16x8_t vmull_high_s8(int8x16_t __p0, int
54760	56418	#else
54761	56419	__ai __attribute__((target("neon"))) int16x8_t vmull_high_s8(int8x16_t __p0, int8x16_t __p1) {
54762	56420	int16x8_t __ret;
54763		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
54764		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	56421	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	56422	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
54765	56423	__ret = __noswap_vmull_s8(__noswap_vget_high_s8(__rev0), __noswap_vget_high_s8(__rev1));
54766		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	56424	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
54767	56425	return __ret;
54768	56426	}
54769	56427	#endif
...	...	@@ -54777,10 +56435,10 @@ __ai __attribute__((target("neon"))) int64x2_t vmull_high_s32(int32x4_t __p0, in
54777	56435	#else
54778	56436	__ai __attribute__((target("neon"))) int64x2_t vmull_high_s32(int32x4_t __p0, int32x4_t __p1) {
54779	56437	int64x2_t __ret;
54780		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
54781		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	56438	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	56439	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
54782	56440	__ret = __noswap_vmull_s32(__noswap_vget_high_s32(__rev0), __noswap_vget_high_s32(__rev1));
54783		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	56441	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
54784	56442	return __ret;
54785	56443	}
54786	56444	#endif
...	...	@@ -54794,179 +56452,179 @@ __ai __attribute__((target("neon"))) int32x4_t vmull_high_s16(int16x8_t __p0, in
54794	56452	#else
54795	56453	__ai __attribute__((target("neon"))) int32x4_t vmull_high_s16(int16x8_t __p0, int16x8_t __p1) {
54796	56454	int32x4_t __ret;
54797		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
54798		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	56455	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	56456	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
54799	56457	__ret = __noswap_vmull_s16(__noswap_vget_high_s16(__rev0), __noswap_vget_high_s16(__rev1));
54800		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	56458	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
54801	56459	return __ret;
54802	56460	}
54803	56461	#endif
54804	56462
54805	56463	#ifdef __LITTLE_ENDIAN__
54806		#define vmull_high_lane_u32(__p0_585, __p1_585, __p2_585) __extension__ ({ \
54807		uint64x2_t __ret_585; \
54808		uint32x4_t __s0_585 = __p0_585; \
54809		uint32x2_t __s1_585 = __p1_585; \
54810		__ret_585 = vmull_u32(vget_high_u32(__s0_585), splat_lane_u32(__s1_585, __p2_585)); \
54811		__ret_585; \
	56464	#define vmull_high_lane_u32(__p0_601, __p1_601, __p2_601) __extension__ ({ \
	56465	uint64x2_t __ret_601; \
	56466	uint32x4_t __s0_601 = __p0_601; \
	56467	uint32x2_t __s1_601 = __p1_601; \
	56468	__ret_601 = vmull_u32(vget_high_u32(__s0_601), splat_lane_u32(__s1_601, __p2_601)); \
	56469	__ret_601; \
54812	56470	})
54813	56471	#else
54814		#define vmull_high_lane_u32(__p0_586, __p1_586, __p2_586) __extension__ ({ \
54815		uint64x2_t __ret_586; \
54816		uint32x4_t __s0_586 = __p0_586; \
54817		uint32x2_t __s1_586 = __p1_586; \
54818		uint32x4_t __rev0_586; __rev0_586 = __builtin_shufflevector(__s0_586, __s0_586, 3, 2, 1, 0); \
54819		uint32x2_t __rev1_586; __rev1_586 = __builtin_shufflevector(__s1_586, __s1_586, 1, 0); \
54820		__ret_586 = __noswap_vmull_u32(__noswap_vget_high_u32(__rev0_586), __noswap_splat_lane_u32(__rev1_586, __p2_586)); \
54821		__ret_586 = __builtin_shufflevector(__ret_586, __ret_586, 1, 0); \
54822		__ret_586; \
	56472	#define vmull_high_lane_u32(__p0_602, __p1_602, __p2_602) __extension__ ({ \
	56473	uint64x2_t __ret_602; \
	56474	uint32x4_t __s0_602 = __p0_602; \
	56475	uint32x2_t __s1_602 = __p1_602; \
	56476	uint32x4_t __rev0_602; __rev0_602 = __builtin_shufflevector(__s0_602, __s0_602, __lane_reverse_128_32); \
	56477	uint32x2_t __rev1_602; __rev1_602 = __builtin_shufflevector(__s1_602, __s1_602, __lane_reverse_64_32); \
	56478	__ret_602 = __noswap_vmull_u32(__noswap_vget_high_u32(__rev0_602), __noswap_splat_lane_u32(__rev1_602, __p2_602)); \
	56479	__ret_602 = __builtin_shufflevector(__ret_602, __ret_602, __lane_reverse_128_64); \
	56480	__ret_602; \
54823	56481	})
54824	56482	#endif
54825	56483
54826	56484	#ifdef __LITTLE_ENDIAN__
54827		#define vmull_high_lane_u16(__p0_587, __p1_587, __p2_587) __extension__ ({ \
54828		uint32x4_t __ret_587; \
54829		uint16x8_t __s0_587 = __p0_587; \
54830		uint16x4_t __s1_587 = __p1_587; \
54831		__ret_587 = vmull_u16(vget_high_u16(__s0_587), splat_lane_u16(__s1_587, __p2_587)); \
54832		__ret_587; \
	56485	#define vmull_high_lane_u16(__p0_603, __p1_603, __p2_603) __extension__ ({ \
	56486	uint32x4_t __ret_603; \
	56487	uint16x8_t __s0_603 = __p0_603; \
	56488	uint16x4_t __s1_603 = __p1_603; \
	56489	__ret_603 = vmull_u16(vget_high_u16(__s0_603), splat_lane_u16(__s1_603, __p2_603)); \
	56490	__ret_603; \
54833	56491	})
54834	56492	#else
54835		#define vmull_high_lane_u16(__p0_588, __p1_588, __p2_588) __extension__ ({ \
54836		uint32x4_t __ret_588; \
54837		uint16x8_t __s0_588 = __p0_588; \
54838		uint16x4_t __s1_588 = __p1_588; \
54839		uint16x8_t __rev0_588; __rev0_588 = __builtin_shufflevector(__s0_588, __s0_588, 7, 6, 5, 4, 3, 2, 1, 0); \
54840		uint16x4_t __rev1_588; __rev1_588 = __builtin_shufflevector(__s1_588, __s1_588, 3, 2, 1, 0); \
54841		__ret_588 = __noswap_vmull_u16(__noswap_vget_high_u16(__rev0_588), __noswap_splat_lane_u16(__rev1_588, __p2_588)); \
54842		__ret_588 = __builtin_shufflevector(__ret_588, __ret_588, 3, 2, 1, 0); \
54843		__ret_588; \
	56493	#define vmull_high_lane_u16(__p0_604, __p1_604, __p2_604) __extension__ ({ \
	56494	uint32x4_t __ret_604; \
	56495	uint16x8_t __s0_604 = __p0_604; \
	56496	uint16x4_t __s1_604 = __p1_604; \
	56497	uint16x8_t __rev0_604; __rev0_604 = __builtin_shufflevector(__s0_604, __s0_604, __lane_reverse_128_16); \
	56498	uint16x4_t __rev1_604; __rev1_604 = __builtin_shufflevector(__s1_604, __s1_604, __lane_reverse_64_16); \
	56499	__ret_604 = __noswap_vmull_u16(__noswap_vget_high_u16(__rev0_604), __noswap_splat_lane_u16(__rev1_604, __p2_604)); \
	56500	__ret_604 = __builtin_shufflevector(__ret_604, __ret_604, __lane_reverse_128_32); \
	56501	__ret_604; \
54844	56502	})
54845	56503	#endif
54846	56504
54847	56505	#ifdef __LITTLE_ENDIAN__
54848		#define vmull_high_lane_s32(__p0_589, __p1_589, __p2_589) __extension__ ({ \
54849		int64x2_t __ret_589; \
54850		int32x4_t __s0_589 = __p0_589; \
54851		int32x2_t __s1_589 = __p1_589; \
54852		__ret_589 = vmull_s32(vget_high_s32(__s0_589), splat_lane_s32(__s1_589, __p2_589)); \
54853		__ret_589; \
	56506	#define vmull_high_lane_s32(__p0_605, __p1_605, __p2_605) __extension__ ({ \
	56507	int64x2_t __ret_605; \
	56508	int32x4_t __s0_605 = __p0_605; \
	56509	int32x2_t __s1_605 = __p1_605; \
	56510	__ret_605 = vmull_s32(vget_high_s32(__s0_605), splat_lane_s32(__s1_605, __p2_605)); \
	56511	__ret_605; \
54854	56512	})
54855	56513	#else
54856		#define vmull_high_lane_s32(__p0_590, __p1_590, __p2_590) __extension__ ({ \
54857		int64x2_t __ret_590; \
54858		int32x4_t __s0_590 = __p0_590; \
54859		int32x2_t __s1_590 = __p1_590; \
54860		int32x4_t __rev0_590; __rev0_590 = __builtin_shufflevector(__s0_590, __s0_590, 3, 2, 1, 0); \
54861		int32x2_t __rev1_590; __rev1_590 = __builtin_shufflevector(__s1_590, __s1_590, 1, 0); \
54862		__ret_590 = __noswap_vmull_s32(__noswap_vget_high_s32(__rev0_590), __noswap_splat_lane_s32(__rev1_590, __p2_590)); \
54863		__ret_590 = __builtin_shufflevector(__ret_590, __ret_590, 1, 0); \
54864		__ret_590; \
	56514	#define vmull_high_lane_s32(__p0_606, __p1_606, __p2_606) __extension__ ({ \
	56515	int64x2_t __ret_606; \
	56516	int32x4_t __s0_606 = __p0_606; \
	56517	int32x2_t __s1_606 = __p1_606; \
	56518	int32x4_t __rev0_606; __rev0_606 = __builtin_shufflevector(__s0_606, __s0_606, __lane_reverse_128_32); \
	56519	int32x2_t __rev1_606; __rev1_606 = __builtin_shufflevector(__s1_606, __s1_606, __lane_reverse_64_32); \
	56520	__ret_606 = __noswap_vmull_s32(__noswap_vget_high_s32(__rev0_606), __noswap_splat_lane_s32(__rev1_606, __p2_606)); \
	56521	__ret_606 = __builtin_shufflevector(__ret_606, __ret_606, __lane_reverse_128_64); \
	56522	__ret_606; \
54865	56523	})
54866	56524	#endif
54867	56525
54868	56526	#ifdef __LITTLE_ENDIAN__
54869		#define vmull_high_lane_s16(__p0_591, __p1_591, __p2_591) __extension__ ({ \
54870		int32x4_t __ret_591; \
54871		int16x8_t __s0_591 = __p0_591; \
54872		int16x4_t __s1_591 = __p1_591; \
54873		__ret_591 = vmull_s16(vget_high_s16(__s0_591), splat_lane_s16(__s1_591, __p2_591)); \
54874		__ret_591; \
	56527	#define vmull_high_lane_s16(__p0_607, __p1_607, __p2_607) __extension__ ({ \
	56528	int32x4_t __ret_607; \
	56529	int16x8_t __s0_607 = __p0_607; \
	56530	int16x4_t __s1_607 = __p1_607; \
	56531	__ret_607 = vmull_s16(vget_high_s16(__s0_607), splat_lane_s16(__s1_607, __p2_607)); \
	56532	__ret_607; \
54875	56533	})
54876	56534	#else
54877		#define vmull_high_lane_s16(__p0_592, __p1_592, __p2_592) __extension__ ({ \
54878		int32x4_t __ret_592; \
54879		int16x8_t __s0_592 = __p0_592; \
54880		int16x4_t __s1_592 = __p1_592; \
54881		int16x8_t __rev0_592; __rev0_592 = __builtin_shufflevector(__s0_592, __s0_592, 7, 6, 5, 4, 3, 2, 1, 0); \
54882		int16x4_t __rev1_592; __rev1_592 = __builtin_shufflevector(__s1_592, __s1_592, 3, 2, 1, 0); \
54883		__ret_592 = __noswap_vmull_s16(__noswap_vget_high_s16(__rev0_592), __noswap_splat_lane_s16(__rev1_592, __p2_592)); \
54884		__ret_592 = __builtin_shufflevector(__ret_592, __ret_592, 3, 2, 1, 0); \
54885		__ret_592; \
	56535	#define vmull_high_lane_s16(__p0_608, __p1_608, __p2_608) __extension__ ({ \
	56536	int32x4_t __ret_608; \
	56537	int16x8_t __s0_608 = __p0_608; \
	56538	int16x4_t __s1_608 = __p1_608; \
	56539	int16x8_t __rev0_608; __rev0_608 = __builtin_shufflevector(__s0_608, __s0_608, __lane_reverse_128_16); \
	56540	int16x4_t __rev1_608; __rev1_608 = __builtin_shufflevector(__s1_608, __s1_608, __lane_reverse_64_16); \
	56541	__ret_608 = __noswap_vmull_s16(__noswap_vget_high_s16(__rev0_608), __noswap_splat_lane_s16(__rev1_608, __p2_608)); \
	56542	__ret_608 = __builtin_shufflevector(__ret_608, __ret_608, __lane_reverse_128_32); \
	56543	__ret_608; \
54886	56544	})
54887	56545	#endif
54888	56546
54889	56547	#ifdef __LITTLE_ENDIAN__
54890		#define vmull_high_laneq_u32(__p0_593, __p1_593, __p2_593) __extension__ ({ \
54891		uint64x2_t __ret_593; \
54892		uint32x4_t __s0_593 = __p0_593; \
54893		uint32x4_t __s1_593 = __p1_593; \
54894		__ret_593 = vmull_u32(vget_high_u32(__s0_593), splat_laneq_u32(__s1_593, __p2_593)); \
54895		__ret_593; \
	56548	#define vmull_high_laneq_u32(__p0_609, __p1_609, __p2_609) __extension__ ({ \
	56549	uint64x2_t __ret_609; \
	56550	uint32x4_t __s0_609 = __p0_609; \
	56551	uint32x4_t __s1_609 = __p1_609; \
	56552	__ret_609 = vmull_u32(vget_high_u32(__s0_609), splat_laneq_u32(__s1_609, __p2_609)); \
	56553	__ret_609; \
54896	56554	})
54897	56555	#else
54898		#define vmull_high_laneq_u32(__p0_594, __p1_594, __p2_594) __extension__ ({ \
54899		uint64x2_t __ret_594; \
54900		uint32x4_t __s0_594 = __p0_594; \
54901		uint32x4_t __s1_594 = __p1_594; \
54902		uint32x4_t __rev0_594; __rev0_594 = __builtin_shufflevector(__s0_594, __s0_594, 3, 2, 1, 0); \
54903		uint32x4_t __rev1_594; __rev1_594 = __builtin_shufflevector(__s1_594, __s1_594, 3, 2, 1, 0); \
54904		__ret_594 = __noswap_vmull_u32(__noswap_vget_high_u32(__rev0_594), __noswap_splat_laneq_u32(__rev1_594, __p2_594)); \
54905		__ret_594 = __builtin_shufflevector(__ret_594, __ret_594, 1, 0); \
54906		__ret_594; \
	56556	#define vmull_high_laneq_u32(__p0_610, __p1_610, __p2_610) __extension__ ({ \
	56557	uint64x2_t __ret_610; \
	56558	uint32x4_t __s0_610 = __p0_610; \
	56559	uint32x4_t __s1_610 = __p1_610; \
	56560	uint32x4_t __rev0_610; __rev0_610 = __builtin_shufflevector(__s0_610, __s0_610, __lane_reverse_128_32); \
	56561	uint32x4_t __rev1_610; __rev1_610 = __builtin_shufflevector(__s1_610, __s1_610, __lane_reverse_128_32); \
	56562	__ret_610 = __noswap_vmull_u32(__noswap_vget_high_u32(__rev0_610), __noswap_splat_laneq_u32(__rev1_610, __p2_610)); \
	56563	__ret_610 = __builtin_shufflevector(__ret_610, __ret_610, __lane_reverse_128_64); \
	56564	__ret_610; \
54907	56565	})
54908	56566	#endif
54909	56567
54910	56568	#ifdef __LITTLE_ENDIAN__
54911		#define vmull_high_laneq_u16(__p0_595, __p1_595, __p2_595) __extension__ ({ \
54912		uint32x4_t __ret_595; \
54913		uint16x8_t __s0_595 = __p0_595; \
54914		uint16x8_t __s1_595 = __p1_595; \
54915		__ret_595 = vmull_u16(vget_high_u16(__s0_595), splat_laneq_u16(__s1_595, __p2_595)); \
54916		__ret_595; \
	56569	#define vmull_high_laneq_u16(__p0_611, __p1_611, __p2_611) __extension__ ({ \
	56570	uint32x4_t __ret_611; \
	56571	uint16x8_t __s0_611 = __p0_611; \
	56572	uint16x8_t __s1_611 = __p1_611; \
	56573	__ret_611 = vmull_u16(vget_high_u16(__s0_611), splat_laneq_u16(__s1_611, __p2_611)); \
	56574	__ret_611; \
54917	56575	})
54918	56576	#else
54919		#define vmull_high_laneq_u16(__p0_596, __p1_596, __p2_596) __extension__ ({ \
54920		uint32x4_t __ret_596; \
54921		uint16x8_t __s0_596 = __p0_596; \
54922		uint16x8_t __s1_596 = __p1_596; \
54923		uint16x8_t __rev0_596; __rev0_596 = __builtin_shufflevector(__s0_596, __s0_596, 7, 6, 5, 4, 3, 2, 1, 0); \
54924		uint16x8_t __rev1_596; __rev1_596 = __builtin_shufflevector(__s1_596, __s1_596, 7, 6, 5, 4, 3, 2, 1, 0); \
54925		__ret_596 = __noswap_vmull_u16(__noswap_vget_high_u16(__rev0_596), __noswap_splat_laneq_u16(__rev1_596, __p2_596)); \
54926		__ret_596 = __builtin_shufflevector(__ret_596, __ret_596, 3, 2, 1, 0); \
54927		__ret_596; \
	56577	#define vmull_high_laneq_u16(__p0_612, __p1_612, __p2_612) __extension__ ({ \
	56578	uint32x4_t __ret_612; \
	56579	uint16x8_t __s0_612 = __p0_612; \
	56580	uint16x8_t __s1_612 = __p1_612; \
	56581	uint16x8_t __rev0_612; __rev0_612 = __builtin_shufflevector(__s0_612, __s0_612, __lane_reverse_128_16); \
	56582	uint16x8_t __rev1_612; __rev1_612 = __builtin_shufflevector(__s1_612, __s1_612, __lane_reverse_128_16); \
	56583	__ret_612 = __noswap_vmull_u16(__noswap_vget_high_u16(__rev0_612), __noswap_splat_laneq_u16(__rev1_612, __p2_612)); \
	56584	__ret_612 = __builtin_shufflevector(__ret_612, __ret_612, __lane_reverse_128_32); \
	56585	__ret_612; \
54928	56586	})
54929	56587	#endif
54930	56588
54931	56589	#ifdef __LITTLE_ENDIAN__
54932		#define vmull_high_laneq_s32(__p0_597, __p1_597, __p2_597) __extension__ ({ \
54933		int64x2_t __ret_597; \
54934		int32x4_t __s0_597 = __p0_597; \
54935		int32x4_t __s1_597 = __p1_597; \
54936		__ret_597 = vmull_s32(vget_high_s32(__s0_597), splat_laneq_s32(__s1_597, __p2_597)); \
54937		__ret_597; \
	56590	#define vmull_high_laneq_s32(__p0_613, __p1_613, __p2_613) __extension__ ({ \
	56591	int64x2_t __ret_613; \
	56592	int32x4_t __s0_613 = __p0_613; \
	56593	int32x4_t __s1_613 = __p1_613; \
	56594	__ret_613 = vmull_s32(vget_high_s32(__s0_613), splat_laneq_s32(__s1_613, __p2_613)); \
	56595	__ret_613; \
54938	56596	})
54939	56597	#else
54940		#define vmull_high_laneq_s32(__p0_598, __p1_598, __p2_598) __extension__ ({ \
54941		int64x2_t __ret_598; \
54942		int32x4_t __s0_598 = __p0_598; \
54943		int32x4_t __s1_598 = __p1_598; \
54944		int32x4_t __rev0_598; __rev0_598 = __builtin_shufflevector(__s0_598, __s0_598, 3, 2, 1, 0); \
54945		int32x4_t __rev1_598; __rev1_598 = __builtin_shufflevector(__s1_598, __s1_598, 3, 2, 1, 0); \
54946		__ret_598 = __noswap_vmull_s32(__noswap_vget_high_s32(__rev0_598), __noswap_splat_laneq_s32(__rev1_598, __p2_598)); \
54947		__ret_598 = __builtin_shufflevector(__ret_598, __ret_598, 1, 0); \
54948		__ret_598; \
	56598	#define vmull_high_laneq_s32(__p0_614, __p1_614, __p2_614) __extension__ ({ \
	56599	int64x2_t __ret_614; \
	56600	int32x4_t __s0_614 = __p0_614; \
	56601	int32x4_t __s1_614 = __p1_614; \
	56602	int32x4_t __rev0_614; __rev0_614 = __builtin_shufflevector(__s0_614, __s0_614, __lane_reverse_128_32); \
	56603	int32x4_t __rev1_614; __rev1_614 = __builtin_shufflevector(__s1_614, __s1_614, __lane_reverse_128_32); \
	56604	__ret_614 = __noswap_vmull_s32(__noswap_vget_high_s32(__rev0_614), __noswap_splat_laneq_s32(__rev1_614, __p2_614)); \
	56605	__ret_614 = __builtin_shufflevector(__ret_614, __ret_614, __lane_reverse_128_64); \
	56606	__ret_614; \
54949	56607	})
54950	56608	#endif
54951	56609
54952	56610	#ifdef __LITTLE_ENDIAN__
54953		#define vmull_high_laneq_s16(__p0_599, __p1_599, __p2_599) __extension__ ({ \
54954		int32x4_t __ret_599; \
54955		int16x8_t __s0_599 = __p0_599; \
54956		int16x8_t __s1_599 = __p1_599; \
54957		__ret_599 = vmull_s16(vget_high_s16(__s0_599), splat_laneq_s16(__s1_599, __p2_599)); \
54958		__ret_599; \
	56611	#define vmull_high_laneq_s16(__p0_615, __p1_615, __p2_615) __extension__ ({ \
	56612	int32x4_t __ret_615; \
	56613	int16x8_t __s0_615 = __p0_615; \
	56614	int16x8_t __s1_615 = __p1_615; \
	56615	__ret_615 = vmull_s16(vget_high_s16(__s0_615), splat_laneq_s16(__s1_615, __p2_615)); \
	56616	__ret_615; \
54959	56617	})
54960	56618	#else
54961		#define vmull_high_laneq_s16(__p0_600, __p1_600, __p2_600) __extension__ ({ \
54962		int32x4_t __ret_600; \
54963		int16x8_t __s0_600 = __p0_600; \
54964		int16x8_t __s1_600 = __p1_600; \
54965		int16x8_t __rev0_600; __rev0_600 = __builtin_shufflevector(__s0_600, __s0_600, 7, 6, 5, 4, 3, 2, 1, 0); \
54966		int16x8_t __rev1_600; __rev1_600 = __builtin_shufflevector(__s1_600, __s1_600, 7, 6, 5, 4, 3, 2, 1, 0); \
54967		__ret_600 = __noswap_vmull_s16(__noswap_vget_high_s16(__rev0_600), __noswap_splat_laneq_s16(__rev1_600, __p2_600)); \
54968		__ret_600 = __builtin_shufflevector(__ret_600, __ret_600, 3, 2, 1, 0); \
54969		__ret_600; \
	56619	#define vmull_high_laneq_s16(__p0_616, __p1_616, __p2_616) __extension__ ({ \
	56620	int32x4_t __ret_616; \
	56621	int16x8_t __s0_616 = __p0_616; \
	56622	int16x8_t __s1_616 = __p1_616; \
	56623	int16x8_t __rev0_616; __rev0_616 = __builtin_shufflevector(__s0_616, __s0_616, __lane_reverse_128_16); \
	56624	int16x8_t __rev1_616; __rev1_616 = __builtin_shufflevector(__s1_616, __s1_616, __lane_reverse_128_16); \
	56625	__ret_616 = __noswap_vmull_s16(__noswap_vget_high_s16(__rev0_616), __noswap_splat_laneq_s16(__rev1_616, __p2_616)); \
	56626	__ret_616 = __builtin_shufflevector(__ret_616, __ret_616, __lane_reverse_128_32); \
	56627	__ret_616; \
54970	56628	})
54971	56629	#endif
54972	56630
...	...	@@ -54979,9 +56637,9 @@ __ai __attribute__((target("neon"))) uint64x2_t vmull_high_n_u32(uint32x4_t __p0
54979	56637	#else
54980	56638	__ai __attribute__((target("neon"))) uint64x2_t vmull_high_n_u32(uint32x4_t __p0, uint32_t __p1) {
54981	56639	uint64x2_t __ret;
54982		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	56640	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
54983	56641	__ret = __noswap_vmull_n_u32(__noswap_vget_high_u32(__rev0), __p1);
54984		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	56642	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
54985	56643	return __ret;
54986	56644	}
54987	56645	#endif
...	...	@@ -54995,9 +56653,9 @@ __ai __attribute__((target("neon"))) uint32x4_t vmull_high_n_u16(uint16x8_t __p0
54995	56653	#else
54996	56654	__ai __attribute__((target("neon"))) uint32x4_t vmull_high_n_u16(uint16x8_t __p0, uint16_t __p1) {
54997	56655	uint32x4_t __ret;
54998		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	56656	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
54999	56657	__ret = __noswap_vmull_n_u16(__noswap_vget_high_u16(__rev0), __p1);
55000		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	56658	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
55001	56659	return __ret;
55002	56660	}
55003	56661	#endif
...	...	@@ -55011,9 +56669,9 @@ __ai __attribute__((target("neon"))) int64x2_t vmull_high_n_s32(int32x4_t __p0,
55011	56669	#else
55012	56670	__ai __attribute__((target("neon"))) int64x2_t vmull_high_n_s32(int32x4_t __p0, int32_t __p1) {
55013	56671	int64x2_t __ret;
55014		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	56672	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
55015	56673	__ret = __noswap_vmull_n_s32(__noswap_vget_high_s32(__rev0), __p1);
55016		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	56674	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
55017	56675	return __ret;
55018	56676	}
55019	56677	#endif
...	...	@@ -55027,115 +56685,115 @@ __ai __attribute__((target("neon"))) int32x4_t vmull_high_n_s16(int16x8_t __p0,
55027	56685	#else
55028	56686	__ai __attribute__((target("neon"))) int32x4_t vmull_high_n_s16(int16x8_t __p0, int16_t __p1) {
55029	56687	int32x4_t __ret;
55030		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	56688	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
55031	56689	__ret = __noswap_vmull_n_s16(__noswap_vget_high_s16(__rev0), __p1);
55032		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	56690	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
55033	56691	return __ret;
55034	56692	}
55035	56693	#endif
55036	56694
55037	56695	#ifdef __LITTLE_ENDIAN__
55038		#define vmull_laneq_u32(__p0_601, __p1_601, __p2_601) __extension__ ({ \
55039		uint64x2_t __ret_601; \
55040		uint32x2_t __s0_601 = __p0_601; \
55041		uint32x4_t __s1_601 = __p1_601; \
55042		__ret_601 = vmull_u32(__s0_601, splat_laneq_u32(__s1_601, __p2_601)); \
55043		__ret_601; \
	56696	#define vmull_laneq_u32(__p0_617, __p1_617, __p2_617) __extension__ ({ \
	56697	uint64x2_t __ret_617; \
	56698	uint32x2_t __s0_617 = __p0_617; \
	56699	uint32x4_t __s1_617 = __p1_617; \
	56700	__ret_617 = vmull_u32(__s0_617, splat_laneq_u32(__s1_617, __p2_617)); \
	56701	__ret_617; \
55044	56702	})
55045	56703	#else
55046		#define vmull_laneq_u32(__p0_602, __p1_602, __p2_602) __extension__ ({ \
55047		uint64x2_t __ret_602; \
55048		uint32x2_t __s0_602 = __p0_602; \
55049		uint32x4_t __s1_602 = __p1_602; \
55050		uint32x2_t __rev0_602; __rev0_602 = __builtin_shufflevector(__s0_602, __s0_602, 1, 0); \
55051		uint32x4_t __rev1_602; __rev1_602 = __builtin_shufflevector(__s1_602, __s1_602, 3, 2, 1, 0); \
55052		__ret_602 = __noswap_vmull_u32(__rev0_602, __noswap_splat_laneq_u32(__rev1_602, __p2_602)); \
55053		__ret_602 = __builtin_shufflevector(__ret_602, __ret_602, 1, 0); \
55054		__ret_602; \
	56704	#define vmull_laneq_u32(__p0_618, __p1_618, __p2_618) __extension__ ({ \
	56705	uint64x2_t __ret_618; \
	56706	uint32x2_t __s0_618 = __p0_618; \
	56707	uint32x4_t __s1_618 = __p1_618; \
	56708	uint32x2_t __rev0_618; __rev0_618 = __builtin_shufflevector(__s0_618, __s0_618, __lane_reverse_64_32); \
	56709	uint32x4_t __rev1_618; __rev1_618 = __builtin_shufflevector(__s1_618, __s1_618, __lane_reverse_128_32); \
	56710	__ret_618 = __noswap_vmull_u32(__rev0_618, __noswap_splat_laneq_u32(__rev1_618, __p2_618)); \
	56711	__ret_618 = __builtin_shufflevector(__ret_618, __ret_618, __lane_reverse_128_64); \
	56712	__ret_618; \
55055	56713	})
55056	56714	#endif
55057	56715
55058	56716	#ifdef __LITTLE_ENDIAN__
55059		#define vmull_laneq_u16(__p0_603, __p1_603, __p2_603) __extension__ ({ \
55060		uint32x4_t __ret_603; \
55061		uint16x4_t __s0_603 = __p0_603; \
55062		uint16x8_t __s1_603 = __p1_603; \
55063		__ret_603 = vmull_u16(__s0_603, splat_laneq_u16(__s1_603, __p2_603)); \
55064		__ret_603; \
	56717	#define vmull_laneq_u16(__p0_619, __p1_619, __p2_619) __extension__ ({ \
	56718	uint32x4_t __ret_619; \
	56719	uint16x4_t __s0_619 = __p0_619; \
	56720	uint16x8_t __s1_619 = __p1_619; \
	56721	__ret_619 = vmull_u16(__s0_619, splat_laneq_u16(__s1_619, __p2_619)); \
	56722	__ret_619; \
55065	56723	})
55066	56724	#else
55067		#define vmull_laneq_u16(__p0_604, __p1_604, __p2_604) __extension__ ({ \
55068		uint32x4_t __ret_604; \
55069		uint16x4_t __s0_604 = __p0_604; \
55070		uint16x8_t __s1_604 = __p1_604; \
55071		uint16x4_t __rev0_604; __rev0_604 = __builtin_shufflevector(__s0_604, __s0_604, 3, 2, 1, 0); \
55072		uint16x8_t __rev1_604; __rev1_604 = __builtin_shufflevector(__s1_604, __s1_604, 7, 6, 5, 4, 3, 2, 1, 0); \
55073		__ret_604 = __noswap_vmull_u16(__rev0_604, __noswap_splat_laneq_u16(__rev1_604, __p2_604)); \
55074		__ret_604 = __builtin_shufflevector(__ret_604, __ret_604, 3, 2, 1, 0); \
55075		__ret_604; \
	56725	#define vmull_laneq_u16(__p0_620, __p1_620, __p2_620) __extension__ ({ \
	56726	uint32x4_t __ret_620; \
	56727	uint16x4_t __s0_620 = __p0_620; \
	56728	uint16x8_t __s1_620 = __p1_620; \
	56729	uint16x4_t __rev0_620; __rev0_620 = __builtin_shufflevector(__s0_620, __s0_620, __lane_reverse_64_16); \
	56730	uint16x8_t __rev1_620; __rev1_620 = __builtin_shufflevector(__s1_620, __s1_620, __lane_reverse_128_16); \
	56731	__ret_620 = __noswap_vmull_u16(__rev0_620, __noswap_splat_laneq_u16(__rev1_620, __p2_620)); \
	56732	__ret_620 = __builtin_shufflevector(__ret_620, __ret_620, __lane_reverse_128_32); \
	56733	__ret_620; \
55076	56734	})
55077	56735	#endif
55078	56736
55079	56737	#ifdef __LITTLE_ENDIAN__
55080		#define vmull_laneq_s32(__p0_605, __p1_605, __p2_605) __extension__ ({ \
55081		int64x2_t __ret_605; \
55082		int32x2_t __s0_605 = __p0_605; \
55083		int32x4_t __s1_605 = __p1_605; \
55084		__ret_605 = vmull_s32(__s0_605, splat_laneq_s32(__s1_605, __p2_605)); \
55085		__ret_605; \
	56738	#define vmull_laneq_s32(__p0_621, __p1_621, __p2_621) __extension__ ({ \
	56739	int64x2_t __ret_621; \
	56740	int32x2_t __s0_621 = __p0_621; \
	56741	int32x4_t __s1_621 = __p1_621; \
	56742	__ret_621 = vmull_s32(__s0_621, splat_laneq_s32(__s1_621, __p2_621)); \
	56743	__ret_621; \
55086	56744	})
55087	56745	#else
55088		#define vmull_laneq_s32(__p0_606, __p1_606, __p2_606) __extension__ ({ \
55089		int64x2_t __ret_606; \
55090		int32x2_t __s0_606 = __p0_606; \
55091		int32x4_t __s1_606 = __p1_606; \
55092		int32x2_t __rev0_606; __rev0_606 = __builtin_shufflevector(__s0_606, __s0_606, 1, 0); \
55093		int32x4_t __rev1_606; __rev1_606 = __builtin_shufflevector(__s1_606, __s1_606, 3, 2, 1, 0); \
55094		__ret_606 = __noswap_vmull_s32(__rev0_606, __noswap_splat_laneq_s32(__rev1_606, __p2_606)); \
55095		__ret_606 = __builtin_shufflevector(__ret_606, __ret_606, 1, 0); \
55096		__ret_606; \
	56746	#define vmull_laneq_s32(__p0_622, __p1_622, __p2_622) __extension__ ({ \
	56747	int64x2_t __ret_622; \
	56748	int32x2_t __s0_622 = __p0_622; \
	56749	int32x4_t __s1_622 = __p1_622; \
	56750	int32x2_t __rev0_622; __rev0_622 = __builtin_shufflevector(__s0_622, __s0_622, __lane_reverse_64_32); \
	56751	int32x4_t __rev1_622; __rev1_622 = __builtin_shufflevector(__s1_622, __s1_622, __lane_reverse_128_32); \
	56752	__ret_622 = __noswap_vmull_s32(__rev0_622, __noswap_splat_laneq_s32(__rev1_622, __p2_622)); \
	56753	__ret_622 = __builtin_shufflevector(__ret_622, __ret_622, __lane_reverse_128_64); \
	56754	__ret_622; \
55097	56755	})
55098	56756	#endif
55099	56757
55100	56758	#ifdef __LITTLE_ENDIAN__
55101		#define vmull_laneq_s16(__p0_607, __p1_607, __p2_607) __extension__ ({ \
55102		int32x4_t __ret_607; \
55103		int16x4_t __s0_607 = __p0_607; \
55104		int16x8_t __s1_607 = __p1_607; \
55105		__ret_607 = vmull_s16(__s0_607, splat_laneq_s16(__s1_607, __p2_607)); \
55106		__ret_607; \
	56759	#define vmull_laneq_s16(__p0_623, __p1_623, __p2_623) __extension__ ({ \
	56760	int32x4_t __ret_623; \
	56761	int16x4_t __s0_623 = __p0_623; \
	56762	int16x8_t __s1_623 = __p1_623; \
	56763	__ret_623 = vmull_s16(__s0_623, splat_laneq_s16(__s1_623, __p2_623)); \
	56764	__ret_623; \
55107	56765	})
55108	56766	#else
55109		#define vmull_laneq_s16(__p0_608, __p1_608, __p2_608) __extension__ ({ \
55110		int32x4_t __ret_608; \
55111		int16x4_t __s0_608 = __p0_608; \
55112		int16x8_t __s1_608 = __p1_608; \
55113		int16x4_t __rev0_608; __rev0_608 = __builtin_shufflevector(__s0_608, __s0_608, 3, 2, 1, 0); \
55114		int16x8_t __rev1_608; __rev1_608 = __builtin_shufflevector(__s1_608, __s1_608, 7, 6, 5, 4, 3, 2, 1, 0); \
55115		__ret_608 = __noswap_vmull_s16(__rev0_608, __noswap_splat_laneq_s16(__rev1_608, __p2_608)); \
55116		__ret_608 = __builtin_shufflevector(__ret_608, __ret_608, 3, 2, 1, 0); \
55117		__ret_608; \
	56767	#define vmull_laneq_s16(__p0_624, __p1_624, __p2_624) __extension__ ({ \
	56768	int32x4_t __ret_624; \
	56769	int16x4_t __s0_624 = __p0_624; \
	56770	int16x8_t __s1_624 = __p1_624; \
	56771	int16x4_t __rev0_624; __rev0_624 = __builtin_shufflevector(__s0_624, __s0_624, __lane_reverse_64_16); \
	56772	int16x8_t __rev1_624; __rev1_624 = __builtin_shufflevector(__s1_624, __s1_624, __lane_reverse_128_16); \
	56773	__ret_624 = __noswap_vmull_s16(__rev0_624, __noswap_splat_laneq_s16(__rev1_624, __p2_624)); \
	56774	__ret_624 = __builtin_shufflevector(__ret_624, __ret_624, __lane_reverse_128_32); \
	56775	__ret_624; \
55118	56776	})
55119	56777	#endif
55120	56778
55121	56779	#ifdef __LITTLE_ENDIAN__
55122	56780	__ai __attribute__((target("neon"))) float64x2_t vmulxq_f64(float64x2_t __p0, float64x2_t __p1) {
55123	56781	float64x2_t __ret;
55124		__ret = (float64x2_t) __builtin_neon_vmulxq_v((int8x16_t)__p0, (int8x16_t)__p1, 42);
	56782	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vmulxq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 42));
55125	56783	return __ret;
55126	56784	}
55127	56785	#else
55128	56786	__ai __attribute__((target("neon"))) float64x2_t vmulxq_f64(float64x2_t __p0, float64x2_t __p1) {
55129	56787	float64x2_t __ret;
55130		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
55131		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
55132		__ret = (float64x2_t) __builtin_neon_vmulxq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 42);
55133		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	56788	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	56789	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	56790	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vmulxq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 42));
	56791	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
55134	56792	return __ret;
55135	56793	}
55136	56794	__ai __attribute__((target("neon"))) float64x2_t __noswap_vmulxq_f64(float64x2_t __p0, float64x2_t __p1) {
55137	56795	float64x2_t __ret;
55138		__ret = (float64x2_t) __builtin_neon_vmulxq_v((int8x16_t)__p0, (int8x16_t)__p1, 42);
	56796	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vmulxq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 42));
55139	56797	return __ret;
55140	56798	}
55141	56799	#endif
...	...	@@ -55143,248 +56801,248 @@ __ai __attribute__((target("neon"))) float64x2_t __noswap_vmulxq_f64(float64x2_t
55143	56801	#ifdef __LITTLE_ENDIAN__
55144	56802	__ai __attribute__((target("neon"))) float32x4_t vmulxq_f32(float32x4_t __p0, float32x4_t __p1) {
55145	56803	float32x4_t __ret;
55146		__ret = (float32x4_t) __builtin_neon_vmulxq_v((int8x16_t)__p0, (int8x16_t)__p1, 41);
	56804	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmulxq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 41));
55147	56805	return __ret;
55148	56806	}
55149	56807	#else
55150	56808	__ai __attribute__((target("neon"))) float32x4_t vmulxq_f32(float32x4_t __p0, float32x4_t __p1) {
55151	56809	float32x4_t __ret;
55152		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
55153		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
55154		__ret = (float32x4_t) __builtin_neon_vmulxq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 41);
55155		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	56810	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	56811	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	56812	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmulxq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 41));
	56813	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
55156	56814	return __ret;
55157	56815	}
55158	56816	__ai __attribute__((target("neon"))) float32x4_t __noswap_vmulxq_f32(float32x4_t __p0, float32x4_t __p1) {
55159	56817	float32x4_t __ret;
55160		__ret = (float32x4_t) __builtin_neon_vmulxq_v((int8x16_t)__p0, (int8x16_t)__p1, 41);
	56818	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vmulxq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 41));
55161	56819	return __ret;
55162	56820	}
55163	56821	#endif
55164	56822
55165	56823	__ai __attribute__((target("neon"))) float64x1_t vmulx_f64(float64x1_t __p0, float64x1_t __p1) {
55166	56824	float64x1_t __ret;
55167		__ret = (float64x1_t) __builtin_neon_vmulx_v((int8x8_t)__p0, (int8x8_t)__p1, 10);
	56825	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vmulx_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 10));
55168	56826	return __ret;
55169	56827	}
55170	56828	#ifdef __LITTLE_ENDIAN__
55171	56829	__ai __attribute__((target("neon"))) float32x2_t vmulx_f32(float32x2_t __p0, float32x2_t __p1) {
55172	56830	float32x2_t __ret;
55173		__ret = (float32x2_t) __builtin_neon_vmulx_v((int8x8_t)__p0, (int8x8_t)__p1, 9);
	56831	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vmulx_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 9));
55174	56832	return __ret;
55175	56833	}
55176	56834	#else
55177	56835	__ai __attribute__((target("neon"))) float32x2_t vmulx_f32(float32x2_t __p0, float32x2_t __p1) {
55178	56836	float32x2_t __ret;
55179		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
55180		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
55181		__ret = (float32x2_t) __builtin_neon_vmulx_v((int8x8_t)__rev0, (int8x8_t)__rev1, 9);
55182		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	56837	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	56838	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	56839	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vmulx_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 9));
	56840	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
55183	56841	return __ret;
55184	56842	}
55185	56843	__ai __attribute__((target("neon"))) float32x2_t __noswap_vmulx_f32(float32x2_t __p0, float32x2_t __p1) {
55186	56844	float32x2_t __ret;
55187		__ret = (float32x2_t) __builtin_neon_vmulx_v((int8x8_t)__p0, (int8x8_t)__p1, 9);
	56845	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vmulx_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 9));
55188	56846	return __ret;
55189	56847	}
55190	56848	#endif
55191	56849
55192	56850	__ai __attribute__((target("neon"))) float64_t vmulxd_f64(float64_t __p0, float64_t __p1) {
55193	56851	float64_t __ret;
55194		__ret = (float64_t) __builtin_neon_vmulxd_f64(__p0, __p1);
	56852	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vmulxd_f64(__p0, __p1));
55195	56853	return __ret;
55196	56854	}
55197	56855	__ai __attribute__((target("neon"))) float32_t vmulxs_f32(float32_t __p0, float32_t __p1) {
55198	56856	float32_t __ret;
55199		__ret = (float32_t) __builtin_neon_vmulxs_f32(__p0, __p1);
	56857	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vmulxs_f32(__p0, __p1));
55200	56858	return __ret;
55201	56859	}
55202		#define vmulxd_lane_f64(__p0_609, __p1_609, __p2_609) __extension__ ({ \
55203		float64_t __ret_609; \
55204		float64_t __s0_609 = __p0_609; \
55205		float64x1_t __s1_609 = __p1_609; \
55206		__ret_609 = vmulxd_f64(__s0_609, vget_lane_f64(__s1_609, __p2_609)); \
55207		__ret_609; \
	56860	#define vmulxd_lane_f64(__p0_625, __p1_625, __p2_625) __extension__ ({ \
	56861	float64_t __ret_625; \
	56862	float64_t __s0_625 = __p0_625; \
	56863	float64x1_t __s1_625 = __p1_625; \
	56864	__ret_625 = vmulxd_f64(__s0_625, vget_lane_f64(__s1_625, __p2_625)); \
	56865	__ret_625; \
55208	56866	})
55209	56867	#ifdef __LITTLE_ENDIAN__
55210		#define vmulxs_lane_f32(__p0_610, __p1_610, __p2_610) __extension__ ({ \
55211		float32_t __ret_610; \
55212		float32_t __s0_610 = __p0_610; \
55213		float32x2_t __s1_610 = __p1_610; \
55214		__ret_610 = vmulxs_f32(__s0_610, vget_lane_f32(__s1_610, __p2_610)); \
55215		__ret_610; \
	56868	#define vmulxs_lane_f32(__p0_626, __p1_626, __p2_626) __extension__ ({ \
	56869	float32_t __ret_626; \
	56870	float32_t __s0_626 = __p0_626; \
	56871	float32x2_t __s1_626 = __p1_626; \
	56872	__ret_626 = vmulxs_f32(__s0_626, vget_lane_f32(__s1_626, __p2_626)); \
	56873	__ret_626; \
55216	56874	})
55217	56875	#else
55218		#define vmulxs_lane_f32(__p0_611, __p1_611, __p2_611) __extension__ ({ \
55219		float32_t __ret_611; \
55220		float32_t __s0_611 = __p0_611; \
55221		float32x2_t __s1_611 = __p1_611; \
55222		float32x2_t __rev1_611; __rev1_611 = __builtin_shufflevector(__s1_611, __s1_611, 1, 0); \
55223		__ret_611 = vmulxs_f32(__s0_611, __noswap_vget_lane_f32(__rev1_611, __p2_611)); \
55224		__ret_611; \
	56876	#define vmulxs_lane_f32(__p0_627, __p1_627, __p2_627) __extension__ ({ \
	56877	float32_t __ret_627; \
	56878	float32_t __s0_627 = __p0_627; \
	56879	float32x2_t __s1_627 = __p1_627; \
	56880	float32x2_t __rev1_627; __rev1_627 = __builtin_shufflevector(__s1_627, __s1_627, __lane_reverse_64_32); \
	56881	__ret_627 = vmulxs_f32(__s0_627, __noswap_vget_lane_f32(__rev1_627, __p2_627)); \
	56882	__ret_627; \
55225	56883	})
55226	56884	#endif
55227	56885
55228	56886	#ifdef __LITTLE_ENDIAN__
55229		#define vmulxq_lane_f64(__p0_612, __p1_612, __p2_612) __extension__ ({ \
55230		float64x2_t __ret_612; \
55231		float64x2_t __s0_612 = __p0_612; \
55232		float64x1_t __s1_612 = __p1_612; \
55233		__ret_612 = vmulxq_f64(__s0_612, splatq_lane_f64(__s1_612, __p2_612)); \
55234		__ret_612; \
	56887	#define vmulxq_lane_f64(__p0_628, __p1_628, __p2_628) __extension__ ({ \
	56888	float64x2_t __ret_628; \
	56889	float64x2_t __s0_628 = __p0_628; \
	56890	float64x1_t __s1_628 = __p1_628; \
	56891	__ret_628 = vmulxq_f64(__s0_628, splatq_lane_f64(__s1_628, __p2_628)); \
	56892	__ret_628; \
55235	56893	})
55236	56894	#else
55237		#define vmulxq_lane_f64(__p0_613, __p1_613, __p2_613) __extension__ ({ \
55238		float64x2_t __ret_613; \
55239		float64x2_t __s0_613 = __p0_613; \
55240		float64x1_t __s1_613 = __p1_613; \
55241		float64x2_t __rev0_613; __rev0_613 = __builtin_shufflevector(__s0_613, __s0_613, 1, 0); \
55242		__ret_613 = __noswap_vmulxq_f64(__rev0_613, __noswap_splatq_lane_f64(__s1_613, __p2_613)); \
55243		__ret_613 = __builtin_shufflevector(__ret_613, __ret_613, 1, 0); \
55244		__ret_613; \
	56895	#define vmulxq_lane_f64(__p0_629, __p1_629, __p2_629) __extension__ ({ \
	56896	float64x2_t __ret_629; \
	56897	float64x2_t __s0_629 = __p0_629; \
	56898	float64x1_t __s1_629 = __p1_629; \
	56899	float64x2_t __rev0_629; __rev0_629 = __builtin_shufflevector(__s0_629, __s0_629, __lane_reverse_128_64); \
	56900	__ret_629 = __noswap_vmulxq_f64(__rev0_629, __noswap_splatq_lane_f64(__s1_629, __p2_629)); \
	56901	__ret_629 = __builtin_shufflevector(__ret_629, __ret_629, __lane_reverse_128_64); \
	56902	__ret_629; \
55245	56903	})
55246	56904	#endif
55247	56905
55248	56906	#ifdef __LITTLE_ENDIAN__
55249		#define vmulxq_lane_f32(__p0_614, __p1_614, __p2_614) __extension__ ({ \
55250		float32x4_t __ret_614; \
55251		float32x4_t __s0_614 = __p0_614; \
55252		float32x2_t __s1_614 = __p1_614; \
55253		__ret_614 = vmulxq_f32(__s0_614, splatq_lane_f32(__s1_614, __p2_614)); \
55254		__ret_614; \
	56907	#define vmulxq_lane_f32(__p0_630, __p1_630, __p2_630) __extension__ ({ \
	56908	float32x4_t __ret_630; \
	56909	float32x4_t __s0_630 = __p0_630; \
	56910	float32x2_t __s1_630 = __p1_630; \
	56911	__ret_630 = vmulxq_f32(__s0_630, splatq_lane_f32(__s1_630, __p2_630)); \
	56912	__ret_630; \
55255	56913	})
55256	56914	#else
55257		#define vmulxq_lane_f32(__p0_615, __p1_615, __p2_615) __extension__ ({ \
55258		float32x4_t __ret_615; \
55259		float32x4_t __s0_615 = __p0_615; \
55260		float32x2_t __s1_615 = __p1_615; \
55261		float32x4_t __rev0_615; __rev0_615 = __builtin_shufflevector(__s0_615, __s0_615, 3, 2, 1, 0); \
55262		float32x2_t __rev1_615; __rev1_615 = __builtin_shufflevector(__s1_615, __s1_615, 1, 0); \
55263		__ret_615 = __noswap_vmulxq_f32(__rev0_615, __noswap_splatq_lane_f32(__rev1_615, __p2_615)); \
55264		__ret_615 = __builtin_shufflevector(__ret_615, __ret_615, 3, 2, 1, 0); \
55265		__ret_615; \
	56915	#define vmulxq_lane_f32(__p0_631, __p1_631, __p2_631) __extension__ ({ \
	56916	float32x4_t __ret_631; \
	56917	float32x4_t __s0_631 = __p0_631; \
	56918	float32x2_t __s1_631 = __p1_631; \
	56919	float32x4_t __rev0_631; __rev0_631 = __builtin_shufflevector(__s0_631, __s0_631, __lane_reverse_128_32); \
	56920	float32x2_t __rev1_631; __rev1_631 = __builtin_shufflevector(__s1_631, __s1_631, __lane_reverse_64_32); \
	56921	__ret_631 = __noswap_vmulxq_f32(__rev0_631, __noswap_splatq_lane_f32(__rev1_631, __p2_631)); \
	56922	__ret_631 = __builtin_shufflevector(__ret_631, __ret_631, __lane_reverse_128_32); \
	56923	__ret_631; \
55266	56924	})
55267	56925	#endif
55268	56926
55269	56927	#ifdef __LITTLE_ENDIAN__
55270		#define vmulx_lane_f32(__p0_616, __p1_616, __p2_616) __extension__ ({ \
55271		float32x2_t __ret_616; \
55272		float32x2_t __s0_616 = __p0_616; \
55273		float32x2_t __s1_616 = __p1_616; \
55274		__ret_616 = vmulx_f32(__s0_616, splat_lane_f32(__s1_616, __p2_616)); \
55275		__ret_616; \
	56928	#define vmulx_lane_f32(__p0_632, __p1_632, __p2_632) __extension__ ({ \
	56929	float32x2_t __ret_632; \
	56930	float32x2_t __s0_632 = __p0_632; \
	56931	float32x2_t __s1_632 = __p1_632; \
	56932	__ret_632 = vmulx_f32(__s0_632, splat_lane_f32(__s1_632, __p2_632)); \
	56933	__ret_632; \
55276	56934	})
55277	56935	#else
55278		#define vmulx_lane_f32(__p0_617, __p1_617, __p2_617) __extension__ ({ \
55279		float32x2_t __ret_617; \
55280		float32x2_t __s0_617 = __p0_617; \
55281		float32x2_t __s1_617 = __p1_617; \
55282		float32x2_t __rev0_617; __rev0_617 = __builtin_shufflevector(__s0_617, __s0_617, 1, 0); \
55283		float32x2_t __rev1_617; __rev1_617 = __builtin_shufflevector(__s1_617, __s1_617, 1, 0); \
55284		__ret_617 = __noswap_vmulx_f32(__rev0_617, __noswap_splat_lane_f32(__rev1_617, __p2_617)); \
55285		__ret_617 = __builtin_shufflevector(__ret_617, __ret_617, 1, 0); \
55286		__ret_617; \
	56936	#define vmulx_lane_f32(__p0_633, __p1_633, __p2_633) __extension__ ({ \
	56937	float32x2_t __ret_633; \
	56938	float32x2_t __s0_633 = __p0_633; \
	56939	float32x2_t __s1_633 = __p1_633; \
	56940	float32x2_t __rev0_633; __rev0_633 = __builtin_shufflevector(__s0_633, __s0_633, __lane_reverse_64_32); \
	56941	float32x2_t __rev1_633; __rev1_633 = __builtin_shufflevector(__s1_633, __s1_633, __lane_reverse_64_32); \
	56942	__ret_633 = __noswap_vmulx_f32(__rev0_633, __noswap_splat_lane_f32(__rev1_633, __p2_633)); \
	56943	__ret_633 = __builtin_shufflevector(__ret_633, __ret_633, __lane_reverse_64_32); \
	56944	__ret_633; \
55287	56945	})
55288	56946	#endif
55289	56947
55290	56948	#ifdef __LITTLE_ENDIAN__
55291		#define vmulxd_laneq_f64(__p0_618, __p1_618, __p2_618) __extension__ ({ \
55292		float64_t __ret_618; \
55293		float64_t __s0_618 = __p0_618; \
55294		float64x2_t __s1_618 = __p1_618; \
55295		__ret_618 = vmulxd_f64(__s0_618, vgetq_lane_f64(__s1_618, __p2_618)); \
55296		__ret_618; \
	56949	#define vmulxd_laneq_f64(__p0_634, __p1_634, __p2_634) __extension__ ({ \
	56950	float64_t __ret_634; \
	56951	float64_t __s0_634 = __p0_634; \
	56952	float64x2_t __s1_634 = __p1_634; \
	56953	__ret_634 = vmulxd_f64(__s0_634, vgetq_lane_f64(__s1_634, __p2_634)); \
	56954	__ret_634; \
55297	56955	})
55298	56956	#else
55299		#define vmulxd_laneq_f64(__p0_619, __p1_619, __p2_619) __extension__ ({ \
55300		float64_t __ret_619; \
55301		float64_t __s0_619 = __p0_619; \
55302		float64x2_t __s1_619 = __p1_619; \
55303		float64x2_t __rev1_619; __rev1_619 = __builtin_shufflevector(__s1_619, __s1_619, 1, 0); \
55304		__ret_619 = vmulxd_f64(__s0_619, __noswap_vgetq_lane_f64(__rev1_619, __p2_619)); \
55305		__ret_619; \
	56957	#define vmulxd_laneq_f64(__p0_635, __p1_635, __p2_635) __extension__ ({ \
	56958	float64_t __ret_635; \
	56959	float64_t __s0_635 = __p0_635; \
	56960	float64x2_t __s1_635 = __p1_635; \
	56961	float64x2_t __rev1_635; __rev1_635 = __builtin_shufflevector(__s1_635, __s1_635, __lane_reverse_128_64); \
	56962	__ret_635 = vmulxd_f64(__s0_635, __noswap_vgetq_lane_f64(__rev1_635, __p2_635)); \
	56963	__ret_635; \
55306	56964	})
55307	56965	#endif
55308	56966
55309	56967	#ifdef __LITTLE_ENDIAN__
55310		#define vmulxs_laneq_f32(__p0_620, __p1_620, __p2_620) __extension__ ({ \
55311		float32_t __ret_620; \
55312		float32_t __s0_620 = __p0_620; \
55313		float32x4_t __s1_620 = __p1_620; \
55314		__ret_620 = vmulxs_f32(__s0_620, vgetq_lane_f32(__s1_620, __p2_620)); \
55315		__ret_620; \
	56968	#define vmulxs_laneq_f32(__p0_636, __p1_636, __p2_636) __extension__ ({ \
	56969	float32_t __ret_636; \
	56970	float32_t __s0_636 = __p0_636; \
	56971	float32x4_t __s1_636 = __p1_636; \
	56972	__ret_636 = vmulxs_f32(__s0_636, vgetq_lane_f32(__s1_636, __p2_636)); \
	56973	__ret_636; \
55316	56974	})
55317	56975	#else
55318		#define vmulxs_laneq_f32(__p0_621, __p1_621, __p2_621) __extension__ ({ \
55319		float32_t __ret_621; \
55320		float32_t __s0_621 = __p0_621; \
55321		float32x4_t __s1_621 = __p1_621; \
55322		float32x4_t __rev1_621; __rev1_621 = __builtin_shufflevector(__s1_621, __s1_621, 3, 2, 1, 0); \
55323		__ret_621 = vmulxs_f32(__s0_621, __noswap_vgetq_lane_f32(__rev1_621, __p2_621)); \
55324		__ret_621; \
	56976	#define vmulxs_laneq_f32(__p0_637, __p1_637, __p2_637) __extension__ ({ \
	56977	float32_t __ret_637; \
	56978	float32_t __s0_637 = __p0_637; \
	56979	float32x4_t __s1_637 = __p1_637; \
	56980	float32x4_t __rev1_637; __rev1_637 = __builtin_shufflevector(__s1_637, __s1_637, __lane_reverse_128_32); \
	56981	__ret_637 = vmulxs_f32(__s0_637, __noswap_vgetq_lane_f32(__rev1_637, __p2_637)); \
	56982	__ret_637; \
55325	56983	})
55326	56984	#endif
55327	56985
55328	56986	#ifdef __LITTLE_ENDIAN__
55329		#define vmulxq_laneq_f64(__p0_622, __p1_622, __p2_622) __extension__ ({ \
55330		float64x2_t __ret_622; \
55331		float64x2_t __s0_622 = __p0_622; \
55332		float64x2_t __s1_622 = __p1_622; \
55333		__ret_622 = vmulxq_f64(__s0_622, splatq_laneq_f64(__s1_622, __p2_622)); \
55334		__ret_622; \
	56987	#define vmulxq_laneq_f64(__p0_638, __p1_638, __p2_638) __extension__ ({ \
	56988	float64x2_t __ret_638; \
	56989	float64x2_t __s0_638 = __p0_638; \
	56990	float64x2_t __s1_638 = __p1_638; \
	56991	__ret_638 = vmulxq_f64(__s0_638, splatq_laneq_f64(__s1_638, __p2_638)); \
	56992	__ret_638; \
55335	56993	})
55336	56994	#else
55337		#define vmulxq_laneq_f64(__p0_623, __p1_623, __p2_623) __extension__ ({ \
55338		float64x2_t __ret_623; \
55339		float64x2_t __s0_623 = __p0_623; \
55340		float64x2_t __s1_623 = __p1_623; \
55341		float64x2_t __rev0_623; __rev0_623 = __builtin_shufflevector(__s0_623, __s0_623, 1, 0); \
55342		float64x2_t __rev1_623; __rev1_623 = __builtin_shufflevector(__s1_623, __s1_623, 1, 0); \
55343		__ret_623 = __noswap_vmulxq_f64(__rev0_623, __noswap_splatq_laneq_f64(__rev1_623, __p2_623)); \
55344		__ret_623 = __builtin_shufflevector(__ret_623, __ret_623, 1, 0); \
55345		__ret_623; \
	56995	#define vmulxq_laneq_f64(__p0_639, __p1_639, __p2_639) __extension__ ({ \
	56996	float64x2_t __ret_639; \
	56997	float64x2_t __s0_639 = __p0_639; \
	56998	float64x2_t __s1_639 = __p1_639; \
	56999	float64x2_t __rev0_639; __rev0_639 = __builtin_shufflevector(__s0_639, __s0_639, __lane_reverse_128_64); \
	57000	float64x2_t __rev1_639; __rev1_639 = __builtin_shufflevector(__s1_639, __s1_639, __lane_reverse_128_64); \
	57001	__ret_639 = __noswap_vmulxq_f64(__rev0_639, __noswap_splatq_laneq_f64(__rev1_639, __p2_639)); \
	57002	__ret_639 = __builtin_shufflevector(__ret_639, __ret_639, __lane_reverse_128_64); \
	57003	__ret_639; \
55346	57004	})
55347	57005	#endif
55348	57006
55349	57007	#ifdef __LITTLE_ENDIAN__
55350		#define vmulxq_laneq_f32(__p0_624, __p1_624, __p2_624) __extension__ ({ \
55351		float32x4_t __ret_624; \
55352		float32x4_t __s0_624 = __p0_624; \
55353		float32x4_t __s1_624 = __p1_624; \
55354		__ret_624 = vmulxq_f32(__s0_624, splatq_laneq_f32(__s1_624, __p2_624)); \
55355		__ret_624; \
	57008	#define vmulxq_laneq_f32(__p0_640, __p1_640, __p2_640) __extension__ ({ \
	57009	float32x4_t __ret_640; \
	57010	float32x4_t __s0_640 = __p0_640; \
	57011	float32x4_t __s1_640 = __p1_640; \
	57012	__ret_640 = vmulxq_f32(__s0_640, splatq_laneq_f32(__s1_640, __p2_640)); \
	57013	__ret_640; \
55356	57014	})
55357	57015	#else
55358		#define vmulxq_laneq_f32(__p0_625, __p1_625, __p2_625) __extension__ ({ \
55359		float32x4_t __ret_625; \
55360		float32x4_t __s0_625 = __p0_625; \
55361		float32x4_t __s1_625 = __p1_625; \
55362		float32x4_t __rev0_625; __rev0_625 = __builtin_shufflevector(__s0_625, __s0_625, 3, 2, 1, 0); \
55363		float32x4_t __rev1_625; __rev1_625 = __builtin_shufflevector(__s1_625, __s1_625, 3, 2, 1, 0); \
55364		__ret_625 = __noswap_vmulxq_f32(__rev0_625, __noswap_splatq_laneq_f32(__rev1_625, __p2_625)); \
55365		__ret_625 = __builtin_shufflevector(__ret_625, __ret_625, 3, 2, 1, 0); \
55366		__ret_625; \
	57016	#define vmulxq_laneq_f32(__p0_641, __p1_641, __p2_641) __extension__ ({ \
	57017	float32x4_t __ret_641; \
	57018	float32x4_t __s0_641 = __p0_641; \
	57019	float32x4_t __s1_641 = __p1_641; \
	57020	float32x4_t __rev0_641; __rev0_641 = __builtin_shufflevector(__s0_641, __s0_641, __lane_reverse_128_32); \
	57021	float32x4_t __rev1_641; __rev1_641 = __builtin_shufflevector(__s1_641, __s1_641, __lane_reverse_128_32); \
	57022	__ret_641 = __noswap_vmulxq_f32(__rev0_641, __noswap_splatq_laneq_f32(__rev1_641, __p2_641)); \
	57023	__ret_641 = __builtin_shufflevector(__ret_641, __ret_641, __lane_reverse_128_32); \
	57024	__ret_641; \
55367	57025	})
55368	57026	#endif
55369	57027
55370	57028	#ifdef __LITTLE_ENDIAN__
55371		#define vmulx_laneq_f32(__p0_626, __p1_626, __p2_626) __extension__ ({ \
55372		float32x2_t __ret_626; \
55373		float32x2_t __s0_626 = __p0_626; \
55374		float32x4_t __s1_626 = __p1_626; \
55375		__ret_626 = vmulx_f32(__s0_626, splat_laneq_f32(__s1_626, __p2_626)); \
55376		__ret_626; \
	57029	#define vmulx_laneq_f32(__p0_642, __p1_642, __p2_642) __extension__ ({ \
	57030	float32x2_t __ret_642; \
	57031	float32x2_t __s0_642 = __p0_642; \
	57032	float32x4_t __s1_642 = __p1_642; \
	57033	__ret_642 = vmulx_f32(__s0_642, splat_laneq_f32(__s1_642, __p2_642)); \
	57034	__ret_642; \
55377	57035	})
55378	57036	#else
55379		#define vmulx_laneq_f32(__p0_627, __p1_627, __p2_627) __extension__ ({ \
55380		float32x2_t __ret_627; \
55381		float32x2_t __s0_627 = __p0_627; \
55382		float32x4_t __s1_627 = __p1_627; \
55383		float32x2_t __rev0_627; __rev0_627 = __builtin_shufflevector(__s0_627, __s0_627, 1, 0); \
55384		float32x4_t __rev1_627; __rev1_627 = __builtin_shufflevector(__s1_627, __s1_627, 3, 2, 1, 0); \
55385		__ret_627 = __noswap_vmulx_f32(__rev0_627, __noswap_splat_laneq_f32(__rev1_627, __p2_627)); \
55386		__ret_627 = __builtin_shufflevector(__ret_627, __ret_627, 1, 0); \
55387		__ret_627; \
	57037	#define vmulx_laneq_f32(__p0_643, __p1_643, __p2_643) __extension__ ({ \
	57038	float32x2_t __ret_643; \
	57039	float32x2_t __s0_643 = __p0_643; \
	57040	float32x4_t __s1_643 = __p1_643; \
	57041	float32x2_t __rev0_643; __rev0_643 = __builtin_shufflevector(__s0_643, __s0_643, __lane_reverse_64_32); \
	57042	float32x4_t __rev1_643; __rev1_643 = __builtin_shufflevector(__s1_643, __s1_643, __lane_reverse_128_32); \
	57043	__ret_643 = __noswap_vmulx_f32(__rev0_643, __noswap_splat_laneq_f32(__rev1_643, __p2_643)); \
	57044	__ret_643 = __builtin_shufflevector(__ret_643, __ret_643, __lane_reverse_64_32); \
	57045	__ret_643; \
55388	57046	})
55389	57047	#endif
55390	57048
...	...	@@ -55397,9 +57055,9 @@ __ai __attribute__((target("neon"))) float64x2_t vnegq_f64(float64x2_t __p0) {
55397	57055	#else
55398	57056	__ai __attribute__((target("neon"))) float64x2_t vnegq_f64(float64x2_t __p0) {
55399	57057	float64x2_t __ret;
55400		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
	57058	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
55401	57059	__ret = -__rev0;
55402		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	57060	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
55403	57061	return __ret;
55404	57062	}
55405	57063	#endif
...	...	@@ -55413,9 +57071,9 @@ __ai __attribute__((target("neon"))) int64x2_t vnegq_s64(int64x2_t __p0) {
55413	57071	#else
55414	57072	__ai __attribute__((target("neon"))) int64x2_t vnegq_s64(int64x2_t __p0) {
55415	57073	int64x2_t __ret;
55416		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
	57074	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
55417	57075	__ret = -__rev0;
55418		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	57076	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
55419	57077	return __ret;
55420	57078	}
55421	57079	#endif
...	...	@@ -55432,22 +57090,22 @@ __ai __attribute__((target("neon"))) int64x1_t vneg_s64(int64x1_t __p0) {
55432	57090	}
55433	57091	__ai __attribute__((target("neon"))) int64_t vnegd_s64(int64_t __p0) {
55434	57092	int64_t __ret;
55435		__ret = (int64_t) __builtin_neon_vnegd_s64(__p0);
	57093	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vnegd_s64(__p0));
55436	57094	return __ret;
55437	57095	}
55438	57096	#ifdef __LITTLE_ENDIAN__
55439	57097	__ai __attribute__((target("neon"))) uint8x16_t vpaddq_u8(uint8x16_t __p0, uint8x16_t __p1) {
55440	57098	uint8x16_t __ret;
55441		__ret = (uint8x16_t) __builtin_neon_vpaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
	57099	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vpaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 48));
55442	57100	return __ret;
55443	57101	}
55444	57102	#else
55445	57103	__ai __attribute__((target("neon"))) uint8x16_t vpaddq_u8(uint8x16_t __p0, uint8x16_t __p1) {
55446	57104	uint8x16_t __ret;
55447		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
55448		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
55449		__ret = (uint8x16_t) __builtin_neon_vpaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
55450		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	57105	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	57106	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	57107	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vpaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 48));
	57108	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
55451	57109	return __ret;
55452	57110	}
55453	57111	#endif
...	...	@@ -55455,16 +57113,16 @@ __ai __attribute__((target("neon"))) uint8x16_t vpaddq_u8(uint8x16_t __p0, uint8
55455	57113	#ifdef __LITTLE_ENDIAN__
55456	57114	__ai __attribute__((target("neon"))) uint32x4_t vpaddq_u32(uint32x4_t __p0, uint32x4_t __p1) {
55457	57115	uint32x4_t __ret;
55458		__ret = (uint32x4_t) __builtin_neon_vpaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
	57116	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vpaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50));
55459	57117	return __ret;
55460	57118	}
55461	57119	#else
55462	57120	__ai __attribute__((target("neon"))) uint32x4_t vpaddq_u32(uint32x4_t __p0, uint32x4_t __p1) {
55463	57121	uint32x4_t __ret;
55464		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
55465		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
55466		__ret = (uint32x4_t) __builtin_neon_vpaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
55467		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	57122	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	57123	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	57124	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vpaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 50));
	57125	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
55468	57126	return __ret;
55469	57127	}
55470	57128	#endif
...	...	@@ -55472,16 +57130,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vpaddq_u32(uint32x4_t __p0, uint
55472	57130	#ifdef __LITTLE_ENDIAN__
55473	57131	__ai __attribute__((target("neon"))) uint64x2_t vpaddq_u64(uint64x2_t __p0, uint64x2_t __p1) {
55474	57132	uint64x2_t __ret;
55475		__ret = (uint64x2_t) __builtin_neon_vpaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 51);
	57133	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vpaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 51));
55476	57134	return __ret;
55477	57135	}
55478	57136	#else
55479	57137	__ai __attribute__((target("neon"))) uint64x2_t vpaddq_u64(uint64x2_t __p0, uint64x2_t __p1) {
55480	57138	uint64x2_t __ret;
55481		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
55482		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
55483		__ret = (uint64x2_t) __builtin_neon_vpaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 51);
55484		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	57139	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	57140	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	57141	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vpaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 51));
	57142	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
55485	57143	return __ret;
55486	57144	}
55487	57145	#endif
...	...	@@ -55489,16 +57147,16 @@ __ai __attribute__((target("neon"))) uint64x2_t vpaddq_u64(uint64x2_t __p0, uint
55489	57147	#ifdef __LITTLE_ENDIAN__
55490	57148	__ai __attribute__((target("neon"))) uint16x8_t vpaddq_u16(uint16x8_t __p0, uint16x8_t __p1) {
55491	57149	uint16x8_t __ret;
55492		__ret = (uint16x8_t) __builtin_neon_vpaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
	57150	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vpaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 49));
55493	57151	return __ret;
55494	57152	}
55495	57153	#else
55496	57154	__ai __attribute__((target("neon"))) uint16x8_t vpaddq_u16(uint16x8_t __p0, uint16x8_t __p1) {
55497	57155	uint16x8_t __ret;
55498		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
55499		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
55500		__ret = (uint16x8_t) __builtin_neon_vpaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
55501		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	57156	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	57157	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	57158	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vpaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 49));
	57159	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
55502	57160	return __ret;
55503	57161	}
55504	57162	#endif
...	...	@@ -55506,16 +57164,16 @@ __ai __attribute__((target("neon"))) uint16x8_t vpaddq_u16(uint16x8_t __p0, uint
55506	57164	#ifdef __LITTLE_ENDIAN__
55507	57165	__ai __attribute__((target("neon"))) int8x16_t vpaddq_s8(int8x16_t __p0, int8x16_t __p1) {
55508	57166	int8x16_t __ret;
55509		__ret = (int8x16_t) __builtin_neon_vpaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
	57167	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vpaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 32));
55510	57168	return __ret;
55511	57169	}
55512	57170	#else
55513	57171	__ai __attribute__((target("neon"))) int8x16_t vpaddq_s8(int8x16_t __p0, int8x16_t __p1) {
55514	57172	int8x16_t __ret;
55515		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
55516		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
55517		__ret = (int8x16_t) __builtin_neon_vpaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 32);
55518		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	57173	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	57174	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	57175	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vpaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 32));
	57176	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
55519	57177	return __ret;
55520	57178	}
55521	57179	#endif
...	...	@@ -55523,16 +57181,16 @@ __ai __attribute__((target("neon"))) int8x16_t vpaddq_s8(int8x16_t __p0, int8x16
55523	57181	#ifdef __LITTLE_ENDIAN__
55524	57182	__ai __attribute__((target("neon"))) float64x2_t vpaddq_f64(float64x2_t __p0, float64x2_t __p1) {
55525	57183	float64x2_t __ret;
55526		__ret = (float64x2_t) __builtin_neon_vpaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 42);
	57184	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vpaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 42));
55527	57185	return __ret;
55528	57186	}
55529	57187	#else
55530	57188	__ai __attribute__((target("neon"))) float64x2_t vpaddq_f64(float64x2_t __p0, float64x2_t __p1) {
55531	57189	float64x2_t __ret;
55532		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
55533		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
55534		__ret = (float64x2_t) __builtin_neon_vpaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 42);
55535		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	57190	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	57191	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	57192	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vpaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 42));
	57193	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
55536	57194	return __ret;
55537	57195	}
55538	57196	#endif
...	...	@@ -55540,16 +57198,16 @@ __ai __attribute__((target("neon"))) float64x2_t vpaddq_f64(float64x2_t __p0, fl
55540	57198	#ifdef __LITTLE_ENDIAN__
55541	57199	__ai __attribute__((target("neon"))) float32x4_t vpaddq_f32(float32x4_t __p0, float32x4_t __p1) {
55542	57200	float32x4_t __ret;
55543		__ret = (float32x4_t) __builtin_neon_vpaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 41);
	57201	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vpaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 41));
55544	57202	return __ret;
55545	57203	}
55546	57204	#else
55547	57205	__ai __attribute__((target("neon"))) float32x4_t vpaddq_f32(float32x4_t __p0, float32x4_t __p1) {
55548	57206	float32x4_t __ret;
55549		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
55550		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
55551		__ret = (float32x4_t) __builtin_neon_vpaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 41);
55552		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	57207	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	57208	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	57209	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vpaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 41));
	57210	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
55553	57211	return __ret;
55554	57212	}
55555	57213	#endif
...	...	@@ -55557,16 +57215,16 @@ __ai __attribute__((target("neon"))) float32x4_t vpaddq_f32(float32x4_t __p0, fl
55557	57215	#ifdef __LITTLE_ENDIAN__
55558	57216	__ai __attribute__((target("neon"))) int32x4_t vpaddq_s32(int32x4_t __p0, int32x4_t __p1) {
55559	57217	int32x4_t __ret;
55560		__ret = (int32x4_t) __builtin_neon_vpaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
	57218	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vpaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 34));
55561	57219	return __ret;
55562	57220	}
55563	57221	#else
55564	57222	__ai __attribute__((target("neon"))) int32x4_t vpaddq_s32(int32x4_t __p0, int32x4_t __p1) {
55565	57223	int32x4_t __ret;
55566		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
55567		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
55568		__ret = (int32x4_t) __builtin_neon_vpaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
55569		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	57224	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	57225	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	57226	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vpaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 34));
	57227	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
55570	57228	return __ret;
55571	57229	}
55572	57230	#endif
...	...	@@ -55574,16 +57232,16 @@ __ai __attribute__((target("neon"))) int32x4_t vpaddq_s32(int32x4_t __p0, int32x
55574	57232	#ifdef __LITTLE_ENDIAN__
55575	57233	__ai __attribute__((target("neon"))) int64x2_t vpaddq_s64(int64x2_t __p0, int64x2_t __p1) {
55576	57234	int64x2_t __ret;
55577		__ret = (int64x2_t) __builtin_neon_vpaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 35);
	57235	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vpaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 35));
55578	57236	return __ret;
55579	57237	}
55580	57238	#else
55581	57239	__ai __attribute__((target("neon"))) int64x2_t vpaddq_s64(int64x2_t __p0, int64x2_t __p1) {
55582	57240	int64x2_t __ret;
55583		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
55584		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
55585		__ret = (int64x2_t) __builtin_neon_vpaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 35);
55586		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	57241	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	57242	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	57243	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vpaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 35));
	57244	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
55587	57245	return __ret;
55588	57246	}
55589	57247	#endif
...	...	@@ -55591,16 +57249,16 @@ __ai __attribute__((target("neon"))) int64x2_t vpaddq_s64(int64x2_t __p0, int64x
55591	57249	#ifdef __LITTLE_ENDIAN__
55592	57250	__ai __attribute__((target("neon"))) int16x8_t vpaddq_s16(int16x8_t __p0, int16x8_t __p1) {
55593	57251	int16x8_t __ret;
55594		__ret = (int16x8_t) __builtin_neon_vpaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
	57252	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vpaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 33));
55595	57253	return __ret;
55596	57254	}
55597	57255	#else
55598	57256	__ai __attribute__((target("neon"))) int16x8_t vpaddq_s16(int16x8_t __p0, int16x8_t __p1) {
55599	57257	int16x8_t __ret;
55600		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
55601		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
55602		__ret = (int16x8_t) __builtin_neon_vpaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
55603		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	57258	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	57259	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	57260	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vpaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 33));
	57261	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
55604	57262	return __ret;
55605	57263	}
55606	57264	#endif
...	...	@@ -55608,14 +57266,14 @@ __ai __attribute__((target("neon"))) int16x8_t vpaddq_s16(int16x8_t __p0, int16x
55608	57266	#ifdef __LITTLE_ENDIAN__
55609	57267	__ai __attribute__((target("neon"))) uint64_t vpaddd_u64(uint64x2_t __p0) {
55610	57268	uint64_t __ret;
55611		__ret = (uint64_t) __builtin_neon_vpaddd_u64(__p0);
	57269	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vpaddd_u64(__p0));
55612	57270	return __ret;
55613	57271	}
55614	57272	#else
55615	57273	__ai __attribute__((target("neon"))) uint64_t vpaddd_u64(uint64x2_t __p0) {
55616	57274	uint64_t __ret;
55617		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
55618		__ret = (uint64_t) __builtin_neon_vpaddd_u64(__rev0);
	57275	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	57276	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vpaddd_u64(__rev0));
55619	57277	return __ret;
55620	57278	}
55621	57279	#endif
...	...	@@ -55623,14 +57281,14 @@ __ai __attribute__((target("neon"))) uint64_t vpaddd_u64(uint64x2_t __p0) {
55623	57281	#ifdef __LITTLE_ENDIAN__
55624	57282	__ai __attribute__((target("neon"))) float64_t vpaddd_f64(float64x2_t __p0) {
55625	57283	float64_t __ret;
55626		__ret = (float64_t) __builtin_neon_vpaddd_f64(__p0);
	57284	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vpaddd_f64(__p0));
55627	57285	return __ret;
55628	57286	}
55629	57287	#else
55630	57288	__ai __attribute__((target("neon"))) float64_t vpaddd_f64(float64x2_t __p0) {
55631	57289	float64_t __ret;
55632		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
55633		__ret = (float64_t) __builtin_neon_vpaddd_f64(__rev0);
	57290	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	57291	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vpaddd_f64(__rev0));
55634	57292	return __ret;
55635	57293	}
55636	57294	#endif
...	...	@@ -55638,14 +57296,14 @@ __ai __attribute__((target("neon"))) float64_t vpaddd_f64(float64x2_t __p0) {
55638	57296	#ifdef __LITTLE_ENDIAN__
55639	57297	__ai __attribute__((target("neon"))) int64_t vpaddd_s64(int64x2_t __p0) {
55640	57298	int64_t __ret;
55641		__ret = (int64_t) __builtin_neon_vpaddd_s64(__p0);
	57299	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vpaddd_s64(__p0));
55642	57300	return __ret;
55643	57301	}
55644	57302	#else
55645	57303	__ai __attribute__((target("neon"))) int64_t vpaddd_s64(int64x2_t __p0) {
55646	57304	int64_t __ret;
55647		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
55648		__ret = (int64_t) __builtin_neon_vpaddd_s64(__rev0);
	57305	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	57306	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vpaddd_s64(__rev0));
55649	57307	return __ret;
55650	57308	}
55651	57309	#endif
...	...	@@ -55653,14 +57311,14 @@ __ai __attribute__((target("neon"))) int64_t vpaddd_s64(int64x2_t __p0) {
55653	57311	#ifdef __LITTLE_ENDIAN__
55654	57312	__ai __attribute__((target("neon"))) float32_t vpadds_f32(float32x2_t __p0) {
55655	57313	float32_t __ret;
55656		__ret = (float32_t) __builtin_neon_vpadds_f32(__p0);
	57314	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vpadds_f32(__p0));
55657	57315	return __ret;
55658	57316	}
55659	57317	#else
55660	57318	__ai __attribute__((target("neon"))) float32_t vpadds_f32(float32x2_t __p0) {
55661	57319	float32_t __ret;
55662		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
55663		__ret = (float32_t) __builtin_neon_vpadds_f32(__rev0);
	57320	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	57321	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vpadds_f32(__rev0));
55664	57322	return __ret;
55665	57323	}
55666	57324	#endif
...	...	@@ -55668,16 +57326,16 @@ __ai __attribute__((target("neon"))) float32_t vpadds_f32(float32x2_t __p0) {
55668	57326	#ifdef __LITTLE_ENDIAN__
55669	57327	__ai __attribute__((target("neon"))) uint8x16_t vpmaxq_u8(uint8x16_t __p0, uint8x16_t __p1) {
55670	57328	uint8x16_t __ret;
55671		__ret = (uint8x16_t) __builtin_neon_vpmaxq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
	57329	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vpmaxq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 48));
55672	57330	return __ret;
55673	57331	}
55674	57332	#else
55675	57333	__ai __attribute__((target("neon"))) uint8x16_t vpmaxq_u8(uint8x16_t __p0, uint8x16_t __p1) {
55676	57334	uint8x16_t __ret;
55677		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
55678		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
55679		__ret = (uint8x16_t) __builtin_neon_vpmaxq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
55680		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	57335	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	57336	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	57337	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vpmaxq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 48));
	57338	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
55681	57339	return __ret;
55682	57340	}
55683	57341	#endif
...	...	@@ -55685,16 +57343,16 @@ __ai __attribute__((target("neon"))) uint8x16_t vpmaxq_u8(uint8x16_t __p0, uint8
55685	57343	#ifdef __LITTLE_ENDIAN__
55686	57344	__ai __attribute__((target("neon"))) uint32x4_t vpmaxq_u32(uint32x4_t __p0, uint32x4_t __p1) {
55687	57345	uint32x4_t __ret;
55688		__ret = (uint32x4_t) __builtin_neon_vpmaxq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
	57346	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vpmaxq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50));
55689	57347	return __ret;
55690	57348	}
55691	57349	#else
55692	57350	__ai __attribute__((target("neon"))) uint32x4_t vpmaxq_u32(uint32x4_t __p0, uint32x4_t __p1) {
55693	57351	uint32x4_t __ret;
55694		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
55695		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
55696		__ret = (uint32x4_t) __builtin_neon_vpmaxq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
55697		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	57352	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	57353	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	57354	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vpmaxq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 50));
	57355	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
55698	57356	return __ret;
55699	57357	}
55700	57358	#endif
...	...	@@ -55702,16 +57360,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vpmaxq_u32(uint32x4_t __p0, uint
55702	57360	#ifdef __LITTLE_ENDIAN__
55703	57361	__ai __attribute__((target("neon"))) uint16x8_t vpmaxq_u16(uint16x8_t __p0, uint16x8_t __p1) {
55704	57362	uint16x8_t __ret;
55705		__ret = (uint16x8_t) __builtin_neon_vpmaxq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
	57363	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vpmaxq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 49));
55706	57364	return __ret;
55707	57365	}
55708	57366	#else
55709	57367	__ai __attribute__((target("neon"))) uint16x8_t vpmaxq_u16(uint16x8_t __p0, uint16x8_t __p1) {
55710	57368	uint16x8_t __ret;
55711		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
55712		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
55713		__ret = (uint16x8_t) __builtin_neon_vpmaxq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
55714		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	57369	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	57370	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	57371	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vpmaxq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 49));
	57372	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
55715	57373	return __ret;
55716	57374	}
55717	57375	#endif
...	...	@@ -55719,16 +57377,16 @@ __ai __attribute__((target("neon"))) uint16x8_t vpmaxq_u16(uint16x8_t __p0, uint
55719	57377	#ifdef __LITTLE_ENDIAN__
55720	57378	__ai __attribute__((target("neon"))) int8x16_t vpmaxq_s8(int8x16_t __p0, int8x16_t __p1) {
55721	57379	int8x16_t __ret;
55722		__ret = (int8x16_t) __builtin_neon_vpmaxq_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
	57380	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vpmaxq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 32));
55723	57381	return __ret;
55724	57382	}
55725	57383	#else
55726	57384	__ai __attribute__((target("neon"))) int8x16_t vpmaxq_s8(int8x16_t __p0, int8x16_t __p1) {
55727	57385	int8x16_t __ret;
55728		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
55729		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
55730		__ret = (int8x16_t) __builtin_neon_vpmaxq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 32);
55731		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	57386	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	57387	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	57388	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vpmaxq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 32));
	57389	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
55732	57390	return __ret;
55733	57391	}
55734	57392	#endif
...	...	@@ -55736,16 +57394,16 @@ __ai __attribute__((target("neon"))) int8x16_t vpmaxq_s8(int8x16_t __p0, int8x16
55736	57394	#ifdef __LITTLE_ENDIAN__
55737	57395	__ai __attribute__((target("neon"))) float64x2_t vpmaxq_f64(float64x2_t __p0, float64x2_t __p1) {
55738	57396	float64x2_t __ret;
55739		__ret = (float64x2_t) __builtin_neon_vpmaxq_v((int8x16_t)__p0, (int8x16_t)__p1, 42);
	57397	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vpmaxq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 42));
55740	57398	return __ret;
55741	57399	}
55742	57400	#else
55743	57401	__ai __attribute__((target("neon"))) float64x2_t vpmaxq_f64(float64x2_t __p0, float64x2_t __p1) {
55744	57402	float64x2_t __ret;
55745		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
55746		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
55747		__ret = (float64x2_t) __builtin_neon_vpmaxq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 42);
55748		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	57403	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	57404	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	57405	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vpmaxq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 42));
	57406	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
55749	57407	return __ret;
55750	57408	}
55751	57409	#endif
...	...	@@ -55753,16 +57411,16 @@ __ai __attribute__((target("neon"))) float64x2_t vpmaxq_f64(float64x2_t __p0, fl
55753	57411	#ifdef __LITTLE_ENDIAN__
55754	57412	__ai __attribute__((target("neon"))) float32x4_t vpmaxq_f32(float32x4_t __p0, float32x4_t __p1) {
55755	57413	float32x4_t __ret;
55756		__ret = (float32x4_t) __builtin_neon_vpmaxq_v((int8x16_t)__p0, (int8x16_t)__p1, 41);
	57414	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vpmaxq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 41));
55757	57415	return __ret;
55758	57416	}
55759	57417	#else
55760	57418	__ai __attribute__((target("neon"))) float32x4_t vpmaxq_f32(float32x4_t __p0, float32x4_t __p1) {
55761	57419	float32x4_t __ret;
55762		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
55763		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
55764		__ret = (float32x4_t) __builtin_neon_vpmaxq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 41);
55765		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	57420	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	57421	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	57422	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vpmaxq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 41));
	57423	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
55766	57424	return __ret;
55767	57425	}
55768	57426	#endif
...	...	@@ -55770,16 +57428,16 @@ __ai __attribute__((target("neon"))) float32x4_t vpmaxq_f32(float32x4_t __p0, fl
55770	57428	#ifdef __LITTLE_ENDIAN__
55771	57429	__ai __attribute__((target("neon"))) int32x4_t vpmaxq_s32(int32x4_t __p0, int32x4_t __p1) {
55772	57430	int32x4_t __ret;
55773		__ret = (int32x4_t) __builtin_neon_vpmaxq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
	57431	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vpmaxq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 34));
55774	57432	return __ret;
55775	57433	}
55776	57434	#else
55777	57435	__ai __attribute__((target("neon"))) int32x4_t vpmaxq_s32(int32x4_t __p0, int32x4_t __p1) {
55778	57436	int32x4_t __ret;
55779		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
55780		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
55781		__ret = (int32x4_t) __builtin_neon_vpmaxq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
55782		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	57437	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	57438	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	57439	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vpmaxq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 34));
	57440	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
55783	57441	return __ret;
55784	57442	}
55785	57443	#endif
...	...	@@ -55787,16 +57445,16 @@ __ai __attribute__((target("neon"))) int32x4_t vpmaxq_s32(int32x4_t __p0, int32x
55787	57445	#ifdef __LITTLE_ENDIAN__
55788	57446	__ai __attribute__((target("neon"))) int16x8_t vpmaxq_s16(int16x8_t __p0, int16x8_t __p1) {
55789	57447	int16x8_t __ret;
55790		__ret = (int16x8_t) __builtin_neon_vpmaxq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
	57448	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vpmaxq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 33));
55791	57449	return __ret;
55792	57450	}
55793	57451	#else
55794	57452	__ai __attribute__((target("neon"))) int16x8_t vpmaxq_s16(int16x8_t __p0, int16x8_t __p1) {
55795	57453	int16x8_t __ret;
55796		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
55797		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
55798		__ret = (int16x8_t) __builtin_neon_vpmaxq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
55799		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	57454	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	57455	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	57456	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vpmaxq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 33));
	57457	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
55800	57458	return __ret;
55801	57459	}
55802	57460	#endif
...	...	@@ -55804,14 +57462,14 @@ __ai __attribute__((target("neon"))) int16x8_t vpmaxq_s16(int16x8_t __p0, int16x
55804	57462	#ifdef __LITTLE_ENDIAN__
55805	57463	__ai __attribute__((target("neon"))) float64_t vpmaxqd_f64(float64x2_t __p0) {
55806	57464	float64_t __ret;
55807		__ret = (float64_t) __builtin_neon_vpmaxqd_f64(__p0);
	57465	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vpmaxqd_f64(__p0));
55808	57466	return __ret;
55809	57467	}
55810	57468	#else
55811	57469	__ai __attribute__((target("neon"))) float64_t vpmaxqd_f64(float64x2_t __p0) {
55812	57470	float64_t __ret;
55813		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
55814		__ret = (float64_t) __builtin_neon_vpmaxqd_f64(__rev0);
	57471	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	57472	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vpmaxqd_f64(__rev0));
55815	57473	return __ret;
55816	57474	}
55817	57475	#endif
...	...	@@ -55819,14 +57477,14 @@ __ai __attribute__((target("neon"))) float64_t vpmaxqd_f64(float64x2_t __p0) {
55819	57477	#ifdef __LITTLE_ENDIAN__
55820	57478	__ai __attribute__((target("neon"))) float32_t vpmaxs_f32(float32x2_t __p0) {
55821	57479	float32_t __ret;
55822		__ret = (float32_t) __builtin_neon_vpmaxs_f32(__p0);
	57480	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vpmaxs_f32(__p0));
55823	57481	return __ret;
55824	57482	}
55825	57483	#else
55826	57484	__ai __attribute__((target("neon"))) float32_t vpmaxs_f32(float32x2_t __p0) {
55827	57485	float32_t __ret;
55828		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
55829		__ret = (float32_t) __builtin_neon_vpmaxs_f32(__rev0);
	57486	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	57487	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vpmaxs_f32(__rev0));
55830	57488	return __ret;
55831	57489	}
55832	57490	#endif
...	...	@@ -55834,16 +57492,16 @@ __ai __attribute__((target("neon"))) float32_t vpmaxs_f32(float32x2_t __p0) {
55834	57492	#ifdef __LITTLE_ENDIAN__
55835	57493	__ai __attribute__((target("neon"))) float64x2_t vpmaxnmq_f64(float64x2_t __p0, float64x2_t __p1) {
55836	57494	float64x2_t __ret;
55837		__ret = (float64x2_t) __builtin_neon_vpmaxnmq_v((int8x16_t)__p0, (int8x16_t)__p1, 42);
	57495	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vpmaxnmq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 42));
55838	57496	return __ret;
55839	57497	}
55840	57498	#else
55841	57499	__ai __attribute__((target("neon"))) float64x2_t vpmaxnmq_f64(float64x2_t __p0, float64x2_t __p1) {
55842	57500	float64x2_t __ret;
55843		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
55844		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
55845		__ret = (float64x2_t) __builtin_neon_vpmaxnmq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 42);
55846		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	57501	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	57502	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	57503	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vpmaxnmq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 42));
	57504	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
55847	57505	return __ret;
55848	57506	}
55849	57507	#endif
...	...	@@ -55851,16 +57509,16 @@ __ai __attribute__((target("neon"))) float64x2_t vpmaxnmq_f64(float64x2_t __p0,
55851	57509	#ifdef __LITTLE_ENDIAN__
55852	57510	__ai __attribute__((target("neon"))) float32x4_t vpmaxnmq_f32(float32x4_t __p0, float32x4_t __p1) {
55853	57511	float32x4_t __ret;
55854		__ret = (float32x4_t) __builtin_neon_vpmaxnmq_v((int8x16_t)__p0, (int8x16_t)__p1, 41);
	57512	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vpmaxnmq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 41));
55855	57513	return __ret;
55856	57514	}
55857	57515	#else
55858	57516	__ai __attribute__((target("neon"))) float32x4_t vpmaxnmq_f32(float32x4_t __p0, float32x4_t __p1) {
55859	57517	float32x4_t __ret;
55860		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
55861		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
55862		__ret = (float32x4_t) __builtin_neon_vpmaxnmq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 41);
55863		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	57518	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	57519	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	57520	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vpmaxnmq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 41));
	57521	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
55864	57522	return __ret;
55865	57523	}
55866	57524	#endif
...	...	@@ -55868,16 +57526,16 @@ __ai __attribute__((target("neon"))) float32x4_t vpmaxnmq_f32(float32x4_t __p0,
55868	57526	#ifdef __LITTLE_ENDIAN__
55869	57527	__ai __attribute__((target("neon"))) float32x2_t vpmaxnm_f32(float32x2_t __p0, float32x2_t __p1) {
55870	57528	float32x2_t __ret;
55871		__ret = (float32x2_t) __builtin_neon_vpmaxnm_v((int8x8_t)__p0, (int8x8_t)__p1, 9);
	57529	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vpmaxnm_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 9));
55872	57530	return __ret;
55873	57531	}
55874	57532	#else
55875	57533	__ai __attribute__((target("neon"))) float32x2_t vpmaxnm_f32(float32x2_t __p0, float32x2_t __p1) {
55876	57534	float32x2_t __ret;
55877		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
55878		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
55879		__ret = (float32x2_t) __builtin_neon_vpmaxnm_v((int8x8_t)__rev0, (int8x8_t)__rev1, 9);
55880		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	57535	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	57536	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	57537	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vpmaxnm_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 9));
	57538	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
55881	57539	return __ret;
55882	57540	}
55883	57541	#endif
...	...	@@ -55885,14 +57543,14 @@ __ai __attribute__((target("neon"))) float32x2_t vpmaxnm_f32(float32x2_t __p0, f
55885	57543	#ifdef __LITTLE_ENDIAN__
55886	57544	__ai __attribute__((target("neon"))) float64_t vpmaxnmqd_f64(float64x2_t __p0) {
55887	57545	float64_t __ret;
55888		__ret = (float64_t) __builtin_neon_vpmaxnmqd_f64(__p0);
	57546	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vpmaxnmqd_f64(__p0));
55889	57547	return __ret;
55890	57548	}
55891	57549	#else
55892	57550	__ai __attribute__((target("neon"))) float64_t vpmaxnmqd_f64(float64x2_t __p0) {
55893	57551	float64_t __ret;
55894		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
55895		__ret = (float64_t) __builtin_neon_vpmaxnmqd_f64(__rev0);
	57552	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	57553	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vpmaxnmqd_f64(__rev0));
55896	57554	return __ret;
55897	57555	}
55898	57556	#endif
...	...	@@ -55900,14 +57558,14 @@ __ai __attribute__((target("neon"))) float64_t vpmaxnmqd_f64(float64x2_t __p0) {
55900	57558	#ifdef __LITTLE_ENDIAN__
55901	57559	__ai __attribute__((target("neon"))) float32_t vpmaxnms_f32(float32x2_t __p0) {
55902	57560	float32_t __ret;
55903		__ret = (float32_t) __builtin_neon_vpmaxnms_f32(__p0);
	57561	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vpmaxnms_f32(__p0));
55904	57562	return __ret;
55905	57563	}
55906	57564	#else
55907	57565	__ai __attribute__((target("neon"))) float32_t vpmaxnms_f32(float32x2_t __p0) {
55908	57566	float32_t __ret;
55909		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
55910		__ret = (float32_t) __builtin_neon_vpmaxnms_f32(__rev0);
	57567	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	57568	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vpmaxnms_f32(__rev0));
55911	57569	return __ret;
55912	57570	}
55913	57571	#endif
...	...	@@ -55915,16 +57573,16 @@ __ai __attribute__((target("neon"))) float32_t vpmaxnms_f32(float32x2_t __p0) {
55915	57573	#ifdef __LITTLE_ENDIAN__
55916	57574	__ai __attribute__((target("neon"))) uint8x16_t vpminq_u8(uint8x16_t __p0, uint8x16_t __p1) {
55917	57575	uint8x16_t __ret;
55918		__ret = (uint8x16_t) __builtin_neon_vpminq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
	57576	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vpminq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 48));
55919	57577	return __ret;
55920	57578	}
55921	57579	#else
55922	57580	__ai __attribute__((target("neon"))) uint8x16_t vpminq_u8(uint8x16_t __p0, uint8x16_t __p1) {
55923	57581	uint8x16_t __ret;
55924		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
55925		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
55926		__ret = (uint8x16_t) __builtin_neon_vpminq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
55927		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	57582	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	57583	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	57584	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vpminq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 48));
	57585	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
55928	57586	return __ret;
55929	57587	}
55930	57588	#endif
...	...	@@ -55932,16 +57590,16 @@ __ai __attribute__((target("neon"))) uint8x16_t vpminq_u8(uint8x16_t __p0, uint8
55932	57590	#ifdef __LITTLE_ENDIAN__
55933	57591	__ai __attribute__((target("neon"))) uint32x4_t vpminq_u32(uint32x4_t __p0, uint32x4_t __p1) {
55934	57592	uint32x4_t __ret;
55935		__ret = (uint32x4_t) __builtin_neon_vpminq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
	57593	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vpminq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50));
55936	57594	return __ret;
55937	57595	}
55938	57596	#else
55939	57597	__ai __attribute__((target("neon"))) uint32x4_t vpminq_u32(uint32x4_t __p0, uint32x4_t __p1) {
55940	57598	uint32x4_t __ret;
55941		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
55942		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
55943		__ret = (uint32x4_t) __builtin_neon_vpminq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
55944		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	57599	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	57600	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	57601	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vpminq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 50));
	57602	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
55945	57603	return __ret;
55946	57604	}
55947	57605	#endif
...	...	@@ -55949,16 +57607,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vpminq_u32(uint32x4_t __p0, uint
55949	57607	#ifdef __LITTLE_ENDIAN__
55950	57608	__ai __attribute__((target("neon"))) uint16x8_t vpminq_u16(uint16x8_t __p0, uint16x8_t __p1) {
55951	57609	uint16x8_t __ret;
55952		__ret = (uint16x8_t) __builtin_neon_vpminq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
	57610	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vpminq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 49));
55953	57611	return __ret;
55954	57612	}
55955	57613	#else
55956	57614	__ai __attribute__((target("neon"))) uint16x8_t vpminq_u16(uint16x8_t __p0, uint16x8_t __p1) {
55957	57615	uint16x8_t __ret;
55958		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
55959		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
55960		__ret = (uint16x8_t) __builtin_neon_vpminq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
55961		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	57616	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	57617	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	57618	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vpminq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 49));
	57619	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
55962	57620	return __ret;
55963	57621	}
55964	57622	#endif
...	...	@@ -55966,16 +57624,16 @@ __ai __attribute__((target("neon"))) uint16x8_t vpminq_u16(uint16x8_t __p0, uint
55966	57624	#ifdef __LITTLE_ENDIAN__
55967	57625	__ai __attribute__((target("neon"))) int8x16_t vpminq_s8(int8x16_t __p0, int8x16_t __p1) {
55968	57626	int8x16_t __ret;
55969		__ret = (int8x16_t) __builtin_neon_vpminq_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
	57627	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vpminq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 32));
55970	57628	return __ret;
55971	57629	}
55972	57630	#else
55973	57631	__ai __attribute__((target("neon"))) int8x16_t vpminq_s8(int8x16_t __p0, int8x16_t __p1) {
55974	57632	int8x16_t __ret;
55975		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
55976		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
55977		__ret = (int8x16_t) __builtin_neon_vpminq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 32);
55978		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	57633	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	57634	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	57635	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vpminq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 32));
	57636	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
55979	57637	return __ret;
55980	57638	}
55981	57639	#endif
...	...	@@ -55983,16 +57641,16 @@ __ai __attribute__((target("neon"))) int8x16_t vpminq_s8(int8x16_t __p0, int8x16
55983	57641	#ifdef __LITTLE_ENDIAN__
55984	57642	__ai __attribute__((target("neon"))) float64x2_t vpminq_f64(float64x2_t __p0, float64x2_t __p1) {
55985	57643	float64x2_t __ret;
55986		__ret = (float64x2_t) __builtin_neon_vpminq_v((int8x16_t)__p0, (int8x16_t)__p1, 42);
	57644	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vpminq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 42));
55987	57645	return __ret;
55988	57646	}
55989	57647	#else
55990	57648	__ai __attribute__((target("neon"))) float64x2_t vpminq_f64(float64x2_t __p0, float64x2_t __p1) {
55991	57649	float64x2_t __ret;
55992		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
55993		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
55994		__ret = (float64x2_t) __builtin_neon_vpminq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 42);
55995		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	57650	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	57651	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	57652	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vpminq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 42));
	57653	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
55996	57654	return __ret;
55997	57655	}
55998	57656	#endif
...	...	@@ -56000,16 +57658,16 @@ __ai __attribute__((target("neon"))) float64x2_t vpminq_f64(float64x2_t __p0, fl
56000	57658	#ifdef __LITTLE_ENDIAN__
56001	57659	__ai __attribute__((target("neon"))) float32x4_t vpminq_f32(float32x4_t __p0, float32x4_t __p1) {
56002	57660	float32x4_t __ret;
56003		__ret = (float32x4_t) __builtin_neon_vpminq_v((int8x16_t)__p0, (int8x16_t)__p1, 41);
	57661	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vpminq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 41));
56004	57662	return __ret;
56005	57663	}
56006	57664	#else
56007	57665	__ai __attribute__((target("neon"))) float32x4_t vpminq_f32(float32x4_t __p0, float32x4_t __p1) {
56008	57666	float32x4_t __ret;
56009		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
56010		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
56011		__ret = (float32x4_t) __builtin_neon_vpminq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 41);
56012		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	57667	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	57668	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	57669	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vpminq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 41));
	57670	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
56013	57671	return __ret;
56014	57672	}
56015	57673	#endif
...	...	@@ -56017,16 +57675,16 @@ __ai __attribute__((target("neon"))) float32x4_t vpminq_f32(float32x4_t __p0, fl
56017	57675	#ifdef __LITTLE_ENDIAN__
56018	57676	__ai __attribute__((target("neon"))) int32x4_t vpminq_s32(int32x4_t __p0, int32x4_t __p1) {
56019	57677	int32x4_t __ret;
56020		__ret = (int32x4_t) __builtin_neon_vpminq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
	57678	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vpminq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 34));
56021	57679	return __ret;
56022	57680	}
56023	57681	#else
56024	57682	__ai __attribute__((target("neon"))) int32x4_t vpminq_s32(int32x4_t __p0, int32x4_t __p1) {
56025	57683	int32x4_t __ret;
56026		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
56027		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
56028		__ret = (int32x4_t) __builtin_neon_vpminq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
56029		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	57684	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	57685	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	57686	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vpminq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 34));
	57687	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
56030	57688	return __ret;
56031	57689	}
56032	57690	#endif
...	...	@@ -56034,16 +57692,16 @@ __ai __attribute__((target("neon"))) int32x4_t vpminq_s32(int32x4_t __p0, int32x
56034	57692	#ifdef __LITTLE_ENDIAN__
56035	57693	__ai __attribute__((target("neon"))) int16x8_t vpminq_s16(int16x8_t __p0, int16x8_t __p1) {
56036	57694	int16x8_t __ret;
56037		__ret = (int16x8_t) __builtin_neon_vpminq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
	57695	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vpminq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 33));
56038	57696	return __ret;
56039	57697	}
56040	57698	#else
56041	57699	__ai __attribute__((target("neon"))) int16x8_t vpminq_s16(int16x8_t __p0, int16x8_t __p1) {
56042	57700	int16x8_t __ret;
56043		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
56044		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
56045		__ret = (int16x8_t) __builtin_neon_vpminq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
56046		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	57701	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	57702	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	57703	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vpminq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 33));
	57704	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
56047	57705	return __ret;
56048	57706	}
56049	57707	#endif
...	...	@@ -56051,14 +57709,14 @@ __ai __attribute__((target("neon"))) int16x8_t vpminq_s16(int16x8_t __p0, int16x
56051	57709	#ifdef __LITTLE_ENDIAN__
56052	57710	__ai __attribute__((target("neon"))) float64_t vpminqd_f64(float64x2_t __p0) {
56053	57711	float64_t __ret;
56054		__ret = (float64_t) __builtin_neon_vpminqd_f64(__p0);
	57712	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vpminqd_f64(__p0));
56055	57713	return __ret;
56056	57714	}
56057	57715	#else
56058	57716	__ai __attribute__((target("neon"))) float64_t vpminqd_f64(float64x2_t __p0) {
56059	57717	float64_t __ret;
56060		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
56061		__ret = (float64_t) __builtin_neon_vpminqd_f64(__rev0);
	57718	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	57719	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vpminqd_f64(__rev0));
56062	57720	return __ret;
56063	57721	}
56064	57722	#endif
...	...	@@ -56066,14 +57724,14 @@ __ai __attribute__((target("neon"))) float64_t vpminqd_f64(float64x2_t __p0) {
56066	57724	#ifdef __LITTLE_ENDIAN__
56067	57725	__ai __attribute__((target("neon"))) float32_t vpmins_f32(float32x2_t __p0) {
56068	57726	float32_t __ret;
56069		__ret = (float32_t) __builtin_neon_vpmins_f32(__p0);
	57727	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vpmins_f32(__p0));
56070	57728	return __ret;
56071	57729	}
56072	57730	#else
56073	57731	__ai __attribute__((target("neon"))) float32_t vpmins_f32(float32x2_t __p0) {
56074	57732	float32_t __ret;
56075		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
56076		__ret = (float32_t) __builtin_neon_vpmins_f32(__rev0);
	57733	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	57734	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vpmins_f32(__rev0));
56077	57735	return __ret;
56078	57736	}
56079	57737	#endif
...	...	@@ -56081,16 +57739,16 @@ __ai __attribute__((target("neon"))) float32_t vpmins_f32(float32x2_t __p0) {
56081	57739	#ifdef __LITTLE_ENDIAN__
56082	57740	__ai __attribute__((target("neon"))) float64x2_t vpminnmq_f64(float64x2_t __p0, float64x2_t __p1) {
56083	57741	float64x2_t __ret;
56084		__ret = (float64x2_t) __builtin_neon_vpminnmq_v((int8x16_t)__p0, (int8x16_t)__p1, 42);
	57742	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vpminnmq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 42));
56085	57743	return __ret;
56086	57744	}
56087	57745	#else
56088	57746	__ai __attribute__((target("neon"))) float64x2_t vpminnmq_f64(float64x2_t __p0, float64x2_t __p1) {
56089	57747	float64x2_t __ret;
56090		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
56091		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
56092		__ret = (float64x2_t) __builtin_neon_vpminnmq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 42);
56093		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	57748	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	57749	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	57750	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vpminnmq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 42));
	57751	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
56094	57752	return __ret;
56095	57753	}
56096	57754	#endif
...	...	@@ -56098,16 +57756,16 @@ __ai __attribute__((target("neon"))) float64x2_t vpminnmq_f64(float64x2_t __p0,
56098	57756	#ifdef __LITTLE_ENDIAN__
56099	57757	__ai __attribute__((target("neon"))) float32x4_t vpminnmq_f32(float32x4_t __p0, float32x4_t __p1) {
56100	57758	float32x4_t __ret;
56101		__ret = (float32x4_t) __builtin_neon_vpminnmq_v((int8x16_t)__p0, (int8x16_t)__p1, 41);
	57759	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vpminnmq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 41));
56102	57760	return __ret;
56103	57761	}
56104	57762	#else
56105	57763	__ai __attribute__((target("neon"))) float32x4_t vpminnmq_f32(float32x4_t __p0, float32x4_t __p1) {
56106	57764	float32x4_t __ret;
56107		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
56108		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
56109		__ret = (float32x4_t) __builtin_neon_vpminnmq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 41);
56110		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	57765	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	57766	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	57767	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vpminnmq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 41));
	57768	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
56111	57769	return __ret;
56112	57770	}
56113	57771	#endif
...	...	@@ -56115,16 +57773,16 @@ __ai __attribute__((target("neon"))) float32x4_t vpminnmq_f32(float32x4_t __p0,
56115	57773	#ifdef __LITTLE_ENDIAN__
56116	57774	__ai __attribute__((target("neon"))) float32x2_t vpminnm_f32(float32x2_t __p0, float32x2_t __p1) {
56117	57775	float32x2_t __ret;
56118		__ret = (float32x2_t) __builtin_neon_vpminnm_v((int8x8_t)__p0, (int8x8_t)__p1, 9);
	57776	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vpminnm_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 9));
56119	57777	return __ret;
56120	57778	}
56121	57779	#else
56122	57780	__ai __attribute__((target("neon"))) float32x2_t vpminnm_f32(float32x2_t __p0, float32x2_t __p1) {
56123	57781	float32x2_t __ret;
56124		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
56125		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
56126		__ret = (float32x2_t) __builtin_neon_vpminnm_v((int8x8_t)__rev0, (int8x8_t)__rev1, 9);
56127		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	57782	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	57783	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	57784	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vpminnm_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 9));
	57785	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
56128	57786	return __ret;
56129	57787	}
56130	57788	#endif
...	...	@@ -56132,14 +57790,14 @@ __ai __attribute__((target("neon"))) float32x2_t vpminnm_f32(float32x2_t __p0, f
56132	57790	#ifdef __LITTLE_ENDIAN__
56133	57791	__ai __attribute__((target("neon"))) float64_t vpminnmqd_f64(float64x2_t __p0) {
56134	57792	float64_t __ret;
56135		__ret = (float64_t) __builtin_neon_vpminnmqd_f64(__p0);
	57793	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vpminnmqd_f64(__p0));
56136	57794	return __ret;
56137	57795	}
56138	57796	#else
56139	57797	__ai __attribute__((target("neon"))) float64_t vpminnmqd_f64(float64x2_t __p0) {
56140	57798	float64_t __ret;
56141		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
56142		__ret = (float64_t) __builtin_neon_vpminnmqd_f64(__rev0);
	57799	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	57800	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vpminnmqd_f64(__rev0));
56143	57801	return __ret;
56144	57802	}
56145	57803	#endif
...	...	@@ -56147,14 +57805,14 @@ __ai __attribute__((target("neon"))) float64_t vpminnmqd_f64(float64x2_t __p0) {
56147	57805	#ifdef __LITTLE_ENDIAN__
56148	57806	__ai __attribute__((target("neon"))) float32_t vpminnms_f32(float32x2_t __p0) {
56149	57807	float32_t __ret;
56150		__ret = (float32_t) __builtin_neon_vpminnms_f32(__p0);
	57808	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vpminnms_f32(__p0));
56151	57809	return __ret;
56152	57810	}
56153	57811	#else
56154	57812	__ai __attribute__((target("neon"))) float32_t vpminnms_f32(float32x2_t __p0) {
56155	57813	float32_t __ret;
56156		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
56157		__ret = (float32_t) __builtin_neon_vpminnms_f32(__rev0);
	57814	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	57815	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vpminnms_f32(__rev0));
56158	57816	return __ret;
56159	57817	}
56160	57818	#endif
...	...	@@ -56162,92 +57820,92 @@ __ai __attribute__((target("neon"))) float32_t vpminnms_f32(float32x2_t __p0) {
56162	57820	#ifdef __LITTLE_ENDIAN__
56163	57821	__ai __attribute__((target("neon"))) int64x2_t vqabsq_s64(int64x2_t __p0) {
56164	57822	int64x2_t __ret;
56165		__ret = (int64x2_t) __builtin_neon_vqabsq_v((int8x16_t)__p0, 35);
	57823	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vqabsq_v(__builtin_bit_cast(int8x16_t, __p0), 35));
56166	57824	return __ret;
56167	57825	}
56168	57826	#else
56169	57827	__ai __attribute__((target("neon"))) int64x2_t vqabsq_s64(int64x2_t __p0) {
56170	57828	int64x2_t __ret;
56171		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
56172		__ret = (int64x2_t) __builtin_neon_vqabsq_v((int8x16_t)__rev0, 35);
56173		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	57829	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	57830	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vqabsq_v(__builtin_bit_cast(int8x16_t, __rev0), 35));
	57831	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
56174	57832	return __ret;
56175	57833	}
56176	57834	#endif
56177	57835
56178	57836	__ai __attribute__((target("neon"))) int64x1_t vqabs_s64(int64x1_t __p0) {
56179	57837	int64x1_t __ret;
56180		__ret = (int64x1_t) __builtin_neon_vqabs_v((int8x8_t)__p0, 3);
	57838	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vqabs_v(__builtin_bit_cast(int8x8_t, __p0), 3));
56181	57839	return __ret;
56182	57840	}
56183	57841	__ai __attribute__((target("neon"))) int8_t vqabsb_s8(int8_t __p0) {
56184	57842	int8_t __ret;
56185		__ret = (int8_t) __builtin_neon_vqabsb_s8(__p0);
	57843	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vqabsb_s8(__p0));
56186	57844	return __ret;
56187	57845	}
56188	57846	__ai __attribute__((target("neon"))) int32_t vqabss_s32(int32_t __p0) {
56189	57847	int32_t __ret;
56190		__ret = (int32_t) __builtin_neon_vqabss_s32(__p0);
	57848	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vqabss_s32(__p0));
56191	57849	return __ret;
56192	57850	}
56193	57851	__ai __attribute__((target("neon"))) int64_t vqabsd_s64(int64_t __p0) {
56194	57852	int64_t __ret;
56195		__ret = (int64_t) __builtin_neon_vqabsd_s64(__p0);
	57853	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vqabsd_s64(__p0));
56196	57854	return __ret;
56197	57855	}
56198	57856	__ai __attribute__((target("neon"))) int16_t vqabsh_s16(int16_t __p0) {
56199	57857	int16_t __ret;
56200		__ret = (int16_t) __builtin_neon_vqabsh_s16(__p0);
	57858	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vqabsh_s16(__p0));
56201	57859	return __ret;
56202	57860	}
56203	57861	__ai __attribute__((target("neon"))) uint8_t vqaddb_u8(uint8_t __p0, uint8_t __p1) {
56204	57862	uint8_t __ret;
56205		__ret = (uint8_t) __builtin_neon_vqaddb_u8(__p0, __p1);
	57863	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vqaddb_u8(__p0, __p1));
56206	57864	return __ret;
56207	57865	}
56208	57866	__ai __attribute__((target("neon"))) uint32_t vqadds_u32(uint32_t __p0, uint32_t __p1) {
56209	57867	uint32_t __ret;
56210		__ret = (uint32_t) __builtin_neon_vqadds_u32(__p0, __p1);
	57868	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vqadds_u32(__p0, __p1));
56211	57869	return __ret;
56212	57870	}
56213	57871	__ai __attribute__((target("neon"))) uint64_t vqaddd_u64(uint64_t __p0, uint64_t __p1) {
56214	57872	uint64_t __ret;
56215		__ret = (uint64_t) __builtin_neon_vqaddd_u64(__p0, __p1);
	57873	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vqaddd_u64(__p0, __p1));
56216	57874	return __ret;
56217	57875	}
56218	57876	__ai __attribute__((target("neon"))) uint16_t vqaddh_u16(uint16_t __p0, uint16_t __p1) {
56219	57877	uint16_t __ret;
56220		__ret = (uint16_t) __builtin_neon_vqaddh_u16(__p0, __p1);
	57878	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vqaddh_u16(__p0, __p1));
56221	57879	return __ret;
56222	57880	}
56223	57881	__ai __attribute__((target("neon"))) int8_t vqaddb_s8(int8_t __p0, int8_t __p1) {
56224	57882	int8_t __ret;
56225		__ret = (int8_t) __builtin_neon_vqaddb_s8(__p0, __p1);
	57883	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vqaddb_s8(__p0, __p1));
56226	57884	return __ret;
56227	57885	}
56228	57886	__ai __attribute__((target("neon"))) int32_t vqadds_s32(int32_t __p0, int32_t __p1) {
56229	57887	int32_t __ret;
56230		__ret = (int32_t) __builtin_neon_vqadds_s32(__p0, __p1);
	57888	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vqadds_s32(__p0, __p1));
56231	57889	return __ret;
56232	57890	}
56233	57891	__ai __attribute__((target("neon"))) int64_t vqaddd_s64(int64_t __p0, int64_t __p1) {
56234	57892	int64_t __ret;
56235		__ret = (int64_t) __builtin_neon_vqaddd_s64(__p0, __p1);
	57893	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vqaddd_s64(__p0, __p1));
56236	57894	return __ret;
56237	57895	}
56238	57896	__ai __attribute__((target("neon"))) int16_t vqaddh_s16(int16_t __p0, int16_t __p1) {
56239	57897	int16_t __ret;
56240		__ret = (int16_t) __builtin_neon_vqaddh_s16(__p0, __p1);
	57898	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vqaddh_s16(__p0, __p1));
56241	57899	return __ret;
56242	57900	}
56243	57901	__ai __attribute__((target("neon"))) int64_t vqdmlals_s32(int64_t __p0, int32_t __p1, int32_t __p2) {
56244	57902	int64_t __ret;
56245		__ret = (int64_t) __builtin_neon_vqdmlals_s32(__p0, __p1, __p2);
	57903	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vqdmlals_s32(__p0, __p1, __p2));
56246	57904	return __ret;
56247	57905	}
56248	57906	__ai __attribute__((target("neon"))) int32_t vqdmlalh_s16(int32_t __p0, int16_t __p1, int16_t __p2) {
56249	57907	int32_t __ret;
56250		__ret = (int32_t) __builtin_neon_vqdmlalh_s16(__p0, __p1, __p2);
	57908	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vqdmlalh_s16(__p0, __p1, __p2));
56251	57909	return __ret;
56252	57910	}
56253	57911	#ifdef __LITTLE_ENDIAN__
...	...	@@ -56259,11 +57917,11 @@ __ai __attribute__((target("neon"))) int64x2_t vqdmlal_high_s32(int64x2_t __p0,
56259	57917	#else
56260	57918	__ai __attribute__((target("neon"))) int64x2_t vqdmlal_high_s32(int64x2_t __p0, int32x4_t __p1, int32x4_t __p2) {
56261	57919	int64x2_t __ret;
56262		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
56263		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
56264		int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	57920	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	57921	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	57922	int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
56265	57923	__ret = __noswap_vqdmlal_s32(__rev0, __noswap_vget_high_s32(__rev1), __noswap_vget_high_s32(__rev2));
56266		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	57924	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
56267	57925	return __ret;
56268	57926	}
56269	57927	#endif
...	...	@@ -56277,108 +57935,108 @@ __ai __attribute__((target("neon"))) int32x4_t vqdmlal_high_s16(int32x4_t __p0,
56277	57935	#else
56278	57936	__ai __attribute__((target("neon"))) int32x4_t vqdmlal_high_s16(int32x4_t __p0, int16x8_t __p1, int16x8_t __p2) {
56279	57937	int32x4_t __ret;
56280		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
56281		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
56282		int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	57938	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	57939	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	57940	int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
56283	57941	__ret = __noswap_vqdmlal_s16(__rev0, __noswap_vget_high_s16(__rev1), __noswap_vget_high_s16(__rev2));
56284		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	57942	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
56285	57943	return __ret;
56286	57944	}
56287	57945	#endif
56288	57946
56289	57947	#ifdef __LITTLE_ENDIAN__
56290		#define vqdmlal_high_lane_s32(__p0_628, __p1_628, __p2_628, __p3_628) __extension__ ({ \
56291		int64x2_t __ret_628; \
56292		int64x2_t __s0_628 = __p0_628; \
56293		int32x4_t __s1_628 = __p1_628; \
56294		int32x2_t __s2_628 = __p2_628; \
56295		__ret_628 = vqdmlal_s32(__s0_628, vget_high_s32(__s1_628), splat_lane_s32(__s2_628, __p3_628)); \
56296		__ret_628; \
	57948	#define vqdmlal_high_lane_s32(__p0_644, __p1_644, __p2_644, __p3_644) __extension__ ({ \
	57949	int64x2_t __ret_644; \
	57950	int64x2_t __s0_644 = __p0_644; \
	57951	int32x4_t __s1_644 = __p1_644; \
	57952	int32x2_t __s2_644 = __p2_644; \
	57953	__ret_644 = vqdmlal_s32(__s0_644, vget_high_s32(__s1_644), splat_lane_s32(__s2_644, __p3_644)); \
	57954	__ret_644; \
56297	57955	})
56298	57956	#else
56299		#define vqdmlal_high_lane_s32(__p0_629, __p1_629, __p2_629, __p3_629) __extension__ ({ \
56300		int64x2_t __ret_629; \
56301		int64x2_t __s0_629 = __p0_629; \
56302		int32x4_t __s1_629 = __p1_629; \
56303		int32x2_t __s2_629 = __p2_629; \
56304		int64x2_t __rev0_629; __rev0_629 = __builtin_shufflevector(__s0_629, __s0_629, 1, 0); \
56305		int32x4_t __rev1_629; __rev1_629 = __builtin_shufflevector(__s1_629, __s1_629, 3, 2, 1, 0); \
56306		int32x2_t __rev2_629; __rev2_629 = __builtin_shufflevector(__s2_629, __s2_629, 1, 0); \
56307		__ret_629 = __noswap_vqdmlal_s32(__rev0_629, __noswap_vget_high_s32(__rev1_629), __noswap_splat_lane_s32(__rev2_629, __p3_629)); \
56308		__ret_629 = __builtin_shufflevector(__ret_629, __ret_629, 1, 0); \
56309		__ret_629; \
	57957	#define vqdmlal_high_lane_s32(__p0_645, __p1_645, __p2_645, __p3_645) __extension__ ({ \
	57958	int64x2_t __ret_645; \
	57959	int64x2_t __s0_645 = __p0_645; \
	57960	int32x4_t __s1_645 = __p1_645; \
	57961	int32x2_t __s2_645 = __p2_645; \
	57962	int64x2_t __rev0_645; __rev0_645 = __builtin_shufflevector(__s0_645, __s0_645, __lane_reverse_128_64); \
	57963	int32x4_t __rev1_645; __rev1_645 = __builtin_shufflevector(__s1_645, __s1_645, __lane_reverse_128_32); \
	57964	int32x2_t __rev2_645; __rev2_645 = __builtin_shufflevector(__s2_645, __s2_645, __lane_reverse_64_32); \
	57965	__ret_645 = __noswap_vqdmlal_s32(__rev0_645, __noswap_vget_high_s32(__rev1_645), __noswap_splat_lane_s32(__rev2_645, __p3_645)); \
	57966	__ret_645 = __builtin_shufflevector(__ret_645, __ret_645, __lane_reverse_128_64); \
	57967	__ret_645; \
56310	57968	})
56311	57969	#endif
56312	57970
56313	57971	#ifdef __LITTLE_ENDIAN__
56314		#define vqdmlal_high_lane_s16(__p0_630, __p1_630, __p2_630, __p3_630) __extension__ ({ \
56315		int32x4_t __ret_630; \
56316		int32x4_t __s0_630 = __p0_630; \
56317		int16x8_t __s1_630 = __p1_630; \
56318		int16x4_t __s2_630 = __p2_630; \
56319		__ret_630 = vqdmlal_s16(__s0_630, vget_high_s16(__s1_630), splat_lane_s16(__s2_630, __p3_630)); \
56320		__ret_630; \
	57972	#define vqdmlal_high_lane_s16(__p0_646, __p1_646, __p2_646, __p3_646) __extension__ ({ \
	57973	int32x4_t __ret_646; \
	57974	int32x4_t __s0_646 = __p0_646; \
	57975	int16x8_t __s1_646 = __p1_646; \
	57976	int16x4_t __s2_646 = __p2_646; \
	57977	__ret_646 = vqdmlal_s16(__s0_646, vget_high_s16(__s1_646), splat_lane_s16(__s2_646, __p3_646)); \
	57978	__ret_646; \
56321	57979	})
56322	57980	#else
56323		#define vqdmlal_high_lane_s16(__p0_631, __p1_631, __p2_631, __p3_631) __extension__ ({ \
56324		int32x4_t __ret_631; \
56325		int32x4_t __s0_631 = __p0_631; \
56326		int16x8_t __s1_631 = __p1_631; \
56327		int16x4_t __s2_631 = __p2_631; \
56328		int32x4_t __rev0_631; __rev0_631 = __builtin_shufflevector(__s0_631, __s0_631, 3, 2, 1, 0); \
56329		int16x8_t __rev1_631; __rev1_631 = __builtin_shufflevector(__s1_631, __s1_631, 7, 6, 5, 4, 3, 2, 1, 0); \
56330		int16x4_t __rev2_631; __rev2_631 = __builtin_shufflevector(__s2_631, __s2_631, 3, 2, 1, 0); \
56331		__ret_631 = __noswap_vqdmlal_s16(__rev0_631, __noswap_vget_high_s16(__rev1_631), __noswap_splat_lane_s16(__rev2_631, __p3_631)); \
56332		__ret_631 = __builtin_shufflevector(__ret_631, __ret_631, 3, 2, 1, 0); \
56333		__ret_631; \
	57981	#define vqdmlal_high_lane_s16(__p0_647, __p1_647, __p2_647, __p3_647) __extension__ ({ \
	57982	int32x4_t __ret_647; \
	57983	int32x4_t __s0_647 = __p0_647; \
	57984	int16x8_t __s1_647 = __p1_647; \
	57985	int16x4_t __s2_647 = __p2_647; \
	57986	int32x4_t __rev0_647; __rev0_647 = __builtin_shufflevector(__s0_647, __s0_647, __lane_reverse_128_32); \
	57987	int16x8_t __rev1_647; __rev1_647 = __builtin_shufflevector(__s1_647, __s1_647, __lane_reverse_128_16); \
	57988	int16x4_t __rev2_647; __rev2_647 = __builtin_shufflevector(__s2_647, __s2_647, __lane_reverse_64_16); \
	57989	__ret_647 = __noswap_vqdmlal_s16(__rev0_647, __noswap_vget_high_s16(__rev1_647), __noswap_splat_lane_s16(__rev2_647, __p3_647)); \
	57990	__ret_647 = __builtin_shufflevector(__ret_647, __ret_647, __lane_reverse_128_32); \
	57991	__ret_647; \
56334	57992	})
56335	57993	#endif
56336	57994
56337	57995	#ifdef __LITTLE_ENDIAN__
56338		#define vqdmlal_high_laneq_s32(__p0_632, __p1_632, __p2_632, __p3_632) __extension__ ({ \
56339		int64x2_t __ret_632; \
56340		int64x2_t __s0_632 = __p0_632; \
56341		int32x4_t __s1_632 = __p1_632; \
56342		int32x4_t __s2_632 = __p2_632; \
56343		__ret_632 = vqdmlal_s32(__s0_632, vget_high_s32(__s1_632), splat_laneq_s32(__s2_632, __p3_632)); \
56344		__ret_632; \
	57996	#define vqdmlal_high_laneq_s32(__p0_648, __p1_648, __p2_648, __p3_648) __extension__ ({ \
	57997	int64x2_t __ret_648; \
	57998	int64x2_t __s0_648 = __p0_648; \
	57999	int32x4_t __s1_648 = __p1_648; \
	58000	int32x4_t __s2_648 = __p2_648; \
	58001	__ret_648 = vqdmlal_s32(__s0_648, vget_high_s32(__s1_648), splat_laneq_s32(__s2_648, __p3_648)); \
	58002	__ret_648; \
56345	58003	})
56346	58004	#else
56347		#define vqdmlal_high_laneq_s32(__p0_633, __p1_633, __p2_633, __p3_633) __extension__ ({ \
56348		int64x2_t __ret_633; \
56349		int64x2_t __s0_633 = __p0_633; \
56350		int32x4_t __s1_633 = __p1_633; \
56351		int32x4_t __s2_633 = __p2_633; \
56352		int64x2_t __rev0_633; __rev0_633 = __builtin_shufflevector(__s0_633, __s0_633, 1, 0); \
56353		int32x4_t __rev1_633; __rev1_633 = __builtin_shufflevector(__s1_633, __s1_633, 3, 2, 1, 0); \
56354		int32x4_t __rev2_633; __rev2_633 = __builtin_shufflevector(__s2_633, __s2_633, 3, 2, 1, 0); \
56355		__ret_633 = __noswap_vqdmlal_s32(__rev0_633, __noswap_vget_high_s32(__rev1_633), __noswap_splat_laneq_s32(__rev2_633, __p3_633)); \
56356		__ret_633 = __builtin_shufflevector(__ret_633, __ret_633, 1, 0); \
56357		__ret_633; \
	58005	#define vqdmlal_high_laneq_s32(__p0_649, __p1_649, __p2_649, __p3_649) __extension__ ({ \
	58006	int64x2_t __ret_649; \
	58007	int64x2_t __s0_649 = __p0_649; \
	58008	int32x4_t __s1_649 = __p1_649; \
	58009	int32x4_t __s2_649 = __p2_649; \
	58010	int64x2_t __rev0_649; __rev0_649 = __builtin_shufflevector(__s0_649, __s0_649, __lane_reverse_128_64); \
	58011	int32x4_t __rev1_649; __rev1_649 = __builtin_shufflevector(__s1_649, __s1_649, __lane_reverse_128_32); \
	58012	int32x4_t __rev2_649; __rev2_649 = __builtin_shufflevector(__s2_649, __s2_649, __lane_reverse_128_32); \
	58013	__ret_649 = __noswap_vqdmlal_s32(__rev0_649, __noswap_vget_high_s32(__rev1_649), __noswap_splat_laneq_s32(__rev2_649, __p3_649)); \
	58014	__ret_649 = __builtin_shufflevector(__ret_649, __ret_649, __lane_reverse_128_64); \
	58015	__ret_649; \
56358	58016	})
56359	58017	#endif
56360	58018
56361	58019	#ifdef __LITTLE_ENDIAN__
56362		#define vqdmlal_high_laneq_s16(__p0_634, __p1_634, __p2_634, __p3_634) __extension__ ({ \
56363		int32x4_t __ret_634; \
56364		int32x4_t __s0_634 = __p0_634; \
56365		int16x8_t __s1_634 = __p1_634; \
56366		int16x8_t __s2_634 = __p2_634; \
56367		__ret_634 = vqdmlal_s16(__s0_634, vget_high_s16(__s1_634), splat_laneq_s16(__s2_634, __p3_634)); \
56368		__ret_634; \
	58020	#define vqdmlal_high_laneq_s16(__p0_650, __p1_650, __p2_650, __p3_650) __extension__ ({ \
	58021	int32x4_t __ret_650; \
	58022	int32x4_t __s0_650 = __p0_650; \
	58023	int16x8_t __s1_650 = __p1_650; \
	58024	int16x8_t __s2_650 = __p2_650; \
	58025	__ret_650 = vqdmlal_s16(__s0_650, vget_high_s16(__s1_650), splat_laneq_s16(__s2_650, __p3_650)); \
	58026	__ret_650; \
56369	58027	})
56370	58028	#else
56371		#define vqdmlal_high_laneq_s16(__p0_635, __p1_635, __p2_635, __p3_635) __extension__ ({ \
56372		int32x4_t __ret_635; \
56373		int32x4_t __s0_635 = __p0_635; \
56374		int16x8_t __s1_635 = __p1_635; \
56375		int16x8_t __s2_635 = __p2_635; \
56376		int32x4_t __rev0_635; __rev0_635 = __builtin_shufflevector(__s0_635, __s0_635, 3, 2, 1, 0); \
56377		int16x8_t __rev1_635; __rev1_635 = __builtin_shufflevector(__s1_635, __s1_635, 7, 6, 5, 4, 3, 2, 1, 0); \
56378		int16x8_t __rev2_635; __rev2_635 = __builtin_shufflevector(__s2_635, __s2_635, 7, 6, 5, 4, 3, 2, 1, 0); \
56379		__ret_635 = __noswap_vqdmlal_s16(__rev0_635, __noswap_vget_high_s16(__rev1_635), __noswap_splat_laneq_s16(__rev2_635, __p3_635)); \
56380		__ret_635 = __builtin_shufflevector(__ret_635, __ret_635, 3, 2, 1, 0); \
56381		__ret_635; \
	58029	#define vqdmlal_high_laneq_s16(__p0_651, __p1_651, __p2_651, __p3_651) __extension__ ({ \
	58030	int32x4_t __ret_651; \
	58031	int32x4_t __s0_651 = __p0_651; \
	58032	int16x8_t __s1_651 = __p1_651; \
	58033	int16x8_t __s2_651 = __p2_651; \
	58034	int32x4_t __rev0_651; __rev0_651 = __builtin_shufflevector(__s0_651, __s0_651, __lane_reverse_128_32); \
	58035	int16x8_t __rev1_651; __rev1_651 = __builtin_shufflevector(__s1_651, __s1_651, __lane_reverse_128_16); \
	58036	int16x8_t __rev2_651; __rev2_651 = __builtin_shufflevector(__s2_651, __s2_651, __lane_reverse_128_16); \
	58037	__ret_651 = __noswap_vqdmlal_s16(__rev0_651, __noswap_vget_high_s16(__rev1_651), __noswap_splat_laneq_s16(__rev2_651, __p3_651)); \
	58038	__ret_651 = __builtin_shufflevector(__ret_651, __ret_651, __lane_reverse_128_32); \
	58039	__ret_651; \
56382	58040	})
56383	58041	#endif
56384	58042
...	...	@@ -56391,10 +58049,10 @@ __ai __attribute__((target("neon"))) int64x2_t vqdmlal_high_n_s32(int64x2_t __p0
56391	58049	#else
56392	58050	__ai __attribute__((target("neon"))) int64x2_t vqdmlal_high_n_s32(int64x2_t __p0, int32x4_t __p1, int32_t __p2) {
56393	58051	int64x2_t __ret;
56394		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
56395		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	58052	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	58053	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
56396	58054	__ret = __noswap_vqdmlal_n_s32(__rev0, __noswap_vget_high_s32(__rev1), __p2);
56397		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	58055	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
56398	58056	return __ret;
56399	58057	}
56400	58058	#endif
...	...	@@ -56408,10 +58066,10 @@ __ai __attribute__((target("neon"))) int32x4_t vqdmlal_high_n_s16(int32x4_t __p0
56408	58066	#else
56409	58067	__ai __attribute__((target("neon"))) int32x4_t vqdmlal_high_n_s16(int32x4_t __p0, int16x8_t __p1, int16_t __p2) {
56410	58068	int32x4_t __ret;
56411		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
56412		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	58069	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	58070	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
56413	58071	__ret = __noswap_vqdmlal_n_s16(__rev0, __noswap_vget_high_s16(__rev1), __p2);
56414		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	58072	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
56415	58073	return __ret;
56416	58074	}
56417	58075	#endif
...	...	@@ -56422,7 +58080,7 @@ __ai __attribute__((target("neon"))) int32x4_t vqdmlal_high_n_s16(int32x4_t __p0
56422	58080	int64_t __s0 = __p0; \
56423	58081	int32_t __s1 = __p1; \
56424	58082	int32x2_t __s2 = __p2; \
56425		__ret = (int64_t) __builtin_neon_vqdmlals_lane_s32(__s0, __s1, __s2, __p3); \
	58083	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vqdmlals_lane_s32(__s0, __s1, __s2, __p3)); \
56426	58084	__ret; \
56427	58085	})
56428	58086	#else
...	...	@@ -56431,8 +58089,8 @@ __ai __attribute__((target("neon"))) int32x4_t vqdmlal_high_n_s16(int32x4_t __p0
56431	58089	int64_t __s0 = __p0; \
56432	58090	int32_t __s1 = __p1; \
56433	58091	int32x2_t __s2 = __p2; \
56434		int32x2_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 1, 0); \
56435		__ret = (int64_t) __builtin_neon_vqdmlals_lane_s32(__s0, __s1, __rev2, __p3); \
	58092	int32x2_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_64_32); \
	58093	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vqdmlals_lane_s32(__s0, __s1, __rev2, __p3)); \
56436	58094	__ret; \
56437	58095	})
56438	58096	#endif
...	...	@@ -56443,7 +58101,7 @@ __ai __attribute__((target("neon"))) int32x4_t vqdmlal_high_n_s16(int32x4_t __p0
56443	58101	int32_t __s0 = __p0; \
56444	58102	int16_t __s1 = __p1; \
56445	58103	int16x4_t __s2 = __p2; \
56446		__ret = (int32_t) __builtin_neon_vqdmlalh_lane_s16(__s0, __s1, __s2, __p3); \
	58104	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vqdmlalh_lane_s16(__s0, __s1, __s2, __p3)); \
56447	58105	__ret; \
56448	58106	})
56449	58107	#else
...	...	@@ -56452,8 +58110,8 @@ __ai __attribute__((target("neon"))) int32x4_t vqdmlal_high_n_s16(int32x4_t __p0
56452	58110	int32_t __s0 = __p0; \
56453	58111	int16_t __s1 = __p1; \
56454	58112	int16x4_t __s2 = __p2; \
56455		int16x4_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 3, 2, 1, 0); \
56456		__ret = (int32_t) __builtin_neon_vqdmlalh_lane_s16(__s0, __s1, __rev2, __p3); \
	58113	int16x4_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_64_16); \
	58114	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vqdmlalh_lane_s16(__s0, __s1, __rev2, __p3)); \
56457	58115	__ret; \
56458	58116	})
56459	58117	#endif
...	...	@@ -56464,7 +58122,7 @@ __ai __attribute__((target("neon"))) int32x4_t vqdmlal_high_n_s16(int32x4_t __p0
56464	58122	int64_t __s0 = __p0; \
56465	58123	int32_t __s1 = __p1; \
56466	58124	int32x4_t __s2 = __p2; \
56467		__ret = (int64_t) __builtin_neon_vqdmlals_laneq_s32(__s0, __s1, __s2, __p3); \
	58125	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vqdmlals_laneq_s32(__s0, __s1, __s2, __p3)); \
56468	58126	__ret; \
56469	58127	})
56470	58128	#else
...	...	@@ -56473,8 +58131,8 @@ __ai __attribute__((target("neon"))) int32x4_t vqdmlal_high_n_s16(int32x4_t __p0
56473	58131	int64_t __s0 = __p0; \
56474	58132	int32_t __s1 = __p1; \
56475	58133	int32x4_t __s2 = __p2; \
56476		int32x4_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 3, 2, 1, 0); \
56477		__ret = (int64_t) __builtin_neon_vqdmlals_laneq_s32(__s0, __s1, __rev2, __p3); \
	58134	int32x4_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_128_32); \
	58135	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vqdmlals_laneq_s32(__s0, __s1, __rev2, __p3)); \
56478	58136	__ret; \
56479	58137	})
56480	58138	#endif
...	...	@@ -56485,7 +58143,7 @@ __ai __attribute__((target("neon"))) int32x4_t vqdmlal_high_n_s16(int32x4_t __p0
56485	58143	int32_t __s0 = __p0; \
56486	58144	int16_t __s1 = __p1; \
56487	58145	int16x8_t __s2 = __p2; \
56488		__ret = (int32_t) __builtin_neon_vqdmlalh_laneq_s16(__s0, __s1, __s2, __p3); \
	58146	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vqdmlalh_laneq_s16(__s0, __s1, __s2, __p3)); \
56489	58147	__ret; \
56490	58148	})
56491	58149	#else
...	...	@@ -56494,68 +58152,68 @@ __ai __attribute__((target("neon"))) int32x4_t vqdmlal_high_n_s16(int32x4_t __p0
56494	58152	int32_t __s0 = __p0; \
56495	58153	int16_t __s1 = __p1; \
56496	58154	int16x8_t __s2 = __p2; \
56497		int16x8_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 7, 6, 5, 4, 3, 2, 1, 0); \
56498		__ret = (int32_t) __builtin_neon_vqdmlalh_laneq_s16(__s0, __s1, __rev2, __p3); \
	58155	int16x8_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_128_16); \
	58156	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vqdmlalh_laneq_s16(__s0, __s1, __rev2, __p3)); \
56499	58157	__ret; \
56500	58158	})
56501	58159	#endif
56502	58160
56503	58161	#ifdef __LITTLE_ENDIAN__
56504		#define vqdmlal_laneq_s32(__p0_636, __p1_636, __p2_636, __p3_636) __extension__ ({ \
56505		int64x2_t __ret_636; \
56506		int64x2_t __s0_636 = __p0_636; \
56507		int32x2_t __s1_636 = __p1_636; \
56508		int32x4_t __s2_636 = __p2_636; \
56509		__ret_636 = vqdmlal_s32(__s0_636, __s1_636, splat_laneq_s32(__s2_636, __p3_636)); \
56510		__ret_636; \
	58162	#define vqdmlal_laneq_s32(__p0_652, __p1_652, __p2_652, __p3_652) __extension__ ({ \
	58163	int64x2_t __ret_652; \
	58164	int64x2_t __s0_652 = __p0_652; \
	58165	int32x2_t __s1_652 = __p1_652; \
	58166	int32x4_t __s2_652 = __p2_652; \
	58167	__ret_652 = vqdmlal_s32(__s0_652, __s1_652, splat_laneq_s32(__s2_652, __p3_652)); \
	58168	__ret_652; \
56511	58169	})
56512	58170	#else
56513		#define vqdmlal_laneq_s32(__p0_637, __p1_637, __p2_637, __p3_637) __extension__ ({ \
56514		int64x2_t __ret_637; \
56515		int64x2_t __s0_637 = __p0_637; \
56516		int32x2_t __s1_637 = __p1_637; \
56517		int32x4_t __s2_637 = __p2_637; \
56518		int64x2_t __rev0_637; __rev0_637 = __builtin_shufflevector(__s0_637, __s0_637, 1, 0); \
56519		int32x2_t __rev1_637; __rev1_637 = __builtin_shufflevector(__s1_637, __s1_637, 1, 0); \
56520		int32x4_t __rev2_637; __rev2_637 = __builtin_shufflevector(__s2_637, __s2_637, 3, 2, 1, 0); \
56521		__ret_637 = __noswap_vqdmlal_s32(__rev0_637, __rev1_637, __noswap_splat_laneq_s32(__rev2_637, __p3_637)); \
56522		__ret_637 = __builtin_shufflevector(__ret_637, __ret_637, 1, 0); \
56523		__ret_637; \
	58171	#define vqdmlal_laneq_s32(__p0_653, __p1_653, __p2_653, __p3_653) __extension__ ({ \
	58172	int64x2_t __ret_653; \
	58173	int64x2_t __s0_653 = __p0_653; \
	58174	int32x2_t __s1_653 = __p1_653; \
	58175	int32x4_t __s2_653 = __p2_653; \
	58176	int64x2_t __rev0_653; __rev0_653 = __builtin_shufflevector(__s0_653, __s0_653, __lane_reverse_128_64); \
	58177	int32x2_t __rev1_653; __rev1_653 = __builtin_shufflevector(__s1_653, __s1_653, __lane_reverse_64_32); \
	58178	int32x4_t __rev2_653; __rev2_653 = __builtin_shufflevector(__s2_653, __s2_653, __lane_reverse_128_32); \
	58179	__ret_653 = __noswap_vqdmlal_s32(__rev0_653, __rev1_653, __noswap_splat_laneq_s32(__rev2_653, __p3_653)); \
	58180	__ret_653 = __builtin_shufflevector(__ret_653, __ret_653, __lane_reverse_128_64); \
	58181	__ret_653; \
56524	58182	})
56525	58183	#endif
56526	58184
56527	58185	#ifdef __LITTLE_ENDIAN__
56528		#define vqdmlal_laneq_s16(__p0_638, __p1_638, __p2_638, __p3_638) __extension__ ({ \
56529		int32x4_t __ret_638; \
56530		int32x4_t __s0_638 = __p0_638; \
56531		int16x4_t __s1_638 = __p1_638; \
56532		int16x8_t __s2_638 = __p2_638; \
56533		__ret_638 = vqdmlal_s16(__s0_638, __s1_638, splat_laneq_s16(__s2_638, __p3_638)); \
56534		__ret_638; \
	58186	#define vqdmlal_laneq_s16(__p0_654, __p1_654, __p2_654, __p3_654) __extension__ ({ \
	58187	int32x4_t __ret_654; \
	58188	int32x4_t __s0_654 = __p0_654; \
	58189	int16x4_t __s1_654 = __p1_654; \
	58190	int16x8_t __s2_654 = __p2_654; \
	58191	__ret_654 = vqdmlal_s16(__s0_654, __s1_654, splat_laneq_s16(__s2_654, __p3_654)); \
	58192	__ret_654; \
56535	58193	})
56536	58194	#else
56537		#define vqdmlal_laneq_s16(__p0_639, __p1_639, __p2_639, __p3_639) __extension__ ({ \
56538		int32x4_t __ret_639; \
56539		int32x4_t __s0_639 = __p0_639; \
56540		int16x4_t __s1_639 = __p1_639; \
56541		int16x8_t __s2_639 = __p2_639; \
56542		int32x4_t __rev0_639; __rev0_639 = __builtin_shufflevector(__s0_639, __s0_639, 3, 2, 1, 0); \
56543		int16x4_t __rev1_639; __rev1_639 = __builtin_shufflevector(__s1_639, __s1_639, 3, 2, 1, 0); \
56544		int16x8_t __rev2_639; __rev2_639 = __builtin_shufflevector(__s2_639, __s2_639, 7, 6, 5, 4, 3, 2, 1, 0); \
56545		__ret_639 = __noswap_vqdmlal_s16(__rev0_639, __rev1_639, __noswap_splat_laneq_s16(__rev2_639, __p3_639)); \
56546		__ret_639 = __builtin_shufflevector(__ret_639, __ret_639, 3, 2, 1, 0); \
56547		__ret_639; \
	58195	#define vqdmlal_laneq_s16(__p0_655, __p1_655, __p2_655, __p3_655) __extension__ ({ \
	58196	int32x4_t __ret_655; \
	58197	int32x4_t __s0_655 = __p0_655; \
	58198	int16x4_t __s1_655 = __p1_655; \
	58199	int16x8_t __s2_655 = __p2_655; \
	58200	int32x4_t __rev0_655; __rev0_655 = __builtin_shufflevector(__s0_655, __s0_655, __lane_reverse_128_32); \
	58201	int16x4_t __rev1_655; __rev1_655 = __builtin_shufflevector(__s1_655, __s1_655, __lane_reverse_64_16); \
	58202	int16x8_t __rev2_655; __rev2_655 = __builtin_shufflevector(__s2_655, __s2_655, __lane_reverse_128_16); \
	58203	__ret_655 = __noswap_vqdmlal_s16(__rev0_655, __rev1_655, __noswap_splat_laneq_s16(__rev2_655, __p3_655)); \
	58204	__ret_655 = __builtin_shufflevector(__ret_655, __ret_655, __lane_reverse_128_32); \
	58205	__ret_655; \
56548	58206	})
56549	58207	#endif
56550	58208
56551	58209	__ai __attribute__((target("neon"))) int64_t vqdmlsls_s32(int64_t __p0, int32_t __p1, int32_t __p2) {
56552	58210	int64_t __ret;
56553		__ret = (int64_t) __builtin_neon_vqdmlsls_s32(__p0, __p1, __p2);
	58211	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vqdmlsls_s32(__p0, __p1, __p2));
56554	58212	return __ret;
56555	58213	}
56556	58214	__ai __attribute__((target("neon"))) int32_t vqdmlslh_s16(int32_t __p0, int16_t __p1, int16_t __p2) {
56557	58215	int32_t __ret;
56558		__ret = (int32_t) __builtin_neon_vqdmlslh_s16(__p0, __p1, __p2);
	58216	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vqdmlslh_s16(__p0, __p1, __p2));
56559	58217	return __ret;
56560	58218	}
56561	58219	#ifdef __LITTLE_ENDIAN__
...	...	@@ -56567,11 +58225,11 @@ __ai __attribute__((target("neon"))) int64x2_t vqdmlsl_high_s32(int64x2_t __p0,
56567	58225	#else
56568	58226	__ai __attribute__((target("neon"))) int64x2_t vqdmlsl_high_s32(int64x2_t __p0, int32x4_t __p1, int32x4_t __p2) {
56569	58227	int64x2_t __ret;
56570		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
56571		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
56572		int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	58228	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	58229	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	58230	int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
56573	58231	__ret = __noswap_vqdmlsl_s32(__rev0, __noswap_vget_high_s32(__rev1), __noswap_vget_high_s32(__rev2));
56574		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	58232	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
56575	58233	return __ret;
56576	58234	}
56577	58235	#endif
...	...	@@ -56585,108 +58243,108 @@ __ai __attribute__((target("neon"))) int32x4_t vqdmlsl_high_s16(int32x4_t __p0,
56585	58243	#else
56586	58244	__ai __attribute__((target("neon"))) int32x4_t vqdmlsl_high_s16(int32x4_t __p0, int16x8_t __p1, int16x8_t __p2) {
56587	58245	int32x4_t __ret;
56588		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
56589		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
56590		int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	58246	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	58247	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	58248	int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
56591	58249	__ret = __noswap_vqdmlsl_s16(__rev0, __noswap_vget_high_s16(__rev1), __noswap_vget_high_s16(__rev2));
56592		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	58250	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
56593	58251	return __ret;
56594	58252	}
56595	58253	#endif
56596	58254
56597	58255	#ifdef __LITTLE_ENDIAN__
56598		#define vqdmlsl_high_lane_s32(__p0_640, __p1_640, __p2_640, __p3_640) __extension__ ({ \
56599		int64x2_t __ret_640; \
56600		int64x2_t __s0_640 = __p0_640; \
56601		int32x4_t __s1_640 = __p1_640; \
56602		int32x2_t __s2_640 = __p2_640; \
56603		__ret_640 = vqdmlsl_s32(__s0_640, vget_high_s32(__s1_640), splat_lane_s32(__s2_640, __p3_640)); \
56604		__ret_640; \
	58256	#define vqdmlsl_high_lane_s32(__p0_656, __p1_656, __p2_656, __p3_656) __extension__ ({ \
	58257	int64x2_t __ret_656; \
	58258	int64x2_t __s0_656 = __p0_656; \
	58259	int32x4_t __s1_656 = __p1_656; \
	58260	int32x2_t __s2_656 = __p2_656; \
	58261	__ret_656 = vqdmlsl_s32(__s0_656, vget_high_s32(__s1_656), splat_lane_s32(__s2_656, __p3_656)); \
	58262	__ret_656; \
56605	58263	})
56606	58264	#else
56607		#define vqdmlsl_high_lane_s32(__p0_641, __p1_641, __p2_641, __p3_641) __extension__ ({ \
56608		int64x2_t __ret_641; \
56609		int64x2_t __s0_641 = __p0_641; \
56610		int32x4_t __s1_641 = __p1_641; \
56611		int32x2_t __s2_641 = __p2_641; \
56612		int64x2_t __rev0_641; __rev0_641 = __builtin_shufflevector(__s0_641, __s0_641, 1, 0); \
56613		int32x4_t __rev1_641; __rev1_641 = __builtin_shufflevector(__s1_641, __s1_641, 3, 2, 1, 0); \
56614		int32x2_t __rev2_641; __rev2_641 = __builtin_shufflevector(__s2_641, __s2_641, 1, 0); \
56615		__ret_641 = __noswap_vqdmlsl_s32(__rev0_641, __noswap_vget_high_s32(__rev1_641), __noswap_splat_lane_s32(__rev2_641, __p3_641)); \
56616		__ret_641 = __builtin_shufflevector(__ret_641, __ret_641, 1, 0); \
56617		__ret_641; \
	58265	#define vqdmlsl_high_lane_s32(__p0_657, __p1_657, __p2_657, __p3_657) __extension__ ({ \
	58266	int64x2_t __ret_657; \
	58267	int64x2_t __s0_657 = __p0_657; \
	58268	int32x4_t __s1_657 = __p1_657; \
	58269	int32x2_t __s2_657 = __p2_657; \
	58270	int64x2_t __rev0_657; __rev0_657 = __builtin_shufflevector(__s0_657, __s0_657, __lane_reverse_128_64); \
	58271	int32x4_t __rev1_657; __rev1_657 = __builtin_shufflevector(__s1_657, __s1_657, __lane_reverse_128_32); \
	58272	int32x2_t __rev2_657; __rev2_657 = __builtin_shufflevector(__s2_657, __s2_657, __lane_reverse_64_32); \
	58273	__ret_657 = __noswap_vqdmlsl_s32(__rev0_657, __noswap_vget_high_s32(__rev1_657), __noswap_splat_lane_s32(__rev2_657, __p3_657)); \
	58274	__ret_657 = __builtin_shufflevector(__ret_657, __ret_657, __lane_reverse_128_64); \
	58275	__ret_657; \
56618	58276	})
56619	58277	#endif
56620	58278
56621	58279	#ifdef __LITTLE_ENDIAN__
56622		#define vqdmlsl_high_lane_s16(__p0_642, __p1_642, __p2_642, __p3_642) __extension__ ({ \
56623		int32x4_t __ret_642; \
56624		int32x4_t __s0_642 = __p0_642; \
56625		int16x8_t __s1_642 = __p1_642; \
56626		int16x4_t __s2_642 = __p2_642; \
56627		__ret_642 = vqdmlsl_s16(__s0_642, vget_high_s16(__s1_642), splat_lane_s16(__s2_642, __p3_642)); \
56628		__ret_642; \
	58280	#define vqdmlsl_high_lane_s16(__p0_658, __p1_658, __p2_658, __p3_658) __extension__ ({ \
	58281	int32x4_t __ret_658; \
	58282	int32x4_t __s0_658 = __p0_658; \
	58283	int16x8_t __s1_658 = __p1_658; \
	58284	int16x4_t __s2_658 = __p2_658; \
	58285	__ret_658 = vqdmlsl_s16(__s0_658, vget_high_s16(__s1_658), splat_lane_s16(__s2_658, __p3_658)); \
	58286	__ret_658; \
56629	58287	})
56630	58288	#else
56631		#define vqdmlsl_high_lane_s16(__p0_643, __p1_643, __p2_643, __p3_643) __extension__ ({ \
56632		int32x4_t __ret_643; \
56633		int32x4_t __s0_643 = __p0_643; \
56634		int16x8_t __s1_643 = __p1_643; \
56635		int16x4_t __s2_643 = __p2_643; \
56636		int32x4_t __rev0_643; __rev0_643 = __builtin_shufflevector(__s0_643, __s0_643, 3, 2, 1, 0); \
56637		int16x8_t __rev1_643; __rev1_643 = __builtin_shufflevector(__s1_643, __s1_643, 7, 6, 5, 4, 3, 2, 1, 0); \
56638		int16x4_t __rev2_643; __rev2_643 = __builtin_shufflevector(__s2_643, __s2_643, 3, 2, 1, 0); \
56639		__ret_643 = __noswap_vqdmlsl_s16(__rev0_643, __noswap_vget_high_s16(__rev1_643), __noswap_splat_lane_s16(__rev2_643, __p3_643)); \
56640		__ret_643 = __builtin_shufflevector(__ret_643, __ret_643, 3, 2, 1, 0); \
56641		__ret_643; \
	58289	#define vqdmlsl_high_lane_s16(__p0_659, __p1_659, __p2_659, __p3_659) __extension__ ({ \
	58290	int32x4_t __ret_659; \
	58291	int32x4_t __s0_659 = __p0_659; \
	58292	int16x8_t __s1_659 = __p1_659; \
	58293	int16x4_t __s2_659 = __p2_659; \
	58294	int32x4_t __rev0_659; __rev0_659 = __builtin_shufflevector(__s0_659, __s0_659, __lane_reverse_128_32); \
	58295	int16x8_t __rev1_659; __rev1_659 = __builtin_shufflevector(__s1_659, __s1_659, __lane_reverse_128_16); \
	58296	int16x4_t __rev2_659; __rev2_659 = __builtin_shufflevector(__s2_659, __s2_659, __lane_reverse_64_16); \
	58297	__ret_659 = __noswap_vqdmlsl_s16(__rev0_659, __noswap_vget_high_s16(__rev1_659), __noswap_splat_lane_s16(__rev2_659, __p3_659)); \
	58298	__ret_659 = __builtin_shufflevector(__ret_659, __ret_659, __lane_reverse_128_32); \
	58299	__ret_659; \
56642	58300	})
56643	58301	#endif
56644	58302
56645	58303	#ifdef __LITTLE_ENDIAN__
56646		#define vqdmlsl_high_laneq_s32(__p0_644, __p1_644, __p2_644, __p3_644) __extension__ ({ \
56647		int64x2_t __ret_644; \
56648		int64x2_t __s0_644 = __p0_644; \
56649		int32x4_t __s1_644 = __p1_644; \
56650		int32x4_t __s2_644 = __p2_644; \
56651		__ret_644 = vqdmlsl_s32(__s0_644, vget_high_s32(__s1_644), splat_laneq_s32(__s2_644, __p3_644)); \
56652		__ret_644; \
	58304	#define vqdmlsl_high_laneq_s32(__p0_660, __p1_660, __p2_660, __p3_660) __extension__ ({ \
	58305	int64x2_t __ret_660; \
	58306	int64x2_t __s0_660 = __p0_660; \
	58307	int32x4_t __s1_660 = __p1_660; \
	58308	int32x4_t __s2_660 = __p2_660; \
	58309	__ret_660 = vqdmlsl_s32(__s0_660, vget_high_s32(__s1_660), splat_laneq_s32(__s2_660, __p3_660)); \
	58310	__ret_660; \
56653	58311	})
56654	58312	#else
56655		#define vqdmlsl_high_laneq_s32(__p0_645, __p1_645, __p2_645, __p3_645) __extension__ ({ \
56656		int64x2_t __ret_645; \
56657		int64x2_t __s0_645 = __p0_645; \
56658		int32x4_t __s1_645 = __p1_645; \
56659		int32x4_t __s2_645 = __p2_645; \
56660		int64x2_t __rev0_645; __rev0_645 = __builtin_shufflevector(__s0_645, __s0_645, 1, 0); \
56661		int32x4_t __rev1_645; __rev1_645 = __builtin_shufflevector(__s1_645, __s1_645, 3, 2, 1, 0); \
56662		int32x4_t __rev2_645; __rev2_645 = __builtin_shufflevector(__s2_645, __s2_645, 3, 2, 1, 0); \
56663		__ret_645 = __noswap_vqdmlsl_s32(__rev0_645, __noswap_vget_high_s32(__rev1_645), __noswap_splat_laneq_s32(__rev2_645, __p3_645)); \
56664		__ret_645 = __builtin_shufflevector(__ret_645, __ret_645, 1, 0); \
56665		__ret_645; \
	58313	#define vqdmlsl_high_laneq_s32(__p0_661, __p1_661, __p2_661, __p3_661) __extension__ ({ \
	58314	int64x2_t __ret_661; \
	58315	int64x2_t __s0_661 = __p0_661; \
	58316	int32x4_t __s1_661 = __p1_661; \
	58317	int32x4_t __s2_661 = __p2_661; \
	58318	int64x2_t __rev0_661; __rev0_661 = __builtin_shufflevector(__s0_661, __s0_661, __lane_reverse_128_64); \
	58319	int32x4_t __rev1_661; __rev1_661 = __builtin_shufflevector(__s1_661, __s1_661, __lane_reverse_128_32); \
	58320	int32x4_t __rev2_661; __rev2_661 = __builtin_shufflevector(__s2_661, __s2_661, __lane_reverse_128_32); \
	58321	__ret_661 = __noswap_vqdmlsl_s32(__rev0_661, __noswap_vget_high_s32(__rev1_661), __noswap_splat_laneq_s32(__rev2_661, __p3_661)); \
	58322	__ret_661 = __builtin_shufflevector(__ret_661, __ret_661, __lane_reverse_128_64); \
	58323	__ret_661; \
56666	58324	})
56667	58325	#endif
56668	58326
56669	58327	#ifdef __LITTLE_ENDIAN__
56670		#define vqdmlsl_high_laneq_s16(__p0_646, __p1_646, __p2_646, __p3_646) __extension__ ({ \
56671		int32x4_t __ret_646; \
56672		int32x4_t __s0_646 = __p0_646; \
56673		int16x8_t __s1_646 = __p1_646; \
56674		int16x8_t __s2_646 = __p2_646; \
56675		__ret_646 = vqdmlsl_s16(__s0_646, vget_high_s16(__s1_646), splat_laneq_s16(__s2_646, __p3_646)); \
56676		__ret_646; \
	58328	#define vqdmlsl_high_laneq_s16(__p0_662, __p1_662, __p2_662, __p3_662) __extension__ ({ \
	58329	int32x4_t __ret_662; \
	58330	int32x4_t __s0_662 = __p0_662; \
	58331	int16x8_t __s1_662 = __p1_662; \
	58332	int16x8_t __s2_662 = __p2_662; \
	58333	__ret_662 = vqdmlsl_s16(__s0_662, vget_high_s16(__s1_662), splat_laneq_s16(__s2_662, __p3_662)); \
	58334	__ret_662; \
56677	58335	})
56678	58336	#else
56679		#define vqdmlsl_high_laneq_s16(__p0_647, __p1_647, __p2_647, __p3_647) __extension__ ({ \
56680		int32x4_t __ret_647; \
56681		int32x4_t __s0_647 = __p0_647; \
56682		int16x8_t __s1_647 = __p1_647; \
56683		int16x8_t __s2_647 = __p2_647; \
56684		int32x4_t __rev0_647; __rev0_647 = __builtin_shufflevector(__s0_647, __s0_647, 3, 2, 1, 0); \
56685		int16x8_t __rev1_647; __rev1_647 = __builtin_shufflevector(__s1_647, __s1_647, 7, 6, 5, 4, 3, 2, 1, 0); \
56686		int16x8_t __rev2_647; __rev2_647 = __builtin_shufflevector(__s2_647, __s2_647, 7, 6, 5, 4, 3, 2, 1, 0); \
56687		__ret_647 = __noswap_vqdmlsl_s16(__rev0_647, __noswap_vget_high_s16(__rev1_647), __noswap_splat_laneq_s16(__rev2_647, __p3_647)); \
56688		__ret_647 = __builtin_shufflevector(__ret_647, __ret_647, 3, 2, 1, 0); \
56689		__ret_647; \
	58337	#define vqdmlsl_high_laneq_s16(__p0_663, __p1_663, __p2_663, __p3_663) __extension__ ({ \
	58338	int32x4_t __ret_663; \
	58339	int32x4_t __s0_663 = __p0_663; \
	58340	int16x8_t __s1_663 = __p1_663; \
	58341	int16x8_t __s2_663 = __p2_663; \
	58342	int32x4_t __rev0_663; __rev0_663 = __builtin_shufflevector(__s0_663, __s0_663, __lane_reverse_128_32); \
	58343	int16x8_t __rev1_663; __rev1_663 = __builtin_shufflevector(__s1_663, __s1_663, __lane_reverse_128_16); \
	58344	int16x8_t __rev2_663; __rev2_663 = __builtin_shufflevector(__s2_663, __s2_663, __lane_reverse_128_16); \
	58345	__ret_663 = __noswap_vqdmlsl_s16(__rev0_663, __noswap_vget_high_s16(__rev1_663), __noswap_splat_laneq_s16(__rev2_663, __p3_663)); \
	58346	__ret_663 = __builtin_shufflevector(__ret_663, __ret_663, __lane_reverse_128_32); \
	58347	__ret_663; \
56690	58348	})
56691	58349	#endif
56692	58350
...	...	@@ -56699,10 +58357,10 @@ __ai __attribute__((target("neon"))) int64x2_t vqdmlsl_high_n_s32(int64x2_t __p0
56699	58357	#else
56700	58358	__ai __attribute__((target("neon"))) int64x2_t vqdmlsl_high_n_s32(int64x2_t __p0, int32x4_t __p1, int32_t __p2) {
56701	58359	int64x2_t __ret;
56702		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
56703		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	58360	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	58361	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
56704	58362	__ret = __noswap_vqdmlsl_n_s32(__rev0, __noswap_vget_high_s32(__rev1), __p2);
56705		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	58363	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
56706	58364	return __ret;
56707	58365	}
56708	58366	#endif
...	...	@@ -56716,10 +58374,10 @@ __ai __attribute__((target("neon"))) int32x4_t vqdmlsl_high_n_s16(int32x4_t __p0
56716	58374	#else
56717	58375	__ai __attribute__((target("neon"))) int32x4_t vqdmlsl_high_n_s16(int32x4_t __p0, int16x8_t __p1, int16_t __p2) {
56718	58376	int32x4_t __ret;
56719		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
56720		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	58377	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	58378	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
56721	58379	__ret = __noswap_vqdmlsl_n_s16(__rev0, __noswap_vget_high_s16(__rev1), __p2);
56722		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	58380	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
56723	58381	return __ret;
56724	58382	}
56725	58383	#endif
...	...	@@ -56730,7 +58388,7 @@ __ai __attribute__((target("neon"))) int32x4_t vqdmlsl_high_n_s16(int32x4_t __p0
56730	58388	int64_t __s0 = __p0; \
56731	58389	int32_t __s1 = __p1; \
56732	58390	int32x2_t __s2 = __p2; \
56733		__ret = (int64_t) __builtin_neon_vqdmlsls_lane_s32(__s0, __s1, __s2, __p3); \
	58391	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vqdmlsls_lane_s32(__s0, __s1, __s2, __p3)); \
56734	58392	__ret; \
56735	58393	})
56736	58394	#else
...	...	@@ -56739,8 +58397,8 @@ __ai __attribute__((target("neon"))) int32x4_t vqdmlsl_high_n_s16(int32x4_t __p0
56739	58397	int64_t __s0 = __p0; \
56740	58398	int32_t __s1 = __p1; \
56741	58399	int32x2_t __s2 = __p2; \
56742		int32x2_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 1, 0); \
56743		__ret = (int64_t) __builtin_neon_vqdmlsls_lane_s32(__s0, __s1, __rev2, __p3); \
	58400	int32x2_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_64_32); \
	58401	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vqdmlsls_lane_s32(__s0, __s1, __rev2, __p3)); \
56744	58402	__ret; \
56745	58403	})
56746	58404	#endif
...	...	@@ -56751,7 +58409,7 @@ __ai __attribute__((target("neon"))) int32x4_t vqdmlsl_high_n_s16(int32x4_t __p0
56751	58409	int32_t __s0 = __p0; \
56752	58410	int16_t __s1 = __p1; \
56753	58411	int16x4_t __s2 = __p2; \
56754		__ret = (int32_t) __builtin_neon_vqdmlslh_lane_s16(__s0, __s1, __s2, __p3); \
	58412	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vqdmlslh_lane_s16(__s0, __s1, __s2, __p3)); \
56755	58413	__ret; \
56756	58414	})
56757	58415	#else
...	...	@@ -56760,8 +58418,8 @@ __ai __attribute__((target("neon"))) int32x4_t vqdmlsl_high_n_s16(int32x4_t __p0
56760	58418	int32_t __s0 = __p0; \
56761	58419	int16_t __s1 = __p1; \
56762	58420	int16x4_t __s2 = __p2; \
56763		int16x4_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 3, 2, 1, 0); \
56764		__ret = (int32_t) __builtin_neon_vqdmlslh_lane_s16(__s0, __s1, __rev2, __p3); \
	58421	int16x4_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_64_16); \
	58422	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vqdmlslh_lane_s16(__s0, __s1, __rev2, __p3)); \
56765	58423	__ret; \
56766	58424	})
56767	58425	#endif
...	...	@@ -56772,7 +58430,7 @@ __ai __attribute__((target("neon"))) int32x4_t vqdmlsl_high_n_s16(int32x4_t __p0
56772	58430	int64_t __s0 = __p0; \
56773	58431	int32_t __s1 = __p1; \
56774	58432	int32x4_t __s2 = __p2; \
56775		__ret = (int64_t) __builtin_neon_vqdmlsls_laneq_s32(__s0, __s1, __s2, __p3); \
	58433	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vqdmlsls_laneq_s32(__s0, __s1, __s2, __p3)); \
56776	58434	__ret; \
56777	58435	})
56778	58436	#else
...	...	@@ -56781,8 +58439,8 @@ __ai __attribute__((target("neon"))) int32x4_t vqdmlsl_high_n_s16(int32x4_t __p0
56781	58439	int64_t __s0 = __p0; \
56782	58440	int32_t __s1 = __p1; \
56783	58441	int32x4_t __s2 = __p2; \
56784		int32x4_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 3, 2, 1, 0); \
56785		__ret = (int64_t) __builtin_neon_vqdmlsls_laneq_s32(__s0, __s1, __rev2, __p3); \
	58442	int32x4_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_128_32); \
	58443	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vqdmlsls_laneq_s32(__s0, __s1, __rev2, __p3)); \
56786	58444	__ret; \
56787	58445	})
56788	58446	#endif
...	...	@@ -56793,7 +58451,7 @@ __ai __attribute__((target("neon"))) int32x4_t vqdmlsl_high_n_s16(int32x4_t __p0
56793	58451	int32_t __s0 = __p0; \
56794	58452	int16_t __s1 = __p1; \
56795	58453	int16x8_t __s2 = __p2; \
56796		__ret = (int32_t) __builtin_neon_vqdmlslh_laneq_s16(__s0, __s1, __s2, __p3); \
	58454	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vqdmlslh_laneq_s16(__s0, __s1, __s2, __p3)); \
56797	58455	__ret; \
56798	58456	})
56799	58457	#else
...	...	@@ -56802,68 +58460,68 @@ __ai __attribute__((target("neon"))) int32x4_t vqdmlsl_high_n_s16(int32x4_t __p0
56802	58460	int32_t __s0 = __p0; \
56803	58461	int16_t __s1 = __p1; \
56804	58462	int16x8_t __s2 = __p2; \
56805		int16x8_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 7, 6, 5, 4, 3, 2, 1, 0); \
56806		__ret = (int32_t) __builtin_neon_vqdmlslh_laneq_s16(__s0, __s1, __rev2, __p3); \
	58463	int16x8_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_128_16); \
	58464	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vqdmlslh_laneq_s16(__s0, __s1, __rev2, __p3)); \
56807	58465	__ret; \
56808	58466	})
56809	58467	#endif
56810	58468
56811	58469	#ifdef __LITTLE_ENDIAN__
56812		#define vqdmlsl_laneq_s32(__p0_648, __p1_648, __p2_648, __p3_648) __extension__ ({ \
56813		int64x2_t __ret_648; \
56814		int64x2_t __s0_648 = __p0_648; \
56815		int32x2_t __s1_648 = __p1_648; \
56816		int32x4_t __s2_648 = __p2_648; \
56817		__ret_648 = vqdmlsl_s32(__s0_648, __s1_648, splat_laneq_s32(__s2_648, __p3_648)); \
56818		__ret_648; \
	58470	#define vqdmlsl_laneq_s32(__p0_664, __p1_664, __p2_664, __p3_664) __extension__ ({ \
	58471	int64x2_t __ret_664; \
	58472	int64x2_t __s0_664 = __p0_664; \
	58473	int32x2_t __s1_664 = __p1_664; \
	58474	int32x4_t __s2_664 = __p2_664; \
	58475	__ret_664 = vqdmlsl_s32(__s0_664, __s1_664, splat_laneq_s32(__s2_664, __p3_664)); \
	58476	__ret_664; \
56819	58477	})
56820	58478	#else
56821		#define vqdmlsl_laneq_s32(__p0_649, __p1_649, __p2_649, __p3_649) __extension__ ({ \
56822		int64x2_t __ret_649; \
56823		int64x2_t __s0_649 = __p0_649; \
56824		int32x2_t __s1_649 = __p1_649; \
56825		int32x4_t __s2_649 = __p2_649; \
56826		int64x2_t __rev0_649; __rev0_649 = __builtin_shufflevector(__s0_649, __s0_649, 1, 0); \
56827		int32x2_t __rev1_649; __rev1_649 = __builtin_shufflevector(__s1_649, __s1_649, 1, 0); \
56828		int32x4_t __rev2_649; __rev2_649 = __builtin_shufflevector(__s2_649, __s2_649, 3, 2, 1, 0); \
56829		__ret_649 = __noswap_vqdmlsl_s32(__rev0_649, __rev1_649, __noswap_splat_laneq_s32(__rev2_649, __p3_649)); \
56830		__ret_649 = __builtin_shufflevector(__ret_649, __ret_649, 1, 0); \
56831		__ret_649; \
	58479	#define vqdmlsl_laneq_s32(__p0_665, __p1_665, __p2_665, __p3_665) __extension__ ({ \
	58480	int64x2_t __ret_665; \
	58481	int64x2_t __s0_665 = __p0_665; \
	58482	int32x2_t __s1_665 = __p1_665; \
	58483	int32x4_t __s2_665 = __p2_665; \
	58484	int64x2_t __rev0_665; __rev0_665 = __builtin_shufflevector(__s0_665, __s0_665, __lane_reverse_128_64); \
	58485	int32x2_t __rev1_665; __rev1_665 = __builtin_shufflevector(__s1_665, __s1_665, __lane_reverse_64_32); \
	58486	int32x4_t __rev2_665; __rev2_665 = __builtin_shufflevector(__s2_665, __s2_665, __lane_reverse_128_32); \
	58487	__ret_665 = __noswap_vqdmlsl_s32(__rev0_665, __rev1_665, __noswap_splat_laneq_s32(__rev2_665, __p3_665)); \
	58488	__ret_665 = __builtin_shufflevector(__ret_665, __ret_665, __lane_reverse_128_64); \
	58489	__ret_665; \
56832	58490	})
56833	58491	#endif
56834	58492
56835	58493	#ifdef __LITTLE_ENDIAN__
56836		#define vqdmlsl_laneq_s16(__p0_650, __p1_650, __p2_650, __p3_650) __extension__ ({ \
56837		int32x4_t __ret_650; \
56838		int32x4_t __s0_650 = __p0_650; \
56839		int16x4_t __s1_650 = __p1_650; \
56840		int16x8_t __s2_650 = __p2_650; \
56841		__ret_650 = vqdmlsl_s16(__s0_650, __s1_650, splat_laneq_s16(__s2_650, __p3_650)); \
56842		__ret_650; \
	58494	#define vqdmlsl_laneq_s16(__p0_666, __p1_666, __p2_666, __p3_666) __extension__ ({ \
	58495	int32x4_t __ret_666; \
	58496	int32x4_t __s0_666 = __p0_666; \
	58497	int16x4_t __s1_666 = __p1_666; \
	58498	int16x8_t __s2_666 = __p2_666; \
	58499	__ret_666 = vqdmlsl_s16(__s0_666, __s1_666, splat_laneq_s16(__s2_666, __p3_666)); \
	58500	__ret_666; \
56843	58501	})
56844	58502	#else
56845		#define vqdmlsl_laneq_s16(__p0_651, __p1_651, __p2_651, __p3_651) __extension__ ({ \
56846		int32x4_t __ret_651; \
56847		int32x4_t __s0_651 = __p0_651; \
56848		int16x4_t __s1_651 = __p1_651; \
56849		int16x8_t __s2_651 = __p2_651; \
56850		int32x4_t __rev0_651; __rev0_651 = __builtin_shufflevector(__s0_651, __s0_651, 3, 2, 1, 0); \
56851		int16x4_t __rev1_651; __rev1_651 = __builtin_shufflevector(__s1_651, __s1_651, 3, 2, 1, 0); \
56852		int16x8_t __rev2_651; __rev2_651 = __builtin_shufflevector(__s2_651, __s2_651, 7, 6, 5, 4, 3, 2, 1, 0); \
56853		__ret_651 = __noswap_vqdmlsl_s16(__rev0_651, __rev1_651, __noswap_splat_laneq_s16(__rev2_651, __p3_651)); \
56854		__ret_651 = __builtin_shufflevector(__ret_651, __ret_651, 3, 2, 1, 0); \
56855		__ret_651; \
	58503	#define vqdmlsl_laneq_s16(__p0_667, __p1_667, __p2_667, __p3_667) __extension__ ({ \
	58504	int32x4_t __ret_667; \
	58505	int32x4_t __s0_667 = __p0_667; \
	58506	int16x4_t __s1_667 = __p1_667; \
	58507	int16x8_t __s2_667 = __p2_667; \
	58508	int32x4_t __rev0_667; __rev0_667 = __builtin_shufflevector(__s0_667, __s0_667, __lane_reverse_128_32); \
	58509	int16x4_t __rev1_667; __rev1_667 = __builtin_shufflevector(__s1_667, __s1_667, __lane_reverse_64_16); \
	58510	int16x8_t __rev2_667; __rev2_667 = __builtin_shufflevector(__s2_667, __s2_667, __lane_reverse_128_16); \
	58511	__ret_667 = __noswap_vqdmlsl_s16(__rev0_667, __rev1_667, __noswap_splat_laneq_s16(__rev2_667, __p3_667)); \
	58512	__ret_667 = __builtin_shufflevector(__ret_667, __ret_667, __lane_reverse_128_32); \
	58513	__ret_667; \
56856	58514	})
56857	58515	#endif
56858	58516
56859	58517	__ai __attribute__((target("neon"))) int32_t vqdmulhs_s32(int32_t __p0, int32_t __p1) {
56860	58518	int32_t __ret;
56861		__ret = (int32_t) __builtin_neon_vqdmulhs_s32(__p0, __p1);
	58519	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vqdmulhs_s32(__p0, __p1));
56862	58520	return __ret;
56863	58521	}
56864	58522	__ai __attribute__((target("neon"))) int16_t vqdmulhh_s16(int16_t __p0, int16_t __p1) {
56865	58523	int16_t __ret;
56866		__ret = (int16_t) __builtin_neon_vqdmulhh_s16(__p0, __p1);
	58524	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vqdmulhh_s16(__p0, __p1));
56867	58525	return __ret;
56868	58526	}
56869	58527	#ifdef __LITTLE_ENDIAN__
...	...	@@ -56871,7 +58529,7 @@ __ai __attribute__((target("neon"))) int16_t vqdmulhh_s16(int16_t __p0, int16_t
56871	58529	int32x4_t __ret; \
56872	58530	int32x4_t __s0 = __p0; \
56873	58531	int32x2_t __s1 = __p1; \
56874		__ret = (int32x4_t) __builtin_neon_vqdmulhq_lane_v((int8x16_t)__s0, (int8x8_t)__s1, __p2, 2); \
	58532	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqdmulhq_lane_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 2)); \
56875	58533	__ret; \
56876	58534	})
56877	58535	#else
...	...	@@ -56879,10 +58537,10 @@ __ai __attribute__((target("neon"))) int16_t vqdmulhh_s16(int16_t __p0, int16_t
56879	58537	int32x4_t __ret; \
56880	58538	int32x4_t __s0 = __p0; \
56881	58539	int32x2_t __s1 = __p1; \
56882		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
56883		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
56884		__ret = (int32x4_t) __builtin_neon_vqdmulhq_lane_v((int8x16_t)__rev0, (int8x8_t)__rev1, __p2, 2); \
56885		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	58540	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	58541	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_32); \
	58542	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqdmulhq_lane_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 2)); \
	58543	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
56886	58544	__ret; \
56887	58545	})
56888	58546	#endif
...	...	@@ -56892,7 +58550,7 @@ __ai __attribute__((target("neon"))) int16_t vqdmulhh_s16(int16_t __p0, int16_t
56892	58550	int16x8_t __ret; \
56893	58551	int16x8_t __s0 = __p0; \
56894	58552	int16x4_t __s1 = __p1; \
56895		__ret = (int16x8_t) __builtin_neon_vqdmulhq_lane_v((int8x16_t)__s0, (int8x8_t)__s1, __p2, 1); \
	58553	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqdmulhq_lane_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 1)); \
56896	58554	__ret; \
56897	58555	})
56898	58556	#else
...	...	@@ -56900,10 +58558,10 @@ __ai __attribute__((target("neon"))) int16_t vqdmulhh_s16(int16_t __p0, int16_t
56900	58558	int16x8_t __ret; \
56901	58559	int16x8_t __s0 = __p0; \
56902	58560	int16x4_t __s1 = __p1; \
56903		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
56904		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
56905		__ret = (int16x8_t) __builtin_neon_vqdmulhq_lane_v((int8x16_t)__rev0, (int8x8_t)__rev1, __p2, 1); \
56906		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	58561	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	58562	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	58563	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqdmulhq_lane_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 1)); \
	58564	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
56907	58565	__ret; \
56908	58566	})
56909	58567	#endif
...	...	@@ -56913,7 +58571,7 @@ __ai __attribute__((target("neon"))) int16_t vqdmulhh_s16(int16_t __p0, int16_t
56913	58571	int32x2_t __ret; \
56914	58572	int32x2_t __s0 = __p0; \
56915	58573	int32x2_t __s1 = __p1; \
56916		__ret = (int32x2_t) __builtin_neon_vqdmulh_lane_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 2); \
	58574	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqdmulh_lane_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 2)); \
56917	58575	__ret; \
56918	58576	})
56919	58577	#else
...	...	@@ -56921,10 +58579,10 @@ __ai __attribute__((target("neon"))) int16_t vqdmulhh_s16(int16_t __p0, int16_t
56921	58579	int32x2_t __ret; \
56922	58580	int32x2_t __s0 = __p0; \
56923	58581	int32x2_t __s1 = __p1; \
56924		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
56925		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
56926		__ret = (int32x2_t) __builtin_neon_vqdmulh_lane_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 2); \
56927		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	58582	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	58583	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_32); \
	58584	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqdmulh_lane_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 2)); \
	58585	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
56928	58586	__ret; \
56929	58587	})
56930	58588	#endif
...	...	@@ -56934,7 +58592,7 @@ __ai __attribute__((target("neon"))) int16_t vqdmulhh_s16(int16_t __p0, int16_t
56934	58592	int16x4_t __ret; \
56935	58593	int16x4_t __s0 = __p0; \
56936	58594	int16x4_t __s1 = __p1; \
56937		__ret = (int16x4_t) __builtin_neon_vqdmulh_lane_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 1); \
	58595	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqdmulh_lane_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 1)); \
56938	58596	__ret; \
56939	58597	})
56940	58598	#else
...	...	@@ -56942,87 +58600,87 @@ __ai __attribute__((target("neon"))) int16_t vqdmulhh_s16(int16_t __p0, int16_t
56942	58600	int16x4_t __ret; \
56943	58601	int16x4_t __s0 = __p0; \
56944	58602	int16x4_t __s1 = __p1; \
56945		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
56946		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
56947		__ret = (int16x4_t) __builtin_neon_vqdmulh_lane_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 1); \
56948		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	58603	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	58604	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	58605	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqdmulh_lane_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 1)); \
	58606	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
56949	58607	__ret; \
56950	58608	})
56951	58609	#endif
56952	58610
56953	58611	#ifdef __LITTLE_ENDIAN__
56954		#define vqdmulhs_lane_s32(__p0_652, __p1_652, __p2_652) __extension__ ({ \
56955		int32_t __ret_652; \
56956		int32_t __s0_652 = __p0_652; \
56957		int32x2_t __s1_652 = __p1_652; \
56958		__ret_652 = vqdmulhs_s32(__s0_652, vget_lane_s32(__s1_652, __p2_652)); \
56959		__ret_652; \
	58612	#define vqdmulhs_lane_s32(__p0_668, __p1_668, __p2_668) __extension__ ({ \
	58613	int32_t __ret_668; \
	58614	int32_t __s0_668 = __p0_668; \
	58615	int32x2_t __s1_668 = __p1_668; \
	58616	__ret_668 = vqdmulhs_s32(__s0_668, vget_lane_s32(__s1_668, __p2_668)); \
	58617	__ret_668; \
56960	58618	})
56961	58619	#else
56962		#define vqdmulhs_lane_s32(__p0_653, __p1_653, __p2_653) __extension__ ({ \
56963		int32_t __ret_653; \
56964		int32_t __s0_653 = __p0_653; \
56965		int32x2_t __s1_653 = __p1_653; \
56966		int32x2_t __rev1_653; __rev1_653 = __builtin_shufflevector(__s1_653, __s1_653, 1, 0); \
56967		__ret_653 = vqdmulhs_s32(__s0_653, __noswap_vget_lane_s32(__rev1_653, __p2_653)); \
56968		__ret_653; \
	58620	#define vqdmulhs_lane_s32(__p0_669, __p1_669, __p2_669) __extension__ ({ \
	58621	int32_t __ret_669; \
	58622	int32_t __s0_669 = __p0_669; \
	58623	int32x2_t __s1_669 = __p1_669; \
	58624	int32x2_t __rev1_669; __rev1_669 = __builtin_shufflevector(__s1_669, __s1_669, __lane_reverse_64_32); \
	58625	__ret_669 = vqdmulhs_s32(__s0_669, __noswap_vget_lane_s32(__rev1_669, __p2_669)); \
	58626	__ret_669; \
56969	58627	})
56970	58628	#endif
56971	58629
56972	58630	#ifdef __LITTLE_ENDIAN__
56973		#define vqdmulhh_lane_s16(__p0_654, __p1_654, __p2_654) __extension__ ({ \
56974		int16_t __ret_654; \
56975		int16_t __s0_654 = __p0_654; \
56976		int16x4_t __s1_654 = __p1_654; \
56977		__ret_654 = vqdmulhh_s16(__s0_654, vget_lane_s16(__s1_654, __p2_654)); \
56978		__ret_654; \
	58631	#define vqdmulhh_lane_s16(__p0_670, __p1_670, __p2_670) __extension__ ({ \
	58632	int16_t __ret_670; \
	58633	int16_t __s0_670 = __p0_670; \
	58634	int16x4_t __s1_670 = __p1_670; \
	58635	__ret_670 = vqdmulhh_s16(__s0_670, vget_lane_s16(__s1_670, __p2_670)); \
	58636	__ret_670; \
56979	58637	})
56980	58638	#else
56981		#define vqdmulhh_lane_s16(__p0_655, __p1_655, __p2_655) __extension__ ({ \
56982		int16_t __ret_655; \
56983		int16_t __s0_655 = __p0_655; \
56984		int16x4_t __s1_655 = __p1_655; \
56985		int16x4_t __rev1_655; __rev1_655 = __builtin_shufflevector(__s1_655, __s1_655, 3, 2, 1, 0); \
56986		__ret_655 = vqdmulhh_s16(__s0_655, __noswap_vget_lane_s16(__rev1_655, __p2_655)); \
56987		__ret_655; \
	58639	#define vqdmulhh_lane_s16(__p0_671, __p1_671, __p2_671) __extension__ ({ \
	58640	int16_t __ret_671; \
	58641	int16_t __s0_671 = __p0_671; \
	58642	int16x4_t __s1_671 = __p1_671; \
	58643	int16x4_t __rev1_671; __rev1_671 = __builtin_shufflevector(__s1_671, __s1_671, __lane_reverse_64_16); \
	58644	__ret_671 = vqdmulhh_s16(__s0_671, __noswap_vget_lane_s16(__rev1_671, __p2_671)); \
	58645	__ret_671; \
56988	58646	})
56989	58647	#endif
56990	58648
56991	58649	#ifdef __LITTLE_ENDIAN__
56992		#define vqdmulhs_laneq_s32(__p0_656, __p1_656, __p2_656) __extension__ ({ \
56993		int32_t __ret_656; \
56994		int32_t __s0_656 = __p0_656; \
56995		int32x4_t __s1_656 = __p1_656; \
56996		__ret_656 = vqdmulhs_s32(__s0_656, vgetq_lane_s32(__s1_656, __p2_656)); \
56997		__ret_656; \
	58650	#define vqdmulhs_laneq_s32(__p0_672, __p1_672, __p2_672) __extension__ ({ \
	58651	int32_t __ret_672; \
	58652	int32_t __s0_672 = __p0_672; \
	58653	int32x4_t __s1_672 = __p1_672; \
	58654	__ret_672 = vqdmulhs_s32(__s0_672, vgetq_lane_s32(__s1_672, __p2_672)); \
	58655	__ret_672; \
56998	58656	})
56999	58657	#else
57000		#define vqdmulhs_laneq_s32(__p0_657, __p1_657, __p2_657) __extension__ ({ \
57001		int32_t __ret_657; \
57002		int32_t __s0_657 = __p0_657; \
57003		int32x4_t __s1_657 = __p1_657; \
57004		int32x4_t __rev1_657; __rev1_657 = __builtin_shufflevector(__s1_657, __s1_657, 3, 2, 1, 0); \
57005		__ret_657 = vqdmulhs_s32(__s0_657, __noswap_vgetq_lane_s32(__rev1_657, __p2_657)); \
57006		__ret_657; \
	58658	#define vqdmulhs_laneq_s32(__p0_673, __p1_673, __p2_673) __extension__ ({ \
	58659	int32_t __ret_673; \
	58660	int32_t __s0_673 = __p0_673; \
	58661	int32x4_t __s1_673 = __p1_673; \
	58662	int32x4_t __rev1_673; __rev1_673 = __builtin_shufflevector(__s1_673, __s1_673, __lane_reverse_128_32); \
	58663	__ret_673 = vqdmulhs_s32(__s0_673, __noswap_vgetq_lane_s32(__rev1_673, __p2_673)); \
	58664	__ret_673; \
57007	58665	})
57008	58666	#endif
57009	58667
57010	58668	#ifdef __LITTLE_ENDIAN__
57011		#define vqdmulhh_laneq_s16(__p0_658, __p1_658, __p2_658) __extension__ ({ \
57012		int16_t __ret_658; \
57013		int16_t __s0_658 = __p0_658; \
57014		int16x8_t __s1_658 = __p1_658; \
57015		__ret_658 = vqdmulhh_s16(__s0_658, vgetq_lane_s16(__s1_658, __p2_658)); \
57016		__ret_658; \
	58669	#define vqdmulhh_laneq_s16(__p0_674, __p1_674, __p2_674) __extension__ ({ \
	58670	int16_t __ret_674; \
	58671	int16_t __s0_674 = __p0_674; \
	58672	int16x8_t __s1_674 = __p1_674; \
	58673	__ret_674 = vqdmulhh_s16(__s0_674, vgetq_lane_s16(__s1_674, __p2_674)); \
	58674	__ret_674; \
57017	58675	})
57018	58676	#else
57019		#define vqdmulhh_laneq_s16(__p0_659, __p1_659, __p2_659) __extension__ ({ \
57020		int16_t __ret_659; \
57021		int16_t __s0_659 = __p0_659; \
57022		int16x8_t __s1_659 = __p1_659; \
57023		int16x8_t __rev1_659; __rev1_659 = __builtin_shufflevector(__s1_659, __s1_659, 7, 6, 5, 4, 3, 2, 1, 0); \
57024		__ret_659 = vqdmulhh_s16(__s0_659, __noswap_vgetq_lane_s16(__rev1_659, __p2_659)); \
57025		__ret_659; \
	58677	#define vqdmulhh_laneq_s16(__p0_675, __p1_675, __p2_675) __extension__ ({ \
	58678	int16_t __ret_675; \
	58679	int16_t __s0_675 = __p0_675; \
	58680	int16x8_t __s1_675 = __p1_675; \
	58681	int16x8_t __rev1_675; __rev1_675 = __builtin_shufflevector(__s1_675, __s1_675, __lane_reverse_128_16); \
	58682	__ret_675 = vqdmulhh_s16(__s0_675, __noswap_vgetq_lane_s16(__rev1_675, __p2_675)); \
	58683	__ret_675; \
57026	58684	})
57027	58685	#endif
57028	58686
...	...	@@ -57031,7 +58689,7 @@ __ai __attribute__((target("neon"))) int16_t vqdmulhh_s16(int16_t __p0, int16_t
57031	58689	int32x4_t __ret; \
57032	58690	int32x4_t __s0 = __p0; \
57033	58691	int32x4_t __s1 = __p1; \
57034		__ret = (int32x4_t) __builtin_neon_vqdmulhq_laneq_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 34); \
	58692	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqdmulhq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 34)); \
57035	58693	__ret; \
57036	58694	})
57037	58695	#else
...	...	@@ -57039,10 +58697,10 @@ __ai __attribute__((target("neon"))) int16_t vqdmulhh_s16(int16_t __p0, int16_t
57039	58697	int32x4_t __ret; \
57040	58698	int32x4_t __s0 = __p0; \
57041	58699	int32x4_t __s1 = __p1; \
57042		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
57043		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
57044		__ret = (int32x4_t) __builtin_neon_vqdmulhq_laneq_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 34); \
57045		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	58700	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	58701	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	58702	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqdmulhq_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 34)); \
	58703	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
57046	58704	__ret; \
57047	58705	})
57048	58706	#endif
...	...	@@ -57052,7 +58710,7 @@ __ai __attribute__((target("neon"))) int16_t vqdmulhh_s16(int16_t __p0, int16_t
57052	58710	int16x8_t __ret; \
57053	58711	int16x8_t __s0 = __p0; \
57054	58712	int16x8_t __s1 = __p1; \
57055		__ret = (int16x8_t) __builtin_neon_vqdmulhq_laneq_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 33); \
	58713	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqdmulhq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 33)); \
57056	58714	__ret; \
57057	58715	})
57058	58716	#else
...	...	@@ -57060,10 +58718,10 @@ __ai __attribute__((target("neon"))) int16_t vqdmulhh_s16(int16_t __p0, int16_t
57060	58718	int16x8_t __ret; \
57061	58719	int16x8_t __s0 = __p0; \
57062	58720	int16x8_t __s1 = __p1; \
57063		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
57064		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
57065		__ret = (int16x8_t) __builtin_neon_vqdmulhq_laneq_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 33); \
57066		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	58721	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	58722	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	58723	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqdmulhq_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 33)); \
	58724	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
57067	58725	__ret; \
57068	58726	})
57069	58727	#endif
...	...	@@ -57073,7 +58731,7 @@ __ai __attribute__((target("neon"))) int16_t vqdmulhh_s16(int16_t __p0, int16_t
57073	58731	int32x2_t __ret; \
57074	58732	int32x2_t __s0 = __p0; \
57075	58733	int32x4_t __s1 = __p1; \
57076		__ret = (int32x2_t) __builtin_neon_vqdmulh_laneq_v((int8x8_t)__s0, (int8x16_t)__s1, __p2, 2); \
	58734	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqdmulh_laneq_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 2)); \
57077	58735	__ret; \
57078	58736	})
57079	58737	#else
...	...	@@ -57081,10 +58739,10 @@ __ai __attribute__((target("neon"))) int16_t vqdmulhh_s16(int16_t __p0, int16_t
57081	58739	int32x2_t __ret; \
57082	58740	int32x2_t __s0 = __p0; \
57083	58741	int32x4_t __s1 = __p1; \
57084		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
57085		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
57086		__ret = (int32x2_t) __builtin_neon_vqdmulh_laneq_v((int8x8_t)__rev0, (int8x16_t)__rev1, __p2, 2); \
57087		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	58742	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	58743	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	58744	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqdmulh_laneq_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 2)); \
	58745	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
57088	58746	__ret; \
57089	58747	})
57090	58748	#endif
...	...	@@ -57094,7 +58752,7 @@ __ai __attribute__((target("neon"))) int16_t vqdmulhh_s16(int16_t __p0, int16_t
57094	58752	int16x4_t __ret; \
57095	58753	int16x4_t __s0 = __p0; \
57096	58754	int16x8_t __s1 = __p1; \
57097		__ret = (int16x4_t) __builtin_neon_vqdmulh_laneq_v((int8x8_t)__s0, (int8x16_t)__s1, __p2, 1); \
	58755	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqdmulh_laneq_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 1)); \
57098	58756	__ret; \
57099	58757	})
57100	58758	#else
...	...	@@ -57102,22 +58760,22 @@ __ai __attribute__((target("neon"))) int16_t vqdmulhh_s16(int16_t __p0, int16_t
57102	58760	int16x4_t __ret; \
57103	58761	int16x4_t __s0 = __p0; \
57104	58762	int16x8_t __s1 = __p1; \
57105		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
57106		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
57107		__ret = (int16x4_t) __builtin_neon_vqdmulh_laneq_v((int8x8_t)__rev0, (int8x16_t)__rev1, __p2, 1); \
57108		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	58763	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	58764	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	58765	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqdmulh_laneq_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 1)); \
	58766	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
57109	58767	__ret; \
57110	58768	})
57111	58769	#endif
57112	58770
57113	58771	__ai __attribute__((target("neon"))) int64_t vqdmulls_s32(int32_t __p0, int32_t __p1) {
57114	58772	int64_t __ret;
57115		__ret = (int64_t) __builtin_neon_vqdmulls_s32(__p0, __p1);
	58773	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vqdmulls_s32(__p0, __p1));
57116	58774	return __ret;
57117	58775	}
57118	58776	__ai __attribute__((target("neon"))) int32_t vqdmullh_s16(int16_t __p0, int16_t __p1) {
57119	58777	int32_t __ret;
57120		__ret = (int32_t) __builtin_neon_vqdmullh_s16(__p0, __p1);
	58778	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vqdmullh_s16(__p0, __p1));
57121	58779	return __ret;
57122	58780	}
57123	58781	#ifdef __LITTLE_ENDIAN__
...	...	@@ -57129,10 +58787,10 @@ __ai __attribute__((target("neon"))) int64x2_t vqdmull_high_s32(int32x4_t __p0,
57129	58787	#else
57130	58788	__ai __attribute__((target("neon"))) int64x2_t vqdmull_high_s32(int32x4_t __p0, int32x4_t __p1) {
57131	58789	int64x2_t __ret;
57132		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
57133		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	58790	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	58791	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
57134	58792	__ret = __noswap_vqdmull_s32(__noswap_vget_high_s32(__rev0), __noswap_vget_high_s32(__rev1));
57135		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	58793	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
57136	58794	return __ret;
57137	58795	}
57138	58796	#endif
...	...	@@ -57146,95 +58804,95 @@ __ai __attribute__((target("neon"))) int32x4_t vqdmull_high_s16(int16x8_t __p0,
57146	58804	#else
57147	58805	__ai __attribute__((target("neon"))) int32x4_t vqdmull_high_s16(int16x8_t __p0, int16x8_t __p1) {
57148	58806	int32x4_t __ret;
57149		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
57150		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	58807	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	58808	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
57151	58809	__ret = __noswap_vqdmull_s16(__noswap_vget_high_s16(__rev0), __noswap_vget_high_s16(__rev1));
57152		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	58810	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
57153	58811	return __ret;
57154	58812	}
57155	58813	#endif
57156	58814
57157	58815	#ifdef __LITTLE_ENDIAN__
57158		#define vqdmull_high_lane_s32(__p0_660, __p1_660, __p2_660) __extension__ ({ \
57159		int64x2_t __ret_660; \
57160		int32x4_t __s0_660 = __p0_660; \
57161		int32x2_t __s1_660 = __p1_660; \
57162		__ret_660 = vqdmull_s32(vget_high_s32(__s0_660), splat_lane_s32(__s1_660, __p2_660)); \
57163		__ret_660; \
	58816	#define vqdmull_high_lane_s32(__p0_676, __p1_676, __p2_676) __extension__ ({ \
	58817	int64x2_t __ret_676; \
	58818	int32x4_t __s0_676 = __p0_676; \
	58819	int32x2_t __s1_676 = __p1_676; \
	58820	__ret_676 = vqdmull_s32(vget_high_s32(__s0_676), splat_lane_s32(__s1_676, __p2_676)); \
	58821	__ret_676; \
57164	58822	})
57165	58823	#else
57166		#define vqdmull_high_lane_s32(__p0_661, __p1_661, __p2_661) __extension__ ({ \
57167		int64x2_t __ret_661; \
57168		int32x4_t __s0_661 = __p0_661; \
57169		int32x2_t __s1_661 = __p1_661; \
57170		int32x4_t __rev0_661; __rev0_661 = __builtin_shufflevector(__s0_661, __s0_661, 3, 2, 1, 0); \
57171		int32x2_t __rev1_661; __rev1_661 = __builtin_shufflevector(__s1_661, __s1_661, 1, 0); \
57172		__ret_661 = __noswap_vqdmull_s32(__noswap_vget_high_s32(__rev0_661), __noswap_splat_lane_s32(__rev1_661, __p2_661)); \
57173		__ret_661 = __builtin_shufflevector(__ret_661, __ret_661, 1, 0); \
57174		__ret_661; \
	58824	#define vqdmull_high_lane_s32(__p0_677, __p1_677, __p2_677) __extension__ ({ \
	58825	int64x2_t __ret_677; \
	58826	int32x4_t __s0_677 = __p0_677; \
	58827	int32x2_t __s1_677 = __p1_677; \
	58828	int32x4_t __rev0_677; __rev0_677 = __builtin_shufflevector(__s0_677, __s0_677, __lane_reverse_128_32); \
	58829	int32x2_t __rev1_677; __rev1_677 = __builtin_shufflevector(__s1_677, __s1_677, __lane_reverse_64_32); \
	58830	__ret_677 = __noswap_vqdmull_s32(__noswap_vget_high_s32(__rev0_677), __noswap_splat_lane_s32(__rev1_677, __p2_677)); \
	58831	__ret_677 = __builtin_shufflevector(__ret_677, __ret_677, __lane_reverse_128_64); \
	58832	__ret_677; \
57175	58833	})
57176	58834	#endif
57177	58835
57178	58836	#ifdef __LITTLE_ENDIAN__
57179		#define vqdmull_high_lane_s16(__p0_662, __p1_662, __p2_662) __extension__ ({ \
57180		int32x4_t __ret_662; \
57181		int16x8_t __s0_662 = __p0_662; \
57182		int16x4_t __s1_662 = __p1_662; \
57183		__ret_662 = vqdmull_s16(vget_high_s16(__s0_662), splat_lane_s16(__s1_662, __p2_662)); \
57184		__ret_662; \
	58837	#define vqdmull_high_lane_s16(__p0_678, __p1_678, __p2_678) __extension__ ({ \
	58838	int32x4_t __ret_678; \
	58839	int16x8_t __s0_678 = __p0_678; \
	58840	int16x4_t __s1_678 = __p1_678; \
	58841	__ret_678 = vqdmull_s16(vget_high_s16(__s0_678), splat_lane_s16(__s1_678, __p2_678)); \
	58842	__ret_678; \
57185	58843	})
57186	58844	#else
57187		#define vqdmull_high_lane_s16(__p0_663, __p1_663, __p2_663) __extension__ ({ \
57188		int32x4_t __ret_663; \
57189		int16x8_t __s0_663 = __p0_663; \
57190		int16x4_t __s1_663 = __p1_663; \
57191		int16x8_t __rev0_663; __rev0_663 = __builtin_shufflevector(__s0_663, __s0_663, 7, 6, 5, 4, 3, 2, 1, 0); \
57192		int16x4_t __rev1_663; __rev1_663 = __builtin_shufflevector(__s1_663, __s1_663, 3, 2, 1, 0); \
57193		__ret_663 = __noswap_vqdmull_s16(__noswap_vget_high_s16(__rev0_663), __noswap_splat_lane_s16(__rev1_663, __p2_663)); \
57194		__ret_663 = __builtin_shufflevector(__ret_663, __ret_663, 3, 2, 1, 0); \
57195		__ret_663; \
	58845	#define vqdmull_high_lane_s16(__p0_679, __p1_679, __p2_679) __extension__ ({ \
	58846	int32x4_t __ret_679; \
	58847	int16x8_t __s0_679 = __p0_679; \
	58848	int16x4_t __s1_679 = __p1_679; \
	58849	int16x8_t __rev0_679; __rev0_679 = __builtin_shufflevector(__s0_679, __s0_679, __lane_reverse_128_16); \
	58850	int16x4_t __rev1_679; __rev1_679 = __builtin_shufflevector(__s1_679, __s1_679, __lane_reverse_64_16); \
	58851	__ret_679 = __noswap_vqdmull_s16(__noswap_vget_high_s16(__rev0_679), __noswap_splat_lane_s16(__rev1_679, __p2_679)); \
	58852	__ret_679 = __builtin_shufflevector(__ret_679, __ret_679, __lane_reverse_128_32); \
	58853	__ret_679; \
57196	58854	})
57197	58855	#endif
57198	58856
57199	58857	#ifdef __LITTLE_ENDIAN__
57200		#define vqdmull_high_laneq_s32(__p0_664, __p1_664, __p2_664) __extension__ ({ \
57201		int64x2_t __ret_664; \
57202		int32x4_t __s0_664 = __p0_664; \
57203		int32x4_t __s1_664 = __p1_664; \
57204		__ret_664 = vqdmull_s32(vget_high_s32(__s0_664), splat_laneq_s32(__s1_664, __p2_664)); \
57205		__ret_664; \
	58858	#define vqdmull_high_laneq_s32(__p0_680, __p1_680, __p2_680) __extension__ ({ \
	58859	int64x2_t __ret_680; \
	58860	int32x4_t __s0_680 = __p0_680; \
	58861	int32x4_t __s1_680 = __p1_680; \
	58862	__ret_680 = vqdmull_s32(vget_high_s32(__s0_680), splat_laneq_s32(__s1_680, __p2_680)); \
	58863	__ret_680; \
57206	58864	})
57207	58865	#else
57208		#define vqdmull_high_laneq_s32(__p0_665, __p1_665, __p2_665) __extension__ ({ \
57209		int64x2_t __ret_665; \
57210		int32x4_t __s0_665 = __p0_665; \
57211		int32x4_t __s1_665 = __p1_665; \
57212		int32x4_t __rev0_665; __rev0_665 = __builtin_shufflevector(__s0_665, __s0_665, 3, 2, 1, 0); \
57213		int32x4_t __rev1_665; __rev1_665 = __builtin_shufflevector(__s1_665, __s1_665, 3, 2, 1, 0); \
57214		__ret_665 = __noswap_vqdmull_s32(__noswap_vget_high_s32(__rev0_665), __noswap_splat_laneq_s32(__rev1_665, __p2_665)); \
57215		__ret_665 = __builtin_shufflevector(__ret_665, __ret_665, 1, 0); \
57216		__ret_665; \
	58866	#define vqdmull_high_laneq_s32(__p0_681, __p1_681, __p2_681) __extension__ ({ \
	58867	int64x2_t __ret_681; \
	58868	int32x4_t __s0_681 = __p0_681; \
	58869	int32x4_t __s1_681 = __p1_681; \
	58870	int32x4_t __rev0_681; __rev0_681 = __builtin_shufflevector(__s0_681, __s0_681, __lane_reverse_128_32); \
	58871	int32x4_t __rev1_681; __rev1_681 = __builtin_shufflevector(__s1_681, __s1_681, __lane_reverse_128_32); \
	58872	__ret_681 = __noswap_vqdmull_s32(__noswap_vget_high_s32(__rev0_681), __noswap_splat_laneq_s32(__rev1_681, __p2_681)); \
	58873	__ret_681 = __builtin_shufflevector(__ret_681, __ret_681, __lane_reverse_128_64); \
	58874	__ret_681; \
57217	58875	})
57218	58876	#endif
57219	58877
57220	58878	#ifdef __LITTLE_ENDIAN__
57221		#define vqdmull_high_laneq_s16(__p0_666, __p1_666, __p2_666) __extension__ ({ \
57222		int32x4_t __ret_666; \
57223		int16x8_t __s0_666 = __p0_666; \
57224		int16x8_t __s1_666 = __p1_666; \
57225		__ret_666 = vqdmull_s16(vget_high_s16(__s0_666), splat_laneq_s16(__s1_666, __p2_666)); \
57226		__ret_666; \
	58879	#define vqdmull_high_laneq_s16(__p0_682, __p1_682, __p2_682) __extension__ ({ \
	58880	int32x4_t __ret_682; \
	58881	int16x8_t __s0_682 = __p0_682; \
	58882	int16x8_t __s1_682 = __p1_682; \
	58883	__ret_682 = vqdmull_s16(vget_high_s16(__s0_682), splat_laneq_s16(__s1_682, __p2_682)); \
	58884	__ret_682; \
57227	58885	})
57228	58886	#else
57229		#define vqdmull_high_laneq_s16(__p0_667, __p1_667, __p2_667) __extension__ ({ \
57230		int32x4_t __ret_667; \
57231		int16x8_t __s0_667 = __p0_667; \
57232		int16x8_t __s1_667 = __p1_667; \
57233		int16x8_t __rev0_667; __rev0_667 = __builtin_shufflevector(__s0_667, __s0_667, 7, 6, 5, 4, 3, 2, 1, 0); \
57234		int16x8_t __rev1_667; __rev1_667 = __builtin_shufflevector(__s1_667, __s1_667, 7, 6, 5, 4, 3, 2, 1, 0); \
57235		__ret_667 = __noswap_vqdmull_s16(__noswap_vget_high_s16(__rev0_667), __noswap_splat_laneq_s16(__rev1_667, __p2_667)); \
57236		__ret_667 = __builtin_shufflevector(__ret_667, __ret_667, 3, 2, 1, 0); \
57237		__ret_667; \
	58887	#define vqdmull_high_laneq_s16(__p0_683, __p1_683, __p2_683) __extension__ ({ \
	58888	int32x4_t __ret_683; \
	58889	int16x8_t __s0_683 = __p0_683; \
	58890	int16x8_t __s1_683 = __p1_683; \
	58891	int16x8_t __rev0_683; __rev0_683 = __builtin_shufflevector(__s0_683, __s0_683, __lane_reverse_128_16); \
	58892	int16x8_t __rev1_683; __rev1_683 = __builtin_shufflevector(__s1_683, __s1_683, __lane_reverse_128_16); \
	58893	__ret_683 = __noswap_vqdmull_s16(__noswap_vget_high_s16(__rev0_683), __noswap_splat_laneq_s16(__rev1_683, __p2_683)); \
	58894	__ret_683 = __builtin_shufflevector(__ret_683, __ret_683, __lane_reverse_128_32); \
	58895	__ret_683; \
57238	58896	})
57239	58897	#endif
57240	58898
...	...	@@ -57247,9 +58905,9 @@ __ai __attribute__((target("neon"))) int64x2_t vqdmull_high_n_s32(int32x4_t __p0
57247	58905	#else
57248	58906	__ai __attribute__((target("neon"))) int64x2_t vqdmull_high_n_s32(int32x4_t __p0, int32_t __p1) {
57249	58907	int64x2_t __ret;
57250		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
	58908	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
57251	58909	__ret = __noswap_vqdmull_n_s32(__noswap_vget_high_s32(__rev0), __p1);
57252		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	58910	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
57253	58911	return __ret;
57254	58912	}
57255	58913	#endif
...	...	@@ -57263,159 +58921,159 @@ __ai __attribute__((target("neon"))) int32x4_t vqdmull_high_n_s16(int16x8_t __p0
57263	58921	#else
57264	58922	__ai __attribute__((target("neon"))) int32x4_t vqdmull_high_n_s16(int16x8_t __p0, int16_t __p1) {
57265	58923	int32x4_t __ret;
57266		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	58924	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
57267	58925	__ret = __noswap_vqdmull_n_s16(__noswap_vget_high_s16(__rev0), __p1);
57268		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	58926	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
57269	58927	return __ret;
57270	58928	}
57271	58929	#endif
57272	58930
57273	58931	#ifdef __LITTLE_ENDIAN__
57274		#define vqdmulls_lane_s32(__p0_668, __p1_668, __p2_668) __extension__ ({ \
57275		int64_t __ret_668; \
57276		int32_t __s0_668 = __p0_668; \
57277		int32x2_t __s1_668 = __p1_668; \
57278		__ret_668 = vqdmulls_s32(__s0_668, vget_lane_s32(__s1_668, __p2_668)); \
57279		__ret_668; \
	58932	#define vqdmulls_lane_s32(__p0_684, __p1_684, __p2_684) __extension__ ({ \
	58933	int64_t __ret_684; \
	58934	int32_t __s0_684 = __p0_684; \
	58935	int32x2_t __s1_684 = __p1_684; \
	58936	__ret_684 = vqdmulls_s32(__s0_684, vget_lane_s32(__s1_684, __p2_684)); \
	58937	__ret_684; \
57280	58938	})
57281	58939	#else
57282		#define vqdmulls_lane_s32(__p0_669, __p1_669, __p2_669) __extension__ ({ \
57283		int64_t __ret_669; \
57284		int32_t __s0_669 = __p0_669; \
57285		int32x2_t __s1_669 = __p1_669; \
57286		int32x2_t __rev1_669; __rev1_669 = __builtin_shufflevector(__s1_669, __s1_669, 1, 0); \
57287		__ret_669 = vqdmulls_s32(__s0_669, __noswap_vget_lane_s32(__rev1_669, __p2_669)); \
57288		__ret_669; \
	58940	#define vqdmulls_lane_s32(__p0_685, __p1_685, __p2_685) __extension__ ({ \
	58941	int64_t __ret_685; \
	58942	int32_t __s0_685 = __p0_685; \
	58943	int32x2_t __s1_685 = __p1_685; \
	58944	int32x2_t __rev1_685; __rev1_685 = __builtin_shufflevector(__s1_685, __s1_685, __lane_reverse_64_32); \
	58945	__ret_685 = vqdmulls_s32(__s0_685, __noswap_vget_lane_s32(__rev1_685, __p2_685)); \
	58946	__ret_685; \
57289	58947	})
57290	58948	#endif
57291	58949
57292	58950	#ifdef __LITTLE_ENDIAN__
57293		#define vqdmullh_lane_s16(__p0_670, __p1_670, __p2_670) __extension__ ({ \
57294		int32_t __ret_670; \
57295		int16_t __s0_670 = __p0_670; \
57296		int16x4_t __s1_670 = __p1_670; \
57297		__ret_670 = vqdmullh_s16(__s0_670, vget_lane_s16(__s1_670, __p2_670)); \
57298		__ret_670; \
	58951	#define vqdmullh_lane_s16(__p0_686, __p1_686, __p2_686) __extension__ ({ \
	58952	int32_t __ret_686; \
	58953	int16_t __s0_686 = __p0_686; \
	58954	int16x4_t __s1_686 = __p1_686; \
	58955	__ret_686 = vqdmullh_s16(__s0_686, vget_lane_s16(__s1_686, __p2_686)); \
	58956	__ret_686; \
57299	58957	})
57300	58958	#else
57301		#define vqdmullh_lane_s16(__p0_671, __p1_671, __p2_671) __extension__ ({ \
57302		int32_t __ret_671; \
57303		int16_t __s0_671 = __p0_671; \
57304		int16x4_t __s1_671 = __p1_671; \
57305		int16x4_t __rev1_671; __rev1_671 = __builtin_shufflevector(__s1_671, __s1_671, 3, 2, 1, 0); \
57306		__ret_671 = vqdmullh_s16(__s0_671, __noswap_vget_lane_s16(__rev1_671, __p2_671)); \
57307		__ret_671; \
	58959	#define vqdmullh_lane_s16(__p0_687, __p1_687, __p2_687) __extension__ ({ \
	58960	int32_t __ret_687; \
	58961	int16_t __s0_687 = __p0_687; \
	58962	int16x4_t __s1_687 = __p1_687; \
	58963	int16x4_t __rev1_687; __rev1_687 = __builtin_shufflevector(__s1_687, __s1_687, __lane_reverse_64_16); \
	58964	__ret_687 = vqdmullh_s16(__s0_687, __noswap_vget_lane_s16(__rev1_687, __p2_687)); \
	58965	__ret_687; \
57308	58966	})
57309	58967	#endif
57310	58968
57311	58969	#ifdef __LITTLE_ENDIAN__
57312		#define vqdmulls_laneq_s32(__p0_672, __p1_672, __p2_672) __extension__ ({ \
57313		int64_t __ret_672; \
57314		int32_t __s0_672 = __p0_672; \
57315		int32x4_t __s1_672 = __p1_672; \
57316		__ret_672 = vqdmulls_s32(__s0_672, vgetq_lane_s32(__s1_672, __p2_672)); \
57317		__ret_672; \
	58970	#define vqdmulls_laneq_s32(__p0_688, __p1_688, __p2_688) __extension__ ({ \
	58971	int64_t __ret_688; \
	58972	int32_t __s0_688 = __p0_688; \
	58973	int32x4_t __s1_688 = __p1_688; \
	58974	__ret_688 = vqdmulls_s32(__s0_688, vgetq_lane_s32(__s1_688, __p2_688)); \
	58975	__ret_688; \
57318	58976	})
57319	58977	#else
57320		#define vqdmulls_laneq_s32(__p0_673, __p1_673, __p2_673) __extension__ ({ \
57321		int64_t __ret_673; \
57322		int32_t __s0_673 = __p0_673; \
57323		int32x4_t __s1_673 = __p1_673; \
57324		int32x4_t __rev1_673; __rev1_673 = __builtin_shufflevector(__s1_673, __s1_673, 3, 2, 1, 0); \
57325		__ret_673 = vqdmulls_s32(__s0_673, __noswap_vgetq_lane_s32(__rev1_673, __p2_673)); \
57326		__ret_673; \
	58978	#define vqdmulls_laneq_s32(__p0_689, __p1_689, __p2_689) __extension__ ({ \
	58979	int64_t __ret_689; \
	58980	int32_t __s0_689 = __p0_689; \
	58981	int32x4_t __s1_689 = __p1_689; \
	58982	int32x4_t __rev1_689; __rev1_689 = __builtin_shufflevector(__s1_689, __s1_689, __lane_reverse_128_32); \
	58983	__ret_689 = vqdmulls_s32(__s0_689, __noswap_vgetq_lane_s32(__rev1_689, __p2_689)); \
	58984	__ret_689; \
57327	58985	})
57328	58986	#endif
57329	58987
57330	58988	#ifdef __LITTLE_ENDIAN__
57331		#define vqdmullh_laneq_s16(__p0_674, __p1_674, __p2_674) __extension__ ({ \
57332		int32_t __ret_674; \
57333		int16_t __s0_674 = __p0_674; \
57334		int16x8_t __s1_674 = __p1_674; \
57335		__ret_674 = vqdmullh_s16(__s0_674, vgetq_lane_s16(__s1_674, __p2_674)); \
57336		__ret_674; \
	58989	#define vqdmullh_laneq_s16(__p0_690, __p1_690, __p2_690) __extension__ ({ \
	58990	int32_t __ret_690; \
	58991	int16_t __s0_690 = __p0_690; \
	58992	int16x8_t __s1_690 = __p1_690; \
	58993	__ret_690 = vqdmullh_s16(__s0_690, vgetq_lane_s16(__s1_690, __p2_690)); \
	58994	__ret_690; \
57337	58995	})
57338	58996	#else
57339		#define vqdmullh_laneq_s16(__p0_675, __p1_675, __p2_675) __extension__ ({ \
57340		int32_t __ret_675; \
57341		int16_t __s0_675 = __p0_675; \
57342		int16x8_t __s1_675 = __p1_675; \
57343		int16x8_t __rev1_675; __rev1_675 = __builtin_shufflevector(__s1_675, __s1_675, 7, 6, 5, 4, 3, 2, 1, 0); \
57344		__ret_675 = vqdmullh_s16(__s0_675, __noswap_vgetq_lane_s16(__rev1_675, __p2_675)); \
57345		__ret_675; \
	58997	#define vqdmullh_laneq_s16(__p0_691, __p1_691, __p2_691) __extension__ ({ \
	58998	int32_t __ret_691; \
	58999	int16_t __s0_691 = __p0_691; \
	59000	int16x8_t __s1_691 = __p1_691; \
	59001	int16x8_t __rev1_691; __rev1_691 = __builtin_shufflevector(__s1_691, __s1_691, __lane_reverse_128_16); \
	59002	__ret_691 = vqdmullh_s16(__s0_691, __noswap_vgetq_lane_s16(__rev1_691, __p2_691)); \
	59003	__ret_691; \
57346	59004	})
57347	59005	#endif
57348	59006
57349	59007	#ifdef __LITTLE_ENDIAN__
57350		#define vqdmull_laneq_s32(__p0_676, __p1_676, __p2_676) __extension__ ({ \
57351		int64x2_t __ret_676; \
57352		int32x2_t __s0_676 = __p0_676; \
57353		int32x4_t __s1_676 = __p1_676; \
57354		__ret_676 = vqdmull_s32(__s0_676, splat_laneq_s32(__s1_676, __p2_676)); \
57355		__ret_676; \
	59008	#define vqdmull_laneq_s32(__p0_692, __p1_692, __p2_692) __extension__ ({ \
	59009	int64x2_t __ret_692; \
	59010	int32x2_t __s0_692 = __p0_692; \
	59011	int32x4_t __s1_692 = __p1_692; \
	59012	__ret_692 = vqdmull_s32(__s0_692, splat_laneq_s32(__s1_692, __p2_692)); \
	59013	__ret_692; \
57356	59014	})
57357	59015	#else
57358		#define vqdmull_laneq_s32(__p0_677, __p1_677, __p2_677) __extension__ ({ \
57359		int64x2_t __ret_677; \
57360		int32x2_t __s0_677 = __p0_677; \
57361		int32x4_t __s1_677 = __p1_677; \
57362		int32x2_t __rev0_677; __rev0_677 = __builtin_shufflevector(__s0_677, __s0_677, 1, 0); \
57363		int32x4_t __rev1_677; __rev1_677 = __builtin_shufflevector(__s1_677, __s1_677, 3, 2, 1, 0); \
57364		__ret_677 = __noswap_vqdmull_s32(__rev0_677, __noswap_splat_laneq_s32(__rev1_677, __p2_677)); \
57365		__ret_677 = __builtin_shufflevector(__ret_677, __ret_677, 1, 0); \
57366		__ret_677; \
	59016	#define vqdmull_laneq_s32(__p0_693, __p1_693, __p2_693) __extension__ ({ \
	59017	int64x2_t __ret_693; \
	59018	int32x2_t __s0_693 = __p0_693; \
	59019	int32x4_t __s1_693 = __p1_693; \
	59020	int32x2_t __rev0_693; __rev0_693 = __builtin_shufflevector(__s0_693, __s0_693, __lane_reverse_64_32); \
	59021	int32x4_t __rev1_693; __rev1_693 = __builtin_shufflevector(__s1_693, __s1_693, __lane_reverse_128_32); \
	59022	__ret_693 = __noswap_vqdmull_s32(__rev0_693, __noswap_splat_laneq_s32(__rev1_693, __p2_693)); \
	59023	__ret_693 = __builtin_shufflevector(__ret_693, __ret_693, __lane_reverse_128_64); \
	59024	__ret_693; \
57367	59025	})
57368	59026	#endif
57369	59027
57370	59028	#ifdef __LITTLE_ENDIAN__
57371		#define vqdmull_laneq_s16(__p0_678, __p1_678, __p2_678) __extension__ ({ \
57372		int32x4_t __ret_678; \
57373		int16x4_t __s0_678 = __p0_678; \
57374		int16x8_t __s1_678 = __p1_678; \
57375		__ret_678 = vqdmull_s16(__s0_678, splat_laneq_s16(__s1_678, __p2_678)); \
57376		__ret_678; \
	59029	#define vqdmull_laneq_s16(__p0_694, __p1_694, __p2_694) __extension__ ({ \
	59030	int32x4_t __ret_694; \
	59031	int16x4_t __s0_694 = __p0_694; \
	59032	int16x8_t __s1_694 = __p1_694; \
	59033	__ret_694 = vqdmull_s16(__s0_694, splat_laneq_s16(__s1_694, __p2_694)); \
	59034	__ret_694; \
57377	59035	})
57378	59036	#else
57379		#define vqdmull_laneq_s16(__p0_679, __p1_679, __p2_679) __extension__ ({ \
57380		int32x4_t __ret_679; \
57381		int16x4_t __s0_679 = __p0_679; \
57382		int16x8_t __s1_679 = __p1_679; \
57383		int16x4_t __rev0_679; __rev0_679 = __builtin_shufflevector(__s0_679, __s0_679, 3, 2, 1, 0); \
57384		int16x8_t __rev1_679; __rev1_679 = __builtin_shufflevector(__s1_679, __s1_679, 7, 6, 5, 4, 3, 2, 1, 0); \
57385		__ret_679 = __noswap_vqdmull_s16(__rev0_679, __noswap_splat_laneq_s16(__rev1_679, __p2_679)); \
57386		__ret_679 = __builtin_shufflevector(__ret_679, __ret_679, 3, 2, 1, 0); \
57387		__ret_679; \
	59037	#define vqdmull_laneq_s16(__p0_695, __p1_695, __p2_695) __extension__ ({ \
	59038	int32x4_t __ret_695; \
	59039	int16x4_t __s0_695 = __p0_695; \
	59040	int16x8_t __s1_695 = __p1_695; \
	59041	int16x4_t __rev0_695; __rev0_695 = __builtin_shufflevector(__s0_695, __s0_695, __lane_reverse_64_16); \
	59042	int16x8_t __rev1_695; __rev1_695 = __builtin_shufflevector(__s1_695, __s1_695, __lane_reverse_128_16); \
	59043	__ret_695 = __noswap_vqdmull_s16(__rev0_695, __noswap_splat_laneq_s16(__rev1_695, __p2_695)); \
	59044	__ret_695 = __builtin_shufflevector(__ret_695, __ret_695, __lane_reverse_128_32); \
	59045	__ret_695; \
57388	59046	})
57389	59047	#endif
57390	59048
57391	59049	__ai __attribute__((target("neon"))) int16_t vqmovns_s32(int32_t __p0) {
57392	59050	int16_t __ret;
57393		__ret = (int16_t) __builtin_neon_vqmovns_s32(__p0);
	59051	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vqmovns_s32(__p0));
57394	59052	return __ret;
57395	59053	}
57396	59054	__ai __attribute__((target("neon"))) int32_t vqmovnd_s64(int64_t __p0) {
57397	59055	int32_t __ret;
57398		__ret = (int32_t) __builtin_neon_vqmovnd_s64(__p0);
	59056	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vqmovnd_s64(__p0));
57399	59057	return __ret;
57400	59058	}
57401	59059	__ai __attribute__((target("neon"))) int8_t vqmovnh_s16(int16_t __p0) {
57402	59060	int8_t __ret;
57403		__ret = (int8_t) __builtin_neon_vqmovnh_s16(__p0);
	59061	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vqmovnh_s16(__p0));
57404	59062	return __ret;
57405	59063	}
57406	59064	__ai __attribute__((target("neon"))) uint16_t vqmovns_u32(uint32_t __p0) {
57407	59065	uint16_t __ret;
57408		__ret = (uint16_t) __builtin_neon_vqmovns_u32(__p0);
	59066	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vqmovns_u32(__p0));
57409	59067	return __ret;
57410	59068	}
57411	59069	__ai __attribute__((target("neon"))) uint32_t vqmovnd_u64(uint64_t __p0) {
57412	59070	uint32_t __ret;
57413		__ret = (uint32_t) __builtin_neon_vqmovnd_u64(__p0);
	59071	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vqmovnd_u64(__p0));
57414	59072	return __ret;
57415	59073	}
57416	59074	__ai __attribute__((target("neon"))) uint8_t vqmovnh_u16(uint16_t __p0) {
57417	59075	uint8_t __ret;
57418		__ret = (uint8_t) __builtin_neon_vqmovnh_u16(__p0);
	59076	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vqmovnh_u16(__p0));
57419	59077	return __ret;
57420	59078	}
57421	59079	#ifdef __LITTLE_ENDIAN__
...	...	@@ -57427,10 +59085,10 @@ __ai __attribute__((target("neon"))) uint16x8_t vqmovn_high_u32(uint16x4_t __p0,
57427	59085	#else
57428	59086	__ai __attribute__((target("neon"))) uint16x8_t vqmovn_high_u32(uint16x4_t __p0, uint32x4_t __p1) {
57429	59087	uint16x8_t __ret;
57430		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
57431		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	59088	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	59089	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
57432	59090	__ret = __noswap_vcombine_u16(__rev0, __noswap_vqmovn_u32(__rev1));
57433		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	59091	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
57434	59092	return __ret;
57435	59093	}
57436	59094	#endif
...	...	@@ -57444,10 +59102,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vqmovn_high_u64(uint32x2_t __p0,
57444	59102	#else
57445	59103	__ai __attribute__((target("neon"))) uint32x4_t vqmovn_high_u64(uint32x2_t __p0, uint64x2_t __p1) {
57446	59104	uint32x4_t __ret;
57447		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
57448		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	59105	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	59106	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
57449	59107	__ret = __noswap_vcombine_u32(__rev0, __noswap_vqmovn_u64(__rev1));
57450		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	59108	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
57451	59109	return __ret;
57452	59110	}
57453	59111	#endif
...	...	@@ -57461,10 +59119,10 @@ __ai __attribute__((target("neon"))) uint8x16_t vqmovn_high_u16(uint8x8_t __p0,
57461	59119	#else
57462	59120	__ai __attribute__((target("neon"))) uint8x16_t vqmovn_high_u16(uint8x8_t __p0, uint16x8_t __p1) {
57463	59121	uint8x16_t __ret;
57464		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
57465		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	59122	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	59123	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
57466	59124	__ret = __noswap_vcombine_u8(__rev0, __noswap_vqmovn_u16(__rev1));
57467		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	59125	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
57468	59126	return __ret;
57469	59127	}
57470	59128	#endif
...	...	@@ -57478,10 +59136,10 @@ __ai __attribute__((target("neon"))) int16x8_t vqmovn_high_s32(int16x4_t __p0, i
57478	59136	#else
57479	59137	__ai __attribute__((target("neon"))) int16x8_t vqmovn_high_s32(int16x4_t __p0, int32x4_t __p1) {
57480	59138	int16x8_t __ret;
57481		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
57482		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	59139	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	59140	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
57483	59141	__ret = __noswap_vcombine_s16(__rev0, __noswap_vqmovn_s32(__rev1));
57484		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	59142	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
57485	59143	return __ret;
57486	59144	}
57487	59145	#endif
...	...	@@ -57495,10 +59153,10 @@ __ai __attribute__((target("neon"))) int32x4_t vqmovn_high_s64(int32x2_t __p0, i
57495	59153	#else
57496	59154	__ai __attribute__((target("neon"))) int32x4_t vqmovn_high_s64(int32x2_t __p0, int64x2_t __p1) {
57497	59155	int32x4_t __ret;
57498		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
57499		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	59156	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	59157	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
57500	59158	__ret = __noswap_vcombine_s32(__rev0, __noswap_vqmovn_s64(__rev1));
57501		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	59159	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
57502	59160	return __ret;
57503	59161	}
57504	59162	#endif
...	...	@@ -57512,42 +59170,42 @@ __ai __attribute__((target("neon"))) int8x16_t vqmovn_high_s16(int8x8_t __p0, in
57512	59170	#else
57513	59171	__ai __attribute__((target("neon"))) int8x16_t vqmovn_high_s16(int8x8_t __p0, int16x8_t __p1) {
57514	59172	int8x16_t __ret;
57515		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
57516		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	59173	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	59174	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
57517	59175	__ret = __noswap_vcombine_s8(__rev0, __noswap_vqmovn_s16(__rev1));
57518		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	59176	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
57519	59177	return __ret;
57520	59178	}
57521	59179	#endif
57522	59180
57523	59181	__ai __attribute__((target("neon"))) uint16_t vqmovuns_s32(int32_t __p0) {
57524	59182	uint16_t __ret;
57525		__ret = (uint16_t) __builtin_neon_vqmovuns_s32(__p0);
	59183	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vqmovuns_s32(__p0));
57526	59184	return __ret;
57527	59185	}
57528	59186	__ai __attribute__((target("neon"))) uint32_t vqmovund_s64(int64_t __p0) {
57529	59187	uint32_t __ret;
57530		__ret = (uint32_t) __builtin_neon_vqmovund_s64(__p0);
	59188	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vqmovund_s64(__p0));
57531	59189	return __ret;
57532	59190	}
57533	59191	__ai __attribute__((target("neon"))) uint8_t vqmovunh_s16(int16_t __p0) {
57534	59192	uint8_t __ret;
57535		__ret = (uint8_t) __builtin_neon_vqmovunh_s16(__p0);
	59193	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vqmovunh_s16(__p0));
57536	59194	return __ret;
57537	59195	}
57538	59196	#ifdef __LITTLE_ENDIAN__
57539	59197	__ai __attribute__((target("neon"))) uint16x8_t vqmovun_high_s32(uint16x4_t __p0, int32x4_t __p1) {
57540	59198	uint16x8_t __ret;
57541		__ret = vcombine_u16((uint16x4_t)(__p0), vqmovun_s32(__p1));
	59199	__ret = vcombine_u16(__builtin_bit_cast(uint16x4_t, __p0), vqmovun_s32(__p1));
57542	59200	return __ret;
57543	59201	}
57544	59202	#else
57545	59203	__ai __attribute__((target("neon"))) uint16x8_t vqmovun_high_s32(uint16x4_t __p0, int32x4_t __p1) {
57546	59204	uint16x8_t __ret;
57547		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
57548		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
57549		__ret = __noswap_vcombine_u16((uint16x4_t)(__rev0), __noswap_vqmovun_s32(__rev1));
57550		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	59205	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	59206	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	59207	__ret = __noswap_vcombine_u16(__builtin_bit_cast(uint16x4_t, __rev0), __noswap_vqmovun_s32(__rev1));
	59208	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
57551	59209	return __ret;
57552	59210	}
57553	59211	#endif
...	...	@@ -57555,16 +59213,16 @@ __ai __attribute__((target("neon"))) uint16x8_t vqmovun_high_s32(uint16x4_t __p0
57555	59213	#ifdef __LITTLE_ENDIAN__
57556	59214	__ai __attribute__((target("neon"))) uint32x4_t vqmovun_high_s64(uint32x2_t __p0, int64x2_t __p1) {
57557	59215	uint32x4_t __ret;
57558		__ret = vcombine_u32((uint32x2_t)(__p0), vqmovun_s64(__p1));
	59216	__ret = vcombine_u32(__builtin_bit_cast(uint32x2_t, __p0), vqmovun_s64(__p1));
57559	59217	return __ret;
57560	59218	}
57561	59219	#else
57562	59220	__ai __attribute__((target("neon"))) uint32x4_t vqmovun_high_s64(uint32x2_t __p0, int64x2_t __p1) {
57563	59221	uint32x4_t __ret;
57564		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
57565		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
57566		__ret = __noswap_vcombine_u32((uint32x2_t)(__rev0), __noswap_vqmovun_s64(__rev1));
57567		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	59222	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	59223	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	59224	__ret = __noswap_vcombine_u32(__builtin_bit_cast(uint32x2_t, __rev0), __noswap_vqmovun_s64(__rev1));
	59225	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
57568	59226	return __ret;
57569	59227	}
57570	59228	#endif
...	...	@@ -57572,16 +59230,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vqmovun_high_s64(uint32x2_t __p0
57572	59230	#ifdef __LITTLE_ENDIAN__
57573	59231	__ai __attribute__((target("neon"))) uint8x16_t vqmovun_high_s16(uint8x8_t __p0, int16x8_t __p1) {
57574	59232	uint8x16_t __ret;
57575		__ret = vcombine_u8((uint8x8_t)(__p0), vqmovun_s16(__p1));
	59233	__ret = vcombine_u8(__builtin_bit_cast(uint8x8_t, __p0), vqmovun_s16(__p1));
57576	59234	return __ret;
57577	59235	}
57578	59236	#else
57579	59237	__ai __attribute__((target("neon"))) uint8x16_t vqmovun_high_s16(uint8x8_t __p0, int16x8_t __p1) {
57580	59238	uint8x16_t __ret;
57581		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
57582		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
57583		__ret = __noswap_vcombine_u8((uint8x8_t)(__rev0), __noswap_vqmovun_s16(__rev1));
57584		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	59239	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	59240	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	59241	__ret = __noswap_vcombine_u8(__builtin_bit_cast(uint8x8_t, __rev0), __noswap_vqmovun_s16(__rev1));
	59242	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
57585	59243	return __ret;
57586	59244	}
57587	59245	#endif
...	...	@@ -57589,52 +59247,52 @@ __ai __attribute__((target("neon"))) uint8x16_t vqmovun_high_s16(uint8x8_t __p0,
57589	59247	#ifdef __LITTLE_ENDIAN__
57590	59248	__ai __attribute__((target("neon"))) int64x2_t vqnegq_s64(int64x2_t __p0) {
57591	59249	int64x2_t __ret;
57592		__ret = (int64x2_t) __builtin_neon_vqnegq_v((int8x16_t)__p0, 35);
	59250	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vqnegq_v(__builtin_bit_cast(int8x16_t, __p0), 35));
57593	59251	return __ret;
57594	59252	}
57595	59253	#else
57596	59254	__ai __attribute__((target("neon"))) int64x2_t vqnegq_s64(int64x2_t __p0) {
57597	59255	int64x2_t __ret;
57598		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
57599		__ret = (int64x2_t) __builtin_neon_vqnegq_v((int8x16_t)__rev0, 35);
57600		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	59256	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	59257	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vqnegq_v(__builtin_bit_cast(int8x16_t, __rev0), 35));
	59258	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
57601	59259	return __ret;
57602	59260	}
57603	59261	#endif
57604	59262
57605	59263	__ai __attribute__((target("neon"))) int64x1_t vqneg_s64(int64x1_t __p0) {
57606	59264	int64x1_t __ret;
57607		__ret = (int64x1_t) __builtin_neon_vqneg_v((int8x8_t)__p0, 3);
	59265	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vqneg_v(__builtin_bit_cast(int8x8_t, __p0), 3));
57608	59266	return __ret;
57609	59267	}
57610	59268	__ai __attribute__((target("neon"))) int8_t vqnegb_s8(int8_t __p0) {
57611	59269	int8_t __ret;
57612		__ret = (int8_t) __builtin_neon_vqnegb_s8(__p0);
	59270	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vqnegb_s8(__p0));
57613	59271	return __ret;
57614	59272	}
57615	59273	__ai __attribute__((target("neon"))) int32_t vqnegs_s32(int32_t __p0) {
57616	59274	int32_t __ret;
57617		__ret = (int32_t) __builtin_neon_vqnegs_s32(__p0);
	59275	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vqnegs_s32(__p0));
57618	59276	return __ret;
57619	59277	}
57620	59278	__ai __attribute__((target("neon"))) int64_t vqnegd_s64(int64_t __p0) {
57621	59279	int64_t __ret;
57622		__ret = (int64_t) __builtin_neon_vqnegd_s64(__p0);
	59280	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vqnegd_s64(__p0));
57623	59281	return __ret;
57624	59282	}
57625	59283	__ai __attribute__((target("neon"))) int16_t vqnegh_s16(int16_t __p0) {
57626	59284	int16_t __ret;
57627		__ret = (int16_t) __builtin_neon_vqnegh_s16(__p0);
	59285	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vqnegh_s16(__p0));
57628	59286	return __ret;
57629	59287	}
57630	59288	__ai __attribute__((target("neon"))) int32_t vqrdmulhs_s32(int32_t __p0, int32_t __p1) {
57631	59289	int32_t __ret;
57632		__ret = (int32_t) __builtin_neon_vqrdmulhs_s32(__p0, __p1);
	59290	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vqrdmulhs_s32(__p0, __p1));
57633	59291	return __ret;
57634	59292	}
57635	59293	__ai __attribute__((target("neon"))) int16_t vqrdmulhh_s16(int16_t __p0, int16_t __p1) {
57636	59294	int16_t __ret;
57637		__ret = (int16_t) __builtin_neon_vqrdmulhh_s16(__p0, __p1);
	59295	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vqrdmulhh_s16(__p0, __p1));
57638	59296	return __ret;
57639	59297	}
57640	59298	#ifdef __LITTLE_ENDIAN__
...	...	@@ -57642,7 +59300,7 @@ __ai __attribute__((target("neon"))) int16_t vqrdmulhh_s16(int16_t __p0, int16_t
57642	59300	int32x4_t __ret; \
57643	59301	int32x4_t __s0 = __p0; \
57644	59302	int32x2_t __s1 = __p1; \
57645		__ret = (int32x4_t) __builtin_neon_vqrdmulhq_lane_v((int8x16_t)__s0, (int8x8_t)__s1, __p2, 2); \
	59303	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqrdmulhq_lane_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 2)); \
57646	59304	__ret; \
57647	59305	})
57648	59306	#else
...	...	@@ -57650,10 +59308,10 @@ __ai __attribute__((target("neon"))) int16_t vqrdmulhh_s16(int16_t __p0, int16_t
57650	59308	int32x4_t __ret; \
57651	59309	int32x4_t __s0 = __p0; \
57652	59310	int32x2_t __s1 = __p1; \
57653		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
57654		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
57655		__ret = (int32x4_t) __builtin_neon_vqrdmulhq_lane_v((int8x16_t)__rev0, (int8x8_t)__rev1, __p2, 2); \
57656		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	59311	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	59312	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_32); \
	59313	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqrdmulhq_lane_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 2)); \
	59314	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
57657	59315	__ret; \
57658	59316	})
57659	59317	#endif
...	...	@@ -57663,7 +59321,7 @@ __ai __attribute__((target("neon"))) int16_t vqrdmulhh_s16(int16_t __p0, int16_t
57663	59321	int16x8_t __ret; \
57664	59322	int16x8_t __s0 = __p0; \
57665	59323	int16x4_t __s1 = __p1; \
57666		__ret = (int16x8_t) __builtin_neon_vqrdmulhq_lane_v((int8x16_t)__s0, (int8x8_t)__s1, __p2, 1); \
	59324	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqrdmulhq_lane_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 1)); \
57667	59325	__ret; \
57668	59326	})
57669	59327	#else
...	...	@@ -57671,10 +59329,10 @@ __ai __attribute__((target("neon"))) int16_t vqrdmulhh_s16(int16_t __p0, int16_t
57671	59329	int16x8_t __ret; \
57672	59330	int16x8_t __s0 = __p0; \
57673	59331	int16x4_t __s1 = __p1; \
57674		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
57675		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
57676		__ret = (int16x8_t) __builtin_neon_vqrdmulhq_lane_v((int8x16_t)__rev0, (int8x8_t)__rev1, __p2, 1); \
57677		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	59332	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	59333	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	59334	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqrdmulhq_lane_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 1)); \
	59335	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
57678	59336	__ret; \
57679	59337	})
57680	59338	#endif
...	...	@@ -57684,7 +59342,7 @@ __ai __attribute__((target("neon"))) int16_t vqrdmulhh_s16(int16_t __p0, int16_t
57684	59342	int32x2_t __ret; \
57685	59343	int32x2_t __s0 = __p0; \
57686	59344	int32x2_t __s1 = __p1; \
57687		__ret = (int32x2_t) __builtin_neon_vqrdmulh_lane_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 2); \
	59345	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqrdmulh_lane_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 2)); \
57688	59346	__ret; \
57689	59347	})
57690	59348	#else
...	...	@@ -57692,10 +59350,10 @@ __ai __attribute__((target("neon"))) int16_t vqrdmulhh_s16(int16_t __p0, int16_t
57692	59350	int32x2_t __ret; \
57693	59351	int32x2_t __s0 = __p0; \
57694	59352	int32x2_t __s1 = __p1; \
57695		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
57696		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
57697		__ret = (int32x2_t) __builtin_neon_vqrdmulh_lane_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 2); \
57698		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	59353	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	59354	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_32); \
	59355	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqrdmulh_lane_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 2)); \
	59356	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
57699	59357	__ret; \
57700	59358	})
57701	59359	#endif
...	...	@@ -57705,7 +59363,7 @@ __ai __attribute__((target("neon"))) int16_t vqrdmulhh_s16(int16_t __p0, int16_t
57705	59363	int16x4_t __ret; \
57706	59364	int16x4_t __s0 = __p0; \
57707	59365	int16x4_t __s1 = __p1; \
57708		__ret = (int16x4_t) __builtin_neon_vqrdmulh_lane_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 1); \
	59366	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqrdmulh_lane_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 1)); \
57709	59367	__ret; \
57710	59368	})
57711	59369	#else
...	...	@@ -57713,87 +59371,87 @@ __ai __attribute__((target("neon"))) int16_t vqrdmulhh_s16(int16_t __p0, int16_t
57713	59371	int16x4_t __ret; \
57714	59372	int16x4_t __s0 = __p0; \
57715	59373	int16x4_t __s1 = __p1; \
57716		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
57717		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
57718		__ret = (int16x4_t) __builtin_neon_vqrdmulh_lane_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 1); \
57719		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	59374	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	59375	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_16); \
	59376	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqrdmulh_lane_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), __p2, 1)); \
	59377	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
57720	59378	__ret; \
57721	59379	})
57722	59380	#endif
57723	59381
57724	59382	#ifdef __LITTLE_ENDIAN__
57725		#define vqrdmulhs_lane_s32(__p0_680, __p1_680, __p2_680) __extension__ ({ \
57726		int32_t __ret_680; \
57727		int32_t __s0_680 = __p0_680; \
57728		int32x2_t __s1_680 = __p1_680; \
57729		__ret_680 = vqrdmulhs_s32(__s0_680, vget_lane_s32(__s1_680, __p2_680)); \
57730		__ret_680; \
	59383	#define vqrdmulhs_lane_s32(__p0_696, __p1_696, __p2_696) __extension__ ({ \
	59384	int32_t __ret_696; \
	59385	int32_t __s0_696 = __p0_696; \
	59386	int32x2_t __s1_696 = __p1_696; \
	59387	__ret_696 = vqrdmulhs_s32(__s0_696, vget_lane_s32(__s1_696, __p2_696)); \
	59388	__ret_696; \
57731	59389	})
57732	59390	#else
57733		#define vqrdmulhs_lane_s32(__p0_681, __p1_681, __p2_681) __extension__ ({ \
57734		int32_t __ret_681; \
57735		int32_t __s0_681 = __p0_681; \
57736		int32x2_t __s1_681 = __p1_681; \
57737		int32x2_t __rev1_681; __rev1_681 = __builtin_shufflevector(__s1_681, __s1_681, 1, 0); \
57738		__ret_681 = vqrdmulhs_s32(__s0_681, __noswap_vget_lane_s32(__rev1_681, __p2_681)); \
57739		__ret_681; \
	59391	#define vqrdmulhs_lane_s32(__p0_697, __p1_697, __p2_697) __extension__ ({ \
	59392	int32_t __ret_697; \
	59393	int32_t __s0_697 = __p0_697; \
	59394	int32x2_t __s1_697 = __p1_697; \
	59395	int32x2_t __rev1_697; __rev1_697 = __builtin_shufflevector(__s1_697, __s1_697, __lane_reverse_64_32); \
	59396	__ret_697 = vqrdmulhs_s32(__s0_697, __noswap_vget_lane_s32(__rev1_697, __p2_697)); \
	59397	__ret_697; \
57740	59398	})
57741	59399	#endif
57742	59400
57743	59401	#ifdef __LITTLE_ENDIAN__
57744		#define vqrdmulhh_lane_s16(__p0_682, __p1_682, __p2_682) __extension__ ({ \
57745		int16_t __ret_682; \
57746		int16_t __s0_682 = __p0_682; \
57747		int16x4_t __s1_682 = __p1_682; \
57748		__ret_682 = vqrdmulhh_s16(__s0_682, vget_lane_s16(__s1_682, __p2_682)); \
57749		__ret_682; \
	59402	#define vqrdmulhh_lane_s16(__p0_698, __p1_698, __p2_698) __extension__ ({ \
	59403	int16_t __ret_698; \
	59404	int16_t __s0_698 = __p0_698; \
	59405	int16x4_t __s1_698 = __p1_698; \
	59406	__ret_698 = vqrdmulhh_s16(__s0_698, vget_lane_s16(__s1_698, __p2_698)); \
	59407	__ret_698; \
57750	59408	})
57751	59409	#else
57752		#define vqrdmulhh_lane_s16(__p0_683, __p1_683, __p2_683) __extension__ ({ \
57753		int16_t __ret_683; \
57754		int16_t __s0_683 = __p0_683; \
57755		int16x4_t __s1_683 = __p1_683; \
57756		int16x4_t __rev1_683; __rev1_683 = __builtin_shufflevector(__s1_683, __s1_683, 3, 2, 1, 0); \
57757		__ret_683 = vqrdmulhh_s16(__s0_683, __noswap_vget_lane_s16(__rev1_683, __p2_683)); \
57758		__ret_683; \
	59410	#define vqrdmulhh_lane_s16(__p0_699, __p1_699, __p2_699) __extension__ ({ \
	59411	int16_t __ret_699; \
	59412	int16_t __s0_699 = __p0_699; \
	59413	int16x4_t __s1_699 = __p1_699; \
	59414	int16x4_t __rev1_699; __rev1_699 = __builtin_shufflevector(__s1_699, __s1_699, __lane_reverse_64_16); \
	59415	__ret_699 = vqrdmulhh_s16(__s0_699, __noswap_vget_lane_s16(__rev1_699, __p2_699)); \
	59416	__ret_699; \
57759	59417	})
57760	59418	#endif
57761	59419
57762	59420	#ifdef __LITTLE_ENDIAN__
57763		#define vqrdmulhs_laneq_s32(__p0_684, __p1_684, __p2_684) __extension__ ({ \
57764		int32_t __ret_684; \
57765		int32_t __s0_684 = __p0_684; \
57766		int32x4_t __s1_684 = __p1_684; \
57767		__ret_684 = vqrdmulhs_s32(__s0_684, vgetq_lane_s32(__s1_684, __p2_684)); \
57768		__ret_684; \
	59421	#define vqrdmulhs_laneq_s32(__p0_700, __p1_700, __p2_700) __extension__ ({ \
	59422	int32_t __ret_700; \
	59423	int32_t __s0_700 = __p0_700; \
	59424	int32x4_t __s1_700 = __p1_700; \
	59425	__ret_700 = vqrdmulhs_s32(__s0_700, vgetq_lane_s32(__s1_700, __p2_700)); \
	59426	__ret_700; \
57769	59427	})
57770	59428	#else
57771		#define vqrdmulhs_laneq_s32(__p0_685, __p1_685, __p2_685) __extension__ ({ \
57772		int32_t __ret_685; \
57773		int32_t __s0_685 = __p0_685; \
57774		int32x4_t __s1_685 = __p1_685; \
57775		int32x4_t __rev1_685; __rev1_685 = __builtin_shufflevector(__s1_685, __s1_685, 3, 2, 1, 0); \
57776		__ret_685 = vqrdmulhs_s32(__s0_685, __noswap_vgetq_lane_s32(__rev1_685, __p2_685)); \
57777		__ret_685; \
	59429	#define vqrdmulhs_laneq_s32(__p0_701, __p1_701, __p2_701) __extension__ ({ \
	59430	int32_t __ret_701; \
	59431	int32_t __s0_701 = __p0_701; \
	59432	int32x4_t __s1_701 = __p1_701; \
	59433	int32x4_t __rev1_701; __rev1_701 = __builtin_shufflevector(__s1_701, __s1_701, __lane_reverse_128_32); \
	59434	__ret_701 = vqrdmulhs_s32(__s0_701, __noswap_vgetq_lane_s32(__rev1_701, __p2_701)); \
	59435	__ret_701; \
57778	59436	})
57779	59437	#endif
57780	59438
57781	59439	#ifdef __LITTLE_ENDIAN__
57782		#define vqrdmulhh_laneq_s16(__p0_686, __p1_686, __p2_686) __extension__ ({ \
57783		int16_t __ret_686; \
57784		int16_t __s0_686 = __p0_686; \
57785		int16x8_t __s1_686 = __p1_686; \
57786		__ret_686 = vqrdmulhh_s16(__s0_686, vgetq_lane_s16(__s1_686, __p2_686)); \
57787		__ret_686; \
	59440	#define vqrdmulhh_laneq_s16(__p0_702, __p1_702, __p2_702) __extension__ ({ \
	59441	int16_t __ret_702; \
	59442	int16_t __s0_702 = __p0_702; \
	59443	int16x8_t __s1_702 = __p1_702; \
	59444	__ret_702 = vqrdmulhh_s16(__s0_702, vgetq_lane_s16(__s1_702, __p2_702)); \
	59445	__ret_702; \
57788	59446	})
57789	59447	#else
57790		#define vqrdmulhh_laneq_s16(__p0_687, __p1_687, __p2_687) __extension__ ({ \
57791		int16_t __ret_687; \
57792		int16_t __s0_687 = __p0_687; \
57793		int16x8_t __s1_687 = __p1_687; \
57794		int16x8_t __rev1_687; __rev1_687 = __builtin_shufflevector(__s1_687, __s1_687, 7, 6, 5, 4, 3, 2, 1, 0); \
57795		__ret_687 = vqrdmulhh_s16(__s0_687, __noswap_vgetq_lane_s16(__rev1_687, __p2_687)); \
57796		__ret_687; \
	59448	#define vqrdmulhh_laneq_s16(__p0_703, __p1_703, __p2_703) __extension__ ({ \
	59449	int16_t __ret_703; \
	59450	int16_t __s0_703 = __p0_703; \
	59451	int16x8_t __s1_703 = __p1_703; \
	59452	int16x8_t __rev1_703; __rev1_703 = __builtin_shufflevector(__s1_703, __s1_703, __lane_reverse_128_16); \
	59453	__ret_703 = vqrdmulhh_s16(__s0_703, __noswap_vgetq_lane_s16(__rev1_703, __p2_703)); \
	59454	__ret_703; \
57797	59455	})
57798	59456	#endif
57799	59457
...	...	@@ -57802,7 +59460,7 @@ __ai __attribute__((target("neon"))) int16_t vqrdmulhh_s16(int16_t __p0, int16_t
57802	59460	int32x4_t __ret; \
57803	59461	int32x4_t __s0 = __p0; \
57804	59462	int32x4_t __s1 = __p1; \
57805		__ret = (int32x4_t) __builtin_neon_vqrdmulhq_laneq_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 34); \
	59463	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqrdmulhq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 34)); \
57806	59464	__ret; \
57807	59465	})
57808	59466	#else
...	...	@@ -57810,10 +59468,10 @@ __ai __attribute__((target("neon"))) int16_t vqrdmulhh_s16(int16_t __p0, int16_t
57810	59468	int32x4_t __ret; \
57811	59469	int32x4_t __s0 = __p0; \
57812	59470	int32x4_t __s1 = __p1; \
57813		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
57814		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
57815		__ret = (int32x4_t) __builtin_neon_vqrdmulhq_laneq_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 34); \
57816		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	59471	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	59472	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	59473	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vqrdmulhq_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 34)); \
	59474	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
57817	59475	__ret; \
57818	59476	})
57819	59477	#endif
...	...	@@ -57823,7 +59481,7 @@ __ai __attribute__((target("neon"))) int16_t vqrdmulhh_s16(int16_t __p0, int16_t
57823	59481	int16x8_t __ret; \
57824	59482	int16x8_t __s0 = __p0; \
57825	59483	int16x8_t __s1 = __p1; \
57826		__ret = (int16x8_t) __builtin_neon_vqrdmulhq_laneq_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 33); \
	59484	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqrdmulhq_laneq_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 33)); \
57827	59485	__ret; \
57828	59486	})
57829	59487	#else
...	...	@@ -57831,10 +59489,10 @@ __ai __attribute__((target("neon"))) int16_t vqrdmulhh_s16(int16_t __p0, int16_t
57831	59489	int16x8_t __ret; \
57832	59490	int16x8_t __s0 = __p0; \
57833	59491	int16x8_t __s1 = __p1; \
57834		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
57835		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
57836		__ret = (int16x8_t) __builtin_neon_vqrdmulhq_laneq_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 33); \
57837		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
	59492	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_16); \
	59493	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	59494	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vqrdmulhq_laneq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 33)); \
	59495	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16); \
57838	59496	__ret; \
57839	59497	})
57840	59498	#endif
...	...	@@ -57844,7 +59502,7 @@ __ai __attribute__((target("neon"))) int16_t vqrdmulhh_s16(int16_t __p0, int16_t
57844	59502	int32x2_t __ret; \
57845	59503	int32x2_t __s0 = __p0; \
57846	59504	int32x4_t __s1 = __p1; \
57847		__ret = (int32x2_t) __builtin_neon_vqrdmulh_laneq_v((int8x8_t)__s0, (int8x16_t)__s1, __p2, 2); \
	59505	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqrdmulh_laneq_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 2)); \
57848	59506	__ret; \
57849	59507	})
57850	59508	#else
...	...	@@ -57852,10 +59510,10 @@ __ai __attribute__((target("neon"))) int16_t vqrdmulhh_s16(int16_t __p0, int16_t
57852	59510	int32x2_t __ret; \
57853	59511	int32x2_t __s0 = __p0; \
57854	59512	int32x4_t __s1 = __p1; \
57855		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
57856		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
57857		__ret = (int32x2_t) __builtin_neon_vqrdmulh_laneq_v((int8x8_t)__rev0, (int8x16_t)__rev1, __p2, 2); \
57858		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	59513	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_32); \
	59514	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	59515	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vqrdmulh_laneq_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 2)); \
	59516	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32); \
57859	59517	__ret; \
57860	59518	})
57861	59519	#endif
...	...	@@ -57865,7 +59523,7 @@ __ai __attribute__((target("neon"))) int16_t vqrdmulhh_s16(int16_t __p0, int16_t
57865	59523	int16x4_t __ret; \
57866	59524	int16x4_t __s0 = __p0; \
57867	59525	int16x8_t __s1 = __p1; \
57868		__ret = (int16x4_t) __builtin_neon_vqrdmulh_laneq_v((int8x8_t)__s0, (int8x16_t)__s1, __p2, 1); \
	59526	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqrdmulh_laneq_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 1)); \
57869	59527	__ret; \
57870	59528	})
57871	59529	#else
...	...	@@ -57873,705 +59531,705 @@ __ai __attribute__((target("neon"))) int16_t vqrdmulhh_s16(int16_t __p0, int16_t
57873	59531	int16x4_t __ret; \
57874	59532	int16x4_t __s0 = __p0; \
57875	59533	int16x8_t __s1 = __p1; \
57876		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
57877		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
57878		__ret = (int16x4_t) __builtin_neon_vqrdmulh_laneq_v((int8x8_t)__rev0, (int8x16_t)__rev1, __p2, 1); \
57879		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	59534	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_64_16); \
	59535	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_16); \
	59536	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vqrdmulh_laneq_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 1)); \
	59537	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16); \
57880	59538	__ret; \
57881	59539	})
57882	59540	#endif
57883	59541
57884	59542	__ai __attribute__((target("neon"))) uint8_t vqrshlb_u8(uint8_t __p0, int8_t __p1) {
57885	59543	uint8_t __ret;
57886		__ret = (uint8_t) __builtin_neon_vqrshlb_u8(__p0, __p1);
	59544	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vqrshlb_u8(__p0, __p1));
57887	59545	return __ret;
57888	59546	}
57889	59547	__ai __attribute__((target("neon"))) uint32_t vqrshls_u32(uint32_t __p0, int32_t __p1) {
57890	59548	uint32_t __ret;
57891		__ret = (uint32_t) __builtin_neon_vqrshls_u32(__p0, __p1);
	59549	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vqrshls_u32(__p0, __p1));
57892	59550	return __ret;
57893	59551	}
57894	59552	__ai __attribute__((target("neon"))) uint64_t vqrshld_u64(uint64_t __p0, int64_t __p1) {
57895	59553	uint64_t __ret;
57896		__ret = (uint64_t) __builtin_neon_vqrshld_u64(__p0, __p1);
	59554	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vqrshld_u64(__p0, __p1));
57897	59555	return __ret;
57898	59556	}
57899	59557	__ai __attribute__((target("neon"))) uint16_t vqrshlh_u16(uint16_t __p0, int16_t __p1) {
57900	59558	uint16_t __ret;
57901		__ret = (uint16_t) __builtin_neon_vqrshlh_u16(__p0, __p1);
	59559	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vqrshlh_u16(__p0, __p1));
57902	59560	return __ret;
57903	59561	}
57904	59562	__ai __attribute__((target("neon"))) int8_t vqrshlb_s8(int8_t __p0, int8_t __p1) {
57905	59563	int8_t __ret;
57906		__ret = (int8_t) __builtin_neon_vqrshlb_s8(__p0, __p1);
	59564	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vqrshlb_s8(__p0, __p1));
57907	59565	return __ret;
57908	59566	}
57909	59567	__ai __attribute__((target("neon"))) int32_t vqrshls_s32(int32_t __p0, int32_t __p1) {
57910	59568	int32_t __ret;
57911		__ret = (int32_t) __builtin_neon_vqrshls_s32(__p0, __p1);
	59569	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vqrshls_s32(__p0, __p1));
57912	59570	return __ret;
57913	59571	}
57914	59572	__ai __attribute__((target("neon"))) int64_t vqrshld_s64(int64_t __p0, int64_t __p1) {
57915	59573	int64_t __ret;
57916		__ret = (int64_t) __builtin_neon_vqrshld_s64(__p0, __p1);
	59574	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vqrshld_s64(__p0, __p1));
57917	59575	return __ret;
57918	59576	}
57919	59577	__ai __attribute__((target("neon"))) int16_t vqrshlh_s16(int16_t __p0, int16_t __p1) {
57920	59578	int16_t __ret;
57921		__ret = (int16_t) __builtin_neon_vqrshlh_s16(__p0, __p1);
	59579	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vqrshlh_s16(__p0, __p1));
57922	59580	return __ret;
57923	59581	}
57924	59582	#ifdef __LITTLE_ENDIAN__
57925		#define vqrshrn_high_n_u32(__p0_688, __p1_688, __p2_688) __extension__ ({ \
57926		uint16x8_t __ret_688; \
57927		uint16x4_t __s0_688 = __p0_688; \
57928		uint32x4_t __s1_688 = __p1_688; \
57929		__ret_688 = (uint16x8_t)(vcombine_u16((uint16x4_t)(__s0_688), (uint16x4_t)(vqrshrn_n_u32(__s1_688, __p2_688)))); \
57930		__ret_688; \
	59583	#define vqrshrn_high_n_u32(__p0_704, __p1_704, __p2_704) __extension__ ({ \
	59584	uint16x8_t __ret_704; \
	59585	uint16x4_t __s0_704 = __p0_704; \
	59586	uint32x4_t __s1_704 = __p1_704; \
	59587	__ret_704 = __builtin_bit_cast(uint16x8_t, vcombine_u16(__builtin_bit_cast(uint16x4_t, __s0_704), __builtin_bit_cast(uint16x4_t, vqrshrn_n_u32(__s1_704, __p2_704)))); \
	59588	__ret_704; \
57931	59589	})
57932	59590	#else
57933		#define vqrshrn_high_n_u32(__p0_689, __p1_689, __p2_689) __extension__ ({ \
57934		uint16x8_t __ret_689; \
57935		uint16x4_t __s0_689 = __p0_689; \
57936		uint32x4_t __s1_689 = __p1_689; \
57937		uint16x4_t __rev0_689; __rev0_689 = __builtin_shufflevector(__s0_689, __s0_689, 3, 2, 1, 0); \
57938		uint32x4_t __rev1_689; __rev1_689 = __builtin_shufflevector(__s1_689, __s1_689, 3, 2, 1, 0); \
57939		__ret_689 = (uint16x8_t)(__noswap_vcombine_u16((uint16x4_t)(__rev0_689), (uint16x4_t)(__noswap_vqrshrn_n_u32(__rev1_689, __p2_689)))); \
57940		__ret_689 = __builtin_shufflevector(__ret_689, __ret_689, 7, 6, 5, 4, 3, 2, 1, 0); \
57941		__ret_689; \
	59591	#define vqrshrn_high_n_u32(__p0_705, __p1_705, __p2_705) __extension__ ({ \
	59592	uint16x8_t __ret_705; \
	59593	uint16x4_t __s0_705 = __p0_705; \
	59594	uint32x4_t __s1_705 = __p1_705; \
	59595	uint16x4_t __rev0_705; __rev0_705 = __builtin_shufflevector(__s0_705, __s0_705, __lane_reverse_64_16); \
	59596	uint32x4_t __rev1_705; __rev1_705 = __builtin_shufflevector(__s1_705, __s1_705, __lane_reverse_128_32); \
	59597	__ret_705 = __builtin_bit_cast(uint16x8_t, __noswap_vcombine_u16(__builtin_bit_cast(uint16x4_t, __rev0_705), __builtin_bit_cast(uint16x4_t, __noswap_vqrshrn_n_u32(__rev1_705, __p2_705)))); \
	59598	__ret_705 = __builtin_shufflevector(__ret_705, __ret_705, __lane_reverse_128_16); \
	59599	__ret_705; \
57942	59600	})
57943	59601	#endif
57944	59602
57945	59603	#ifdef __LITTLE_ENDIAN__
57946		#define vqrshrn_high_n_u64(__p0_690, __p1_690, __p2_690) __extension__ ({ \
57947		uint32x4_t __ret_690; \
57948		uint32x2_t __s0_690 = __p0_690; \
57949		uint64x2_t __s1_690 = __p1_690; \
57950		__ret_690 = (uint32x4_t)(vcombine_u32((uint32x2_t)(__s0_690), (uint32x2_t)(vqrshrn_n_u64(__s1_690, __p2_690)))); \
57951		__ret_690; \
	59604	#define vqrshrn_high_n_u64(__p0_706, __p1_706, __p2_706) __extension__ ({ \
	59605	uint32x4_t __ret_706; \
	59606	uint32x2_t __s0_706 = __p0_706; \
	59607	uint64x2_t __s1_706 = __p1_706; \
	59608	__ret_706 = __builtin_bit_cast(uint32x4_t, vcombine_u32(__builtin_bit_cast(uint32x2_t, __s0_706), __builtin_bit_cast(uint32x2_t, vqrshrn_n_u64(__s1_706, __p2_706)))); \
	59609	__ret_706; \
57952	59610	})
57953	59611	#else
57954		#define vqrshrn_high_n_u64(__p0_691, __p1_691, __p2_691) __extension__ ({ \
57955		uint32x4_t __ret_691; \
57956		uint32x2_t __s0_691 = __p0_691; \
57957		uint64x2_t __s1_691 = __p1_691; \
57958		uint32x2_t __rev0_691; __rev0_691 = __builtin_shufflevector(__s0_691, __s0_691, 1, 0); \
57959		uint64x2_t __rev1_691; __rev1_691 = __builtin_shufflevector(__s1_691, __s1_691, 1, 0); \
57960		__ret_691 = (uint32x4_t)(__noswap_vcombine_u32((uint32x2_t)(__rev0_691), (uint32x2_t)(__noswap_vqrshrn_n_u64(__rev1_691, __p2_691)))); \
57961		__ret_691 = __builtin_shufflevector(__ret_691, __ret_691, 3, 2, 1, 0); \
57962		__ret_691; \
	59612	#define vqrshrn_high_n_u64(__p0_707, __p1_707, __p2_707) __extension__ ({ \
	59613	uint32x4_t __ret_707; \
	59614	uint32x2_t __s0_707 = __p0_707; \
	59615	uint64x2_t __s1_707 = __p1_707; \
	59616	uint32x2_t __rev0_707; __rev0_707 = __builtin_shufflevector(__s0_707, __s0_707, __lane_reverse_64_32); \
	59617	uint64x2_t __rev1_707; __rev1_707 = __builtin_shufflevector(__s1_707, __s1_707, __lane_reverse_128_64); \
	59618	__ret_707 = __builtin_bit_cast(uint32x4_t, __noswap_vcombine_u32(__builtin_bit_cast(uint32x2_t, __rev0_707), __builtin_bit_cast(uint32x2_t, __noswap_vqrshrn_n_u64(__rev1_707, __p2_707)))); \
	59619	__ret_707 = __builtin_shufflevector(__ret_707, __ret_707, __lane_reverse_128_32); \
	59620	__ret_707; \
57963	59621	})
57964	59622	#endif
57965	59623
57966	59624	#ifdef __LITTLE_ENDIAN__
57967		#define vqrshrn_high_n_u16(__p0_692, __p1_692, __p2_692) __extension__ ({ \
57968		uint8x16_t __ret_692; \
57969		uint8x8_t __s0_692 = __p0_692; \
57970		uint16x8_t __s1_692 = __p1_692; \
57971		__ret_692 = (uint8x16_t)(vcombine_u8((uint8x8_t)(__s0_692), (uint8x8_t)(vqrshrn_n_u16(__s1_692, __p2_692)))); \
57972		__ret_692; \
	59625	#define vqrshrn_high_n_u16(__p0_708, __p1_708, __p2_708) __extension__ ({ \
	59626	uint8x16_t __ret_708; \
	59627	uint8x8_t __s0_708 = __p0_708; \
	59628	uint16x8_t __s1_708 = __p1_708; \
	59629	__ret_708 = __builtin_bit_cast(uint8x16_t, vcombine_u8(__builtin_bit_cast(uint8x8_t, __s0_708), __builtin_bit_cast(uint8x8_t, vqrshrn_n_u16(__s1_708, __p2_708)))); \
	59630	__ret_708; \
57973	59631	})
57974	59632	#else
57975		#define vqrshrn_high_n_u16(__p0_693, __p1_693, __p2_693) __extension__ ({ \
57976		uint8x16_t __ret_693; \
57977		uint8x8_t __s0_693 = __p0_693; \
57978		uint16x8_t __s1_693 = __p1_693; \
57979		uint8x8_t __rev0_693; __rev0_693 = __builtin_shufflevector(__s0_693, __s0_693, 7, 6, 5, 4, 3, 2, 1, 0); \
57980		uint16x8_t __rev1_693; __rev1_693 = __builtin_shufflevector(__s1_693, __s1_693, 7, 6, 5, 4, 3, 2, 1, 0); \
57981		__ret_693 = (uint8x16_t)(__noswap_vcombine_u8((uint8x8_t)(__rev0_693), (uint8x8_t)(__noswap_vqrshrn_n_u16(__rev1_693, __p2_693)))); \
57982		__ret_693 = __builtin_shufflevector(__ret_693, __ret_693, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
57983		__ret_693; \
	59633	#define vqrshrn_high_n_u16(__p0_709, __p1_709, __p2_709) __extension__ ({ \
	59634	uint8x16_t __ret_709; \
	59635	uint8x8_t __s0_709 = __p0_709; \
	59636	uint16x8_t __s1_709 = __p1_709; \
	59637	uint8x8_t __rev0_709; __rev0_709 = __builtin_shufflevector(__s0_709, __s0_709, __lane_reverse_64_8); \
	59638	uint16x8_t __rev1_709; __rev1_709 = __builtin_shufflevector(__s1_709, __s1_709, __lane_reverse_128_16); \
	59639	__ret_709 = __builtin_bit_cast(uint8x16_t, __noswap_vcombine_u8(__builtin_bit_cast(uint8x8_t, __rev0_709), __builtin_bit_cast(uint8x8_t, __noswap_vqrshrn_n_u16(__rev1_709, __p2_709)))); \
	59640	__ret_709 = __builtin_shufflevector(__ret_709, __ret_709, __lane_reverse_128_8); \
	59641	__ret_709; \
57984	59642	})
57985	59643	#endif
57986	59644
57987	59645	#ifdef __LITTLE_ENDIAN__
57988		#define vqrshrn_high_n_s32(__p0_694, __p1_694, __p2_694) __extension__ ({ \
57989		int16x8_t __ret_694; \
57990		int16x4_t __s0_694 = __p0_694; \
57991		int32x4_t __s1_694 = __p1_694; \
57992		__ret_694 = (int16x8_t)(vcombine_s16((int16x4_t)(__s0_694), (int16x4_t)(vqrshrn_n_s32(__s1_694, __p2_694)))); \
57993		__ret_694; \
	59646	#define vqrshrn_high_n_s32(__p0_710, __p1_710, __p2_710) __extension__ ({ \
	59647	int16x8_t __ret_710; \
	59648	int16x4_t __s0_710 = __p0_710; \
	59649	int32x4_t __s1_710 = __p1_710; \
	59650	__ret_710 = __builtin_bit_cast(int16x8_t, vcombine_s16(__builtin_bit_cast(int16x4_t, __s0_710), __builtin_bit_cast(int16x4_t, vqrshrn_n_s32(__s1_710, __p2_710)))); \
	59651	__ret_710; \
57994	59652	})
57995	59653	#else
57996		#define vqrshrn_high_n_s32(__p0_695, __p1_695, __p2_695) __extension__ ({ \
57997		int16x8_t __ret_695; \
57998		int16x4_t __s0_695 = __p0_695; \
57999		int32x4_t __s1_695 = __p1_695; \
58000		int16x4_t __rev0_695; __rev0_695 = __builtin_shufflevector(__s0_695, __s0_695, 3, 2, 1, 0); \
58001		int32x4_t __rev1_695; __rev1_695 = __builtin_shufflevector(__s1_695, __s1_695, 3, 2, 1, 0); \
58002		__ret_695 = (int16x8_t)(__noswap_vcombine_s16((int16x4_t)(__rev0_695), (int16x4_t)(__noswap_vqrshrn_n_s32(__rev1_695, __p2_695)))); \
58003		__ret_695 = __builtin_shufflevector(__ret_695, __ret_695, 7, 6, 5, 4, 3, 2, 1, 0); \
58004		__ret_695; \
	59654	#define vqrshrn_high_n_s32(__p0_711, __p1_711, __p2_711) __extension__ ({ \
	59655	int16x8_t __ret_711; \
	59656	int16x4_t __s0_711 = __p0_711; \
	59657	int32x4_t __s1_711 = __p1_711; \
	59658	int16x4_t __rev0_711; __rev0_711 = __builtin_shufflevector(__s0_711, __s0_711, __lane_reverse_64_16); \
	59659	int32x4_t __rev1_711; __rev1_711 = __builtin_shufflevector(__s1_711, __s1_711, __lane_reverse_128_32); \
	59660	__ret_711 = __builtin_bit_cast(int16x8_t, __noswap_vcombine_s16(__builtin_bit_cast(int16x4_t, __rev0_711), __builtin_bit_cast(int16x4_t, __noswap_vqrshrn_n_s32(__rev1_711, __p2_711)))); \
	59661	__ret_711 = __builtin_shufflevector(__ret_711, __ret_711, __lane_reverse_128_16); \
	59662	__ret_711; \
58005	59663	})
58006	59664	#endif
58007	59665
58008	59666	#ifdef __LITTLE_ENDIAN__
58009		#define vqrshrn_high_n_s64(__p0_696, __p1_696, __p2_696) __extension__ ({ \
58010		int32x4_t __ret_696; \
58011		int32x2_t __s0_696 = __p0_696; \
58012		int64x2_t __s1_696 = __p1_696; \
58013		__ret_696 = (int32x4_t)(vcombine_s32((int32x2_t)(__s0_696), (int32x2_t)(vqrshrn_n_s64(__s1_696, __p2_696)))); \
58014		__ret_696; \
	59667	#define vqrshrn_high_n_s64(__p0_712, __p1_712, __p2_712) __extension__ ({ \
	59668	int32x4_t __ret_712; \
	59669	int32x2_t __s0_712 = __p0_712; \
	59670	int64x2_t __s1_712 = __p1_712; \
	59671	__ret_712 = __builtin_bit_cast(int32x4_t, vcombine_s32(__builtin_bit_cast(int32x2_t, __s0_712), __builtin_bit_cast(int32x2_t, vqrshrn_n_s64(__s1_712, __p2_712)))); \
	59672	__ret_712; \
58015	59673	})
58016	59674	#else
58017		#define vqrshrn_high_n_s64(__p0_697, __p1_697, __p2_697) __extension__ ({ \
58018		int32x4_t __ret_697; \
58019		int32x2_t __s0_697 = __p0_697; \
58020		int64x2_t __s1_697 = __p1_697; \
58021		int32x2_t __rev0_697; __rev0_697 = __builtin_shufflevector(__s0_697, __s0_697, 1, 0); \
58022		int64x2_t __rev1_697; __rev1_697 = __builtin_shufflevector(__s1_697, __s1_697, 1, 0); \
58023		__ret_697 = (int32x4_t)(__noswap_vcombine_s32((int32x2_t)(__rev0_697), (int32x2_t)(__noswap_vqrshrn_n_s64(__rev1_697, __p2_697)))); \
58024		__ret_697 = __builtin_shufflevector(__ret_697, __ret_697, 3, 2, 1, 0); \
58025		__ret_697; \
	59675	#define vqrshrn_high_n_s64(__p0_713, __p1_713, __p2_713) __extension__ ({ \
	59676	int32x4_t __ret_713; \
	59677	int32x2_t __s0_713 = __p0_713; \
	59678	int64x2_t __s1_713 = __p1_713; \
	59679	int32x2_t __rev0_713; __rev0_713 = __builtin_shufflevector(__s0_713, __s0_713, __lane_reverse_64_32); \
	59680	int64x2_t __rev1_713; __rev1_713 = __builtin_shufflevector(__s1_713, __s1_713, __lane_reverse_128_64); \
	59681	__ret_713 = __builtin_bit_cast(int32x4_t, __noswap_vcombine_s32(__builtin_bit_cast(int32x2_t, __rev0_713), __builtin_bit_cast(int32x2_t, __noswap_vqrshrn_n_s64(__rev1_713, __p2_713)))); \
	59682	__ret_713 = __builtin_shufflevector(__ret_713, __ret_713, __lane_reverse_128_32); \
	59683	__ret_713; \
58026	59684	})
58027	59685	#endif
58028	59686
58029	59687	#ifdef __LITTLE_ENDIAN__
58030		#define vqrshrn_high_n_s16(__p0_698, __p1_698, __p2_698) __extension__ ({ \
58031		int8x16_t __ret_698; \
58032		int8x8_t __s0_698 = __p0_698; \
58033		int16x8_t __s1_698 = __p1_698; \
58034		__ret_698 = (int8x16_t)(vcombine_s8((int8x8_t)(__s0_698), (int8x8_t)(vqrshrn_n_s16(__s1_698, __p2_698)))); \
58035		__ret_698; \
	59688	#define vqrshrn_high_n_s16(__p0_714, __p1_714, __p2_714) __extension__ ({ \
	59689	int8x16_t __ret_714; \
	59690	int8x8_t __s0_714 = __p0_714; \
	59691	int16x8_t __s1_714 = __p1_714; \
	59692	__ret_714 = __builtin_bit_cast(int8x16_t, vcombine_s8(__builtin_bit_cast(int8x8_t, __s0_714), __builtin_bit_cast(int8x8_t, vqrshrn_n_s16(__s1_714, __p2_714)))); \
	59693	__ret_714; \
58036	59694	})
58037	59695	#else
58038		#define vqrshrn_high_n_s16(__p0_699, __p1_699, __p2_699) __extension__ ({ \
58039		int8x16_t __ret_699; \
58040		int8x8_t __s0_699 = __p0_699; \
58041		int16x8_t __s1_699 = __p1_699; \
58042		int8x8_t __rev0_699; __rev0_699 = __builtin_shufflevector(__s0_699, __s0_699, 7, 6, 5, 4, 3, 2, 1, 0); \
58043		int16x8_t __rev1_699; __rev1_699 = __builtin_shufflevector(__s1_699, __s1_699, 7, 6, 5, 4, 3, 2, 1, 0); \
58044		__ret_699 = (int8x16_t)(__noswap_vcombine_s8((int8x8_t)(__rev0_699), (int8x8_t)(__noswap_vqrshrn_n_s16(__rev1_699, __p2_699)))); \
58045		__ret_699 = __builtin_shufflevector(__ret_699, __ret_699, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
58046		__ret_699; \
	59696	#define vqrshrn_high_n_s16(__p0_715, __p1_715, __p2_715) __extension__ ({ \
	59697	int8x16_t __ret_715; \
	59698	int8x8_t __s0_715 = __p0_715; \
	59699	int16x8_t __s1_715 = __p1_715; \
	59700	int8x8_t __rev0_715; __rev0_715 = __builtin_shufflevector(__s0_715, __s0_715, __lane_reverse_64_8); \
	59701	int16x8_t __rev1_715; __rev1_715 = __builtin_shufflevector(__s1_715, __s1_715, __lane_reverse_128_16); \
	59702	__ret_715 = __builtin_bit_cast(int8x16_t, __noswap_vcombine_s8(__builtin_bit_cast(int8x8_t, __rev0_715), __builtin_bit_cast(int8x8_t, __noswap_vqrshrn_n_s16(__rev1_715, __p2_715)))); \
	59703	__ret_715 = __builtin_shufflevector(__ret_715, __ret_715, __lane_reverse_128_8); \
	59704	__ret_715; \
58047	59705	})
58048	59706	#endif
58049	59707
58050	59708	#define vqrshrns_n_u32(__p0, __p1) __extension__ ({ \
58051	59709	uint16_t __ret; \
58052	59710	uint32_t __s0 = __p0; \
58053		__ret = (uint16_t) __builtin_neon_vqrshrns_n_u32(__s0, __p1); \
	59711	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vqrshrns_n_u32(__s0, __p1)); \
58054	59712	__ret; \
58055	59713	})
58056	59714	#define vqrshrnd_n_u64(__p0, __p1) __extension__ ({ \
58057	59715	uint32_t __ret; \
58058	59716	uint64_t __s0 = __p0; \
58059		__ret = (uint32_t) __builtin_neon_vqrshrnd_n_u64(__s0, __p1); \
	59717	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vqrshrnd_n_u64(__s0, __p1)); \
58060	59718	__ret; \
58061	59719	})
58062	59720	#define vqrshrnh_n_u16(__p0, __p1) __extension__ ({ \
58063	59721	uint8_t __ret; \
58064	59722	uint16_t __s0 = __p0; \
58065		__ret = (uint8_t) __builtin_neon_vqrshrnh_n_u16(__s0, __p1); \
	59723	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vqrshrnh_n_u16(__s0, __p1)); \
58066	59724	__ret; \
58067	59725	})
58068	59726	#define vqrshrns_n_s32(__p0, __p1) __extension__ ({ \
58069	59727	int16_t __ret; \
58070	59728	int32_t __s0 = __p0; \
58071		__ret = (int16_t) __builtin_neon_vqrshrns_n_s32(__s0, __p1); \
	59729	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vqrshrns_n_s32(__s0, __p1)); \
58072	59730	__ret; \
58073	59731	})
58074	59732	#define vqrshrnd_n_s64(__p0, __p1) __extension__ ({ \
58075	59733	int32_t __ret; \
58076	59734	int64_t __s0 = __p0; \
58077		__ret = (int32_t) __builtin_neon_vqrshrnd_n_s64(__s0, __p1); \
	59735	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vqrshrnd_n_s64(__s0, __p1)); \
58078	59736	__ret; \
58079	59737	})
58080	59738	#define vqrshrnh_n_s16(__p0, __p1) __extension__ ({ \
58081	59739	int8_t __ret; \
58082	59740	int16_t __s0 = __p0; \
58083		__ret = (int8_t) __builtin_neon_vqrshrnh_n_s16(__s0, __p1); \
	59741	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vqrshrnh_n_s16(__s0, __p1)); \
58084	59742	__ret; \
58085	59743	})
58086	59744	#ifdef __LITTLE_ENDIAN__
58087		#define vqrshrun_high_n_s32(__p0_700, __p1_700, __p2_700) __extension__ ({ \
58088		int16x8_t __ret_700; \
58089		int16x4_t __s0_700 = __p0_700; \
58090		int32x4_t __s1_700 = __p1_700; \
58091		__ret_700 = (int16x8_t)(vcombine_s16((int16x4_t)(__s0_700), (int16x4_t)(vqrshrun_n_s32(__s1_700, __p2_700)))); \
58092		__ret_700; \
	59745	#define vqrshrun_high_n_s32(__p0_716, __p1_716, __p2_716) __extension__ ({ \
	59746	int16x8_t __ret_716; \
	59747	int16x4_t __s0_716 = __p0_716; \
	59748	int32x4_t __s1_716 = __p1_716; \
	59749	__ret_716 = __builtin_bit_cast(int16x8_t, vcombine_s16(__builtin_bit_cast(int16x4_t, __s0_716), __builtin_bit_cast(int16x4_t, vqrshrun_n_s32(__s1_716, __p2_716)))); \
	59750	__ret_716; \
58093	59751	})
58094	59752	#else
58095		#define vqrshrun_high_n_s32(__p0_701, __p1_701, __p2_701) __extension__ ({ \
58096		int16x8_t __ret_701; \
58097		int16x4_t __s0_701 = __p0_701; \
58098		int32x4_t __s1_701 = __p1_701; \
58099		int16x4_t __rev0_701; __rev0_701 = __builtin_shufflevector(__s0_701, __s0_701, 3, 2, 1, 0); \
58100		int32x4_t __rev1_701; __rev1_701 = __builtin_shufflevector(__s1_701, __s1_701, 3, 2, 1, 0); \
58101		__ret_701 = (int16x8_t)(__noswap_vcombine_s16((int16x4_t)(__rev0_701), (int16x4_t)(__noswap_vqrshrun_n_s32(__rev1_701, __p2_701)))); \
58102		__ret_701 = __builtin_shufflevector(__ret_701, __ret_701, 7, 6, 5, 4, 3, 2, 1, 0); \
58103		__ret_701; \
	59753	#define vqrshrun_high_n_s32(__p0_717, __p1_717, __p2_717) __extension__ ({ \
	59754	int16x8_t __ret_717; \
	59755	int16x4_t __s0_717 = __p0_717; \
	59756	int32x4_t __s1_717 = __p1_717; \
	59757	int16x4_t __rev0_717; __rev0_717 = __builtin_shufflevector(__s0_717, __s0_717, __lane_reverse_64_16); \
	59758	int32x4_t __rev1_717; __rev1_717 = __builtin_shufflevector(__s1_717, __s1_717, __lane_reverse_128_32); \
	59759	__ret_717 = __builtin_bit_cast(int16x8_t, __noswap_vcombine_s16(__builtin_bit_cast(int16x4_t, __rev0_717), __builtin_bit_cast(int16x4_t, __noswap_vqrshrun_n_s32(__rev1_717, __p2_717)))); \
	59760	__ret_717 = __builtin_shufflevector(__ret_717, __ret_717, __lane_reverse_128_16); \
	59761	__ret_717; \
58104	59762	})
58105	59763	#endif
58106	59764
58107	59765	#ifdef __LITTLE_ENDIAN__
58108		#define vqrshrun_high_n_s64(__p0_702, __p1_702, __p2_702) __extension__ ({ \
58109		int32x4_t __ret_702; \
58110		int32x2_t __s0_702 = __p0_702; \
58111		int64x2_t __s1_702 = __p1_702; \
58112		__ret_702 = (int32x4_t)(vcombine_s32((int32x2_t)(__s0_702), (int32x2_t)(vqrshrun_n_s64(__s1_702, __p2_702)))); \
58113		__ret_702; \
	59766	#define vqrshrun_high_n_s64(__p0_718, __p1_718, __p2_718) __extension__ ({ \
	59767	int32x4_t __ret_718; \
	59768	int32x2_t __s0_718 = __p0_718; \
	59769	int64x2_t __s1_718 = __p1_718; \
	59770	__ret_718 = __builtin_bit_cast(int32x4_t, vcombine_s32(__builtin_bit_cast(int32x2_t, __s0_718), __builtin_bit_cast(int32x2_t, vqrshrun_n_s64(__s1_718, __p2_718)))); \
	59771	__ret_718; \
58114	59772	})
58115	59773	#else
58116		#define vqrshrun_high_n_s64(__p0_703, __p1_703, __p2_703) __extension__ ({ \
58117		int32x4_t __ret_703; \
58118		int32x2_t __s0_703 = __p0_703; \
58119		int64x2_t __s1_703 = __p1_703; \
58120		int32x2_t __rev0_703; __rev0_703 = __builtin_shufflevector(__s0_703, __s0_703, 1, 0); \
58121		int64x2_t __rev1_703; __rev1_703 = __builtin_shufflevector(__s1_703, __s1_703, 1, 0); \
58122		__ret_703 = (int32x4_t)(__noswap_vcombine_s32((int32x2_t)(__rev0_703), (int32x2_t)(__noswap_vqrshrun_n_s64(__rev1_703, __p2_703)))); \
58123		__ret_703 = __builtin_shufflevector(__ret_703, __ret_703, 3, 2, 1, 0); \
58124		__ret_703; \
	59774	#define vqrshrun_high_n_s64(__p0_719, __p1_719, __p2_719) __extension__ ({ \
	59775	int32x4_t __ret_719; \
	59776	int32x2_t __s0_719 = __p0_719; \
	59777	int64x2_t __s1_719 = __p1_719; \
	59778	int32x2_t __rev0_719; __rev0_719 = __builtin_shufflevector(__s0_719, __s0_719, __lane_reverse_64_32); \
	59779	int64x2_t __rev1_719; __rev1_719 = __builtin_shufflevector(__s1_719, __s1_719, __lane_reverse_128_64); \
	59780	__ret_719 = __builtin_bit_cast(int32x4_t, __noswap_vcombine_s32(__builtin_bit_cast(int32x2_t, __rev0_719), __builtin_bit_cast(int32x2_t, __noswap_vqrshrun_n_s64(__rev1_719, __p2_719)))); \
	59781	__ret_719 = __builtin_shufflevector(__ret_719, __ret_719, __lane_reverse_128_32); \
	59782	__ret_719; \
58125	59783	})
58126	59784	#endif
58127	59785
58128	59786	#ifdef __LITTLE_ENDIAN__
58129		#define vqrshrun_high_n_s16(__p0_704, __p1_704, __p2_704) __extension__ ({ \
58130		int8x16_t __ret_704; \
58131		int8x8_t __s0_704 = __p0_704; \
58132		int16x8_t __s1_704 = __p1_704; \
58133		__ret_704 = (int8x16_t)(vcombine_s8((int8x8_t)(__s0_704), (int8x8_t)(vqrshrun_n_s16(__s1_704, __p2_704)))); \
58134		__ret_704; \
	59787	#define vqrshrun_high_n_s16(__p0_720, __p1_720, __p2_720) __extension__ ({ \
	59788	int8x16_t __ret_720; \
	59789	int8x8_t __s0_720 = __p0_720; \
	59790	int16x8_t __s1_720 = __p1_720; \
	59791	__ret_720 = __builtin_bit_cast(int8x16_t, vcombine_s8(__builtin_bit_cast(int8x8_t, __s0_720), __builtin_bit_cast(int8x8_t, vqrshrun_n_s16(__s1_720, __p2_720)))); \
	59792	__ret_720; \
58135	59793	})
58136	59794	#else
58137		#define vqrshrun_high_n_s16(__p0_705, __p1_705, __p2_705) __extension__ ({ \
58138		int8x16_t __ret_705; \
58139		int8x8_t __s0_705 = __p0_705; \
58140		int16x8_t __s1_705 = __p1_705; \
58141		int8x8_t __rev0_705; __rev0_705 = __builtin_shufflevector(__s0_705, __s0_705, 7, 6, 5, 4, 3, 2, 1, 0); \
58142		int16x8_t __rev1_705; __rev1_705 = __builtin_shufflevector(__s1_705, __s1_705, 7, 6, 5, 4, 3, 2, 1, 0); \
58143		__ret_705 = (int8x16_t)(__noswap_vcombine_s8((int8x8_t)(__rev0_705), (int8x8_t)(__noswap_vqrshrun_n_s16(__rev1_705, __p2_705)))); \
58144		__ret_705 = __builtin_shufflevector(__ret_705, __ret_705, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
58145		__ret_705; \
	59795	#define vqrshrun_high_n_s16(__p0_721, __p1_721, __p2_721) __extension__ ({ \
	59796	int8x16_t __ret_721; \
	59797	int8x8_t __s0_721 = __p0_721; \
	59798	int16x8_t __s1_721 = __p1_721; \
	59799	int8x8_t __rev0_721; __rev0_721 = __builtin_shufflevector(__s0_721, __s0_721, __lane_reverse_64_8); \
	59800	int16x8_t __rev1_721; __rev1_721 = __builtin_shufflevector(__s1_721, __s1_721, __lane_reverse_128_16); \
	59801	__ret_721 = __builtin_bit_cast(int8x16_t, __noswap_vcombine_s8(__builtin_bit_cast(int8x8_t, __rev0_721), __builtin_bit_cast(int8x8_t, __noswap_vqrshrun_n_s16(__rev1_721, __p2_721)))); \
	59802	__ret_721 = __builtin_shufflevector(__ret_721, __ret_721, __lane_reverse_128_8); \
	59803	__ret_721; \
58146	59804	})
58147	59805	#endif
58148	59806
58149	59807	#define vqrshruns_n_s32(__p0, __p1) __extension__ ({ \
58150	59808	uint16_t __ret; \
58151	59809	int32_t __s0 = __p0; \
58152		__ret = (uint16_t) __builtin_neon_vqrshruns_n_s32(__s0, __p1); \
	59810	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vqrshruns_n_s32(__s0, __p1)); \
58153	59811	__ret; \
58154	59812	})
58155	59813	#define vqrshrund_n_s64(__p0, __p1) __extension__ ({ \
58156	59814	uint32_t __ret; \
58157	59815	int64_t __s0 = __p0; \
58158		__ret = (uint32_t) __builtin_neon_vqrshrund_n_s64(__s0, __p1); \
	59816	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vqrshrund_n_s64(__s0, __p1)); \
58159	59817	__ret; \
58160	59818	})
58161	59819	#define vqrshrunh_n_s16(__p0, __p1) __extension__ ({ \
58162	59820	uint8_t __ret; \
58163	59821	int16_t __s0 = __p0; \
58164		__ret = (uint8_t) __builtin_neon_vqrshrunh_n_s16(__s0, __p1); \
	59822	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vqrshrunh_n_s16(__s0, __p1)); \
58165	59823	__ret; \
58166	59824	})
58167	59825	__ai __attribute__((target("neon"))) uint8_t vqshlb_u8(uint8_t __p0, int8_t __p1) {
58168	59826	uint8_t __ret;
58169		__ret = (uint8_t) __builtin_neon_vqshlb_u8(__p0, __p1);
	59827	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vqshlb_u8(__p0, __p1));
58170	59828	return __ret;
58171	59829	}
58172	59830	__ai __attribute__((target("neon"))) uint32_t vqshls_u32(uint32_t __p0, int32_t __p1) {
58173	59831	uint32_t __ret;
58174		__ret = (uint32_t) __builtin_neon_vqshls_u32(__p0, __p1);
	59832	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vqshls_u32(__p0, __p1));
58175	59833	return __ret;
58176	59834	}
58177	59835	__ai __attribute__((target("neon"))) uint64_t vqshld_u64(uint64_t __p0, int64_t __p1) {
58178	59836	uint64_t __ret;
58179		__ret = (uint64_t) __builtin_neon_vqshld_u64(__p0, __p1);
	59837	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vqshld_u64(__p0, __p1));
58180	59838	return __ret;
58181	59839	}
58182	59840	__ai __attribute__((target("neon"))) uint16_t vqshlh_u16(uint16_t __p0, int16_t __p1) {
58183	59841	uint16_t __ret;
58184		__ret = (uint16_t) __builtin_neon_vqshlh_u16(__p0, __p1);
	59842	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vqshlh_u16(__p0, __p1));
58185	59843	return __ret;
58186	59844	}
58187	59845	__ai __attribute__((target("neon"))) int8_t vqshlb_s8(int8_t __p0, int8_t __p1) {
58188	59846	int8_t __ret;
58189		__ret = (int8_t) __builtin_neon_vqshlb_s8(__p0, __p1);
	59847	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vqshlb_s8(__p0, __p1));
58190	59848	return __ret;
58191	59849	}
58192	59850	__ai __attribute__((target("neon"))) int32_t vqshls_s32(int32_t __p0, int32_t __p1) {
58193	59851	int32_t __ret;
58194		__ret = (int32_t) __builtin_neon_vqshls_s32(__p0, __p1);
	59852	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vqshls_s32(__p0, __p1));
58195	59853	return __ret;
58196	59854	}
58197	59855	__ai __attribute__((target("neon"))) int64_t vqshld_s64(int64_t __p0, int64_t __p1) {
58198	59856	int64_t __ret;
58199		__ret = (int64_t) __builtin_neon_vqshld_s64(__p0, __p1);
	59857	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vqshld_s64(__p0, __p1));
58200	59858	return __ret;
58201	59859	}
58202	59860	__ai __attribute__((target("neon"))) int16_t vqshlh_s16(int16_t __p0, int16_t __p1) {
58203	59861	int16_t __ret;
58204		__ret = (int16_t) __builtin_neon_vqshlh_s16(__p0, __p1);
	59862	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vqshlh_s16(__p0, __p1));
58205	59863	return __ret;
58206	59864	}
58207	59865	#define vqshlb_n_u8(__p0, __p1) __extension__ ({ \
58208	59866	uint8_t __ret; \
58209	59867	uint8_t __s0 = __p0; \
58210		__ret = (uint8_t) __builtin_neon_vqshlb_n_u8(__s0, __p1); \
	59868	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vqshlb_n_u8(__s0, __p1)); \
58211	59869	__ret; \
58212	59870	})
58213	59871	#define vqshls_n_u32(__p0, __p1) __extension__ ({ \
58214	59872	uint32_t __ret; \
58215	59873	uint32_t __s0 = __p0; \
58216		__ret = (uint32_t) __builtin_neon_vqshls_n_u32(__s0, __p1); \
	59874	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vqshls_n_u32(__s0, __p1)); \
58217	59875	__ret; \
58218	59876	})
58219	59877	#define vqshld_n_u64(__p0, __p1) __extension__ ({ \
58220	59878	uint64_t __ret; \
58221	59879	uint64_t __s0 = __p0; \
58222		__ret = (uint64_t) __builtin_neon_vqshld_n_u64(__s0, __p1); \
	59880	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vqshld_n_u64(__s0, __p1)); \
58223	59881	__ret; \
58224	59882	})
58225	59883	#define vqshlh_n_u16(__p0, __p1) __extension__ ({ \
58226	59884	uint16_t __ret; \
58227	59885	uint16_t __s0 = __p0; \
58228		__ret = (uint16_t) __builtin_neon_vqshlh_n_u16(__s0, __p1); \
	59886	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vqshlh_n_u16(__s0, __p1)); \
58229	59887	__ret; \
58230	59888	})
58231	59889	#define vqshlb_n_s8(__p0, __p1) __extension__ ({ \
58232	59890	int8_t __ret; \
58233	59891	int8_t __s0 = __p0; \
58234		__ret = (int8_t) __builtin_neon_vqshlb_n_s8(__s0, __p1); \
	59892	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vqshlb_n_s8(__s0, __p1)); \
58235	59893	__ret; \
58236	59894	})
58237	59895	#define vqshls_n_s32(__p0, __p1) __extension__ ({ \
58238	59896	int32_t __ret; \
58239	59897	int32_t __s0 = __p0; \
58240		__ret = (int32_t) __builtin_neon_vqshls_n_s32(__s0, __p1); \
	59898	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vqshls_n_s32(__s0, __p1)); \
58241	59899	__ret; \
58242	59900	})
58243	59901	#define vqshld_n_s64(__p0, __p1) __extension__ ({ \
58244	59902	int64_t __ret; \
58245	59903	int64_t __s0 = __p0; \
58246		__ret = (int64_t) __builtin_neon_vqshld_n_s64(__s0, __p1); \
	59904	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vqshld_n_s64(__s0, __p1)); \
58247	59905	__ret; \
58248	59906	})
58249	59907	#define vqshlh_n_s16(__p0, __p1) __extension__ ({ \
58250	59908	int16_t __ret; \
58251	59909	int16_t __s0 = __p0; \
58252		__ret = (int16_t) __builtin_neon_vqshlh_n_s16(__s0, __p1); \
	59910	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vqshlh_n_s16(__s0, __p1)); \
58253	59911	__ret; \
58254	59912	})
58255	59913	#define vqshlub_n_s8(__p0, __p1) __extension__ ({ \
58256	59914	int8_t __ret; \
58257	59915	int8_t __s0 = __p0; \
58258		__ret = (int8_t) __builtin_neon_vqshlub_n_s8(__s0, __p1); \
	59916	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vqshlub_n_s8(__s0, __p1)); \
58259	59917	__ret; \
58260	59918	})
58261	59919	#define vqshlus_n_s32(__p0, __p1) __extension__ ({ \
58262	59920	int32_t __ret; \
58263	59921	int32_t __s0 = __p0; \
58264		__ret = (int32_t) __builtin_neon_vqshlus_n_s32(__s0, __p1); \
	59922	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vqshlus_n_s32(__s0, __p1)); \
58265	59923	__ret; \
58266	59924	})
58267	59925	#define vqshlud_n_s64(__p0, __p1) __extension__ ({ \
58268	59926	int64_t __ret; \
58269	59927	int64_t __s0 = __p0; \
58270		__ret = (int64_t) __builtin_neon_vqshlud_n_s64(__s0, __p1); \
	59928	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vqshlud_n_s64(__s0, __p1)); \
58271	59929	__ret; \
58272	59930	})
58273	59931	#define vqshluh_n_s16(__p0, __p1) __extension__ ({ \
58274	59932	int16_t __ret; \
58275	59933	int16_t __s0 = __p0; \
58276		__ret = (int16_t) __builtin_neon_vqshluh_n_s16(__s0, __p1); \
	59934	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vqshluh_n_s16(__s0, __p1)); \
58277	59935	__ret; \
58278	59936	})
58279	59937	#ifdef __LITTLE_ENDIAN__
58280		#define vqshrn_high_n_u32(__p0_706, __p1_706, __p2_706) __extension__ ({ \
58281		uint16x8_t __ret_706; \
58282		uint16x4_t __s0_706 = __p0_706; \
58283		uint32x4_t __s1_706 = __p1_706; \
58284		__ret_706 = (uint16x8_t)(vcombine_u16((uint16x4_t)(__s0_706), (uint16x4_t)(vqshrn_n_u32(__s1_706, __p2_706)))); \
58285		__ret_706; \
	59938	#define vqshrn_high_n_u32(__p0_722, __p1_722, __p2_722) __extension__ ({ \
	59939	uint16x8_t __ret_722; \
	59940	uint16x4_t __s0_722 = __p0_722; \
	59941	uint32x4_t __s1_722 = __p1_722; \
	59942	__ret_722 = __builtin_bit_cast(uint16x8_t, vcombine_u16(__builtin_bit_cast(uint16x4_t, __s0_722), __builtin_bit_cast(uint16x4_t, vqshrn_n_u32(__s1_722, __p2_722)))); \
	59943	__ret_722; \
58286	59944	})
58287	59945	#else
58288		#define vqshrn_high_n_u32(__p0_707, __p1_707, __p2_707) __extension__ ({ \
58289		uint16x8_t __ret_707; \
58290		uint16x4_t __s0_707 = __p0_707; \
58291		uint32x4_t __s1_707 = __p1_707; \
58292		uint16x4_t __rev0_707; __rev0_707 = __builtin_shufflevector(__s0_707, __s0_707, 3, 2, 1, 0); \
58293		uint32x4_t __rev1_707; __rev1_707 = __builtin_shufflevector(__s1_707, __s1_707, 3, 2, 1, 0); \
58294		__ret_707 = (uint16x8_t)(__noswap_vcombine_u16((uint16x4_t)(__rev0_707), (uint16x4_t)(__noswap_vqshrn_n_u32(__rev1_707, __p2_707)))); \
58295		__ret_707 = __builtin_shufflevector(__ret_707, __ret_707, 7, 6, 5, 4, 3, 2, 1, 0); \
58296		__ret_707; \
	59946	#define vqshrn_high_n_u32(__p0_723, __p1_723, __p2_723) __extension__ ({ \
	59947	uint16x8_t __ret_723; \
	59948	uint16x4_t __s0_723 = __p0_723; \
	59949	uint32x4_t __s1_723 = __p1_723; \
	59950	uint16x4_t __rev0_723; __rev0_723 = __builtin_shufflevector(__s0_723, __s0_723, __lane_reverse_64_16); \
	59951	uint32x4_t __rev1_723; __rev1_723 = __builtin_shufflevector(__s1_723, __s1_723, __lane_reverse_128_32); \
	59952	__ret_723 = __builtin_bit_cast(uint16x8_t, __noswap_vcombine_u16(__builtin_bit_cast(uint16x4_t, __rev0_723), __builtin_bit_cast(uint16x4_t, __noswap_vqshrn_n_u32(__rev1_723, __p2_723)))); \
	59953	__ret_723 = __builtin_shufflevector(__ret_723, __ret_723, __lane_reverse_128_16); \
	59954	__ret_723; \
58297	59955	})
58298	59956	#endif
58299	59957
58300	59958	#ifdef __LITTLE_ENDIAN__
58301		#define vqshrn_high_n_u64(__p0_708, __p1_708, __p2_708) __extension__ ({ \
58302		uint32x4_t __ret_708; \
58303		uint32x2_t __s0_708 = __p0_708; \
58304		uint64x2_t __s1_708 = __p1_708; \
58305		__ret_708 = (uint32x4_t)(vcombine_u32((uint32x2_t)(__s0_708), (uint32x2_t)(vqshrn_n_u64(__s1_708, __p2_708)))); \
58306		__ret_708; \
	59959	#define vqshrn_high_n_u64(__p0_724, __p1_724, __p2_724) __extension__ ({ \
	59960	uint32x4_t __ret_724; \
	59961	uint32x2_t __s0_724 = __p0_724; \
	59962	uint64x2_t __s1_724 = __p1_724; \
	59963	__ret_724 = __builtin_bit_cast(uint32x4_t, vcombine_u32(__builtin_bit_cast(uint32x2_t, __s0_724), __builtin_bit_cast(uint32x2_t, vqshrn_n_u64(__s1_724, __p2_724)))); \
	59964	__ret_724; \
58307	59965	})
58308	59966	#else
58309		#define vqshrn_high_n_u64(__p0_709, __p1_709, __p2_709) __extension__ ({ \
58310		uint32x4_t __ret_709; \
58311		uint32x2_t __s0_709 = __p0_709; \
58312		uint64x2_t __s1_709 = __p1_709; \
58313		uint32x2_t __rev0_709; __rev0_709 = __builtin_shufflevector(__s0_709, __s0_709, 1, 0); \
58314		uint64x2_t __rev1_709; __rev1_709 = __builtin_shufflevector(__s1_709, __s1_709, 1, 0); \
58315		__ret_709 = (uint32x4_t)(__noswap_vcombine_u32((uint32x2_t)(__rev0_709), (uint32x2_t)(__noswap_vqshrn_n_u64(__rev1_709, __p2_709)))); \
58316		__ret_709 = __builtin_shufflevector(__ret_709, __ret_709, 3, 2, 1, 0); \
58317		__ret_709; \
	59967	#define vqshrn_high_n_u64(__p0_725, __p1_725, __p2_725) __extension__ ({ \
	59968	uint32x4_t __ret_725; \
	59969	uint32x2_t __s0_725 = __p0_725; \
	59970	uint64x2_t __s1_725 = __p1_725; \
	59971	uint32x2_t __rev0_725; __rev0_725 = __builtin_shufflevector(__s0_725, __s0_725, __lane_reverse_64_32); \
	59972	uint64x2_t __rev1_725; __rev1_725 = __builtin_shufflevector(__s1_725, __s1_725, __lane_reverse_128_64); \
	59973	__ret_725 = __builtin_bit_cast(uint32x4_t, __noswap_vcombine_u32(__builtin_bit_cast(uint32x2_t, __rev0_725), __builtin_bit_cast(uint32x2_t, __noswap_vqshrn_n_u64(__rev1_725, __p2_725)))); \
	59974	__ret_725 = __builtin_shufflevector(__ret_725, __ret_725, __lane_reverse_128_32); \
	59975	__ret_725; \
58318	59976	})
58319	59977	#endif
58320	59978
58321	59979	#ifdef __LITTLE_ENDIAN__
58322		#define vqshrn_high_n_u16(__p0_710, __p1_710, __p2_710) __extension__ ({ \
58323		uint8x16_t __ret_710; \
58324		uint8x8_t __s0_710 = __p0_710; \
58325		uint16x8_t __s1_710 = __p1_710; \
58326		__ret_710 = (uint8x16_t)(vcombine_u8((uint8x8_t)(__s0_710), (uint8x8_t)(vqshrn_n_u16(__s1_710, __p2_710)))); \
58327		__ret_710; \
	59980	#define vqshrn_high_n_u16(__p0_726, __p1_726, __p2_726) __extension__ ({ \
	59981	uint8x16_t __ret_726; \
	59982	uint8x8_t __s0_726 = __p0_726; \
	59983	uint16x8_t __s1_726 = __p1_726; \
	59984	__ret_726 = __builtin_bit_cast(uint8x16_t, vcombine_u8(__builtin_bit_cast(uint8x8_t, __s0_726), __builtin_bit_cast(uint8x8_t, vqshrn_n_u16(__s1_726, __p2_726)))); \
	59985	__ret_726; \
58328	59986	})
58329	59987	#else
58330		#define vqshrn_high_n_u16(__p0_711, __p1_711, __p2_711) __extension__ ({ \
58331		uint8x16_t __ret_711; \
58332		uint8x8_t __s0_711 = __p0_711; \
58333		uint16x8_t __s1_711 = __p1_711; \
58334		uint8x8_t __rev0_711; __rev0_711 = __builtin_shufflevector(__s0_711, __s0_711, 7, 6, 5, 4, 3, 2, 1, 0); \
58335		uint16x8_t __rev1_711; __rev1_711 = __builtin_shufflevector(__s1_711, __s1_711, 7, 6, 5, 4, 3, 2, 1, 0); \
58336		__ret_711 = (uint8x16_t)(__noswap_vcombine_u8((uint8x8_t)(__rev0_711), (uint8x8_t)(__noswap_vqshrn_n_u16(__rev1_711, __p2_711)))); \
58337		__ret_711 = __builtin_shufflevector(__ret_711, __ret_711, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
58338		__ret_711; \
	59988	#define vqshrn_high_n_u16(__p0_727, __p1_727, __p2_727) __extension__ ({ \
	59989	uint8x16_t __ret_727; \
	59990	uint8x8_t __s0_727 = __p0_727; \
	59991	uint16x8_t __s1_727 = __p1_727; \
	59992	uint8x8_t __rev0_727; __rev0_727 = __builtin_shufflevector(__s0_727, __s0_727, __lane_reverse_64_8); \
	59993	uint16x8_t __rev1_727; __rev1_727 = __builtin_shufflevector(__s1_727, __s1_727, __lane_reverse_128_16); \
	59994	__ret_727 = __builtin_bit_cast(uint8x16_t, __noswap_vcombine_u8(__builtin_bit_cast(uint8x8_t, __rev0_727), __builtin_bit_cast(uint8x8_t, __noswap_vqshrn_n_u16(__rev1_727, __p2_727)))); \
	59995	__ret_727 = __builtin_shufflevector(__ret_727, __ret_727, __lane_reverse_128_8); \
	59996	__ret_727; \
58339	59997	})
58340	59998	#endif
58341	59999
58342	60000	#ifdef __LITTLE_ENDIAN__
58343		#define vqshrn_high_n_s32(__p0_712, __p1_712, __p2_712) __extension__ ({ \
58344		int16x8_t __ret_712; \
58345		int16x4_t __s0_712 = __p0_712; \
58346		int32x4_t __s1_712 = __p1_712; \
58347		__ret_712 = (int16x8_t)(vcombine_s16((int16x4_t)(__s0_712), (int16x4_t)(vqshrn_n_s32(__s1_712, __p2_712)))); \
58348		__ret_712; \
	60001	#define vqshrn_high_n_s32(__p0_728, __p1_728, __p2_728) __extension__ ({ \
	60002	int16x8_t __ret_728; \
	60003	int16x4_t __s0_728 = __p0_728; \
	60004	int32x4_t __s1_728 = __p1_728; \
	60005	__ret_728 = __builtin_bit_cast(int16x8_t, vcombine_s16(__builtin_bit_cast(int16x4_t, __s0_728), __builtin_bit_cast(int16x4_t, vqshrn_n_s32(__s1_728, __p2_728)))); \
	60006	__ret_728; \
58349	60007	})
58350	60008	#else
58351		#define vqshrn_high_n_s32(__p0_713, __p1_713, __p2_713) __extension__ ({ \
58352		int16x8_t __ret_713; \
58353		int16x4_t __s0_713 = __p0_713; \
58354		int32x4_t __s1_713 = __p1_713; \
58355		int16x4_t __rev0_713; __rev0_713 = __builtin_shufflevector(__s0_713, __s0_713, 3, 2, 1, 0); \
58356		int32x4_t __rev1_713; __rev1_713 = __builtin_shufflevector(__s1_713, __s1_713, 3, 2, 1, 0); \
58357		__ret_713 = (int16x8_t)(__noswap_vcombine_s16((int16x4_t)(__rev0_713), (int16x4_t)(__noswap_vqshrn_n_s32(__rev1_713, __p2_713)))); \
58358		__ret_713 = __builtin_shufflevector(__ret_713, __ret_713, 7, 6, 5, 4, 3, 2, 1, 0); \
58359		__ret_713; \
	60009	#define vqshrn_high_n_s32(__p0_729, __p1_729, __p2_729) __extension__ ({ \
	60010	int16x8_t __ret_729; \
	60011	int16x4_t __s0_729 = __p0_729; \
	60012	int32x4_t __s1_729 = __p1_729; \
	60013	int16x4_t __rev0_729; __rev0_729 = __builtin_shufflevector(__s0_729, __s0_729, __lane_reverse_64_16); \
	60014	int32x4_t __rev1_729; __rev1_729 = __builtin_shufflevector(__s1_729, __s1_729, __lane_reverse_128_32); \
	60015	__ret_729 = __builtin_bit_cast(int16x8_t, __noswap_vcombine_s16(__builtin_bit_cast(int16x4_t, __rev0_729), __builtin_bit_cast(int16x4_t, __noswap_vqshrn_n_s32(__rev1_729, __p2_729)))); \
	60016	__ret_729 = __builtin_shufflevector(__ret_729, __ret_729, __lane_reverse_128_16); \
	60017	__ret_729; \
58360	60018	})
58361	60019	#endif
58362	60020
58363	60021	#ifdef __LITTLE_ENDIAN__
58364		#define vqshrn_high_n_s64(__p0_714, __p1_714, __p2_714) __extension__ ({ \
58365		int32x4_t __ret_714; \
58366		int32x2_t __s0_714 = __p0_714; \
58367		int64x2_t __s1_714 = __p1_714; \
58368		__ret_714 = (int32x4_t)(vcombine_s32((int32x2_t)(__s0_714), (int32x2_t)(vqshrn_n_s64(__s1_714, __p2_714)))); \
58369		__ret_714; \
	60022	#define vqshrn_high_n_s64(__p0_730, __p1_730, __p2_730) __extension__ ({ \
	60023	int32x4_t __ret_730; \
	60024	int32x2_t __s0_730 = __p0_730; \
	60025	int64x2_t __s1_730 = __p1_730; \
	60026	__ret_730 = __builtin_bit_cast(int32x4_t, vcombine_s32(__builtin_bit_cast(int32x2_t, __s0_730), __builtin_bit_cast(int32x2_t, vqshrn_n_s64(__s1_730, __p2_730)))); \
	60027	__ret_730; \
58370	60028	})
58371	60029	#else
58372		#define vqshrn_high_n_s64(__p0_715, __p1_715, __p2_715) __extension__ ({ \
58373		int32x4_t __ret_715; \
58374		int32x2_t __s0_715 = __p0_715; \
58375		int64x2_t __s1_715 = __p1_715; \
58376		int32x2_t __rev0_715; __rev0_715 = __builtin_shufflevector(__s0_715, __s0_715, 1, 0); \
58377		int64x2_t __rev1_715; __rev1_715 = __builtin_shufflevector(__s1_715, __s1_715, 1, 0); \
58378		__ret_715 = (int32x4_t)(__noswap_vcombine_s32((int32x2_t)(__rev0_715), (int32x2_t)(__noswap_vqshrn_n_s64(__rev1_715, __p2_715)))); \
58379		__ret_715 = __builtin_shufflevector(__ret_715, __ret_715, 3, 2, 1, 0); \
58380		__ret_715; \
	60030	#define vqshrn_high_n_s64(__p0_731, __p1_731, __p2_731) __extension__ ({ \
	60031	int32x4_t __ret_731; \
	60032	int32x2_t __s0_731 = __p0_731; \
	60033	int64x2_t __s1_731 = __p1_731; \
	60034	int32x2_t __rev0_731; __rev0_731 = __builtin_shufflevector(__s0_731, __s0_731, __lane_reverse_64_32); \
	60035	int64x2_t __rev1_731; __rev1_731 = __builtin_shufflevector(__s1_731, __s1_731, __lane_reverse_128_64); \
	60036	__ret_731 = __builtin_bit_cast(int32x4_t, __noswap_vcombine_s32(__builtin_bit_cast(int32x2_t, __rev0_731), __builtin_bit_cast(int32x2_t, __noswap_vqshrn_n_s64(__rev1_731, __p2_731)))); \
	60037	__ret_731 = __builtin_shufflevector(__ret_731, __ret_731, __lane_reverse_128_32); \
	60038	__ret_731; \
58381	60039	})
58382	60040	#endif
58383	60041
58384	60042	#ifdef __LITTLE_ENDIAN__
58385		#define vqshrn_high_n_s16(__p0_716, __p1_716, __p2_716) __extension__ ({ \
58386		int8x16_t __ret_716; \
58387		int8x8_t __s0_716 = __p0_716; \
58388		int16x8_t __s1_716 = __p1_716; \
58389		__ret_716 = (int8x16_t)(vcombine_s8((int8x8_t)(__s0_716), (int8x8_t)(vqshrn_n_s16(__s1_716, __p2_716)))); \
58390		__ret_716; \
	60043	#define vqshrn_high_n_s16(__p0_732, __p1_732, __p2_732) __extension__ ({ \
	60044	int8x16_t __ret_732; \
	60045	int8x8_t __s0_732 = __p0_732; \
	60046	int16x8_t __s1_732 = __p1_732; \
	60047	__ret_732 = __builtin_bit_cast(int8x16_t, vcombine_s8(__builtin_bit_cast(int8x8_t, __s0_732), __builtin_bit_cast(int8x8_t, vqshrn_n_s16(__s1_732, __p2_732)))); \
	60048	__ret_732; \
58391	60049	})
58392	60050	#else
58393		#define vqshrn_high_n_s16(__p0_717, __p1_717, __p2_717) __extension__ ({ \
58394		int8x16_t __ret_717; \
58395		int8x8_t __s0_717 = __p0_717; \
58396		int16x8_t __s1_717 = __p1_717; \
58397		int8x8_t __rev0_717; __rev0_717 = __builtin_shufflevector(__s0_717, __s0_717, 7, 6, 5, 4, 3, 2, 1, 0); \
58398		int16x8_t __rev1_717; __rev1_717 = __builtin_shufflevector(__s1_717, __s1_717, 7, 6, 5, 4, 3, 2, 1, 0); \
58399		__ret_717 = (int8x16_t)(__noswap_vcombine_s8((int8x8_t)(__rev0_717), (int8x8_t)(__noswap_vqshrn_n_s16(__rev1_717, __p2_717)))); \
58400		__ret_717 = __builtin_shufflevector(__ret_717, __ret_717, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
58401		__ret_717; \
	60051	#define vqshrn_high_n_s16(__p0_733, __p1_733, __p2_733) __extension__ ({ \
	60052	int8x16_t __ret_733; \
	60053	int8x8_t __s0_733 = __p0_733; \
	60054	int16x8_t __s1_733 = __p1_733; \
	60055	int8x8_t __rev0_733; __rev0_733 = __builtin_shufflevector(__s0_733, __s0_733, __lane_reverse_64_8); \
	60056	int16x8_t __rev1_733; __rev1_733 = __builtin_shufflevector(__s1_733, __s1_733, __lane_reverse_128_16); \
	60057	__ret_733 = __builtin_bit_cast(int8x16_t, __noswap_vcombine_s8(__builtin_bit_cast(int8x8_t, __rev0_733), __builtin_bit_cast(int8x8_t, __noswap_vqshrn_n_s16(__rev1_733, __p2_733)))); \
	60058	__ret_733 = __builtin_shufflevector(__ret_733, __ret_733, __lane_reverse_128_8); \
	60059	__ret_733; \
58402	60060	})
58403	60061	#endif
58404	60062
58405	60063	#define vqshrns_n_u32(__p0, __p1) __extension__ ({ \
58406	60064	uint16_t __ret; \
58407	60065	uint32_t __s0 = __p0; \
58408		__ret = (uint16_t) __builtin_neon_vqshrns_n_u32(__s0, __p1); \
	60066	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vqshrns_n_u32(__s0, __p1)); \
58409	60067	__ret; \
58410	60068	})
58411	60069	#define vqshrnd_n_u64(__p0, __p1) __extension__ ({ \
58412	60070	uint32_t __ret; \
58413	60071	uint64_t __s0 = __p0; \
58414		__ret = (uint32_t) __builtin_neon_vqshrnd_n_u64(__s0, __p1); \
	60072	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vqshrnd_n_u64(__s0, __p1)); \
58415	60073	__ret; \
58416	60074	})
58417	60075	#define vqshrnh_n_u16(__p0, __p1) __extension__ ({ \
58418	60076	uint8_t __ret; \
58419	60077	uint16_t __s0 = __p0; \
58420		__ret = (uint8_t) __builtin_neon_vqshrnh_n_u16(__s0, __p1); \
	60078	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vqshrnh_n_u16(__s0, __p1)); \
58421	60079	__ret; \
58422	60080	})
58423	60081	#define vqshrns_n_s32(__p0, __p1) __extension__ ({ \
58424	60082	int16_t __ret; \
58425	60083	int32_t __s0 = __p0; \
58426		__ret = (int16_t) __builtin_neon_vqshrns_n_s32(__s0, __p1); \
	60084	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vqshrns_n_s32(__s0, __p1)); \
58427	60085	__ret; \
58428	60086	})
58429	60087	#define vqshrnd_n_s64(__p0, __p1) __extension__ ({ \
58430	60088	int32_t __ret; \
58431	60089	int64_t __s0 = __p0; \
58432		__ret = (int32_t) __builtin_neon_vqshrnd_n_s64(__s0, __p1); \
	60090	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vqshrnd_n_s64(__s0, __p1)); \
58433	60091	__ret; \
58434	60092	})
58435	60093	#define vqshrnh_n_s16(__p0, __p1) __extension__ ({ \
58436	60094	int8_t __ret; \
58437	60095	int16_t __s0 = __p0; \
58438		__ret = (int8_t) __builtin_neon_vqshrnh_n_s16(__s0, __p1); \
	60096	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vqshrnh_n_s16(__s0, __p1)); \
58439	60097	__ret; \
58440	60098	})
58441	60099	#ifdef __LITTLE_ENDIAN__
58442		#define vqshrun_high_n_s32(__p0_718, __p1_718, __p2_718) __extension__ ({ \
58443		int16x8_t __ret_718; \
58444		int16x4_t __s0_718 = __p0_718; \
58445		int32x4_t __s1_718 = __p1_718; \
58446		__ret_718 = (int16x8_t)(vcombine_s16((int16x4_t)(__s0_718), (int16x4_t)(vqshrun_n_s32(__s1_718, __p2_718)))); \
58447		__ret_718; \
	60100	#define vqshrun_high_n_s32(__p0_734, __p1_734, __p2_734) __extension__ ({ \
	60101	int16x8_t __ret_734; \
	60102	int16x4_t __s0_734 = __p0_734; \
	60103	int32x4_t __s1_734 = __p1_734; \
	60104	__ret_734 = __builtin_bit_cast(int16x8_t, vcombine_s16(__builtin_bit_cast(int16x4_t, __s0_734), __builtin_bit_cast(int16x4_t, vqshrun_n_s32(__s1_734, __p2_734)))); \
	60105	__ret_734; \
58448	60106	})
58449	60107	#else
58450		#define vqshrun_high_n_s32(__p0_719, __p1_719, __p2_719) __extension__ ({ \
58451		int16x8_t __ret_719; \
58452		int16x4_t __s0_719 = __p0_719; \
58453		int32x4_t __s1_719 = __p1_719; \
58454		int16x4_t __rev0_719; __rev0_719 = __builtin_shufflevector(__s0_719, __s0_719, 3, 2, 1, 0); \
58455		int32x4_t __rev1_719; __rev1_719 = __builtin_shufflevector(__s1_719, __s1_719, 3, 2, 1, 0); \
58456		__ret_719 = (int16x8_t)(__noswap_vcombine_s16((int16x4_t)(__rev0_719), (int16x4_t)(__noswap_vqshrun_n_s32(__rev1_719, __p2_719)))); \
58457		__ret_719 = __builtin_shufflevector(__ret_719, __ret_719, 7, 6, 5, 4, 3, 2, 1, 0); \
58458		__ret_719; \
	60108	#define vqshrun_high_n_s32(__p0_735, __p1_735, __p2_735) __extension__ ({ \
	60109	int16x8_t __ret_735; \
	60110	int16x4_t __s0_735 = __p0_735; \
	60111	int32x4_t __s1_735 = __p1_735; \
	60112	int16x4_t __rev0_735; __rev0_735 = __builtin_shufflevector(__s0_735, __s0_735, __lane_reverse_64_16); \
	60113	int32x4_t __rev1_735; __rev1_735 = __builtin_shufflevector(__s1_735, __s1_735, __lane_reverse_128_32); \
	60114	__ret_735 = __builtin_bit_cast(int16x8_t, __noswap_vcombine_s16(__builtin_bit_cast(int16x4_t, __rev0_735), __builtin_bit_cast(int16x4_t, __noswap_vqshrun_n_s32(__rev1_735, __p2_735)))); \
	60115	__ret_735 = __builtin_shufflevector(__ret_735, __ret_735, __lane_reverse_128_16); \
	60116	__ret_735; \
58459	60117	})
58460	60118	#endif
58461	60119
58462	60120	#ifdef __LITTLE_ENDIAN__
58463		#define vqshrun_high_n_s64(__p0_720, __p1_720, __p2_720) __extension__ ({ \
58464		int32x4_t __ret_720; \
58465		int32x2_t __s0_720 = __p0_720; \
58466		int64x2_t __s1_720 = __p1_720; \
58467		__ret_720 = (int32x4_t)(vcombine_s32((int32x2_t)(__s0_720), (int32x2_t)(vqshrun_n_s64(__s1_720, __p2_720)))); \
58468		__ret_720; \
	60121	#define vqshrun_high_n_s64(__p0_736, __p1_736, __p2_736) __extension__ ({ \
	60122	int32x4_t __ret_736; \
	60123	int32x2_t __s0_736 = __p0_736; \
	60124	int64x2_t __s1_736 = __p1_736; \
	60125	__ret_736 = __builtin_bit_cast(int32x4_t, vcombine_s32(__builtin_bit_cast(int32x2_t, __s0_736), __builtin_bit_cast(int32x2_t, vqshrun_n_s64(__s1_736, __p2_736)))); \
	60126	__ret_736; \
58469	60127	})
58470	60128	#else
58471		#define vqshrun_high_n_s64(__p0_721, __p1_721, __p2_721) __extension__ ({ \
58472		int32x4_t __ret_721; \
58473		int32x2_t __s0_721 = __p0_721; \
58474		int64x2_t __s1_721 = __p1_721; \
58475		int32x2_t __rev0_721; __rev0_721 = __builtin_shufflevector(__s0_721, __s0_721, 1, 0); \
58476		int64x2_t __rev1_721; __rev1_721 = __builtin_shufflevector(__s1_721, __s1_721, 1, 0); \
58477		__ret_721 = (int32x4_t)(__noswap_vcombine_s32((int32x2_t)(__rev0_721), (int32x2_t)(__noswap_vqshrun_n_s64(__rev1_721, __p2_721)))); \
58478		__ret_721 = __builtin_shufflevector(__ret_721, __ret_721, 3, 2, 1, 0); \
58479		__ret_721; \
	60129	#define vqshrun_high_n_s64(__p0_737, __p1_737, __p2_737) __extension__ ({ \
	60130	int32x4_t __ret_737; \
	60131	int32x2_t __s0_737 = __p0_737; \
	60132	int64x2_t __s1_737 = __p1_737; \
	60133	int32x2_t __rev0_737; __rev0_737 = __builtin_shufflevector(__s0_737, __s0_737, __lane_reverse_64_32); \
	60134	int64x2_t __rev1_737; __rev1_737 = __builtin_shufflevector(__s1_737, __s1_737, __lane_reverse_128_64); \
	60135	__ret_737 = __builtin_bit_cast(int32x4_t, __noswap_vcombine_s32(__builtin_bit_cast(int32x2_t, __rev0_737), __builtin_bit_cast(int32x2_t, __noswap_vqshrun_n_s64(__rev1_737, __p2_737)))); \
	60136	__ret_737 = __builtin_shufflevector(__ret_737, __ret_737, __lane_reverse_128_32); \
	60137	__ret_737; \
58480	60138	})
58481	60139	#endif
58482	60140
58483	60141	#ifdef __LITTLE_ENDIAN__
58484		#define vqshrun_high_n_s16(__p0_722, __p1_722, __p2_722) __extension__ ({ \
58485		int8x16_t __ret_722; \
58486		int8x8_t __s0_722 = __p0_722; \
58487		int16x8_t __s1_722 = __p1_722; \
58488		__ret_722 = (int8x16_t)(vcombine_s8((int8x8_t)(__s0_722), (int8x8_t)(vqshrun_n_s16(__s1_722, __p2_722)))); \
58489		__ret_722; \
	60142	#define vqshrun_high_n_s16(__p0_738, __p1_738, __p2_738) __extension__ ({ \
	60143	int8x16_t __ret_738; \
	60144	int8x8_t __s0_738 = __p0_738; \
	60145	int16x8_t __s1_738 = __p1_738; \
	60146	__ret_738 = __builtin_bit_cast(int8x16_t, vcombine_s8(__builtin_bit_cast(int8x8_t, __s0_738), __builtin_bit_cast(int8x8_t, vqshrun_n_s16(__s1_738, __p2_738)))); \
	60147	__ret_738; \
58490	60148	})
58491	60149	#else
58492		#define vqshrun_high_n_s16(__p0_723, __p1_723, __p2_723) __extension__ ({ \
58493		int8x16_t __ret_723; \
58494		int8x8_t __s0_723 = __p0_723; \
58495		int16x8_t __s1_723 = __p1_723; \
58496		int8x8_t __rev0_723; __rev0_723 = __builtin_shufflevector(__s0_723, __s0_723, 7, 6, 5, 4, 3, 2, 1, 0); \
58497		int16x8_t __rev1_723; __rev1_723 = __builtin_shufflevector(__s1_723, __s1_723, 7, 6, 5, 4, 3, 2, 1, 0); \
58498		__ret_723 = (int8x16_t)(__noswap_vcombine_s8((int8x8_t)(__rev0_723), (int8x8_t)(__noswap_vqshrun_n_s16(__rev1_723, __p2_723)))); \
58499		__ret_723 = __builtin_shufflevector(__ret_723, __ret_723, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
58500		__ret_723; \
	60150	#define vqshrun_high_n_s16(__p0_739, __p1_739, __p2_739) __extension__ ({ \
	60151	int8x16_t __ret_739; \
	60152	int8x8_t __s0_739 = __p0_739; \
	60153	int16x8_t __s1_739 = __p1_739; \
	60154	int8x8_t __rev0_739; __rev0_739 = __builtin_shufflevector(__s0_739, __s0_739, __lane_reverse_64_8); \
	60155	int16x8_t __rev1_739; __rev1_739 = __builtin_shufflevector(__s1_739, __s1_739, __lane_reverse_128_16); \
	60156	__ret_739 = __builtin_bit_cast(int8x16_t, __noswap_vcombine_s8(__builtin_bit_cast(int8x8_t, __rev0_739), __builtin_bit_cast(int8x8_t, __noswap_vqshrun_n_s16(__rev1_739, __p2_739)))); \
	60157	__ret_739 = __builtin_shufflevector(__ret_739, __ret_739, __lane_reverse_128_8); \
	60158	__ret_739; \
58501	60159	})
58502	60160	#endif
58503	60161
58504	60162	#define vqshruns_n_s32(__p0, __p1) __extension__ ({ \
58505	60163	uint16_t __ret; \
58506	60164	int32_t __s0 = __p0; \
58507		__ret = (uint16_t) __builtin_neon_vqshruns_n_s32(__s0, __p1); \
	60165	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vqshruns_n_s32(__s0, __p1)); \
58508	60166	__ret; \
58509	60167	})
58510	60168	#define vqshrund_n_s64(__p0, __p1) __extension__ ({ \
58511	60169	uint32_t __ret; \
58512	60170	int64_t __s0 = __p0; \
58513		__ret = (uint32_t) __builtin_neon_vqshrund_n_s64(__s0, __p1); \
	60171	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vqshrund_n_s64(__s0, __p1)); \
58514	60172	__ret; \
58515	60173	})
58516	60174	#define vqshrunh_n_s16(__p0, __p1) __extension__ ({ \
58517	60175	uint8_t __ret; \
58518	60176	int16_t __s0 = __p0; \
58519		__ret = (uint8_t) __builtin_neon_vqshrunh_n_s16(__s0, __p1); \
	60177	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vqshrunh_n_s16(__s0, __p1)); \
58520	60178	__ret; \
58521	60179	})
58522	60180	__ai __attribute__((target("neon"))) uint8_t vqsubb_u8(uint8_t __p0, uint8_t __p1) {
58523	60181	uint8_t __ret;
58524		__ret = (uint8_t) __builtin_neon_vqsubb_u8(__p0, __p1);
	60182	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vqsubb_u8(__p0, __p1));
58525	60183	return __ret;
58526	60184	}
58527	60185	__ai __attribute__((target("neon"))) uint32_t vqsubs_u32(uint32_t __p0, uint32_t __p1) {
58528	60186	uint32_t __ret;
58529		__ret = (uint32_t) __builtin_neon_vqsubs_u32(__p0, __p1);
	60187	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vqsubs_u32(__p0, __p1));
58530	60188	return __ret;
58531	60189	}
58532	60190	__ai __attribute__((target("neon"))) uint64_t vqsubd_u64(uint64_t __p0, uint64_t __p1) {
58533	60191	uint64_t __ret;
58534		__ret = (uint64_t) __builtin_neon_vqsubd_u64(__p0, __p1);
	60192	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vqsubd_u64(__p0, __p1));
58535	60193	return __ret;
58536	60194	}
58537	60195	__ai __attribute__((target("neon"))) uint16_t vqsubh_u16(uint16_t __p0, uint16_t __p1) {
58538	60196	uint16_t __ret;
58539		__ret = (uint16_t) __builtin_neon_vqsubh_u16(__p0, __p1);
	60197	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vqsubh_u16(__p0, __p1));
58540	60198	return __ret;
58541	60199	}
58542	60200	__ai __attribute__((target("neon"))) int8_t vqsubb_s8(int8_t __p0, int8_t __p1) {
58543	60201	int8_t __ret;
58544		__ret = (int8_t) __builtin_neon_vqsubb_s8(__p0, __p1);
	60202	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vqsubb_s8(__p0, __p1));
58545	60203	return __ret;
58546	60204	}
58547	60205	__ai __attribute__((target("neon"))) int32_t vqsubs_s32(int32_t __p0, int32_t __p1) {
58548	60206	int32_t __ret;
58549		__ret = (int32_t) __builtin_neon_vqsubs_s32(__p0, __p1);
	60207	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vqsubs_s32(__p0, __p1));
58550	60208	return __ret;
58551	60209	}
58552	60210	__ai __attribute__((target("neon"))) int64_t vqsubd_s64(int64_t __p0, int64_t __p1) {
58553	60211	int64_t __ret;
58554		__ret = (int64_t) __builtin_neon_vqsubd_s64(__p0, __p1);
	60212	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vqsubd_s64(__p0, __p1));
58555	60213	return __ret;
58556	60214	}
58557	60215	__ai __attribute__((target("neon"))) int16_t vqsubh_s16(int16_t __p0, int16_t __p1) {
58558	60216	int16_t __ret;
58559		__ret = (int16_t) __builtin_neon_vqsubh_s16(__p0, __p1);
	60217	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vqsubh_s16(__p0, __p1));
58560	60218	return __ret;
58561	60219	}
58562	60220	#ifdef __LITTLE_ENDIAN__
58563	60221	__ai __attribute__((target("neon"))) poly8x8_t vqtbl1_p8(poly8x16_t __p0, uint8x8_t __p1) {
58564	60222	poly8x8_t __ret;
58565		__ret = (poly8x8_t) __builtin_neon_vqtbl1_v((int8x16_t)__p0, (int8x8_t)__p1, 4);
	60223	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vqtbl1_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 4));
58566	60224	return __ret;
58567	60225	}
58568	60226	#else
58569	60227	__ai __attribute__((target("neon"))) poly8x8_t vqtbl1_p8(poly8x16_t __p0, uint8x8_t __p1) {
58570	60228	poly8x8_t __ret;
58571		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58572		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
58573		__ret = (poly8x8_t) __builtin_neon_vqtbl1_v((int8x16_t)__rev0, (int8x8_t)__rev1, 4);
58574		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	60229	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	60230	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	60231	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vqtbl1_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 4));
	60232	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
58575	60233	return __ret;
58576	60234	}
58577	60235	#endif
...	...	@@ -58579,16 +60237,16 @@ __ai __attribute__((target("neon"))) poly8x8_t vqtbl1_p8(poly8x16_t __p0, uint8x
58579	60237	#ifdef __LITTLE_ENDIAN__
58580	60238	__ai __attribute__((target("neon"))) poly8x16_t vqtbl1q_p8(poly8x16_t __p0, uint8x16_t __p1) {
58581	60239	poly8x16_t __ret;
58582		__ret = (poly8x16_t) __builtin_neon_vqtbl1q_v((int8x16_t)__p0, (int8x16_t)__p1, 36);
	60240	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vqtbl1q_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 36));
58583	60241	return __ret;
58584	60242	}
58585	60243	#else
58586	60244	__ai __attribute__((target("neon"))) poly8x16_t vqtbl1q_p8(poly8x16_t __p0, uint8x16_t __p1) {
58587	60245	poly8x16_t __ret;
58588		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58589		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58590		__ret = (poly8x16_t) __builtin_neon_vqtbl1q_v((int8x16_t)__rev0, (int8x16_t)__rev1, 36);
58591		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	60246	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	60247	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	60248	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vqtbl1q_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 36));
	60249	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
58592	60250	return __ret;
58593	60251	}
58594	60252	#endif
...	...	@@ -58596,16 +60254,16 @@ __ai __attribute__((target("neon"))) poly8x16_t vqtbl1q_p8(poly8x16_t __p0, uint
58596	60254	#ifdef __LITTLE_ENDIAN__
58597	60255	__ai __attribute__((target("neon"))) uint8x16_t vqtbl1q_u8(uint8x16_t __p0, uint8x16_t __p1) {
58598	60256	uint8x16_t __ret;
58599		__ret = (uint8x16_t) __builtin_neon_vqtbl1q_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
	60257	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vqtbl1q_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 48));
58600	60258	return __ret;
58601	60259	}
58602	60260	#else
58603	60261	__ai __attribute__((target("neon"))) uint8x16_t vqtbl1q_u8(uint8x16_t __p0, uint8x16_t __p1) {
58604	60262	uint8x16_t __ret;
58605		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58606		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58607		__ret = (uint8x16_t) __builtin_neon_vqtbl1q_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
58608		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	60263	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	60264	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	60265	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vqtbl1q_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 48));
	60266	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
58609	60267	return __ret;
58610	60268	}
58611	60269	#endif
...	...	@@ -58613,16 +60271,33 @@ __ai __attribute__((target("neon"))) uint8x16_t vqtbl1q_u8(uint8x16_t __p0, uint
58613	60271	#ifdef __LITTLE_ENDIAN__
58614	60272	__ai __attribute__((target("neon"))) int8x16_t vqtbl1q_s8(int8x16_t __p0, uint8x16_t __p1) {
58615	60273	int8x16_t __ret;
58616		__ret = (int8x16_t) __builtin_neon_vqtbl1q_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
	60274	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vqtbl1q_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 32));
58617	60275	return __ret;
58618	60276	}
58619	60277	#else
58620	60278	__ai __attribute__((target("neon"))) int8x16_t vqtbl1q_s8(int8x16_t __p0, uint8x16_t __p1) {
58621	60279	int8x16_t __ret;
58622		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58623		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58624		__ret = (int8x16_t) __builtin_neon_vqtbl1q_v((int8x16_t)__rev0, (int8x16_t)__rev1, 32);
58625		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	60280	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	60281	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	60282	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vqtbl1q_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 32));
	60283	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
	60284	return __ret;
	60285	}
	60286	#endif
	60287
	60288	#ifdef __LITTLE_ENDIAN__
	60289	__ai __attribute__((target("neon"))) mfloat8x16_t vqtbl1q_mf8(mfloat8x16_t __p0, uint8x16_t __p1) {
	60290	mfloat8x16_t __ret;
	60291	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vqtbl1q_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 44));
	60292	return __ret;
	60293	}
	60294	#else
	60295	__ai __attribute__((target("neon"))) mfloat8x16_t vqtbl1q_mf8(mfloat8x16_t __p0, uint8x16_t __p1) {
	60296	mfloat8x16_t __ret;
	60297	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	60298	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	60299	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vqtbl1q_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 44));
	60300	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
58626	60301	return __ret;
58627	60302	}
58628	60303	#endif
...	...	@@ -58630,16 +60305,16 @@ __ai __attribute__((target("neon"))) int8x16_t vqtbl1q_s8(int8x16_t __p0, uint8x
58630	60305	#ifdef __LITTLE_ENDIAN__
58631	60306	__ai __attribute__((target("neon"))) uint8x8_t vqtbl1_u8(uint8x16_t __p0, uint8x8_t __p1) {
58632	60307	uint8x8_t __ret;
58633		__ret = (uint8x8_t) __builtin_neon_vqtbl1_v((int8x16_t)__p0, (int8x8_t)__p1, 16);
	60308	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqtbl1_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 16));
58634	60309	return __ret;
58635	60310	}
58636	60311	#else
58637	60312	__ai __attribute__((target("neon"))) uint8x8_t vqtbl1_u8(uint8x16_t __p0, uint8x8_t __p1) {
58638	60313	uint8x8_t __ret;
58639		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58640		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
58641		__ret = (uint8x8_t) __builtin_neon_vqtbl1_v((int8x16_t)__rev0, (int8x8_t)__rev1, 16);
58642		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	60314	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	60315	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	60316	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqtbl1_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 16));
	60317	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
58643	60318	return __ret;
58644	60319	}
58645	60320	#endif
...	...	@@ -58647,16 +60322,33 @@ __ai __attribute__((target("neon"))) uint8x8_t vqtbl1_u8(uint8x16_t __p0, uint8x
58647	60322	#ifdef __LITTLE_ENDIAN__
58648	60323	__ai __attribute__((target("neon"))) int8x8_t vqtbl1_s8(int8x16_t __p0, uint8x8_t __p1) {
58649	60324	int8x8_t __ret;
58650		__ret = (int8x8_t) __builtin_neon_vqtbl1_v((int8x16_t)__p0, (int8x8_t)__p1, 0);
	60325	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqtbl1_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 0));
58651	60326	return __ret;
58652	60327	}
58653	60328	#else
58654	60329	__ai __attribute__((target("neon"))) int8x8_t vqtbl1_s8(int8x16_t __p0, uint8x8_t __p1) {
58655	60330	int8x8_t __ret;
58656		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58657		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
58658		__ret = (int8x8_t) __builtin_neon_vqtbl1_v((int8x16_t)__rev0, (int8x8_t)__rev1, 0);
58659		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	60331	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	60332	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	60333	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqtbl1_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 0));
	60334	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
	60335	return __ret;
	60336	}
	60337	#endif
	60338
	60339	#ifdef __LITTLE_ENDIAN__
	60340	__ai __attribute__((target("neon"))) mfloat8x8_t vqtbl1_mf8(mfloat8x16_t __p0, uint8x8_t __p1) {
	60341	mfloat8x8_t __ret;
	60342	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vqtbl1_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 12));
	60343	return __ret;
	60344	}
	60345	#else
	60346	__ai __attribute__((target("neon"))) mfloat8x8_t vqtbl1_mf8(mfloat8x16_t __p0, uint8x8_t __p1) {
	60347	mfloat8x8_t __ret;
	60348	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	60349	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	60350	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vqtbl1_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 12));
	60351	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
58660	60352	return __ret;
58661	60353	}
58662	60354	#endif
...	...	@@ -58664,18 +60356,18 @@ __ai __attribute__((target("neon"))) int8x8_t vqtbl1_s8(int8x16_t __p0, uint8x8_
58664	60356	#ifdef __LITTLE_ENDIAN__
58665	60357	__ai __attribute__((target("neon"))) poly8x8_t vqtbl2_p8(poly8x16x2_t __p0, uint8x8_t __p1) {
58666	60358	poly8x8_t __ret;
58667		__ret = (poly8x8_t) __builtin_neon_vqtbl2_v((int8x16_t)__p0.val[0], (int8x16_t)__p0.val[1], (int8x8_t)__p1, 4);
	60359	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vqtbl2_v(__builtin_bit_cast(int8x16_t, __p0.val[0]), __builtin_bit_cast(int8x16_t, __p0.val[1]), __builtin_bit_cast(int8x8_t, __p1), 4));
58668	60360	return __ret;
58669	60361	}
58670	60362	#else
58671	60363	__ai __attribute__((target("neon"))) poly8x8_t vqtbl2_p8(poly8x16x2_t __p0, uint8x8_t __p1) {
58672	60364	poly8x8_t __ret;
58673	60365	poly8x16x2_t __rev0;
58674		__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58675		__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58676		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
58677		__ret = (poly8x8_t) __builtin_neon_vqtbl2_v((int8x16_t)__rev0.val[0], (int8x16_t)__rev0.val[1], (int8x8_t)__rev1, 4);
58678		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	60366	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_128_8);
	60367	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_128_8);
	60368	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	60369	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vqtbl2_v(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x8_t, __rev1), 4));
	60370	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
58679	60371	return __ret;
58680	60372	}
58681	60373	#endif
...	...	@@ -58683,18 +60375,18 @@ __ai __attribute__((target("neon"))) poly8x8_t vqtbl2_p8(poly8x16x2_t __p0, uint
58683	60375	#ifdef __LITTLE_ENDIAN__
58684	60376	__ai __attribute__((target("neon"))) poly8x16_t vqtbl2q_p8(poly8x16x2_t __p0, uint8x16_t __p1) {
58685	60377	poly8x16_t __ret;
58686		__ret = (poly8x16_t) __builtin_neon_vqtbl2q_v((int8x16_t)__p0.val[0], (int8x16_t)__p0.val[1], (int8x16_t)__p1, 36);
	60378	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vqtbl2q_v(__builtin_bit_cast(int8x16_t, __p0.val[0]), __builtin_bit_cast(int8x16_t, __p0.val[1]), __builtin_bit_cast(int8x16_t, __p1), 36));
58687	60379	return __ret;
58688	60380	}
58689	60381	#else
58690	60382	__ai __attribute__((target("neon"))) poly8x16_t vqtbl2q_p8(poly8x16x2_t __p0, uint8x16_t __p1) {
58691	60383	poly8x16_t __ret;
58692	60384	poly8x16x2_t __rev0;
58693		__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58694		__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58695		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58696		__ret = (poly8x16_t) __builtin_neon_vqtbl2q_v((int8x16_t)__rev0.val[0], (int8x16_t)__rev0.val[1], (int8x16_t)__rev1, 36);
58697		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	60385	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_128_8);
	60386	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_128_8);
	60387	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	60388	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vqtbl2q_v(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x16_t, __rev1), 36));
	60389	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
58698	60390	return __ret;
58699	60391	}
58700	60392	#endif
...	...	@@ -58702,18 +60394,18 @@ __ai __attribute__((target("neon"))) poly8x16_t vqtbl2q_p8(poly8x16x2_t __p0, ui
58702	60394	#ifdef __LITTLE_ENDIAN__
58703	60395	__ai __attribute__((target("neon"))) uint8x16_t vqtbl2q_u8(uint8x16x2_t __p0, uint8x16_t __p1) {
58704	60396	uint8x16_t __ret;
58705		__ret = (uint8x16_t) __builtin_neon_vqtbl2q_v((int8x16_t)__p0.val[0], (int8x16_t)__p0.val[1], (int8x16_t)__p1, 48);
	60397	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vqtbl2q_v(__builtin_bit_cast(int8x16_t, __p0.val[0]), __builtin_bit_cast(int8x16_t, __p0.val[1]), __builtin_bit_cast(int8x16_t, __p1), 48));
58706	60398	return __ret;
58707	60399	}
58708	60400	#else
58709	60401	__ai __attribute__((target("neon"))) uint8x16_t vqtbl2q_u8(uint8x16x2_t __p0, uint8x16_t __p1) {
58710	60402	uint8x16_t __ret;
58711	60403	uint8x16x2_t __rev0;
58712		__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58713		__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58714		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58715		__ret = (uint8x16_t) __builtin_neon_vqtbl2q_v((int8x16_t)__rev0.val[0], (int8x16_t)__rev0.val[1], (int8x16_t)__rev1, 48);
58716		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	60404	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_128_8);
	60405	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_128_8);
	60406	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	60407	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vqtbl2q_v(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x16_t, __rev1), 48));
	60408	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
58717	60409	return __ret;
58718	60410	}
58719	60411	#endif
...	...	@@ -58721,18 +60413,37 @@ __ai __attribute__((target("neon"))) uint8x16_t vqtbl2q_u8(uint8x16x2_t __p0, ui
58721	60413	#ifdef __LITTLE_ENDIAN__
58722	60414	__ai __attribute__((target("neon"))) int8x16_t vqtbl2q_s8(int8x16x2_t __p0, uint8x16_t __p1) {
58723	60415	int8x16_t __ret;
58724		__ret = (int8x16_t) __builtin_neon_vqtbl2q_v((int8x16_t)__p0.val[0], (int8x16_t)__p0.val[1], (int8x16_t)__p1, 32);
	60416	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vqtbl2q_v(__builtin_bit_cast(int8x16_t, __p0.val[0]), __builtin_bit_cast(int8x16_t, __p0.val[1]), __builtin_bit_cast(int8x16_t, __p1), 32));
58725	60417	return __ret;
58726	60418	}
58727	60419	#else
58728	60420	__ai __attribute__((target("neon"))) int8x16_t vqtbl2q_s8(int8x16x2_t __p0, uint8x16_t __p1) {
58729	60421	int8x16_t __ret;
58730	60422	int8x16x2_t __rev0;
58731		__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58732		__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58733		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58734		__ret = (int8x16_t) __builtin_neon_vqtbl2q_v((int8x16_t)__rev0.val[0], (int8x16_t)__rev0.val[1], (int8x16_t)__rev1, 32);
58735		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	60423	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_128_8);
	60424	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_128_8);
	60425	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	60426	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vqtbl2q_v(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x16_t, __rev1), 32));
	60427	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
	60428	return __ret;
	60429	}
	60430	#endif
	60431
	60432	#ifdef __LITTLE_ENDIAN__
	60433	__ai __attribute__((target("neon"))) mfloat8x16_t vqtbl2q_mf8(mfloat8x16x2_t __p0, uint8x16_t __p1) {
	60434	mfloat8x16_t __ret;
	60435	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vqtbl2q_v(__builtin_bit_cast(int8x16_t, __p0.val[0]), __builtin_bit_cast(int8x16_t, __p0.val[1]), __builtin_bit_cast(int8x16_t, __p1), 44));
	60436	return __ret;
	60437	}
	60438	#else
	60439	__ai __attribute__((target("neon"))) mfloat8x16_t vqtbl2q_mf8(mfloat8x16x2_t __p0, uint8x16_t __p1) {
	60440	mfloat8x16_t __ret;
	60441	mfloat8x16x2_t __rev0;
	60442	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_128_8);
	60443	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_128_8);
	60444	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	60445	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vqtbl2q_v(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x16_t, __rev1), 44));
	60446	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
58736	60447	return __ret;
58737	60448	}
58738	60449	#endif
...	...	@@ -58740,18 +60451,18 @@ __ai __attribute__((target("neon"))) int8x16_t vqtbl2q_s8(int8x16x2_t __p0, uint
58740	60451	#ifdef __LITTLE_ENDIAN__
58741	60452	__ai __attribute__((target("neon"))) uint8x8_t vqtbl2_u8(uint8x16x2_t __p0, uint8x8_t __p1) {
58742	60453	uint8x8_t __ret;
58743		__ret = (uint8x8_t) __builtin_neon_vqtbl2_v((int8x16_t)__p0.val[0], (int8x16_t)__p0.val[1], (int8x8_t)__p1, 16);
	60454	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqtbl2_v(__builtin_bit_cast(int8x16_t, __p0.val[0]), __builtin_bit_cast(int8x16_t, __p0.val[1]), __builtin_bit_cast(int8x8_t, __p1), 16));
58744	60455	return __ret;
58745	60456	}
58746	60457	#else
58747	60458	__ai __attribute__((target("neon"))) uint8x8_t vqtbl2_u8(uint8x16x2_t __p0, uint8x8_t __p1) {
58748	60459	uint8x8_t __ret;
58749	60460	uint8x16x2_t __rev0;
58750		__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58751		__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58752		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
58753		__ret = (uint8x8_t) __builtin_neon_vqtbl2_v((int8x16_t)__rev0.val[0], (int8x16_t)__rev0.val[1], (int8x8_t)__rev1, 16);
58754		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	60461	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_128_8);
	60462	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_128_8);
	60463	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	60464	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqtbl2_v(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x8_t, __rev1), 16));
	60465	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
58755	60466	return __ret;
58756	60467	}
58757	60468	#endif
...	...	@@ -58759,18 +60470,37 @@ __ai __attribute__((target("neon"))) uint8x8_t vqtbl2_u8(uint8x16x2_t __p0, uint
58759	60470	#ifdef __LITTLE_ENDIAN__
58760	60471	__ai __attribute__((target("neon"))) int8x8_t vqtbl2_s8(int8x16x2_t __p0, uint8x8_t __p1) {
58761	60472	int8x8_t __ret;
58762		__ret = (int8x8_t) __builtin_neon_vqtbl2_v((int8x16_t)__p0.val[0], (int8x16_t)__p0.val[1], (int8x8_t)__p1, 0);
	60473	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqtbl2_v(__builtin_bit_cast(int8x16_t, __p0.val[0]), __builtin_bit_cast(int8x16_t, __p0.val[1]), __builtin_bit_cast(int8x8_t, __p1), 0));
58763	60474	return __ret;
58764	60475	}
58765	60476	#else
58766	60477	__ai __attribute__((target("neon"))) int8x8_t vqtbl2_s8(int8x16x2_t __p0, uint8x8_t __p1) {
58767	60478	int8x8_t __ret;
58768	60479	int8x16x2_t __rev0;
58769		__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58770		__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58771		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
58772		__ret = (int8x8_t) __builtin_neon_vqtbl2_v((int8x16_t)__rev0.val[0], (int8x16_t)__rev0.val[1], (int8x8_t)__rev1, 0);
58773		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	60480	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_128_8);
	60481	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_128_8);
	60482	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	60483	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqtbl2_v(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x8_t, __rev1), 0));
	60484	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
	60485	return __ret;
	60486	}
	60487	#endif
	60488
	60489	#ifdef __LITTLE_ENDIAN__
	60490	__ai __attribute__((target("neon"))) mfloat8x8_t vqtbl2_mf8(mfloat8x16x2_t __p0, uint8x8_t __p1) {
	60491	mfloat8x8_t __ret;
	60492	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vqtbl2_v(__builtin_bit_cast(int8x16_t, __p0.val[0]), __builtin_bit_cast(int8x16_t, __p0.val[1]), __builtin_bit_cast(int8x8_t, __p1), 12));
	60493	return __ret;
	60494	}
	60495	#else
	60496	__ai __attribute__((target("neon"))) mfloat8x8_t vqtbl2_mf8(mfloat8x16x2_t __p0, uint8x8_t __p1) {
	60497	mfloat8x8_t __ret;
	60498	mfloat8x16x2_t __rev0;
	60499	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_128_8);
	60500	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_128_8);
	60501	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	60502	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vqtbl2_v(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x8_t, __rev1), 12));
	60503	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
58774	60504	return __ret;
58775	60505	}
58776	60506	#endif
...	...	@@ -58778,19 +60508,19 @@ __ai __attribute__((target("neon"))) int8x8_t vqtbl2_s8(int8x16x2_t __p0, uint8x
58778	60508	#ifdef __LITTLE_ENDIAN__
58779	60509	__ai __attribute__((target("neon"))) poly8x8_t vqtbl3_p8(poly8x16x3_t __p0, uint8x8_t __p1) {
58780	60510	poly8x8_t __ret;
58781		__ret = (poly8x8_t) __builtin_neon_vqtbl3_v((int8x16_t)__p0.val[0], (int8x16_t)__p0.val[1], (int8x16_t)__p0.val[2], (int8x8_t)__p1, 4);
	60511	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vqtbl3_v(__builtin_bit_cast(int8x16_t, __p0.val[0]), __builtin_bit_cast(int8x16_t, __p0.val[1]), __builtin_bit_cast(int8x16_t, __p0.val[2]), __builtin_bit_cast(int8x8_t, __p1), 4));
58782	60512	return __ret;
58783	60513	}
58784	60514	#else
58785	60515	__ai __attribute__((target("neon"))) poly8x8_t vqtbl3_p8(poly8x16x3_t __p0, uint8x8_t __p1) {
58786	60516	poly8x8_t __ret;
58787	60517	poly8x16x3_t __rev0;
58788		__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58789		__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58790		__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58791		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
58792		__ret = (poly8x8_t) __builtin_neon_vqtbl3_v((int8x16_t)__rev0.val[0], (int8x16_t)__rev0.val[1], (int8x16_t)__rev0.val[2], (int8x8_t)__rev1, 4);
58793		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	60518	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_128_8);
	60519	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_128_8);
	60520	__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], __lane_reverse_128_8);
	60521	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	60522	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vqtbl3_v(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x16_t, __rev0.val[2]), __builtin_bit_cast(int8x8_t, __rev1), 4));
	60523	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
58794	60524	return __ret;
58795	60525	}
58796	60526	#endif
...	...	@@ -58798,19 +60528,19 @@ __ai __attribute__((target("neon"))) poly8x8_t vqtbl3_p8(poly8x16x3_t __p0, uint
58798	60528	#ifdef __LITTLE_ENDIAN__
58799	60529	__ai __attribute__((target("neon"))) poly8x16_t vqtbl3q_p8(poly8x16x3_t __p0, uint8x16_t __p1) {
58800	60530	poly8x16_t __ret;
58801		__ret = (poly8x16_t) __builtin_neon_vqtbl3q_v((int8x16_t)__p0.val[0], (int8x16_t)__p0.val[1], (int8x16_t)__p0.val[2], (int8x16_t)__p1, 36);
	60531	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vqtbl3q_v(__builtin_bit_cast(int8x16_t, __p0.val[0]), __builtin_bit_cast(int8x16_t, __p0.val[1]), __builtin_bit_cast(int8x16_t, __p0.val[2]), __builtin_bit_cast(int8x16_t, __p1), 36));
58802	60532	return __ret;
58803	60533	}
58804	60534	#else
58805	60535	__ai __attribute__((target("neon"))) poly8x16_t vqtbl3q_p8(poly8x16x3_t __p0, uint8x16_t __p1) {
58806	60536	poly8x16_t __ret;
58807	60537	poly8x16x3_t __rev0;
58808		__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58809		__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58810		__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58811		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58812		__ret = (poly8x16_t) __builtin_neon_vqtbl3q_v((int8x16_t)__rev0.val[0], (int8x16_t)__rev0.val[1], (int8x16_t)__rev0.val[2], (int8x16_t)__rev1, 36);
58813		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	60538	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_128_8);
	60539	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_128_8);
	60540	__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], __lane_reverse_128_8);
	60541	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	60542	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vqtbl3q_v(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x16_t, __rev0.val[2]), __builtin_bit_cast(int8x16_t, __rev1), 36));
	60543	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
58814	60544	return __ret;
58815	60545	}
58816	60546	#endif
...	...	@@ -58818,19 +60548,19 @@ __ai __attribute__((target("neon"))) poly8x16_t vqtbl3q_p8(poly8x16x3_t __p0, ui
58818	60548	#ifdef __LITTLE_ENDIAN__
58819	60549	__ai __attribute__((target("neon"))) uint8x16_t vqtbl3q_u8(uint8x16x3_t __p0, uint8x16_t __p1) {
58820	60550	uint8x16_t __ret;
58821		__ret = (uint8x16_t) __builtin_neon_vqtbl3q_v((int8x16_t)__p0.val[0], (int8x16_t)__p0.val[1], (int8x16_t)__p0.val[2], (int8x16_t)__p1, 48);
	60551	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vqtbl3q_v(__builtin_bit_cast(int8x16_t, __p0.val[0]), __builtin_bit_cast(int8x16_t, __p0.val[1]), __builtin_bit_cast(int8x16_t, __p0.val[2]), __builtin_bit_cast(int8x16_t, __p1), 48));
58822	60552	return __ret;
58823	60553	}
58824	60554	#else
58825	60555	__ai __attribute__((target("neon"))) uint8x16_t vqtbl3q_u8(uint8x16x3_t __p0, uint8x16_t __p1) {
58826	60556	uint8x16_t __ret;
58827	60557	uint8x16x3_t __rev0;
58828		__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58829		__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58830		__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58831		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58832		__ret = (uint8x16_t) __builtin_neon_vqtbl3q_v((int8x16_t)__rev0.val[0], (int8x16_t)__rev0.val[1], (int8x16_t)__rev0.val[2], (int8x16_t)__rev1, 48);
58833		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	60558	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_128_8);
	60559	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_128_8);
	60560	__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], __lane_reverse_128_8);
	60561	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	60562	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vqtbl3q_v(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x16_t, __rev0.val[2]), __builtin_bit_cast(int8x16_t, __rev1), 48));
	60563	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
58834	60564	return __ret;
58835	60565	}
58836	60566	#endif
...	...	@@ -58838,19 +60568,39 @@ __ai __attribute__((target("neon"))) uint8x16_t vqtbl3q_u8(uint8x16x3_t __p0, ui
58838	60568	#ifdef __LITTLE_ENDIAN__
58839	60569	__ai __attribute__((target("neon"))) int8x16_t vqtbl3q_s8(int8x16x3_t __p0, uint8x16_t __p1) {
58840	60570	int8x16_t __ret;
58841		__ret = (int8x16_t) __builtin_neon_vqtbl3q_v((int8x16_t)__p0.val[0], (int8x16_t)__p0.val[1], (int8x16_t)__p0.val[2], (int8x16_t)__p1, 32);
	60571	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vqtbl3q_v(__builtin_bit_cast(int8x16_t, __p0.val[0]), __builtin_bit_cast(int8x16_t, __p0.val[1]), __builtin_bit_cast(int8x16_t, __p0.val[2]), __builtin_bit_cast(int8x16_t, __p1), 32));
58842	60572	return __ret;
58843	60573	}
58844	60574	#else
58845	60575	__ai __attribute__((target("neon"))) int8x16_t vqtbl3q_s8(int8x16x3_t __p0, uint8x16_t __p1) {
58846	60576	int8x16_t __ret;
58847	60577	int8x16x3_t __rev0;
58848		__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58849		__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58850		__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58851		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58852		__ret = (int8x16_t) __builtin_neon_vqtbl3q_v((int8x16_t)__rev0.val[0], (int8x16_t)__rev0.val[1], (int8x16_t)__rev0.val[2], (int8x16_t)__rev1, 32);
58853		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	60578	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_128_8);
	60579	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_128_8);
	60580	__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], __lane_reverse_128_8);
	60581	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	60582	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vqtbl3q_v(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x16_t, __rev0.val[2]), __builtin_bit_cast(int8x16_t, __rev1), 32));
	60583	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
	60584	return __ret;
	60585	}
	60586	#endif
	60587
	60588	#ifdef __LITTLE_ENDIAN__
	60589	__ai __attribute__((target("neon"))) mfloat8x16_t vqtbl3q_mf8(mfloat8x16x3_t __p0, uint8x16_t __p1) {
	60590	mfloat8x16_t __ret;
	60591	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vqtbl3q_v(__builtin_bit_cast(int8x16_t, __p0.val[0]), __builtin_bit_cast(int8x16_t, __p0.val[1]), __builtin_bit_cast(int8x16_t, __p0.val[2]), __builtin_bit_cast(int8x16_t, __p1), 44));
	60592	return __ret;
	60593	}
	60594	#else
	60595	__ai __attribute__((target("neon"))) mfloat8x16_t vqtbl3q_mf8(mfloat8x16x3_t __p0, uint8x16_t __p1) {
	60596	mfloat8x16_t __ret;
	60597	mfloat8x16x3_t __rev0;
	60598	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_128_8);
	60599	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_128_8);
	60600	__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], __lane_reverse_128_8);
	60601	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	60602	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vqtbl3q_v(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x16_t, __rev0.val[2]), __builtin_bit_cast(int8x16_t, __rev1), 44));
	60603	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
58854	60604	return __ret;
58855	60605	}
58856	60606	#endif
...	...	@@ -58858,19 +60608,19 @@ __ai __attribute__((target("neon"))) int8x16_t vqtbl3q_s8(int8x16x3_t __p0, uint
58858	60608	#ifdef __LITTLE_ENDIAN__
58859	60609	__ai __attribute__((target("neon"))) uint8x8_t vqtbl3_u8(uint8x16x3_t __p0, uint8x8_t __p1) {
58860	60610	uint8x8_t __ret;
58861		__ret = (uint8x8_t) __builtin_neon_vqtbl3_v((int8x16_t)__p0.val[0], (int8x16_t)__p0.val[1], (int8x16_t)__p0.val[2], (int8x8_t)__p1, 16);
	60611	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqtbl3_v(__builtin_bit_cast(int8x16_t, __p0.val[0]), __builtin_bit_cast(int8x16_t, __p0.val[1]), __builtin_bit_cast(int8x16_t, __p0.val[2]), __builtin_bit_cast(int8x8_t, __p1), 16));
58862	60612	return __ret;
58863	60613	}
58864	60614	#else
58865	60615	__ai __attribute__((target("neon"))) uint8x8_t vqtbl3_u8(uint8x16x3_t __p0, uint8x8_t __p1) {
58866	60616	uint8x8_t __ret;
58867	60617	uint8x16x3_t __rev0;
58868		__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58869		__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58870		__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58871		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
58872		__ret = (uint8x8_t) __builtin_neon_vqtbl3_v((int8x16_t)__rev0.val[0], (int8x16_t)__rev0.val[1], (int8x16_t)__rev0.val[2], (int8x8_t)__rev1, 16);
58873		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	60618	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_128_8);
	60619	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_128_8);
	60620	__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], __lane_reverse_128_8);
	60621	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	60622	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqtbl3_v(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x16_t, __rev0.val[2]), __builtin_bit_cast(int8x8_t, __rev1), 16));
	60623	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
58874	60624	return __ret;
58875	60625	}
58876	60626	#endif
...	...	@@ -58878,19 +60628,39 @@ __ai __attribute__((target("neon"))) uint8x8_t vqtbl3_u8(uint8x16x3_t __p0, uint
58878	60628	#ifdef __LITTLE_ENDIAN__
58879	60629	__ai __attribute__((target("neon"))) int8x8_t vqtbl3_s8(int8x16x3_t __p0, uint8x8_t __p1) {
58880	60630	int8x8_t __ret;
58881		__ret = (int8x8_t) __builtin_neon_vqtbl3_v((int8x16_t)__p0.val[0], (int8x16_t)__p0.val[1], (int8x16_t)__p0.val[2], (int8x8_t)__p1, 0);
	60631	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqtbl3_v(__builtin_bit_cast(int8x16_t, __p0.val[0]), __builtin_bit_cast(int8x16_t, __p0.val[1]), __builtin_bit_cast(int8x16_t, __p0.val[2]), __builtin_bit_cast(int8x8_t, __p1), 0));
58882	60632	return __ret;
58883	60633	}
58884	60634	#else
58885	60635	__ai __attribute__((target("neon"))) int8x8_t vqtbl3_s8(int8x16x3_t __p0, uint8x8_t __p1) {
58886	60636	int8x8_t __ret;
58887	60637	int8x16x3_t __rev0;
58888		__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58889		__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58890		__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58891		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
58892		__ret = (int8x8_t) __builtin_neon_vqtbl3_v((int8x16_t)__rev0.val[0], (int8x16_t)__rev0.val[1], (int8x16_t)__rev0.val[2], (int8x8_t)__rev1, 0);
58893		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	60638	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_128_8);
	60639	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_128_8);
	60640	__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], __lane_reverse_128_8);
	60641	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	60642	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqtbl3_v(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x16_t, __rev0.val[2]), __builtin_bit_cast(int8x8_t, __rev1), 0));
	60643	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
	60644	return __ret;
	60645	}
	60646	#endif
	60647
	60648	#ifdef __LITTLE_ENDIAN__
	60649	__ai __attribute__((target("neon"))) mfloat8x8_t vqtbl3_mf8(mfloat8x16x3_t __p0, uint8x8_t __p1) {
	60650	mfloat8x8_t __ret;
	60651	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vqtbl3_v(__builtin_bit_cast(int8x16_t, __p0.val[0]), __builtin_bit_cast(int8x16_t, __p0.val[1]), __builtin_bit_cast(int8x16_t, __p0.val[2]), __builtin_bit_cast(int8x8_t, __p1), 12));
	60652	return __ret;
	60653	}
	60654	#else
	60655	__ai __attribute__((target("neon"))) mfloat8x8_t vqtbl3_mf8(mfloat8x16x3_t __p0, uint8x8_t __p1) {
	60656	mfloat8x8_t __ret;
	60657	mfloat8x16x3_t __rev0;
	60658	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_128_8);
	60659	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_128_8);
	60660	__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], __lane_reverse_128_8);
	60661	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	60662	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vqtbl3_v(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x16_t, __rev0.val[2]), __builtin_bit_cast(int8x8_t, __rev1), 12));
	60663	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
58894	60664	return __ret;
58895	60665	}
58896	60666	#endif
...	...	@@ -58898,20 +60668,20 @@ __ai __attribute__((target("neon"))) int8x8_t vqtbl3_s8(int8x16x3_t __p0, uint8x
58898	60668	#ifdef __LITTLE_ENDIAN__
58899	60669	__ai __attribute__((target("neon"))) poly8x8_t vqtbl4_p8(poly8x16x4_t __p0, uint8x8_t __p1) {
58900	60670	poly8x8_t __ret;
58901		__ret = (poly8x8_t) __builtin_neon_vqtbl4_v((int8x16_t)__p0.val[0], (int8x16_t)__p0.val[1], (int8x16_t)__p0.val[2], (int8x16_t)__p0.val[3], (int8x8_t)__p1, 4);
	60671	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vqtbl4_v(__builtin_bit_cast(int8x16_t, __p0.val[0]), __builtin_bit_cast(int8x16_t, __p0.val[1]), __builtin_bit_cast(int8x16_t, __p0.val[2]), __builtin_bit_cast(int8x16_t, __p0.val[3]), __builtin_bit_cast(int8x8_t, __p1), 4));
58902	60672	return __ret;
58903	60673	}
58904	60674	#else
58905	60675	__ai __attribute__((target("neon"))) poly8x8_t vqtbl4_p8(poly8x16x4_t __p0, uint8x8_t __p1) {
58906	60676	poly8x8_t __ret;
58907	60677	poly8x16x4_t __rev0;
58908		__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58909		__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58910		__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58911		__rev0.val[3] = __builtin_shufflevector(__p0.val[3], __p0.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58912		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
58913		__ret = (poly8x8_t) __builtin_neon_vqtbl4_v((int8x16_t)__rev0.val[0], (int8x16_t)__rev0.val[1], (int8x16_t)__rev0.val[2], (int8x16_t)__rev0.val[3], (int8x8_t)__rev1, 4);
58914		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	60678	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_128_8);
	60679	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_128_8);
	60680	__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], __lane_reverse_128_8);
	60681	__rev0.val[3] = __builtin_shufflevector(__p0.val[3], __p0.val[3], __lane_reverse_128_8);
	60682	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	60683	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vqtbl4_v(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x16_t, __rev0.val[2]), __builtin_bit_cast(int8x16_t, __rev0.val[3]), __builtin_bit_cast(int8x8_t, __rev1), 4));
	60684	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
58915	60685	return __ret;
58916	60686	}
58917	60687	#endif
...	...	@@ -58919,20 +60689,20 @@ __ai __attribute__((target("neon"))) poly8x8_t vqtbl4_p8(poly8x16x4_t __p0, uint
58919	60689	#ifdef __LITTLE_ENDIAN__
58920	60690	__ai __attribute__((target("neon"))) poly8x16_t vqtbl4q_p8(poly8x16x4_t __p0, uint8x16_t __p1) {
58921	60691	poly8x16_t __ret;
58922		__ret = (poly8x16_t) __builtin_neon_vqtbl4q_v((int8x16_t)__p0.val[0], (int8x16_t)__p0.val[1], (int8x16_t)__p0.val[2], (int8x16_t)__p0.val[3], (int8x16_t)__p1, 36);
	60692	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vqtbl4q_v(__builtin_bit_cast(int8x16_t, __p0.val[0]), __builtin_bit_cast(int8x16_t, __p0.val[1]), __builtin_bit_cast(int8x16_t, __p0.val[2]), __builtin_bit_cast(int8x16_t, __p0.val[3]), __builtin_bit_cast(int8x16_t, __p1), 36));
58923	60693	return __ret;
58924	60694	}
58925	60695	#else
58926	60696	__ai __attribute__((target("neon"))) poly8x16_t vqtbl4q_p8(poly8x16x4_t __p0, uint8x16_t __p1) {
58927	60697	poly8x16_t __ret;
58928	60698	poly8x16x4_t __rev0;
58929		__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58930		__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58931		__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58932		__rev0.val[3] = __builtin_shufflevector(__p0.val[3], __p0.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58933		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58934		__ret = (poly8x16_t) __builtin_neon_vqtbl4q_v((int8x16_t)__rev0.val[0], (int8x16_t)__rev0.val[1], (int8x16_t)__rev0.val[2], (int8x16_t)__rev0.val[3], (int8x16_t)__rev1, 36);
58935		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	60699	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_128_8);
	60700	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_128_8);
	60701	__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], __lane_reverse_128_8);
	60702	__rev0.val[3] = __builtin_shufflevector(__p0.val[3], __p0.val[3], __lane_reverse_128_8);
	60703	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	60704	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vqtbl4q_v(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x16_t, __rev0.val[2]), __builtin_bit_cast(int8x16_t, __rev0.val[3]), __builtin_bit_cast(int8x16_t, __rev1), 36));
	60705	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
58936	60706	return __ret;
58937	60707	}
58938	60708	#endif
...	...	@@ -58940,20 +60710,20 @@ __ai __attribute__((target("neon"))) poly8x16_t vqtbl4q_p8(poly8x16x4_t __p0, ui
58940	60710	#ifdef __LITTLE_ENDIAN__
58941	60711	__ai __attribute__((target("neon"))) uint8x16_t vqtbl4q_u8(uint8x16x4_t __p0, uint8x16_t __p1) {
58942	60712	uint8x16_t __ret;
58943		__ret = (uint8x16_t) __builtin_neon_vqtbl4q_v((int8x16_t)__p0.val[0], (int8x16_t)__p0.val[1], (int8x16_t)__p0.val[2], (int8x16_t)__p0.val[3], (int8x16_t)__p1, 48);
	60713	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vqtbl4q_v(__builtin_bit_cast(int8x16_t, __p0.val[0]), __builtin_bit_cast(int8x16_t, __p0.val[1]), __builtin_bit_cast(int8x16_t, __p0.val[2]), __builtin_bit_cast(int8x16_t, __p0.val[3]), __builtin_bit_cast(int8x16_t, __p1), 48));
58944	60714	return __ret;
58945	60715	}
58946	60716	#else
58947	60717	__ai __attribute__((target("neon"))) uint8x16_t vqtbl4q_u8(uint8x16x4_t __p0, uint8x16_t __p1) {
58948	60718	uint8x16_t __ret;
58949	60719	uint8x16x4_t __rev0;
58950		__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58951		__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58952		__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58953		__rev0.val[3] = __builtin_shufflevector(__p0.val[3], __p0.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58954		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58955		__ret = (uint8x16_t) __builtin_neon_vqtbl4q_v((int8x16_t)__rev0.val[0], (int8x16_t)__rev0.val[1], (int8x16_t)__rev0.val[2], (int8x16_t)__rev0.val[3], (int8x16_t)__rev1, 48);
58956		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	60720	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_128_8);
	60721	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_128_8);
	60722	__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], __lane_reverse_128_8);
	60723	__rev0.val[3] = __builtin_shufflevector(__p0.val[3], __p0.val[3], __lane_reverse_128_8);
	60724	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	60725	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vqtbl4q_v(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x16_t, __rev0.val[2]), __builtin_bit_cast(int8x16_t, __rev0.val[3]), __builtin_bit_cast(int8x16_t, __rev1), 48));
	60726	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
58957	60727	return __ret;
58958	60728	}
58959	60729	#endif
...	...	@@ -58961,20 +60731,41 @@ __ai __attribute__((target("neon"))) uint8x16_t vqtbl4q_u8(uint8x16x4_t __p0, ui
58961	60731	#ifdef __LITTLE_ENDIAN__
58962	60732	__ai __attribute__((target("neon"))) int8x16_t vqtbl4q_s8(int8x16x4_t __p0, uint8x16_t __p1) {
58963	60733	int8x16_t __ret;
58964		__ret = (int8x16_t) __builtin_neon_vqtbl4q_v((int8x16_t)__p0.val[0], (int8x16_t)__p0.val[1], (int8x16_t)__p0.val[2], (int8x16_t)__p0.val[3], (int8x16_t)__p1, 32);
	60734	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vqtbl4q_v(__builtin_bit_cast(int8x16_t, __p0.val[0]), __builtin_bit_cast(int8x16_t, __p0.val[1]), __builtin_bit_cast(int8x16_t, __p0.val[2]), __builtin_bit_cast(int8x16_t, __p0.val[3]), __builtin_bit_cast(int8x16_t, __p1), 32));
58965	60735	return __ret;
58966	60736	}
58967	60737	#else
58968	60738	__ai __attribute__((target("neon"))) int8x16_t vqtbl4q_s8(int8x16x4_t __p0, uint8x16_t __p1) {
58969	60739	int8x16_t __ret;
58970	60740	int8x16x4_t __rev0;
58971		__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58972		__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58973		__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58974		__rev0.val[3] = __builtin_shufflevector(__p0.val[3], __p0.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58975		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58976		__ret = (int8x16_t) __builtin_neon_vqtbl4q_v((int8x16_t)__rev0.val[0], (int8x16_t)__rev0.val[1], (int8x16_t)__rev0.val[2], (int8x16_t)__rev0.val[3], (int8x16_t)__rev1, 32);
58977		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	60741	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_128_8);
	60742	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_128_8);
	60743	__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], __lane_reverse_128_8);
	60744	__rev0.val[3] = __builtin_shufflevector(__p0.val[3], __p0.val[3], __lane_reverse_128_8);
	60745	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	60746	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vqtbl4q_v(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x16_t, __rev0.val[2]), __builtin_bit_cast(int8x16_t, __rev0.val[3]), __builtin_bit_cast(int8x16_t, __rev1), 32));
	60747	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
	60748	return __ret;
	60749	}
	60750	#endif
	60751
	60752	#ifdef __LITTLE_ENDIAN__
	60753	__ai __attribute__((target("neon"))) mfloat8x16_t vqtbl4q_mf8(mfloat8x16x4_t __p0, uint8x16_t __p1) {
	60754	mfloat8x16_t __ret;
	60755	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vqtbl4q_v(__builtin_bit_cast(int8x16_t, __p0.val[0]), __builtin_bit_cast(int8x16_t, __p0.val[1]), __builtin_bit_cast(int8x16_t, __p0.val[2]), __builtin_bit_cast(int8x16_t, __p0.val[3]), __builtin_bit_cast(int8x16_t, __p1), 44));
	60756	return __ret;
	60757	}
	60758	#else
	60759	__ai __attribute__((target("neon"))) mfloat8x16_t vqtbl4q_mf8(mfloat8x16x4_t __p0, uint8x16_t __p1) {
	60760	mfloat8x16_t __ret;
	60761	mfloat8x16x4_t __rev0;
	60762	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_128_8);
	60763	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_128_8);
	60764	__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], __lane_reverse_128_8);
	60765	__rev0.val[3] = __builtin_shufflevector(__p0.val[3], __p0.val[3], __lane_reverse_128_8);
	60766	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	60767	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vqtbl4q_v(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x16_t, __rev0.val[2]), __builtin_bit_cast(int8x16_t, __rev0.val[3]), __builtin_bit_cast(int8x16_t, __rev1), 44));
	60768	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
58978	60769	return __ret;
58979	60770	}
58980	60771	#endif
...	...	@@ -58982,20 +60773,20 @@ __ai __attribute__((target("neon"))) int8x16_t vqtbl4q_s8(int8x16x4_t __p0, uint
58982	60773	#ifdef __LITTLE_ENDIAN__
58983	60774	__ai __attribute__((target("neon"))) uint8x8_t vqtbl4_u8(uint8x16x4_t __p0, uint8x8_t __p1) {
58984	60775	uint8x8_t __ret;
58985		__ret = (uint8x8_t) __builtin_neon_vqtbl4_v((int8x16_t)__p0.val[0], (int8x16_t)__p0.val[1], (int8x16_t)__p0.val[2], (int8x16_t)__p0.val[3], (int8x8_t)__p1, 16);
	60776	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqtbl4_v(__builtin_bit_cast(int8x16_t, __p0.val[0]), __builtin_bit_cast(int8x16_t, __p0.val[1]), __builtin_bit_cast(int8x16_t, __p0.val[2]), __builtin_bit_cast(int8x16_t, __p0.val[3]), __builtin_bit_cast(int8x8_t, __p1), 16));
58986	60777	return __ret;
58987	60778	}
58988	60779	#else
58989	60780	__ai __attribute__((target("neon"))) uint8x8_t vqtbl4_u8(uint8x16x4_t __p0, uint8x8_t __p1) {
58990	60781	uint8x8_t __ret;
58991	60782	uint8x16x4_t __rev0;
58992		__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58993		__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58994		__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58995		__rev0.val[3] = __builtin_shufflevector(__p0.val[3], __p0.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
58996		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
58997		__ret = (uint8x8_t) __builtin_neon_vqtbl4_v((int8x16_t)__rev0.val[0], (int8x16_t)__rev0.val[1], (int8x16_t)__rev0.val[2], (int8x16_t)__rev0.val[3], (int8x8_t)__rev1, 16);
58998		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	60783	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_128_8);
	60784	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_128_8);
	60785	__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], __lane_reverse_128_8);
	60786	__rev0.val[3] = __builtin_shufflevector(__p0.val[3], __p0.val[3], __lane_reverse_128_8);
	60787	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	60788	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqtbl4_v(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x16_t, __rev0.val[2]), __builtin_bit_cast(int8x16_t, __rev0.val[3]), __builtin_bit_cast(int8x8_t, __rev1), 16));
	60789	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
58999	60790	return __ret;
59000	60791	}
59001	60792	#endif
...	...	@@ -59003,20 +60794,41 @@ __ai __attribute__((target("neon"))) uint8x8_t vqtbl4_u8(uint8x16x4_t __p0, uint
59003	60794	#ifdef __LITTLE_ENDIAN__
59004	60795	__ai __attribute__((target("neon"))) int8x8_t vqtbl4_s8(int8x16x4_t __p0, uint8x8_t __p1) {
59005	60796	int8x8_t __ret;
59006		__ret = (int8x8_t) __builtin_neon_vqtbl4_v((int8x16_t)__p0.val[0], (int8x16_t)__p0.val[1], (int8x16_t)__p0.val[2], (int8x16_t)__p0.val[3], (int8x8_t)__p1, 0);
	60797	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqtbl4_v(__builtin_bit_cast(int8x16_t, __p0.val[0]), __builtin_bit_cast(int8x16_t, __p0.val[1]), __builtin_bit_cast(int8x16_t, __p0.val[2]), __builtin_bit_cast(int8x16_t, __p0.val[3]), __builtin_bit_cast(int8x8_t, __p1), 0));
59007	60798	return __ret;
59008	60799	}
59009	60800	#else
59010	60801	__ai __attribute__((target("neon"))) int8x8_t vqtbl4_s8(int8x16x4_t __p0, uint8x8_t __p1) {
59011	60802	int8x8_t __ret;
59012	60803	int8x16x4_t __rev0;
59013		__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59014		__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59015		__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59016		__rev0.val[3] = __builtin_shufflevector(__p0.val[3], __p0.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59017		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
59018		__ret = (int8x8_t) __builtin_neon_vqtbl4_v((int8x16_t)__rev0.val[0], (int8x16_t)__rev0.val[1], (int8x16_t)__rev0.val[2], (int8x16_t)__rev0.val[3], (int8x8_t)__rev1, 0);
59019		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	60804	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_128_8);
	60805	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_128_8);
	60806	__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], __lane_reverse_128_8);
	60807	__rev0.val[3] = __builtin_shufflevector(__p0.val[3], __p0.val[3], __lane_reverse_128_8);
	60808	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	60809	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqtbl4_v(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x16_t, __rev0.val[2]), __builtin_bit_cast(int8x16_t, __rev0.val[3]), __builtin_bit_cast(int8x8_t, __rev1), 0));
	60810	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
	60811	return __ret;
	60812	}
	60813	#endif
	60814
	60815	#ifdef __LITTLE_ENDIAN__
	60816	__ai __attribute__((target("neon"))) mfloat8x8_t vqtbl4_mf8(mfloat8x16x4_t __p0, uint8x8_t __p1) {
	60817	mfloat8x8_t __ret;
	60818	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vqtbl4_v(__builtin_bit_cast(int8x16_t, __p0.val[0]), __builtin_bit_cast(int8x16_t, __p0.val[1]), __builtin_bit_cast(int8x16_t, __p0.val[2]), __builtin_bit_cast(int8x16_t, __p0.val[3]), __builtin_bit_cast(int8x8_t, __p1), 12));
	60819	return __ret;
	60820	}
	60821	#else
	60822	__ai __attribute__((target("neon"))) mfloat8x8_t vqtbl4_mf8(mfloat8x16x4_t __p0, uint8x8_t __p1) {
	60823	mfloat8x8_t __ret;
	60824	mfloat8x16x4_t __rev0;
	60825	__rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], __lane_reverse_128_8);
	60826	__rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], __lane_reverse_128_8);
	60827	__rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], __lane_reverse_128_8);
	60828	__rev0.val[3] = __builtin_shufflevector(__p0.val[3], __p0.val[3], __lane_reverse_128_8);
	60829	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	60830	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vqtbl4_v(__builtin_bit_cast(int8x16_t, __rev0.val[0]), __builtin_bit_cast(int8x16_t, __rev0.val[1]), __builtin_bit_cast(int8x16_t, __rev0.val[2]), __builtin_bit_cast(int8x16_t, __rev0.val[3]), __builtin_bit_cast(int8x8_t, __rev1), 12));
	60831	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
59020	60832	return __ret;
59021	60833	}
59022	60834	#endif
...	...	@@ -59024,17 +60836,17 @@ __ai __attribute__((target("neon"))) int8x8_t vqtbl4_s8(int8x16x4_t __p0, uint8x
59024	60836	#ifdef __LITTLE_ENDIAN__
59025	60837	__ai __attribute__((target("neon"))) poly8x8_t vqtbx1_p8(poly8x8_t __p0, poly8x16_t __p1, uint8x8_t __p2) {
59026	60838	poly8x8_t __ret;
59027		__ret = (poly8x8_t) __builtin_neon_vqtbx1_v((int8x8_t)__p0, (int8x16_t)__p1, (int8x8_t)__p2, 4);
	60839	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vqtbx1_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 4));
59028	60840	return __ret;
59029	60841	}
59030	60842	#else
59031	60843	__ai __attribute__((target("neon"))) poly8x8_t vqtbx1_p8(poly8x8_t __p0, poly8x16_t __p1, uint8x8_t __p2) {
59032	60844	poly8x8_t __ret;
59033		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
59034		poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59035		uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
59036		__ret = (poly8x8_t) __builtin_neon_vqtbx1_v((int8x8_t)__rev0, (int8x16_t)__rev1, (int8x8_t)__rev2, 4);
59037		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	60845	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	60846	poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	60847	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	60848	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vqtbx1_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 4));
	60849	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
59038	60850	return __ret;
59039	60851	}
59040	60852	#endif
...	...	@@ -59042,17 +60854,17 @@ __ai __attribute__((target("neon"))) poly8x8_t vqtbx1_p8(poly8x8_t __p0, poly8x1
59042	60854	#ifdef __LITTLE_ENDIAN__
59043	60855	__ai __attribute__((target("neon"))) poly8x16_t vqtbx1q_p8(poly8x16_t __p0, poly8x16_t __p1, uint8x16_t __p2) {
59044	60856	poly8x16_t __ret;
59045		__ret = (poly8x16_t) __builtin_neon_vqtbx1q_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 36);
	60857	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vqtbx1q_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 36));
59046	60858	return __ret;
59047	60859	}
59048	60860	#else
59049	60861	__ai __attribute__((target("neon"))) poly8x16_t vqtbx1q_p8(poly8x16_t __p0, poly8x16_t __p1, uint8x16_t __p2) {
59050	60862	poly8x16_t __ret;
59051		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59052		poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59053		uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59054		__ret = (poly8x16_t) __builtin_neon_vqtbx1q_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 36);
59055		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	60863	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	60864	poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	60865	uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	60866	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vqtbx1q_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 36));
	60867	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
59056	60868	return __ret;
59057	60869	}
59058	60870	#endif
...	...	@@ -59060,17 +60872,17 @@ __ai __attribute__((target("neon"))) poly8x16_t vqtbx1q_p8(poly8x16_t __p0, poly
59060	60872	#ifdef __LITTLE_ENDIAN__
59061	60873	__ai __attribute__((target("neon"))) uint8x16_t vqtbx1q_u8(uint8x16_t __p0, uint8x16_t __p1, uint8x16_t __p2) {
59062	60874	uint8x16_t __ret;
59063		__ret = (uint8x16_t) __builtin_neon_vqtbx1q_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 48);
	60875	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vqtbx1q_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 48));
59064	60876	return __ret;
59065	60877	}
59066	60878	#else
59067	60879	__ai __attribute__((target("neon"))) uint8x16_t vqtbx1q_u8(uint8x16_t __p0, uint8x16_t __p1, uint8x16_t __p2) {
59068	60880	uint8x16_t __ret;
59069		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59070		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59071		uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59072		__ret = (uint8x16_t) __builtin_neon_vqtbx1q_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 48);
59073		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	60881	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	60882	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	60883	uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	60884	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vqtbx1q_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 48));
	60885	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
59074	60886	return __ret;
59075	60887	}
59076	60888	#endif
...	...	@@ -59078,17 +60890,35 @@ __ai __attribute__((target("neon"))) uint8x16_t vqtbx1q_u8(uint8x16_t __p0, uint
59078	60890	#ifdef __LITTLE_ENDIAN__
59079	60891	__ai __attribute__((target("neon"))) int8x16_t vqtbx1q_s8(int8x16_t __p0, int8x16_t __p1, uint8x16_t __p2) {
59080	60892	int8x16_t __ret;
59081		__ret = (int8x16_t) __builtin_neon_vqtbx1q_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 32);
	60893	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vqtbx1q_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 32));
59082	60894	return __ret;
59083	60895	}
59084	60896	#else
59085	60897	__ai __attribute__((target("neon"))) int8x16_t vqtbx1q_s8(int8x16_t __p0, int8x16_t __p1, uint8x16_t __p2) {
59086	60898	int8x16_t __ret;
59087		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59088		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59089		uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59090		__ret = (int8x16_t) __builtin_neon_vqtbx1q_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 32);
59091		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	60899	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	60900	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	60901	uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	60902	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vqtbx1q_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 32));
	60903	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
	60904	return __ret;
	60905	}
	60906	#endif
	60907
	60908	#ifdef __LITTLE_ENDIAN__
	60909	__ai __attribute__((target("neon"))) mfloat8x16_t vqtbx1q_mf8(mfloat8x16_t __p0, mfloat8x16_t __p1, uint8x16_t __p2) {
	60910	mfloat8x16_t __ret;
	60911	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vqtbx1q_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 44));
	60912	return __ret;
	60913	}
	60914	#else
	60915	__ai __attribute__((target("neon"))) mfloat8x16_t vqtbx1q_mf8(mfloat8x16_t __p0, mfloat8x16_t __p1, uint8x16_t __p2) {
	60916	mfloat8x16_t __ret;
	60917	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	60918	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	60919	uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	60920	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vqtbx1q_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 44));
	60921	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
59092	60922	return __ret;
59093	60923	}
59094	60924	#endif
...	...	@@ -59096,17 +60926,17 @@ __ai __attribute__((target("neon"))) int8x16_t vqtbx1q_s8(int8x16_t __p0, int8x1
59096	60926	#ifdef __LITTLE_ENDIAN__
59097	60927	__ai __attribute__((target("neon"))) uint8x8_t vqtbx1_u8(uint8x8_t __p0, uint8x16_t __p1, uint8x8_t __p2) {
59098	60928	uint8x8_t __ret;
59099		__ret = (uint8x8_t) __builtin_neon_vqtbx1_v((int8x8_t)__p0, (int8x16_t)__p1, (int8x8_t)__p2, 16);
	60929	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqtbx1_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 16));
59100	60930	return __ret;
59101	60931	}
59102	60932	#else
59103	60933	__ai __attribute__((target("neon"))) uint8x8_t vqtbx1_u8(uint8x8_t __p0, uint8x16_t __p1, uint8x8_t __p2) {
59104	60934	uint8x8_t __ret;
59105		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
59106		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59107		uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
59108		__ret = (uint8x8_t) __builtin_neon_vqtbx1_v((int8x8_t)__rev0, (int8x16_t)__rev1, (int8x8_t)__rev2, 16);
59109		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	60935	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	60936	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	60937	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	60938	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqtbx1_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 16));
	60939	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
59110	60940	return __ret;
59111	60941	}
59112	60942	#endif
...	...	@@ -59114,17 +60944,35 @@ __ai __attribute__((target("neon"))) uint8x8_t vqtbx1_u8(uint8x8_t __p0, uint8x1
59114	60944	#ifdef __LITTLE_ENDIAN__
59115	60945	__ai __attribute__((target("neon"))) int8x8_t vqtbx1_s8(int8x8_t __p0, int8x16_t __p1, uint8x8_t __p2) {
59116	60946	int8x8_t __ret;
59117		__ret = (int8x8_t) __builtin_neon_vqtbx1_v((int8x8_t)__p0, (int8x16_t)__p1, (int8x8_t)__p2, 0);
	60947	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqtbx1_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 0));
59118	60948	return __ret;
59119	60949	}
59120	60950	#else
59121	60951	__ai __attribute__((target("neon"))) int8x8_t vqtbx1_s8(int8x8_t __p0, int8x16_t __p1, uint8x8_t __p2) {
59122	60952	int8x8_t __ret;
59123		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
59124		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59125		uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
59126		__ret = (int8x8_t) __builtin_neon_vqtbx1_v((int8x8_t)__rev0, (int8x16_t)__rev1, (int8x8_t)__rev2, 0);
59127		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	60953	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	60954	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	60955	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	60956	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqtbx1_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 0));
	60957	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
	60958	return __ret;
	60959	}
	60960	#endif
	60961
	60962	#ifdef __LITTLE_ENDIAN__
	60963	__ai __attribute__((target("neon"))) mfloat8x8_t vqtbx1_mf8(mfloat8x8_t __p0, mfloat8x16_t __p1, uint8x8_t __p2) {
	60964	mfloat8x8_t __ret;
	60965	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vqtbx1_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x8_t, __p2), 12));
	60966	return __ret;
	60967	}
	60968	#else
	60969	__ai __attribute__((target("neon"))) mfloat8x8_t vqtbx1_mf8(mfloat8x8_t __p0, mfloat8x16_t __p1, uint8x8_t __p2) {
	60970	mfloat8x8_t __ret;
	60971	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	60972	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	60973	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	60974	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vqtbx1_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x8_t, __rev2), 12));
	60975	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
59128	60976	return __ret;
59129	60977	}
59130	60978	#endif
...	...	@@ -59132,19 +60980,19 @@ __ai __attribute__((target("neon"))) int8x8_t vqtbx1_s8(int8x8_t __p0, int8x16_t
59132	60980	#ifdef __LITTLE_ENDIAN__
59133	60981	__ai __attribute__((target("neon"))) poly8x8_t vqtbx2_p8(poly8x8_t __p0, poly8x16x2_t __p1, uint8x8_t __p2) {
59134	60982	poly8x8_t __ret;
59135		__ret = (poly8x8_t) __builtin_neon_vqtbx2_v((int8x8_t)__p0, (int8x16_t)__p1.val[0], (int8x16_t)__p1.val[1], (int8x8_t)__p2, 4);
	60983	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vqtbx2_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x16_t, __p1.val[0]), __builtin_bit_cast(int8x16_t, __p1.val[1]), __builtin_bit_cast(int8x8_t, __p2), 4));
59136	60984	return __ret;
59137	60985	}
59138	60986	#else
59139	60987	__ai __attribute__((target("neon"))) poly8x8_t vqtbx2_p8(poly8x8_t __p0, poly8x16x2_t __p1, uint8x8_t __p2) {
59140	60988	poly8x8_t __ret;
59141		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	60989	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
59142	60990	poly8x16x2_t __rev1;
59143		__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59144		__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59145		uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
59146		__ret = (poly8x8_t) __builtin_neon_vqtbx2_v((int8x8_t)__rev0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x8_t)__rev2, 4);
59147		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	60991	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_128_8);
	60992	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_128_8);
	60993	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	60994	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vqtbx2_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev2), 4));
	60995	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
59148	60996	return __ret;
59149	60997	}
59150	60998	#endif
...	...	@@ -59152,19 +61000,19 @@ __ai __attribute__((target("neon"))) poly8x8_t vqtbx2_p8(poly8x8_t __p0, poly8x1
59152	61000	#ifdef __LITTLE_ENDIAN__
59153	61001	__ai __attribute__((target("neon"))) poly8x16_t vqtbx2q_p8(poly8x16_t __p0, poly8x16x2_t __p1, uint8x16_t __p2) {
59154	61002	poly8x16_t __ret;
59155		__ret = (poly8x16_t) __builtin_neon_vqtbx2q_v((int8x16_t)__p0, (int8x16_t)__p1.val[0], (int8x16_t)__p1.val[1], (int8x16_t)__p2, 36);
	61003	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vqtbx2q_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1.val[0]), __builtin_bit_cast(int8x16_t, __p1.val[1]), __builtin_bit_cast(int8x16_t, __p2), 36));
59156	61004	return __ret;
59157	61005	}
59158	61006	#else
59159	61007	__ai __attribute__((target("neon"))) poly8x16_t vqtbx2q_p8(poly8x16_t __p0, poly8x16x2_t __p1, uint8x16_t __p2) {
59160	61008	poly8x16_t __ret;
59161		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	61009	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
59162	61010	poly8x16x2_t __rev1;
59163		__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59164		__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59165		uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59166		__ret = (poly8x16_t) __builtin_neon_vqtbx2q_v((int8x16_t)__rev0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev2, 36);
59167		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	61011	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_128_8);
	61012	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_128_8);
	61013	uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	61014	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vqtbx2q_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev2), 36));
	61015	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
59168	61016	return __ret;
59169	61017	}
59170	61018	#endif
...	...	@@ -59172,19 +61020,19 @@ __ai __attribute__((target("neon"))) poly8x16_t vqtbx2q_p8(poly8x16_t __p0, poly
59172	61020	#ifdef __LITTLE_ENDIAN__
59173	61021	__ai __attribute__((target("neon"))) uint8x16_t vqtbx2q_u8(uint8x16_t __p0, uint8x16x2_t __p1, uint8x16_t __p2) {
59174	61022	uint8x16_t __ret;
59175		__ret = (uint8x16_t) __builtin_neon_vqtbx2q_v((int8x16_t)__p0, (int8x16_t)__p1.val[0], (int8x16_t)__p1.val[1], (int8x16_t)__p2, 48);
	61023	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vqtbx2q_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1.val[0]), __builtin_bit_cast(int8x16_t, __p1.val[1]), __builtin_bit_cast(int8x16_t, __p2), 48));
59176	61024	return __ret;
59177	61025	}
59178	61026	#else
59179	61027	__ai __attribute__((target("neon"))) uint8x16_t vqtbx2q_u8(uint8x16_t __p0, uint8x16x2_t __p1, uint8x16_t __p2) {
59180	61028	uint8x16_t __ret;
59181		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	61029	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
59182	61030	uint8x16x2_t __rev1;
59183		__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59184		__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59185		uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59186		__ret = (uint8x16_t) __builtin_neon_vqtbx2q_v((int8x16_t)__rev0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev2, 48);
59187		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	61031	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_128_8);
	61032	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_128_8);
	61033	uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	61034	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vqtbx2q_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev2), 48));
	61035	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
59188	61036	return __ret;
59189	61037	}
59190	61038	#endif
...	...	@@ -59192,19 +61040,39 @@ __ai __attribute__((target("neon"))) uint8x16_t vqtbx2q_u8(uint8x16_t __p0, uint
59192	61040	#ifdef __LITTLE_ENDIAN__
59193	61041	__ai __attribute__((target("neon"))) int8x16_t vqtbx2q_s8(int8x16_t __p0, int8x16x2_t __p1, uint8x16_t __p2) {
59194	61042	int8x16_t __ret;
59195		__ret = (int8x16_t) __builtin_neon_vqtbx2q_v((int8x16_t)__p0, (int8x16_t)__p1.val[0], (int8x16_t)__p1.val[1], (int8x16_t)__p2, 32);
	61043	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vqtbx2q_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1.val[0]), __builtin_bit_cast(int8x16_t, __p1.val[1]), __builtin_bit_cast(int8x16_t, __p2), 32));
59196	61044	return __ret;
59197	61045	}
59198	61046	#else
59199	61047	__ai __attribute__((target("neon"))) int8x16_t vqtbx2q_s8(int8x16_t __p0, int8x16x2_t __p1, uint8x16_t __p2) {
59200	61048	int8x16_t __ret;
59201		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	61049	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
59202	61050	int8x16x2_t __rev1;
59203		__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59204		__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59205		uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59206		__ret = (int8x16_t) __builtin_neon_vqtbx2q_v((int8x16_t)__rev0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev2, 32);
59207		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	61051	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_128_8);
	61052	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_128_8);
	61053	uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	61054	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vqtbx2q_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev2), 32));
	61055	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
	61056	return __ret;
	61057	}
	61058	#endif
	61059
	61060	#ifdef __LITTLE_ENDIAN__
	61061	__ai __attribute__((target("neon"))) mfloat8x16_t vqtbx2q_mf8(mfloat8x16_t __p0, mfloat8x16x2_t __p1, uint8x16_t __p2) {
	61062	mfloat8x16_t __ret;
	61063	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vqtbx2q_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1.val[0]), __builtin_bit_cast(int8x16_t, __p1.val[1]), __builtin_bit_cast(int8x16_t, __p2), 44));
	61064	return __ret;
	61065	}
	61066	#else
	61067	__ai __attribute__((target("neon"))) mfloat8x16_t vqtbx2q_mf8(mfloat8x16_t __p0, mfloat8x16x2_t __p1, uint8x16_t __p2) {
	61068	mfloat8x16_t __ret;
	61069	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	61070	mfloat8x16x2_t __rev1;
	61071	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_128_8);
	61072	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_128_8);
	61073	uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	61074	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vqtbx2q_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev2), 44));
	61075	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
59208	61076	return __ret;
59209	61077	}
59210	61078	#endif
...	...	@@ -59212,19 +61080,19 @@ __ai __attribute__((target("neon"))) int8x16_t vqtbx2q_s8(int8x16_t __p0, int8x1
59212	61080	#ifdef __LITTLE_ENDIAN__
59213	61081	__ai __attribute__((target("neon"))) uint8x8_t vqtbx2_u8(uint8x8_t __p0, uint8x16x2_t __p1, uint8x8_t __p2) {
59214	61082	uint8x8_t __ret;
59215		__ret = (uint8x8_t) __builtin_neon_vqtbx2_v((int8x8_t)__p0, (int8x16_t)__p1.val[0], (int8x16_t)__p1.val[1], (int8x8_t)__p2, 16);
	61083	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqtbx2_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x16_t, __p1.val[0]), __builtin_bit_cast(int8x16_t, __p1.val[1]), __builtin_bit_cast(int8x8_t, __p2), 16));
59216	61084	return __ret;
59217	61085	}
59218	61086	#else
59219	61087	__ai __attribute__((target("neon"))) uint8x8_t vqtbx2_u8(uint8x8_t __p0, uint8x16x2_t __p1, uint8x8_t __p2) {
59220	61088	uint8x8_t __ret;
59221		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	61089	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
59222	61090	uint8x16x2_t __rev1;
59223		__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59224		__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59225		uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
59226		__ret = (uint8x8_t) __builtin_neon_vqtbx2_v((int8x8_t)__rev0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x8_t)__rev2, 16);
59227		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	61091	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_128_8);
	61092	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_128_8);
	61093	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	61094	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqtbx2_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev2), 16));
	61095	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
59228	61096	return __ret;
59229	61097	}
59230	61098	#endif
...	...	@@ -59232,19 +61100,39 @@ __ai __attribute__((target("neon"))) uint8x8_t vqtbx2_u8(uint8x8_t __p0, uint8x1
59232	61100	#ifdef __LITTLE_ENDIAN__
59233	61101	__ai __attribute__((target("neon"))) int8x8_t vqtbx2_s8(int8x8_t __p0, int8x16x2_t __p1, uint8x8_t __p2) {
59234	61102	int8x8_t __ret;
59235		__ret = (int8x8_t) __builtin_neon_vqtbx2_v((int8x8_t)__p0, (int8x16_t)__p1.val[0], (int8x16_t)__p1.val[1], (int8x8_t)__p2, 0);
	61103	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqtbx2_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x16_t, __p1.val[0]), __builtin_bit_cast(int8x16_t, __p1.val[1]), __builtin_bit_cast(int8x8_t, __p2), 0));
59236	61104	return __ret;
59237	61105	}
59238	61106	#else
59239	61107	__ai __attribute__((target("neon"))) int8x8_t vqtbx2_s8(int8x8_t __p0, int8x16x2_t __p1, uint8x8_t __p2) {
59240	61108	int8x8_t __ret;
59241		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	61109	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
59242	61110	int8x16x2_t __rev1;
59243		__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59244		__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59245		uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
59246		__ret = (int8x8_t) __builtin_neon_vqtbx2_v((int8x8_t)__rev0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x8_t)__rev2, 0);
59247		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	61111	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_128_8);
	61112	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_128_8);
	61113	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	61114	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqtbx2_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev2), 0));
	61115	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
	61116	return __ret;
	61117	}
	61118	#endif
	61119
	61120	#ifdef __LITTLE_ENDIAN__
	61121	__ai __attribute__((target("neon"))) mfloat8x8_t vqtbx2_mf8(mfloat8x8_t __p0, mfloat8x16x2_t __p1, uint8x8_t __p2) {
	61122	mfloat8x8_t __ret;
	61123	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vqtbx2_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x16_t, __p1.val[0]), __builtin_bit_cast(int8x16_t, __p1.val[1]), __builtin_bit_cast(int8x8_t, __p2), 12));
	61124	return __ret;
	61125	}
	61126	#else
	61127	__ai __attribute__((target("neon"))) mfloat8x8_t vqtbx2_mf8(mfloat8x8_t __p0, mfloat8x16x2_t __p1, uint8x8_t __p2) {
	61128	mfloat8x8_t __ret;
	61129	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	61130	mfloat8x16x2_t __rev1;
	61131	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_128_8);
	61132	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_128_8);
	61133	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	61134	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vqtbx2_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev2), 12));
	61135	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
59248	61136	return __ret;
59249	61137	}
59250	61138	#endif
...	...	@@ -59252,20 +61140,20 @@ __ai __attribute__((target("neon"))) int8x8_t vqtbx2_s8(int8x8_t __p0, int8x16x2
59252	61140	#ifdef __LITTLE_ENDIAN__
59253	61141	__ai __attribute__((target("neon"))) poly8x8_t vqtbx3_p8(poly8x8_t __p0, poly8x16x3_t __p1, uint8x8_t __p2) {
59254	61142	poly8x8_t __ret;
59255		__ret = (poly8x8_t) __builtin_neon_vqtbx3_v((int8x8_t)__p0, (int8x16_t)__p1.val[0], (int8x16_t)__p1.val[1], (int8x16_t)__p1.val[2], (int8x8_t)__p2, 4);
	61143	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vqtbx3_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x16_t, __p1.val[0]), __builtin_bit_cast(int8x16_t, __p1.val[1]), __builtin_bit_cast(int8x16_t, __p1.val[2]), __builtin_bit_cast(int8x8_t, __p2), 4));
59256	61144	return __ret;
59257	61145	}
59258	61146	#else
59259	61147	__ai __attribute__((target("neon"))) poly8x8_t vqtbx3_p8(poly8x8_t __p0, poly8x16x3_t __p1, uint8x8_t __p2) {
59260	61148	poly8x8_t __ret;
59261		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	61149	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
59262	61150	poly8x16x3_t __rev1;
59263		__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59264		__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59265		__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59266		uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
59267		__ret = (poly8x8_t) __builtin_neon_vqtbx3_v((int8x8_t)__rev0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x8_t)__rev2, 4);
59268		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	61151	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_128_8);
	61152	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_128_8);
	61153	__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], __lane_reverse_128_8);
	61154	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	61155	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vqtbx3_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev2), 4));
	61156	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
59269	61157	return __ret;
59270	61158	}
59271	61159	#endif
...	...	@@ -59273,20 +61161,20 @@ __ai __attribute__((target("neon"))) poly8x8_t vqtbx3_p8(poly8x8_t __p0, poly8x1
59273	61161	#ifdef __LITTLE_ENDIAN__
59274	61162	__ai __attribute__((target("neon"))) poly8x16_t vqtbx3q_p8(poly8x16_t __p0, poly8x16x3_t __p1, uint8x16_t __p2) {
59275	61163	poly8x16_t __ret;
59276		__ret = (poly8x16_t) __builtin_neon_vqtbx3q_v((int8x16_t)__p0, (int8x16_t)__p1.val[0], (int8x16_t)__p1.val[1], (int8x16_t)__p1.val[2], (int8x16_t)__p2, 36);
	61164	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vqtbx3q_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1.val[0]), __builtin_bit_cast(int8x16_t, __p1.val[1]), __builtin_bit_cast(int8x16_t, __p1.val[2]), __builtin_bit_cast(int8x16_t, __p2), 36));
59277	61165	return __ret;
59278	61166	}
59279	61167	#else
59280	61168	__ai __attribute__((target("neon"))) poly8x16_t vqtbx3q_p8(poly8x16_t __p0, poly8x16x3_t __p1, uint8x16_t __p2) {
59281	61169	poly8x16_t __ret;
59282		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	61170	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
59283	61171	poly8x16x3_t __rev1;
59284		__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59285		__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59286		__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59287		uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59288		__ret = (poly8x16_t) __builtin_neon_vqtbx3q_v((int8x16_t)__rev0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev2, 36);
59289		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	61172	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_128_8);
	61173	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_128_8);
	61174	__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], __lane_reverse_128_8);
	61175	uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	61176	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vqtbx3q_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev2), 36));
	61177	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
59290	61178	return __ret;
59291	61179	}
59292	61180	#endif
...	...	@@ -59294,20 +61182,20 @@ __ai __attribute__((target("neon"))) poly8x16_t vqtbx3q_p8(poly8x16_t __p0, poly
59294	61182	#ifdef __LITTLE_ENDIAN__
59295	61183	__ai __attribute__((target("neon"))) uint8x16_t vqtbx3q_u8(uint8x16_t __p0, uint8x16x3_t __p1, uint8x16_t __p2) {
59296	61184	uint8x16_t __ret;
59297		__ret = (uint8x16_t) __builtin_neon_vqtbx3q_v((int8x16_t)__p0, (int8x16_t)__p1.val[0], (int8x16_t)__p1.val[1], (int8x16_t)__p1.val[2], (int8x16_t)__p2, 48);
	61185	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vqtbx3q_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1.val[0]), __builtin_bit_cast(int8x16_t, __p1.val[1]), __builtin_bit_cast(int8x16_t, __p1.val[2]), __builtin_bit_cast(int8x16_t, __p2), 48));
59298	61186	return __ret;
59299	61187	}
59300	61188	#else
59301	61189	__ai __attribute__((target("neon"))) uint8x16_t vqtbx3q_u8(uint8x16_t __p0, uint8x16x3_t __p1, uint8x16_t __p2) {
59302	61190	uint8x16_t __ret;
59303		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	61191	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
59304	61192	uint8x16x3_t __rev1;
59305		__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59306		__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59307		__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59308		uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59309		__ret = (uint8x16_t) __builtin_neon_vqtbx3q_v((int8x16_t)__rev0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev2, 48);
59310		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	61193	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_128_8);
	61194	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_128_8);
	61195	__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], __lane_reverse_128_8);
	61196	uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	61197	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vqtbx3q_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev2), 48));
	61198	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
59311	61199	return __ret;
59312	61200	}
59313	61201	#endif
...	...	@@ -59315,20 +61203,41 @@ __ai __attribute__((target("neon"))) uint8x16_t vqtbx3q_u8(uint8x16_t __p0, uint
59315	61203	#ifdef __LITTLE_ENDIAN__
59316	61204	__ai __attribute__((target("neon"))) int8x16_t vqtbx3q_s8(int8x16_t __p0, int8x16x3_t __p1, uint8x16_t __p2) {
59317	61205	int8x16_t __ret;
59318		__ret = (int8x16_t) __builtin_neon_vqtbx3q_v((int8x16_t)__p0, (int8x16_t)__p1.val[0], (int8x16_t)__p1.val[1], (int8x16_t)__p1.val[2], (int8x16_t)__p2, 32);
	61206	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vqtbx3q_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1.val[0]), __builtin_bit_cast(int8x16_t, __p1.val[1]), __builtin_bit_cast(int8x16_t, __p1.val[2]), __builtin_bit_cast(int8x16_t, __p2), 32));
59319	61207	return __ret;
59320	61208	}
59321	61209	#else
59322	61210	__ai __attribute__((target("neon"))) int8x16_t vqtbx3q_s8(int8x16_t __p0, int8x16x3_t __p1, uint8x16_t __p2) {
59323	61211	int8x16_t __ret;
59324		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	61212	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
59325	61213	int8x16x3_t __rev1;
59326		__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59327		__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59328		__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59329		uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59330		__ret = (int8x16_t) __builtin_neon_vqtbx3q_v((int8x16_t)__rev0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev2, 32);
59331		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	61214	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_128_8);
	61215	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_128_8);
	61216	__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], __lane_reverse_128_8);
	61217	uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	61218	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vqtbx3q_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev2), 32));
	61219	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
	61220	return __ret;
	61221	}
	61222	#endif
	61223
	61224	#ifdef __LITTLE_ENDIAN__
	61225	__ai __attribute__((target("neon"))) mfloat8x16_t vqtbx3q_mf8(mfloat8x16_t __p0, mfloat8x16x3_t __p1, uint8x16_t __p2) {
	61226	mfloat8x16_t __ret;
	61227	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vqtbx3q_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1.val[0]), __builtin_bit_cast(int8x16_t, __p1.val[1]), __builtin_bit_cast(int8x16_t, __p1.val[2]), __builtin_bit_cast(int8x16_t, __p2), 44));
	61228	return __ret;
	61229	}
	61230	#else
	61231	__ai __attribute__((target("neon"))) mfloat8x16_t vqtbx3q_mf8(mfloat8x16_t __p0, mfloat8x16x3_t __p1, uint8x16_t __p2) {
	61232	mfloat8x16_t __ret;
	61233	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	61234	mfloat8x16x3_t __rev1;
	61235	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_128_8);
	61236	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_128_8);
	61237	__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], __lane_reverse_128_8);
	61238	uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	61239	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vqtbx3q_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev2), 44));
	61240	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
59332	61241	return __ret;
59333	61242	}
59334	61243	#endif
...	...	@@ -59336,20 +61245,20 @@ __ai __attribute__((target("neon"))) int8x16_t vqtbx3q_s8(int8x16_t __p0, int8x1
59336	61245	#ifdef __LITTLE_ENDIAN__
59337	61246	__ai __attribute__((target("neon"))) uint8x8_t vqtbx3_u8(uint8x8_t __p0, uint8x16x3_t __p1, uint8x8_t __p2) {
59338	61247	uint8x8_t __ret;
59339		__ret = (uint8x8_t) __builtin_neon_vqtbx3_v((int8x8_t)__p0, (int8x16_t)__p1.val[0], (int8x16_t)__p1.val[1], (int8x16_t)__p1.val[2], (int8x8_t)__p2, 16);
	61248	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqtbx3_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x16_t, __p1.val[0]), __builtin_bit_cast(int8x16_t, __p1.val[1]), __builtin_bit_cast(int8x16_t, __p1.val[2]), __builtin_bit_cast(int8x8_t, __p2), 16));
59340	61249	return __ret;
59341	61250	}
59342	61251	#else
59343	61252	__ai __attribute__((target("neon"))) uint8x8_t vqtbx3_u8(uint8x8_t __p0, uint8x16x3_t __p1, uint8x8_t __p2) {
59344	61253	uint8x8_t __ret;
59345		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	61254	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
59346	61255	uint8x16x3_t __rev1;
59347		__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59348		__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59349		__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59350		uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
59351		__ret = (uint8x8_t) __builtin_neon_vqtbx3_v((int8x8_t)__rev0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x8_t)__rev2, 16);
59352		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	61256	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_128_8);
	61257	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_128_8);
	61258	__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], __lane_reverse_128_8);
	61259	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	61260	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqtbx3_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev2), 16));
	61261	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
59353	61262	return __ret;
59354	61263	}
59355	61264	#endif
...	...	@@ -59357,20 +61266,41 @@ __ai __attribute__((target("neon"))) uint8x8_t vqtbx3_u8(uint8x8_t __p0, uint8x1
59357	61266	#ifdef __LITTLE_ENDIAN__
59358	61267	__ai __attribute__((target("neon"))) int8x8_t vqtbx3_s8(int8x8_t __p0, int8x16x3_t __p1, uint8x8_t __p2) {
59359	61268	int8x8_t __ret;
59360		__ret = (int8x8_t) __builtin_neon_vqtbx3_v((int8x8_t)__p0, (int8x16_t)__p1.val[0], (int8x16_t)__p1.val[1], (int8x16_t)__p1.val[2], (int8x8_t)__p2, 0);
	61269	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqtbx3_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x16_t, __p1.val[0]), __builtin_bit_cast(int8x16_t, __p1.val[1]), __builtin_bit_cast(int8x16_t, __p1.val[2]), __builtin_bit_cast(int8x8_t, __p2), 0));
59361	61270	return __ret;
59362	61271	}
59363	61272	#else
59364	61273	__ai __attribute__((target("neon"))) int8x8_t vqtbx3_s8(int8x8_t __p0, int8x16x3_t __p1, uint8x8_t __p2) {
59365	61274	int8x8_t __ret;
59366		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	61275	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
59367	61276	int8x16x3_t __rev1;
59368		__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59369		__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59370		__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59371		uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
59372		__ret = (int8x8_t) __builtin_neon_vqtbx3_v((int8x8_t)__rev0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x8_t)__rev2, 0);
59373		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	61277	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_128_8);
	61278	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_128_8);
	61279	__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], __lane_reverse_128_8);
	61280	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	61281	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqtbx3_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev2), 0));
	61282	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
	61283	return __ret;
	61284	}
	61285	#endif
	61286
	61287	#ifdef __LITTLE_ENDIAN__
	61288	__ai __attribute__((target("neon"))) mfloat8x8_t vqtbx3_mf8(mfloat8x8_t __p0, mfloat8x16x3_t __p1, uint8x8_t __p2) {
	61289	mfloat8x8_t __ret;
	61290	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vqtbx3_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x16_t, __p1.val[0]), __builtin_bit_cast(int8x16_t, __p1.val[1]), __builtin_bit_cast(int8x16_t, __p1.val[2]), __builtin_bit_cast(int8x8_t, __p2), 12));
	61291	return __ret;
	61292	}
	61293	#else
	61294	__ai __attribute__((target("neon"))) mfloat8x8_t vqtbx3_mf8(mfloat8x8_t __p0, mfloat8x16x3_t __p1, uint8x8_t __p2) {
	61295	mfloat8x8_t __ret;
	61296	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	61297	mfloat8x16x3_t __rev1;
	61298	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_128_8);
	61299	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_128_8);
	61300	__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], __lane_reverse_128_8);
	61301	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	61302	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vqtbx3_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev2), 12));
	61303	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
59374	61304	return __ret;
59375	61305	}
59376	61306	#endif
...	...	@@ -59378,21 +61308,21 @@ __ai __attribute__((target("neon"))) int8x8_t vqtbx3_s8(int8x8_t __p0, int8x16x3
59378	61308	#ifdef __LITTLE_ENDIAN__
59379	61309	__ai __attribute__((target("neon"))) poly8x8_t vqtbx4_p8(poly8x8_t __p0, poly8x16x4_t __p1, uint8x8_t __p2) {
59380	61310	poly8x8_t __ret;
59381		__ret = (poly8x8_t) __builtin_neon_vqtbx4_v((int8x8_t)__p0, (int8x16_t)__p1.val[0], (int8x16_t)__p1.val[1], (int8x16_t)__p1.val[2], (int8x16_t)__p1.val[3], (int8x8_t)__p2, 4);
	61311	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vqtbx4_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x16_t, __p1.val[0]), __builtin_bit_cast(int8x16_t, __p1.val[1]), __builtin_bit_cast(int8x16_t, __p1.val[2]), __builtin_bit_cast(int8x16_t, __p1.val[3]), __builtin_bit_cast(int8x8_t, __p2), 4));
59382	61312	return __ret;
59383	61313	}
59384	61314	#else
59385	61315	__ai __attribute__((target("neon"))) poly8x8_t vqtbx4_p8(poly8x8_t __p0, poly8x16x4_t __p1, uint8x8_t __p2) {
59386	61316	poly8x8_t __ret;
59387		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	61317	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
59388	61318	poly8x16x4_t __rev1;
59389		__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59390		__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59391		__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59392		__rev1.val[3] = __builtin_shufflevector(__p1.val[3], __p1.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59393		uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
59394		__ret = (poly8x8_t) __builtin_neon_vqtbx4_v((int8x8_t)__rev0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], (int8x8_t)__rev2, 4);
59395		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	61319	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_128_8);
	61320	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_128_8);
	61321	__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], __lane_reverse_128_8);
	61322	__rev1.val[3] = __builtin_shufflevector(__p1.val[3], __p1.val[3], __lane_reverse_128_8);
	61323	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	61324	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vqtbx4_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __builtin_bit_cast(int8x8_t, __rev2), 4));
	61325	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
59396	61326	return __ret;
59397	61327	}
59398	61328	#endif
...	...	@@ -59400,21 +61330,21 @@ __ai __attribute__((target("neon"))) poly8x8_t vqtbx4_p8(poly8x8_t __p0, poly8x1
59400	61330	#ifdef __LITTLE_ENDIAN__
59401	61331	__ai __attribute__((target("neon"))) poly8x16_t vqtbx4q_p8(poly8x16_t __p0, poly8x16x4_t __p1, uint8x16_t __p2) {
59402	61332	poly8x16_t __ret;
59403		__ret = (poly8x16_t) __builtin_neon_vqtbx4q_v((int8x16_t)__p0, (int8x16_t)__p1.val[0], (int8x16_t)__p1.val[1], (int8x16_t)__p1.val[2], (int8x16_t)__p1.val[3], (int8x16_t)__p2, 36);
	61333	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vqtbx4q_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1.val[0]), __builtin_bit_cast(int8x16_t, __p1.val[1]), __builtin_bit_cast(int8x16_t, __p1.val[2]), __builtin_bit_cast(int8x16_t, __p1.val[3]), __builtin_bit_cast(int8x16_t, __p2), 36));
59404	61334	return __ret;
59405	61335	}
59406	61336	#else
59407	61337	__ai __attribute__((target("neon"))) poly8x16_t vqtbx4q_p8(poly8x16_t __p0, poly8x16x4_t __p1, uint8x16_t __p2) {
59408	61338	poly8x16_t __ret;
59409		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	61339	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
59410	61340	poly8x16x4_t __rev1;
59411		__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59412		__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59413		__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59414		__rev1.val[3] = __builtin_shufflevector(__p1.val[3], __p1.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59415		uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59416		__ret = (poly8x16_t) __builtin_neon_vqtbx4q_v((int8x16_t)__rev0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], (int8x16_t)__rev2, 36);
59417		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	61341	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_128_8);
	61342	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_128_8);
	61343	__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], __lane_reverse_128_8);
	61344	__rev1.val[3] = __builtin_shufflevector(__p1.val[3], __p1.val[3], __lane_reverse_128_8);
	61345	uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	61346	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vqtbx4q_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __builtin_bit_cast(int8x16_t, __rev2), 36));
	61347	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
59418	61348	return __ret;
59419	61349	}
59420	61350	#endif
...	...	@@ -59422,21 +61352,21 @@ __ai __attribute__((target("neon"))) poly8x16_t vqtbx4q_p8(poly8x16_t __p0, poly
59422	61352	#ifdef __LITTLE_ENDIAN__
59423	61353	__ai __attribute__((target("neon"))) uint8x16_t vqtbx4q_u8(uint8x16_t __p0, uint8x16x4_t __p1, uint8x16_t __p2) {
59424	61354	uint8x16_t __ret;
59425		__ret = (uint8x16_t) __builtin_neon_vqtbx4q_v((int8x16_t)__p0, (int8x16_t)__p1.val[0], (int8x16_t)__p1.val[1], (int8x16_t)__p1.val[2], (int8x16_t)__p1.val[3], (int8x16_t)__p2, 48);
	61355	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vqtbx4q_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1.val[0]), __builtin_bit_cast(int8x16_t, __p1.val[1]), __builtin_bit_cast(int8x16_t, __p1.val[2]), __builtin_bit_cast(int8x16_t, __p1.val[3]), __builtin_bit_cast(int8x16_t, __p2), 48));
59426	61356	return __ret;
59427	61357	}
59428	61358	#else
59429	61359	__ai __attribute__((target("neon"))) uint8x16_t vqtbx4q_u8(uint8x16_t __p0, uint8x16x4_t __p1, uint8x16_t __p2) {
59430	61360	uint8x16_t __ret;
59431		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	61361	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
59432	61362	uint8x16x4_t __rev1;
59433		__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59434		__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59435		__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59436		__rev1.val[3] = __builtin_shufflevector(__p1.val[3], __p1.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59437		uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59438		__ret = (uint8x16_t) __builtin_neon_vqtbx4q_v((int8x16_t)__rev0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], (int8x16_t)__rev2, 48);
59439		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	61363	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_128_8);
	61364	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_128_8);
	61365	__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], __lane_reverse_128_8);
	61366	__rev1.val[3] = __builtin_shufflevector(__p1.val[3], __p1.val[3], __lane_reverse_128_8);
	61367	uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	61368	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vqtbx4q_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __builtin_bit_cast(int8x16_t, __rev2), 48));
	61369	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
59440	61370	return __ret;
59441	61371	}
59442	61372	#endif
...	...	@@ -59444,21 +61374,43 @@ __ai __attribute__((target("neon"))) uint8x16_t vqtbx4q_u8(uint8x16_t __p0, uint
59444	61374	#ifdef __LITTLE_ENDIAN__
59445	61375	__ai __attribute__((target("neon"))) int8x16_t vqtbx4q_s8(int8x16_t __p0, int8x16x4_t __p1, uint8x16_t __p2) {
59446	61376	int8x16_t __ret;
59447		__ret = (int8x16_t) __builtin_neon_vqtbx4q_v((int8x16_t)__p0, (int8x16_t)__p1.val[0], (int8x16_t)__p1.val[1], (int8x16_t)__p1.val[2], (int8x16_t)__p1.val[3], (int8x16_t)__p2, 32);
	61377	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vqtbx4q_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1.val[0]), __builtin_bit_cast(int8x16_t, __p1.val[1]), __builtin_bit_cast(int8x16_t, __p1.val[2]), __builtin_bit_cast(int8x16_t, __p1.val[3]), __builtin_bit_cast(int8x16_t, __p2), 32));
59448	61378	return __ret;
59449	61379	}
59450	61380	#else
59451	61381	__ai __attribute__((target("neon"))) int8x16_t vqtbx4q_s8(int8x16_t __p0, int8x16x4_t __p1, uint8x16_t __p2) {
59452	61382	int8x16_t __ret;
59453		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	61383	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
59454	61384	int8x16x4_t __rev1;
59455		__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59456		__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59457		__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59458		__rev1.val[3] = __builtin_shufflevector(__p1.val[3], __p1.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59459		uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59460		__ret = (int8x16_t) __builtin_neon_vqtbx4q_v((int8x16_t)__rev0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], (int8x16_t)__rev2, 32);
59461		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	61385	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_128_8);
	61386	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_128_8);
	61387	__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], __lane_reverse_128_8);
	61388	__rev1.val[3] = __builtin_shufflevector(__p1.val[3], __p1.val[3], __lane_reverse_128_8);
	61389	uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	61390	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vqtbx4q_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __builtin_bit_cast(int8x16_t, __rev2), 32));
	61391	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
	61392	return __ret;
	61393	}
	61394	#endif
	61395
	61396	#ifdef __LITTLE_ENDIAN__
	61397	__ai __attribute__((target("neon"))) mfloat8x16_t vqtbx4q_mf8(mfloat8x16_t __p0, mfloat8x16x4_t __p1, uint8x16_t __p2) {
	61398	mfloat8x16_t __ret;
	61399	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vqtbx4q_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1.val[0]), __builtin_bit_cast(int8x16_t, __p1.val[1]), __builtin_bit_cast(int8x16_t, __p1.val[2]), __builtin_bit_cast(int8x16_t, __p1.val[3]), __builtin_bit_cast(int8x16_t, __p2), 44));
	61400	return __ret;
	61401	}
	61402	#else
	61403	__ai __attribute__((target("neon"))) mfloat8x16_t vqtbx4q_mf8(mfloat8x16_t __p0, mfloat8x16x4_t __p1, uint8x16_t __p2) {
	61404	mfloat8x16_t __ret;
	61405	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	61406	mfloat8x16x4_t __rev1;
	61407	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_128_8);
	61408	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_128_8);
	61409	__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], __lane_reverse_128_8);
	61410	__rev1.val[3] = __builtin_shufflevector(__p1.val[3], __p1.val[3], __lane_reverse_128_8);
	61411	uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	61412	__ret = __builtin_bit_cast(mfloat8x16_t, __builtin_neon_vqtbx4q_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __builtin_bit_cast(int8x16_t, __rev2), 44));
	61413	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
59462	61414	return __ret;
59463	61415	}
59464	61416	#endif
...	...	@@ -59466,21 +61418,21 @@ __ai __attribute__((target("neon"))) int8x16_t vqtbx4q_s8(int8x16_t __p0, int8x1
59466	61418	#ifdef __LITTLE_ENDIAN__
59467	61419	__ai __attribute__((target("neon"))) uint8x8_t vqtbx4_u8(uint8x8_t __p0, uint8x16x4_t __p1, uint8x8_t __p2) {
59468	61420	uint8x8_t __ret;
59469		__ret = (uint8x8_t) __builtin_neon_vqtbx4_v((int8x8_t)__p0, (int8x16_t)__p1.val[0], (int8x16_t)__p1.val[1], (int8x16_t)__p1.val[2], (int8x16_t)__p1.val[3], (int8x8_t)__p2, 16);
	61421	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqtbx4_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x16_t, __p1.val[0]), __builtin_bit_cast(int8x16_t, __p1.val[1]), __builtin_bit_cast(int8x16_t, __p1.val[2]), __builtin_bit_cast(int8x16_t, __p1.val[3]), __builtin_bit_cast(int8x8_t, __p2), 16));
59470	61422	return __ret;
59471	61423	}
59472	61424	#else
59473	61425	__ai __attribute__((target("neon"))) uint8x8_t vqtbx4_u8(uint8x8_t __p0, uint8x16x4_t __p1, uint8x8_t __p2) {
59474	61426	uint8x8_t __ret;
59475		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	61427	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
59476	61428	uint8x16x4_t __rev1;
59477		__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59478		__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59479		__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59480		__rev1.val[3] = __builtin_shufflevector(__p1.val[3], __p1.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59481		uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
59482		__ret = (uint8x8_t) __builtin_neon_vqtbx4_v((int8x8_t)__rev0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], (int8x8_t)__rev2, 16);
59483		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	61429	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_128_8);
	61430	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_128_8);
	61431	__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], __lane_reverse_128_8);
	61432	__rev1.val[3] = __builtin_shufflevector(__p1.val[3], __p1.val[3], __lane_reverse_128_8);
	61433	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	61434	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vqtbx4_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __builtin_bit_cast(int8x8_t, __rev2), 16));
	61435	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
59484	61436	return __ret;
59485	61437	}
59486	61438	#endif
...	...	@@ -59488,21 +61440,43 @@ __ai __attribute__((target("neon"))) uint8x8_t vqtbx4_u8(uint8x8_t __p0, uint8x1
59488	61440	#ifdef __LITTLE_ENDIAN__
59489	61441	__ai __attribute__((target("neon"))) int8x8_t vqtbx4_s8(int8x8_t __p0, int8x16x4_t __p1, uint8x8_t __p2) {
59490	61442	int8x8_t __ret;
59491		__ret = (int8x8_t) __builtin_neon_vqtbx4_v((int8x8_t)__p0, (int8x16_t)__p1.val[0], (int8x16_t)__p1.val[1], (int8x16_t)__p1.val[2], (int8x16_t)__p1.val[3], (int8x8_t)__p2, 0);
	61443	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqtbx4_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x16_t, __p1.val[0]), __builtin_bit_cast(int8x16_t, __p1.val[1]), __builtin_bit_cast(int8x16_t, __p1.val[2]), __builtin_bit_cast(int8x16_t, __p1.val[3]), __builtin_bit_cast(int8x8_t, __p2), 0));
59492	61444	return __ret;
59493	61445	}
59494	61446	#else
59495	61447	__ai __attribute__((target("neon"))) int8x8_t vqtbx4_s8(int8x8_t __p0, int8x16x4_t __p1, uint8x8_t __p2) {
59496	61448	int8x8_t __ret;
59497		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	61449	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
59498	61450	int8x16x4_t __rev1;
59499		__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59500		__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59501		__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59502		__rev1.val[3] = __builtin_shufflevector(__p1.val[3], __p1.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59503		uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
59504		__ret = (int8x8_t) __builtin_neon_vqtbx4_v((int8x8_t)__rev0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], (int8x8_t)__rev2, 0);
59505		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	61451	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_128_8);
	61452	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_128_8);
	61453	__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], __lane_reverse_128_8);
	61454	__rev1.val[3] = __builtin_shufflevector(__p1.val[3], __p1.val[3], __lane_reverse_128_8);
	61455	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	61456	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vqtbx4_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __builtin_bit_cast(int8x8_t, __rev2), 0));
	61457	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
	61458	return __ret;
	61459	}
	61460	#endif
	61461
	61462	#ifdef __LITTLE_ENDIAN__
	61463	__ai __attribute__((target("neon"))) mfloat8x8_t vqtbx4_mf8(mfloat8x8_t __p0, mfloat8x16x4_t __p1, uint8x8_t __p2) {
	61464	mfloat8x8_t __ret;
	61465	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vqtbx4_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x16_t, __p1.val[0]), __builtin_bit_cast(int8x16_t, __p1.val[1]), __builtin_bit_cast(int8x16_t, __p1.val[2]), __builtin_bit_cast(int8x16_t, __p1.val[3]), __builtin_bit_cast(int8x8_t, __p2), 12));
	61466	return __ret;
	61467	}
	61468	#else
	61469	__ai __attribute__((target("neon"))) mfloat8x8_t vqtbx4_mf8(mfloat8x8_t __p0, mfloat8x16x4_t __p1, uint8x8_t __p2) {
	61470	mfloat8x8_t __ret;
	61471	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	61472	mfloat8x16x4_t __rev1;
	61473	__rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], __lane_reverse_128_8);
	61474	__rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], __lane_reverse_128_8);
	61475	__rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], __lane_reverse_128_8);
	61476	__rev1.val[3] = __builtin_shufflevector(__p1.val[3], __p1.val[3], __lane_reverse_128_8);
	61477	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
	61478	__ret = __builtin_bit_cast(mfloat8x8_t, __builtin_neon_vqtbx4_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __builtin_bit_cast(int8x8_t, __rev2), 12));
	61479	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
59506	61480	return __ret;
59507	61481	}
59508	61482	#endif
...	...	@@ -59516,11 +61490,11 @@ __ai __attribute__((target("neon"))) uint16x8_t vraddhn_high_u32(uint16x4_t __p0
59516	61490	#else
59517	61491	__ai __attribute__((target("neon"))) uint16x8_t vraddhn_high_u32(uint16x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
59518	61492	uint16x8_t __ret;
59519		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
59520		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
59521		uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	61493	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	61494	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	61495	uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
59522	61496	__ret = __noswap_vcombine_u16(__rev0, __noswap_vraddhn_u32(__rev1, __rev2));
59523		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	61497	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
59524	61498	return __ret;
59525	61499	}
59526	61500	#endif
...	...	@@ -59534,11 +61508,11 @@ __ai __attribute__((target("neon"))) uint32x4_t vraddhn_high_u64(uint32x2_t __p0
59534	61508	#else
59535	61509	__ai __attribute__((target("neon"))) uint32x4_t vraddhn_high_u64(uint32x2_t __p0, uint64x2_t __p1, uint64x2_t __p2) {
59536	61510	uint32x4_t __ret;
59537		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
59538		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
59539		uint64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
	61511	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	61512	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	61513	uint64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_64);
59540	61514	__ret = __noswap_vcombine_u32(__rev0, __noswap_vraddhn_u64(__rev1, __rev2));
59541		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	61515	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
59542	61516	return __ret;
59543	61517	}
59544	61518	#endif
...	...	@@ -59552,11 +61526,11 @@ __ai __attribute__((target("neon"))) uint8x16_t vraddhn_high_u16(uint8x8_t __p0,
59552	61526	#else
59553	61527	__ai __attribute__((target("neon"))) uint8x16_t vraddhn_high_u16(uint8x8_t __p0, uint16x8_t __p1, uint16x8_t __p2) {
59554	61528	uint8x16_t __ret;
59555		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
59556		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
59557		uint16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	61529	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	61530	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	61531	uint16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
59558	61532	__ret = __noswap_vcombine_u8(__rev0, __noswap_vraddhn_u16(__rev1, __rev2));
59559		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	61533	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
59560	61534	return __ret;
59561	61535	}
59562	61536	#endif
...	...	@@ -59570,11 +61544,11 @@ __ai __attribute__((target("neon"))) int16x8_t vraddhn_high_s32(int16x4_t __p0,
59570	61544	#else
59571	61545	__ai __attribute__((target("neon"))) int16x8_t vraddhn_high_s32(int16x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
59572	61546	int16x8_t __ret;
59573		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
59574		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
59575		int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	61547	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	61548	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	61549	int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
59576	61550	__ret = __noswap_vcombine_s16(__rev0, __noswap_vraddhn_s32(__rev1, __rev2));
59577		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	61551	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
59578	61552	return __ret;
59579	61553	}
59580	61554	#endif
...	...	@@ -59588,11 +61562,11 @@ __ai __attribute__((target("neon"))) int32x4_t vraddhn_high_s64(int32x2_t __p0,
59588	61562	#else
59589	61563	__ai __attribute__((target("neon"))) int32x4_t vraddhn_high_s64(int32x2_t __p0, int64x2_t __p1, int64x2_t __p2) {
59590	61564	int32x4_t __ret;
59591		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
59592		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
59593		int64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
	61565	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	61566	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	61567	int64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_64);
59594	61568	__ret = __noswap_vcombine_s32(__rev0, __noswap_vraddhn_s64(__rev1, __rev2));
59595		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	61569	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
59596	61570	return __ret;
59597	61571	}
59598	61572	#endif
...	...	@@ -59606,11 +61580,11 @@ __ai __attribute__((target("neon"))) int8x16_t vraddhn_high_s16(int8x8_t __p0, i
59606	61580	#else
59607	61581	__ai __attribute__((target("neon"))) int8x16_t vraddhn_high_s16(int8x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
59608	61582	int8x16_t __ret;
59609		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
59610		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
59611		int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	61583	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	61584	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	61585	int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
59612	61586	__ret = __noswap_vcombine_s8(__rev0, __noswap_vraddhn_s16(__rev1, __rev2));
59613		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	61587	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
59614	61588	return __ret;
59615	61589	}
59616	61590	#endif
...	...	@@ -59618,15 +61592,15 @@ __ai __attribute__((target("neon"))) int8x16_t vraddhn_high_s16(int8x8_t __p0, i
59618	61592	#ifdef __LITTLE_ENDIAN__
59619	61593	__ai __attribute__((target("neon"))) poly8x8_t vrbit_p8(poly8x8_t __p0) {
59620	61594	poly8x8_t __ret;
59621		__ret = (poly8x8_t) __builtin_neon_vrbit_v((int8x8_t)__p0, 4);
	61595	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vrbit_v(__builtin_bit_cast(int8x8_t, __p0), 4));
59622	61596	return __ret;
59623	61597	}
59624	61598	#else
59625	61599	__ai __attribute__((target("neon"))) poly8x8_t vrbit_p8(poly8x8_t __p0) {
59626	61600	poly8x8_t __ret;
59627		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
59628		__ret = (poly8x8_t) __builtin_neon_vrbit_v((int8x8_t)__rev0, 4);
59629		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	61601	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	61602	__ret = __builtin_bit_cast(poly8x8_t, __builtin_neon_vrbit_v(__builtin_bit_cast(int8x8_t, __rev0), 4));
	61603	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
59630	61604	return __ret;
59631	61605	}
59632	61606	#endif
...	...	@@ -59634,15 +61608,15 @@ __ai __attribute__((target("neon"))) poly8x8_t vrbit_p8(poly8x8_t __p0) {
59634	61608	#ifdef __LITTLE_ENDIAN__
59635	61609	__ai __attribute__((target("neon"))) poly8x16_t vrbitq_p8(poly8x16_t __p0) {
59636	61610	poly8x16_t __ret;
59637		__ret = (poly8x16_t) __builtin_neon_vrbitq_v((int8x16_t)__p0, 36);
	61611	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vrbitq_v(__builtin_bit_cast(int8x16_t, __p0), 36));
59638	61612	return __ret;
59639	61613	}
59640	61614	#else
59641	61615	__ai __attribute__((target("neon"))) poly8x16_t vrbitq_p8(poly8x16_t __p0) {
59642	61616	poly8x16_t __ret;
59643		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59644		__ret = (poly8x16_t) __builtin_neon_vrbitq_v((int8x16_t)__rev0, 36);
59645		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	61617	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	61618	__ret = __builtin_bit_cast(poly8x16_t, __builtin_neon_vrbitq_v(__builtin_bit_cast(int8x16_t, __rev0), 36));
	61619	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
59646	61620	return __ret;
59647	61621	}
59648	61622	#endif
...	...	@@ -59650,15 +61624,15 @@ __ai __attribute__((target("neon"))) poly8x16_t vrbitq_p8(poly8x16_t __p0) {
59650	61624	#ifdef __LITTLE_ENDIAN__
59651	61625	__ai __attribute__((target("neon"))) uint8x16_t vrbitq_u8(uint8x16_t __p0) {
59652	61626	uint8x16_t __ret;
59653		__ret = (uint8x16_t) __builtin_neon_vrbitq_v((int8x16_t)__p0, 48);
	61627	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vrbitq_v(__builtin_bit_cast(int8x16_t, __p0), 48));
59654	61628	return __ret;
59655	61629	}
59656	61630	#else
59657	61631	__ai __attribute__((target("neon"))) uint8x16_t vrbitq_u8(uint8x16_t __p0) {
59658	61632	uint8x16_t __ret;
59659		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59660		__ret = (uint8x16_t) __builtin_neon_vrbitq_v((int8x16_t)__rev0, 48);
59661		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	61633	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	61634	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vrbitq_v(__builtin_bit_cast(int8x16_t, __rev0), 48));
	61635	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
59662	61636	return __ret;
59663	61637	}
59664	61638	#endif
...	...	@@ -59666,15 +61640,15 @@ __ai __attribute__((target("neon"))) uint8x16_t vrbitq_u8(uint8x16_t __p0) {
59666	61640	#ifdef __LITTLE_ENDIAN__
59667	61641	__ai __attribute__((target("neon"))) int8x16_t vrbitq_s8(int8x16_t __p0) {
59668	61642	int8x16_t __ret;
59669		__ret = (int8x16_t) __builtin_neon_vrbitq_v((int8x16_t)__p0, 32);
	61643	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vrbitq_v(__builtin_bit_cast(int8x16_t, __p0), 32));
59670	61644	return __ret;
59671	61645	}
59672	61646	#else
59673	61647	__ai __attribute__((target("neon"))) int8x16_t vrbitq_s8(int8x16_t __p0) {
59674	61648	int8x16_t __ret;
59675		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
59676		__ret = (int8x16_t) __builtin_neon_vrbitq_v((int8x16_t)__rev0, 32);
59677		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	61649	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	61650	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vrbitq_v(__builtin_bit_cast(int8x16_t, __rev0), 32));
	61651	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
59678	61652	return __ret;
59679	61653	}
59680	61654	#endif
...	...	@@ -59682,15 +61656,15 @@ __ai __attribute__((target("neon"))) int8x16_t vrbitq_s8(int8x16_t __p0) {
59682	61656	#ifdef __LITTLE_ENDIAN__
59683	61657	__ai __attribute__((target("neon"))) uint8x8_t vrbit_u8(uint8x8_t __p0) {
59684	61658	uint8x8_t __ret;
59685		__ret = (uint8x8_t) __builtin_neon_vrbit_v((int8x8_t)__p0, 16);
	61659	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vrbit_v(__builtin_bit_cast(int8x8_t, __p0), 16));
59686	61660	return __ret;
59687	61661	}
59688	61662	#else
59689	61663	__ai __attribute__((target("neon"))) uint8x8_t vrbit_u8(uint8x8_t __p0) {
59690	61664	uint8x8_t __ret;
59691		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
59692		__ret = (uint8x8_t) __builtin_neon_vrbit_v((int8x8_t)__rev0, 16);
59693		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	61665	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	61666	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vrbit_v(__builtin_bit_cast(int8x8_t, __rev0), 16));
	61667	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
59694	61668	return __ret;
59695	61669	}
59696	61670	#endif
...	...	@@ -59698,15 +61672,15 @@ __ai __attribute__((target("neon"))) uint8x8_t vrbit_u8(uint8x8_t __p0) {
59698	61672	#ifdef __LITTLE_ENDIAN__
59699	61673	__ai __attribute__((target("neon"))) int8x8_t vrbit_s8(int8x8_t __p0) {
59700	61674	int8x8_t __ret;
59701		__ret = (int8x8_t) __builtin_neon_vrbit_v((int8x8_t)__p0, 0);
	61675	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vrbit_v(__builtin_bit_cast(int8x8_t, __p0), 0));
59702	61676	return __ret;
59703	61677	}
59704	61678	#else
59705	61679	__ai __attribute__((target("neon"))) int8x8_t vrbit_s8(int8x8_t __p0) {
59706	61680	int8x8_t __ret;
59707		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
59708		__ret = (int8x8_t) __builtin_neon_vrbit_v((int8x8_t)__rev0, 0);
59709		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	61681	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	61682	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vrbit_v(__builtin_bit_cast(int8x8_t, __rev0), 0));
	61683	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
59710	61684	return __ret;
59711	61685	}
59712	61686	#endif
...	...	@@ -59714,2549 +61688,2549 @@ __ai __attribute__((target("neon"))) int8x8_t vrbit_s8(int8x8_t __p0) {
59714	61688	#ifdef __LITTLE_ENDIAN__
59715	61689	__ai __attribute__((target("neon"))) float64x2_t vrecpeq_f64(float64x2_t __p0) {
59716	61690	float64x2_t __ret;
59717		__ret = (float64x2_t) __builtin_neon_vrecpeq_v((int8x16_t)__p0, 42);
	61691	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vrecpeq_v(__builtin_bit_cast(int8x16_t, __p0), 42));
59718	61692	return __ret;
59719	61693	}
59720	61694	#else
59721	61695	__ai __attribute__((target("neon"))) float64x2_t vrecpeq_f64(float64x2_t __p0) {
59722	61696	float64x2_t __ret;
59723		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
59724		__ret = (float64x2_t) __builtin_neon_vrecpeq_v((int8x16_t)__rev0, 42);
59725		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	61697	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	61698	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vrecpeq_v(__builtin_bit_cast(int8x16_t, __rev0), 42));
	61699	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
59726	61700	return __ret;
59727	61701	}
59728	61702	#endif
59729	61703
59730	61704	__ai __attribute__((target("neon"))) float64x1_t vrecpe_f64(float64x1_t __p0) {
59731	61705	float64x1_t __ret;
59732		__ret = (float64x1_t) __builtin_neon_vrecpe_v((int8x8_t)__p0, 10);
	61706	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vrecpe_v(__builtin_bit_cast(int8x8_t, __p0), 10));
59733	61707	return __ret;
59734	61708	}
59735	61709	__ai __attribute__((target("neon"))) float64_t vrecped_f64(float64_t __p0) {
59736	61710	float64_t __ret;
59737		__ret = (float64_t) __builtin_neon_vrecped_f64(__p0);
	61711	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vrecped_f64(__p0));
59738	61712	return __ret;
59739	61713	}
59740	61714	__ai __attribute__((target("neon"))) float32_t vrecpes_f32(float32_t __p0) {
59741	61715	float32_t __ret;
59742		__ret = (float32_t) __builtin_neon_vrecpes_f32(__p0);
	61716	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vrecpes_f32(__p0));
59743	61717	return __ret;
59744	61718	}
59745	61719	#ifdef __LITTLE_ENDIAN__
59746	61720	__ai __attribute__((target("neon"))) float64x2_t vrecpsq_f64(float64x2_t __p0, float64x2_t __p1) {
59747	61721	float64x2_t __ret;
59748		__ret = (float64x2_t) __builtin_neon_vrecpsq_v((int8x16_t)__p0, (int8x16_t)__p1, 42);
	61722	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vrecpsq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 42));
59749	61723	return __ret;
59750	61724	}
59751	61725	#else
59752	61726	__ai __attribute__((target("neon"))) float64x2_t vrecpsq_f64(float64x2_t __p0, float64x2_t __p1) {
59753	61727	float64x2_t __ret;
59754		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
59755		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
59756		__ret = (float64x2_t) __builtin_neon_vrecpsq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 42);
59757		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	61728	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	61729	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	61730	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vrecpsq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 42));
	61731	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
59758	61732	return __ret;
59759	61733	}
59760	61734	#endif
59761	61735
59762	61736	__ai __attribute__((target("neon"))) float64x1_t vrecps_f64(float64x1_t __p0, float64x1_t __p1) {
59763	61737	float64x1_t __ret;
59764		__ret = (float64x1_t) __builtin_neon_vrecps_v((int8x8_t)__p0, (int8x8_t)__p1, 10);
	61738	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vrecps_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 10));
59765	61739	return __ret;
59766	61740	}
59767	61741	__ai __attribute__((target("neon"))) float64_t vrecpsd_f64(float64_t __p0, float64_t __p1) {
59768	61742	float64_t __ret;
59769		__ret = (float64_t) __builtin_neon_vrecpsd_f64(__p0, __p1);
	61743	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vrecpsd_f64(__p0, __p1));
59770	61744	return __ret;
59771	61745	}
59772	61746	__ai __attribute__((target("neon"))) float32_t vrecpss_f32(float32_t __p0, float32_t __p1) {
59773	61747	float32_t __ret;
59774		__ret = (float32_t) __builtin_neon_vrecpss_f32(__p0, __p1);
	61748	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vrecpss_f32(__p0, __p1));
59775	61749	return __ret;
59776	61750	}
59777	61751	__ai __attribute__((target("neon"))) float64_t vrecpxd_f64(float64_t __p0) {
59778	61752	float64_t __ret;
59779		__ret = (float64_t) __builtin_neon_vrecpxd_f64(__p0);
	61753	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vrecpxd_f64(__p0));
59780	61754	return __ret;
59781	61755	}
59782	61756	__ai __attribute__((target("neon"))) float32_t vrecpxs_f32(float32_t __p0) {
59783	61757	float32_t __ret;
59784		__ret = (float32_t) __builtin_neon_vrecpxs_f32(__p0);
	61758	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vrecpxs_f32(__p0));
59785	61759	return __ret;
59786	61760	}
59787	61761	__ai __attribute__((target("neon"))) poly8x8_t vreinterpret_p8_p64(poly64x1_t __p0) {
59788	61762	poly8x8_t __ret;
59789		__ret = (poly8x8_t)(__p0);
	61763	__ret = __builtin_bit_cast(poly8x8_t, __p0);
59790	61764	return __ret;
59791	61765	}
59792	61766	__ai __attribute__((target("neon"))) poly8x8_t vreinterpret_p8_p16(poly16x4_t __p0) {
59793	61767	poly8x8_t __ret;
59794		__ret = (poly8x8_t)(__p0);
	61768	__ret = __builtin_bit_cast(poly8x8_t, __p0);
59795	61769	return __ret;
59796	61770	}
59797	61771	__ai __attribute__((target("neon"))) poly8x8_t vreinterpret_p8_u8(uint8x8_t __p0) {
59798	61772	poly8x8_t __ret;
59799		__ret = (poly8x8_t)(__p0);
	61773	__ret = __builtin_bit_cast(poly8x8_t, __p0);
59800	61774	return __ret;
59801	61775	}
59802	61776	__ai __attribute__((target("neon"))) poly8x8_t vreinterpret_p8_u32(uint32x2_t __p0) {
59803	61777	poly8x8_t __ret;
59804		__ret = (poly8x8_t)(__p0);
	61778	__ret = __builtin_bit_cast(poly8x8_t, __p0);
59805	61779	return __ret;
59806	61780	}
59807	61781	__ai __attribute__((target("neon"))) poly8x8_t vreinterpret_p8_u64(uint64x1_t __p0) {
59808	61782	poly8x8_t __ret;
59809		__ret = (poly8x8_t)(__p0);
	61783	__ret = __builtin_bit_cast(poly8x8_t, __p0);
59810	61784	return __ret;
59811	61785	}
59812	61786	__ai __attribute__((target("neon"))) poly8x8_t vreinterpret_p8_u16(uint16x4_t __p0) {
59813	61787	poly8x8_t __ret;
59814		__ret = (poly8x8_t)(__p0);
	61788	__ret = __builtin_bit_cast(poly8x8_t, __p0);
59815	61789	return __ret;
59816	61790	}
59817	61791	__ai __attribute__((target("neon"))) poly8x8_t vreinterpret_p8_s8(int8x8_t __p0) {
59818	61792	poly8x8_t __ret;
59819		__ret = (poly8x8_t)(__p0);
	61793	__ret = __builtin_bit_cast(poly8x8_t, __p0);
59820	61794	return __ret;
59821	61795	}
59822	61796	__ai __attribute__((target("neon"))) poly8x8_t vreinterpret_p8_f64(float64x1_t __p0) {
59823	61797	poly8x8_t __ret;
59824		__ret = (poly8x8_t)(__p0);
	61798	__ret = __builtin_bit_cast(poly8x8_t, __p0);
59825	61799	return __ret;
59826	61800	}
59827	61801	__ai __attribute__((target("neon"))) poly8x8_t vreinterpret_p8_f32(float32x2_t __p0) {
59828	61802	poly8x8_t __ret;
59829		__ret = (poly8x8_t)(__p0);
	61803	__ret = __builtin_bit_cast(poly8x8_t, __p0);
59830	61804	return __ret;
59831	61805	}
59832	61806	__ai __attribute__((target("neon"))) poly8x8_t vreinterpret_p8_f16(float16x4_t __p0) {
59833	61807	poly8x8_t __ret;
59834		__ret = (poly8x8_t)(__p0);
	61808	__ret = __builtin_bit_cast(poly8x8_t, __p0);
59835	61809	return __ret;
59836	61810	}
59837	61811	__ai __attribute__((target("neon"))) poly8x8_t vreinterpret_p8_s32(int32x2_t __p0) {
59838	61812	poly8x8_t __ret;
59839		__ret = (poly8x8_t)(__p0);
	61813	__ret = __builtin_bit_cast(poly8x8_t, __p0);
59840	61814	return __ret;
59841	61815	}
59842	61816	__ai __attribute__((target("neon"))) poly8x8_t vreinterpret_p8_s64(int64x1_t __p0) {
59843	61817	poly8x8_t __ret;
59844		__ret = (poly8x8_t)(__p0);
	61818	__ret = __builtin_bit_cast(poly8x8_t, __p0);
59845	61819	return __ret;
59846	61820	}
59847	61821	__ai __attribute__((target("neon"))) poly8x8_t vreinterpret_p8_mf8(mfloat8x8_t __p0) {
59848	61822	poly8x8_t __ret;
59849		__ret = (poly8x8_t)(__p0);
	61823	__ret = __builtin_bit_cast(poly8x8_t, __p0);
59850	61824	return __ret;
59851	61825	}
59852	61826	__ai __attribute__((target("neon"))) poly8x8_t vreinterpret_p8_s16(int16x4_t __p0) {
59853	61827	poly8x8_t __ret;
59854		__ret = (poly8x8_t)(__p0);
	61828	__ret = __builtin_bit_cast(poly8x8_t, __p0);
59855	61829	return __ret;
59856	61830	}
59857	61831	__ai __attribute__((target("neon"))) poly64x1_t vreinterpret_p64_p8(poly8x8_t __p0) {
59858	61832	poly64x1_t __ret;
59859		__ret = (poly64x1_t)(__p0);
	61833	__ret = __builtin_bit_cast(poly64x1_t, __p0);
59860	61834	return __ret;
59861	61835	}
59862	61836	__ai __attribute__((target("neon"))) poly64x1_t vreinterpret_p64_p16(poly16x4_t __p0) {
59863	61837	poly64x1_t __ret;
59864		__ret = (poly64x1_t)(__p0);
	61838	__ret = __builtin_bit_cast(poly64x1_t, __p0);
59865	61839	return __ret;
59866	61840	}
59867	61841	__ai __attribute__((target("neon"))) poly64x1_t vreinterpret_p64_u8(uint8x8_t __p0) {
59868	61842	poly64x1_t __ret;
59869		__ret = (poly64x1_t)(__p0);
	61843	__ret = __builtin_bit_cast(poly64x1_t, __p0);
59870	61844	return __ret;
59871	61845	}
59872	61846	__ai __attribute__((target("neon"))) poly64x1_t vreinterpret_p64_u32(uint32x2_t __p0) {
59873	61847	poly64x1_t __ret;
59874		__ret = (poly64x1_t)(__p0);
	61848	__ret = __builtin_bit_cast(poly64x1_t, __p0);
59875	61849	return __ret;
59876	61850	}
59877	61851	__ai __attribute__((target("neon"))) poly64x1_t vreinterpret_p64_u64(uint64x1_t __p0) {
59878	61852	poly64x1_t __ret;
59879		__ret = (poly64x1_t)(__p0);
	61853	__ret = __builtin_bit_cast(poly64x1_t, __p0);
59880	61854	return __ret;
59881	61855	}
59882	61856	__ai __attribute__((target("neon"))) poly64x1_t vreinterpret_p64_u16(uint16x4_t __p0) {
59883	61857	poly64x1_t __ret;
59884		__ret = (poly64x1_t)(__p0);
	61858	__ret = __builtin_bit_cast(poly64x1_t, __p0);
59885	61859	return __ret;
59886	61860	}
59887	61861	__ai __attribute__((target("neon"))) poly64x1_t vreinterpret_p64_s8(int8x8_t __p0) {
59888	61862	poly64x1_t __ret;
59889		__ret = (poly64x1_t)(__p0);
	61863	__ret = __builtin_bit_cast(poly64x1_t, __p0);
59890	61864	return __ret;
59891	61865	}
59892	61866	__ai __attribute__((target("neon"))) poly64x1_t vreinterpret_p64_f64(float64x1_t __p0) {
59893	61867	poly64x1_t __ret;
59894		__ret = (poly64x1_t)(__p0);
	61868	__ret = __builtin_bit_cast(poly64x1_t, __p0);
59895	61869	return __ret;
59896	61870	}
59897	61871	__ai __attribute__((target("neon"))) poly64x1_t vreinterpret_p64_f32(float32x2_t __p0) {
59898	61872	poly64x1_t __ret;
59899		__ret = (poly64x1_t)(__p0);
	61873	__ret = __builtin_bit_cast(poly64x1_t, __p0);
59900	61874	return __ret;
59901	61875	}
59902	61876	__ai __attribute__((target("neon"))) poly64x1_t vreinterpret_p64_f16(float16x4_t __p0) {
59903	61877	poly64x1_t __ret;
59904		__ret = (poly64x1_t)(__p0);
	61878	__ret = __builtin_bit_cast(poly64x1_t, __p0);
59905	61879	return __ret;
59906	61880	}
59907	61881	__ai __attribute__((target("neon"))) poly64x1_t vreinterpret_p64_s32(int32x2_t __p0) {
59908	61882	poly64x1_t __ret;
59909		__ret = (poly64x1_t)(__p0);
	61883	__ret = __builtin_bit_cast(poly64x1_t, __p0);
59910	61884	return __ret;
59911	61885	}
59912	61886	__ai __attribute__((target("neon"))) poly64x1_t vreinterpret_p64_s64(int64x1_t __p0) {
59913	61887	poly64x1_t __ret;
59914		__ret = (poly64x1_t)(__p0);
	61888	__ret = __builtin_bit_cast(poly64x1_t, __p0);
59915	61889	return __ret;
59916	61890	}
59917	61891	__ai __attribute__((target("neon"))) poly64x1_t vreinterpret_p64_mf8(mfloat8x8_t __p0) {
59918	61892	poly64x1_t __ret;
59919		__ret = (poly64x1_t)(__p0);
	61893	__ret = __builtin_bit_cast(poly64x1_t, __p0);
59920	61894	return __ret;
59921	61895	}
59922	61896	__ai __attribute__((target("neon"))) poly64x1_t vreinterpret_p64_s16(int16x4_t __p0) {
59923	61897	poly64x1_t __ret;
59924		__ret = (poly64x1_t)(__p0);
	61898	__ret = __builtin_bit_cast(poly64x1_t, __p0);
59925	61899	return __ret;
59926	61900	}
59927	61901	__ai __attribute__((target("neon"))) poly16x4_t vreinterpret_p16_p8(poly8x8_t __p0) {
59928	61902	poly16x4_t __ret;
59929		__ret = (poly16x4_t)(__p0);
	61903	__ret = __builtin_bit_cast(poly16x4_t, __p0);
59930	61904	return __ret;
59931	61905	}
59932	61906	__ai __attribute__((target("neon"))) poly16x4_t vreinterpret_p16_p64(poly64x1_t __p0) {
59933	61907	poly16x4_t __ret;
59934		__ret = (poly16x4_t)(__p0);
	61908	__ret = __builtin_bit_cast(poly16x4_t, __p0);
59935	61909	return __ret;
59936	61910	}
59937	61911	__ai __attribute__((target("neon"))) poly16x4_t vreinterpret_p16_u8(uint8x8_t __p0) {
59938	61912	poly16x4_t __ret;
59939		__ret = (poly16x4_t)(__p0);
	61913	__ret = __builtin_bit_cast(poly16x4_t, __p0);
59940	61914	return __ret;
59941	61915	}
59942	61916	__ai __attribute__((target("neon"))) poly16x4_t vreinterpret_p16_u32(uint32x2_t __p0) {
59943	61917	poly16x4_t __ret;
59944		__ret = (poly16x4_t)(__p0);
	61918	__ret = __builtin_bit_cast(poly16x4_t, __p0);
59945	61919	return __ret;
59946	61920	}
59947	61921	__ai __attribute__((target("neon"))) poly16x4_t vreinterpret_p16_u64(uint64x1_t __p0) {
59948	61922	poly16x4_t __ret;
59949		__ret = (poly16x4_t)(__p0);
	61923	__ret = __builtin_bit_cast(poly16x4_t, __p0);
59950	61924	return __ret;
59951	61925	}
59952	61926	__ai __attribute__((target("neon"))) poly16x4_t vreinterpret_p16_u16(uint16x4_t __p0) {
59953	61927	poly16x4_t __ret;
59954		__ret = (poly16x4_t)(__p0);
	61928	__ret = __builtin_bit_cast(poly16x4_t, __p0);
59955	61929	return __ret;
59956	61930	}
59957	61931	__ai __attribute__((target("neon"))) poly16x4_t vreinterpret_p16_s8(int8x8_t __p0) {
59958	61932	poly16x4_t __ret;
59959		__ret = (poly16x4_t)(__p0);
	61933	__ret = __builtin_bit_cast(poly16x4_t, __p0);
59960	61934	return __ret;
59961	61935	}
59962	61936	__ai __attribute__((target("neon"))) poly16x4_t vreinterpret_p16_f64(float64x1_t __p0) {
59963	61937	poly16x4_t __ret;
59964		__ret = (poly16x4_t)(__p0);
	61938	__ret = __builtin_bit_cast(poly16x4_t, __p0);
59965	61939	return __ret;
59966	61940	}
59967	61941	__ai __attribute__((target("neon"))) poly16x4_t vreinterpret_p16_f32(float32x2_t __p0) {
59968	61942	poly16x4_t __ret;
59969		__ret = (poly16x4_t)(__p0);
	61943	__ret = __builtin_bit_cast(poly16x4_t, __p0);
59970	61944	return __ret;
59971	61945	}
59972	61946	__ai __attribute__((target("neon"))) poly16x4_t vreinterpret_p16_f16(float16x4_t __p0) {
59973	61947	poly16x4_t __ret;
59974		__ret = (poly16x4_t)(__p0);
	61948	__ret = __builtin_bit_cast(poly16x4_t, __p0);
59975	61949	return __ret;
59976	61950	}
59977	61951	__ai __attribute__((target("neon"))) poly16x4_t vreinterpret_p16_s32(int32x2_t __p0) {
59978	61952	poly16x4_t __ret;
59979		__ret = (poly16x4_t)(__p0);
	61953	__ret = __builtin_bit_cast(poly16x4_t, __p0);
59980	61954	return __ret;
59981	61955	}
59982	61956	__ai __attribute__((target("neon"))) poly16x4_t vreinterpret_p16_s64(int64x1_t __p0) {
59983	61957	poly16x4_t __ret;
59984		__ret = (poly16x4_t)(__p0);
	61958	__ret = __builtin_bit_cast(poly16x4_t, __p0);
59985	61959	return __ret;
59986	61960	}
59987	61961	__ai __attribute__((target("neon"))) poly16x4_t vreinterpret_p16_mf8(mfloat8x8_t __p0) {
59988	61962	poly16x4_t __ret;
59989		__ret = (poly16x4_t)(__p0);
	61963	__ret = __builtin_bit_cast(poly16x4_t, __p0);
59990	61964	return __ret;
59991	61965	}
59992	61966	__ai __attribute__((target("neon"))) poly16x4_t vreinterpret_p16_s16(int16x4_t __p0) {
59993	61967	poly16x4_t __ret;
59994		__ret = (poly16x4_t)(__p0);
	61968	__ret = __builtin_bit_cast(poly16x4_t, __p0);
59995	61969	return __ret;
59996	61970	}
59997	61971	__ai __attribute__((target("neon"))) poly8x16_t vreinterpretq_p8_p128(poly128_t __p0) {
59998	61972	poly8x16_t __ret;
59999		__ret = (poly8x16_t)(__p0);
	61973	__ret = __builtin_bit_cast(poly8x16_t, __p0);
60000	61974	return __ret;
60001	61975	}
60002	61976	__ai __attribute__((target("neon"))) poly8x16_t vreinterpretq_p8_p64(poly64x2_t __p0) {
60003	61977	poly8x16_t __ret;
60004		__ret = (poly8x16_t)(__p0);
	61978	__ret = __builtin_bit_cast(poly8x16_t, __p0);
60005	61979	return __ret;
60006	61980	}
60007	61981	__ai __attribute__((target("neon"))) poly8x16_t vreinterpretq_p8_p16(poly16x8_t __p0) {
60008	61982	poly8x16_t __ret;
60009		__ret = (poly8x16_t)(__p0);
	61983	__ret = __builtin_bit_cast(poly8x16_t, __p0);
60010	61984	return __ret;
60011	61985	}
60012	61986	__ai __attribute__((target("neon"))) poly8x16_t vreinterpretq_p8_u8(uint8x16_t __p0) {
60013	61987	poly8x16_t __ret;
60014		__ret = (poly8x16_t)(__p0);
	61988	__ret = __builtin_bit_cast(poly8x16_t, __p0);
60015	61989	return __ret;
60016	61990	}
60017	61991	__ai __attribute__((target("neon"))) poly8x16_t vreinterpretq_p8_u32(uint32x4_t __p0) {
60018	61992	poly8x16_t __ret;
60019		__ret = (poly8x16_t)(__p0);
	61993	__ret = __builtin_bit_cast(poly8x16_t, __p0);
60020	61994	return __ret;
60021	61995	}
60022	61996	__ai __attribute__((target("neon"))) poly8x16_t vreinterpretq_p8_u64(uint64x2_t __p0) {
60023	61997	poly8x16_t __ret;
60024		__ret = (poly8x16_t)(__p0);
	61998	__ret = __builtin_bit_cast(poly8x16_t, __p0);
60025	61999	return __ret;
60026	62000	}
60027	62001	__ai __attribute__((target("neon"))) poly8x16_t vreinterpretq_p8_u16(uint16x8_t __p0) {
60028	62002	poly8x16_t __ret;
60029		__ret = (poly8x16_t)(__p0);
	62003	__ret = __builtin_bit_cast(poly8x16_t, __p0);
60030	62004	return __ret;
60031	62005	}
60032	62006	__ai __attribute__((target("neon"))) poly8x16_t vreinterpretq_p8_s8(int8x16_t __p0) {
60033	62007	poly8x16_t __ret;
60034		__ret = (poly8x16_t)(__p0);
	62008	__ret = __builtin_bit_cast(poly8x16_t, __p0);
60035	62009	return __ret;
60036	62010	}
60037	62011	__ai __attribute__((target("neon"))) poly8x16_t vreinterpretq_p8_f64(float64x2_t __p0) {
60038	62012	poly8x16_t __ret;
60039		__ret = (poly8x16_t)(__p0);
	62013	__ret = __builtin_bit_cast(poly8x16_t, __p0);
60040	62014	return __ret;
60041	62015	}
60042	62016	__ai __attribute__((target("neon"))) poly8x16_t vreinterpretq_p8_f32(float32x4_t __p0) {
60043	62017	poly8x16_t __ret;
60044		__ret = (poly8x16_t)(__p0);
	62018	__ret = __builtin_bit_cast(poly8x16_t, __p0);
60045	62019	return __ret;
60046	62020	}
60047	62021	__ai __attribute__((target("neon"))) poly8x16_t vreinterpretq_p8_f16(float16x8_t __p0) {
60048	62022	poly8x16_t __ret;
60049		__ret = (poly8x16_t)(__p0);
	62023	__ret = __builtin_bit_cast(poly8x16_t, __p0);
60050	62024	return __ret;
60051	62025	}
60052	62026	__ai __attribute__((target("neon"))) poly8x16_t vreinterpretq_p8_s32(int32x4_t __p0) {
60053	62027	poly8x16_t __ret;
60054		__ret = (poly8x16_t)(__p0);
	62028	__ret = __builtin_bit_cast(poly8x16_t, __p0);
60055	62029	return __ret;
60056	62030	}
60057	62031	__ai __attribute__((target("neon"))) poly8x16_t vreinterpretq_p8_s64(int64x2_t __p0) {
60058	62032	poly8x16_t __ret;
60059		__ret = (poly8x16_t)(__p0);
	62033	__ret = __builtin_bit_cast(poly8x16_t, __p0);
60060	62034	return __ret;
60061	62035	}
60062	62036	__ai __attribute__((target("neon"))) poly8x16_t vreinterpretq_p8_mf8(mfloat8x16_t __p0) {
60063	62037	poly8x16_t __ret;
60064		__ret = (poly8x16_t)(__p0);
	62038	__ret = __builtin_bit_cast(poly8x16_t, __p0);
60065	62039	return __ret;
60066	62040	}
60067	62041	__ai __attribute__((target("neon"))) poly8x16_t vreinterpretq_p8_s16(int16x8_t __p0) {
60068	62042	poly8x16_t __ret;
60069		__ret = (poly8x16_t)(__p0);
	62043	__ret = __builtin_bit_cast(poly8x16_t, __p0);
60070	62044	return __ret;
60071	62045	}
60072	62046	__ai __attribute__((target("neon"))) poly128_t vreinterpretq_p128_p8(poly8x16_t __p0) {
60073	62047	poly128_t __ret;
60074		__ret = (poly128_t)(__p0);
	62048	__ret = __builtin_bit_cast(poly128_t, __p0);
60075	62049	return __ret;
60076	62050	}
60077	62051	__ai __attribute__((target("neon"))) poly128_t vreinterpretq_p128_p64(poly64x2_t __p0) {
60078	62052	poly128_t __ret;
60079		__ret = (poly128_t)(__p0);
	62053	__ret = __builtin_bit_cast(poly128_t, __p0);
60080	62054	return __ret;
60081	62055	}
60082	62056	__ai __attribute__((target("neon"))) poly128_t vreinterpretq_p128_p16(poly16x8_t __p0) {
60083	62057	poly128_t __ret;
60084		__ret = (poly128_t)(__p0);
	62058	__ret = __builtin_bit_cast(poly128_t, __p0);
60085	62059	return __ret;
60086	62060	}
60087	62061	__ai __attribute__((target("neon"))) poly128_t vreinterpretq_p128_u8(uint8x16_t __p0) {
60088	62062	poly128_t __ret;
60089		__ret = (poly128_t)(__p0);
	62063	__ret = __builtin_bit_cast(poly128_t, __p0);
60090	62064	return __ret;
60091	62065	}
60092	62066	__ai __attribute__((target("neon"))) poly128_t vreinterpretq_p128_u32(uint32x4_t __p0) {
60093	62067	poly128_t __ret;
60094		__ret = (poly128_t)(__p0);
	62068	__ret = __builtin_bit_cast(poly128_t, __p0);
60095	62069	return __ret;
60096	62070	}
60097	62071	__ai __attribute__((target("neon"))) poly128_t vreinterpretq_p128_u64(uint64x2_t __p0) {
60098	62072	poly128_t __ret;
60099		__ret = (poly128_t)(__p0);
	62073	__ret = __builtin_bit_cast(poly128_t, __p0);
60100	62074	return __ret;
60101	62075	}
60102	62076	__ai __attribute__((target("neon"))) poly128_t vreinterpretq_p128_u16(uint16x8_t __p0) {
60103	62077	poly128_t __ret;
60104		__ret = (poly128_t)(__p0);
	62078	__ret = __builtin_bit_cast(poly128_t, __p0);
60105	62079	return __ret;
60106	62080	}
60107	62081	__ai __attribute__((target("neon"))) poly128_t vreinterpretq_p128_s8(int8x16_t __p0) {
60108	62082	poly128_t __ret;
60109		__ret = (poly128_t)(__p0);
	62083	__ret = __builtin_bit_cast(poly128_t, __p0);
60110	62084	return __ret;
60111	62085	}
60112	62086	__ai __attribute__((target("neon"))) poly128_t vreinterpretq_p128_f64(float64x2_t __p0) {
60113	62087	poly128_t __ret;
60114		__ret = (poly128_t)(__p0);
	62088	__ret = __builtin_bit_cast(poly128_t, __p0);
60115	62089	return __ret;
60116	62090	}
60117	62091	__ai __attribute__((target("neon"))) poly128_t vreinterpretq_p128_f32(float32x4_t __p0) {
60118	62092	poly128_t __ret;
60119		__ret = (poly128_t)(__p0);
	62093	__ret = __builtin_bit_cast(poly128_t, __p0);
60120	62094	return __ret;
60121	62095	}
60122	62096	__ai __attribute__((target("neon"))) poly128_t vreinterpretq_p128_f16(float16x8_t __p0) {
60123	62097	poly128_t __ret;
60124		__ret = (poly128_t)(__p0);
	62098	__ret = __builtin_bit_cast(poly128_t, __p0);
60125	62099	return __ret;
60126	62100	}
60127	62101	__ai __attribute__((target("neon"))) poly128_t vreinterpretq_p128_s32(int32x4_t __p0) {
60128	62102	poly128_t __ret;
60129		__ret = (poly128_t)(__p0);
	62103	__ret = __builtin_bit_cast(poly128_t, __p0);
60130	62104	return __ret;
60131	62105	}
60132	62106	__ai __attribute__((target("neon"))) poly128_t vreinterpretq_p128_s64(int64x2_t __p0) {
60133	62107	poly128_t __ret;
60134		__ret = (poly128_t)(__p0);
	62108	__ret = __builtin_bit_cast(poly128_t, __p0);
60135	62109	return __ret;
60136	62110	}
60137	62111	__ai __attribute__((target("neon"))) poly128_t vreinterpretq_p128_mf8(mfloat8x16_t __p0) {
60138	62112	poly128_t __ret;
60139		__ret = (poly128_t)(__p0);
	62113	__ret = __builtin_bit_cast(poly128_t, __p0);
60140	62114	return __ret;
60141	62115	}
60142	62116	__ai __attribute__((target("neon"))) poly128_t vreinterpretq_p128_s16(int16x8_t __p0) {
60143	62117	poly128_t __ret;
60144		__ret = (poly128_t)(__p0);
	62118	__ret = __builtin_bit_cast(poly128_t, __p0);
60145	62119	return __ret;
60146	62120	}
60147	62121	__ai __attribute__((target("neon"))) poly64x2_t vreinterpretq_p64_p8(poly8x16_t __p0) {
60148	62122	poly64x2_t __ret;
60149		__ret = (poly64x2_t)(__p0);
	62123	__ret = __builtin_bit_cast(poly64x2_t, __p0);
60150	62124	return __ret;
60151	62125	}
60152	62126	__ai __attribute__((target("neon"))) poly64x2_t vreinterpretq_p64_p128(poly128_t __p0) {
60153	62127	poly64x2_t __ret;
60154		__ret = (poly64x2_t)(__p0);
	62128	__ret = __builtin_bit_cast(poly64x2_t, __p0);
60155	62129	return __ret;
60156	62130	}
60157	62131	__ai __attribute__((target("neon"))) poly64x2_t vreinterpretq_p64_p16(poly16x8_t __p0) {
60158	62132	poly64x2_t __ret;
60159		__ret = (poly64x2_t)(__p0);
	62133	__ret = __builtin_bit_cast(poly64x2_t, __p0);
60160	62134	return __ret;
60161	62135	}
60162	62136	__ai __attribute__((target("neon"))) poly64x2_t vreinterpretq_p64_u8(uint8x16_t __p0) {
60163	62137	poly64x2_t __ret;
60164		__ret = (poly64x2_t)(__p0);
	62138	__ret = __builtin_bit_cast(poly64x2_t, __p0);
60165	62139	return __ret;
60166	62140	}
60167	62141	__ai __attribute__((target("neon"))) poly64x2_t vreinterpretq_p64_u32(uint32x4_t __p0) {
60168	62142	poly64x2_t __ret;
60169		__ret = (poly64x2_t)(__p0);
	62143	__ret = __builtin_bit_cast(poly64x2_t, __p0);
60170	62144	return __ret;
60171	62145	}
60172	62146	__ai __attribute__((target("neon"))) poly64x2_t vreinterpretq_p64_u64(uint64x2_t __p0) {
60173	62147	poly64x2_t __ret;
60174		__ret = (poly64x2_t)(__p0);
	62148	__ret = __builtin_bit_cast(poly64x2_t, __p0);
60175	62149	return __ret;
60176	62150	}
60177	62151	__ai __attribute__((target("neon"))) poly64x2_t vreinterpretq_p64_u16(uint16x8_t __p0) {
60178	62152	poly64x2_t __ret;
60179		__ret = (poly64x2_t)(__p0);
	62153	__ret = __builtin_bit_cast(poly64x2_t, __p0);
60180	62154	return __ret;
60181	62155	}
60182	62156	__ai __attribute__((target("neon"))) poly64x2_t vreinterpretq_p64_s8(int8x16_t __p0) {
60183	62157	poly64x2_t __ret;
60184		__ret = (poly64x2_t)(__p0);
	62158	__ret = __builtin_bit_cast(poly64x2_t, __p0);
60185	62159	return __ret;
60186	62160	}
60187	62161	__ai __attribute__((target("neon"))) poly64x2_t vreinterpretq_p64_f64(float64x2_t __p0) {
60188	62162	poly64x2_t __ret;
60189		__ret = (poly64x2_t)(__p0);
	62163	__ret = __builtin_bit_cast(poly64x2_t, __p0);
60190	62164	return __ret;
60191	62165	}
60192	62166	__ai __attribute__((target("neon"))) poly64x2_t vreinterpretq_p64_f32(float32x4_t __p0) {
60193	62167	poly64x2_t __ret;
60194		__ret = (poly64x2_t)(__p0);
	62168	__ret = __builtin_bit_cast(poly64x2_t, __p0);
60195	62169	return __ret;
60196	62170	}
60197	62171	__ai __attribute__((target("neon"))) poly64x2_t vreinterpretq_p64_f16(float16x8_t __p0) {
60198	62172	poly64x2_t __ret;
60199		__ret = (poly64x2_t)(__p0);
	62173	__ret = __builtin_bit_cast(poly64x2_t, __p0);
60200	62174	return __ret;
60201	62175	}
60202	62176	__ai __attribute__((target("neon"))) poly64x2_t vreinterpretq_p64_s32(int32x4_t __p0) {
60203	62177	poly64x2_t __ret;
60204		__ret = (poly64x2_t)(__p0);
	62178	__ret = __builtin_bit_cast(poly64x2_t, __p0);
60205	62179	return __ret;
60206	62180	}
60207	62181	__ai __attribute__((target("neon"))) poly64x2_t vreinterpretq_p64_s64(int64x2_t __p0) {
60208	62182	poly64x2_t __ret;
60209		__ret = (poly64x2_t)(__p0);
	62183	__ret = __builtin_bit_cast(poly64x2_t, __p0);
60210	62184	return __ret;
60211	62185	}
60212	62186	__ai __attribute__((target("neon"))) poly64x2_t vreinterpretq_p64_mf8(mfloat8x16_t __p0) {
60213	62187	poly64x2_t __ret;
60214		__ret = (poly64x2_t)(__p0);
	62188	__ret = __builtin_bit_cast(poly64x2_t, __p0);
60215	62189	return __ret;
60216	62190	}
60217	62191	__ai __attribute__((target("neon"))) poly64x2_t vreinterpretq_p64_s16(int16x8_t __p0) {
60218	62192	poly64x2_t __ret;
60219		__ret = (poly64x2_t)(__p0);
	62193	__ret = __builtin_bit_cast(poly64x2_t, __p0);
60220	62194	return __ret;
60221	62195	}
60222	62196	__ai __attribute__((target("neon"))) poly16x8_t vreinterpretq_p16_p8(poly8x16_t __p0) {
60223	62197	poly16x8_t __ret;
60224		__ret = (poly16x8_t)(__p0);
	62198	__ret = __builtin_bit_cast(poly16x8_t, __p0);
60225	62199	return __ret;
60226	62200	}
60227	62201	__ai __attribute__((target("neon"))) poly16x8_t vreinterpretq_p16_p128(poly128_t __p0) {
60228	62202	poly16x8_t __ret;
60229		__ret = (poly16x8_t)(__p0);
	62203	__ret = __builtin_bit_cast(poly16x8_t, __p0);
60230	62204	return __ret;
60231	62205	}
60232	62206	__ai __attribute__((target("neon"))) poly16x8_t vreinterpretq_p16_p64(poly64x2_t __p0) {
60233	62207	poly16x8_t __ret;
60234		__ret = (poly16x8_t)(__p0);
	62208	__ret = __builtin_bit_cast(poly16x8_t, __p0);
60235	62209	return __ret;
60236	62210	}
60237	62211	__ai __attribute__((target("neon"))) poly16x8_t vreinterpretq_p16_u8(uint8x16_t __p0) {
60238	62212	poly16x8_t __ret;
60239		__ret = (poly16x8_t)(__p0);
	62213	__ret = __builtin_bit_cast(poly16x8_t, __p0);
60240	62214	return __ret;
60241	62215	}
60242	62216	__ai __attribute__((target("neon"))) poly16x8_t vreinterpretq_p16_u32(uint32x4_t __p0) {
60243	62217	poly16x8_t __ret;
60244		__ret = (poly16x8_t)(__p0);
	62218	__ret = __builtin_bit_cast(poly16x8_t, __p0);
60245	62219	return __ret;
60246	62220	}
60247	62221	__ai __attribute__((target("neon"))) poly16x8_t vreinterpretq_p16_u64(uint64x2_t __p0) {
60248	62222	poly16x8_t __ret;
60249		__ret = (poly16x8_t)(__p0);
	62223	__ret = __builtin_bit_cast(poly16x8_t, __p0);
60250	62224	return __ret;
60251	62225	}
60252	62226	__ai __attribute__((target("neon"))) poly16x8_t vreinterpretq_p16_u16(uint16x8_t __p0) {
60253	62227	poly16x8_t __ret;
60254		__ret = (poly16x8_t)(__p0);
	62228	__ret = __builtin_bit_cast(poly16x8_t, __p0);
60255	62229	return __ret;
60256	62230	}
60257	62231	__ai __attribute__((target("neon"))) poly16x8_t vreinterpretq_p16_s8(int8x16_t __p0) {
60258	62232	poly16x8_t __ret;
60259		__ret = (poly16x8_t)(__p0);
	62233	__ret = __builtin_bit_cast(poly16x8_t, __p0);
60260	62234	return __ret;
60261	62235	}
60262	62236	__ai __attribute__((target("neon"))) poly16x8_t vreinterpretq_p16_f64(float64x2_t __p0) {
60263	62237	poly16x8_t __ret;
60264		__ret = (poly16x8_t)(__p0);
	62238	__ret = __builtin_bit_cast(poly16x8_t, __p0);
60265	62239	return __ret;
60266	62240	}
60267	62241	__ai __attribute__((target("neon"))) poly16x8_t vreinterpretq_p16_f32(float32x4_t __p0) {
60268	62242	poly16x8_t __ret;
60269		__ret = (poly16x8_t)(__p0);
	62243	__ret = __builtin_bit_cast(poly16x8_t, __p0);
60270	62244	return __ret;
60271	62245	}
60272	62246	__ai __attribute__((target("neon"))) poly16x8_t vreinterpretq_p16_f16(float16x8_t __p0) {
60273	62247	poly16x8_t __ret;
60274		__ret = (poly16x8_t)(__p0);
	62248	__ret = __builtin_bit_cast(poly16x8_t, __p0);
60275	62249	return __ret;
60276	62250	}
60277	62251	__ai __attribute__((target("neon"))) poly16x8_t vreinterpretq_p16_s32(int32x4_t __p0) {
60278	62252	poly16x8_t __ret;
60279		__ret = (poly16x8_t)(__p0);
	62253	__ret = __builtin_bit_cast(poly16x8_t, __p0);
60280	62254	return __ret;
60281	62255	}
60282	62256	__ai __attribute__((target("neon"))) poly16x8_t vreinterpretq_p16_s64(int64x2_t __p0) {
60283	62257	poly16x8_t __ret;
60284		__ret = (poly16x8_t)(__p0);
	62258	__ret = __builtin_bit_cast(poly16x8_t, __p0);
60285	62259	return __ret;
60286	62260	}
60287	62261	__ai __attribute__((target("neon"))) poly16x8_t vreinterpretq_p16_mf8(mfloat8x16_t __p0) {
60288	62262	poly16x8_t __ret;
60289		__ret = (poly16x8_t)(__p0);
	62263	__ret = __builtin_bit_cast(poly16x8_t, __p0);
60290	62264	return __ret;
60291	62265	}
60292	62266	__ai __attribute__((target("neon"))) poly16x8_t vreinterpretq_p16_s16(int16x8_t __p0) {
60293	62267	poly16x8_t __ret;
60294		__ret = (poly16x8_t)(__p0);
	62268	__ret = __builtin_bit_cast(poly16x8_t, __p0);
60295	62269	return __ret;
60296	62270	}
60297	62271	__ai __attribute__((target("neon"))) uint8x16_t vreinterpretq_u8_p8(poly8x16_t __p0) {
60298	62272	uint8x16_t __ret;
60299		__ret = (uint8x16_t)(__p0);
	62273	__ret = __builtin_bit_cast(uint8x16_t, __p0);
60300	62274	return __ret;
60301	62275	}
60302	62276	__ai __attribute__((target("neon"))) uint8x16_t vreinterpretq_u8_p128(poly128_t __p0) {
60303	62277	uint8x16_t __ret;
60304		__ret = (uint8x16_t)(__p0);
	62278	__ret = __builtin_bit_cast(uint8x16_t, __p0);
60305	62279	return __ret;
60306	62280	}
60307	62281	__ai __attribute__((target("neon"))) uint8x16_t vreinterpretq_u8_p64(poly64x2_t __p0) {
60308	62282	uint8x16_t __ret;
60309		__ret = (uint8x16_t)(__p0);
	62283	__ret = __builtin_bit_cast(uint8x16_t, __p0);
60310	62284	return __ret;
60311	62285	}
60312	62286	__ai __attribute__((target("neon"))) uint8x16_t vreinterpretq_u8_p16(poly16x8_t __p0) {
60313	62287	uint8x16_t __ret;
60314		__ret = (uint8x16_t)(__p0);
	62288	__ret = __builtin_bit_cast(uint8x16_t, __p0);
60315	62289	return __ret;
60316	62290	}
60317	62291	__ai __attribute__((target("neon"))) uint8x16_t vreinterpretq_u8_u32(uint32x4_t __p0) {
60318	62292	uint8x16_t __ret;
60319		__ret = (uint8x16_t)(__p0);
	62293	__ret = __builtin_bit_cast(uint8x16_t, __p0);
60320	62294	return __ret;
60321	62295	}
60322	62296	__ai __attribute__((target("neon"))) uint8x16_t vreinterpretq_u8_u64(uint64x2_t __p0) {
60323	62297	uint8x16_t __ret;
60324		__ret = (uint8x16_t)(__p0);
	62298	__ret = __builtin_bit_cast(uint8x16_t, __p0);
60325	62299	return __ret;
60326	62300	}
60327	62301	__ai __attribute__((target("neon"))) uint8x16_t vreinterpretq_u8_u16(uint16x8_t __p0) {
60328	62302	uint8x16_t __ret;
60329		__ret = (uint8x16_t)(__p0);
	62303	__ret = __builtin_bit_cast(uint8x16_t, __p0);
60330	62304	return __ret;
60331	62305	}
60332	62306	__ai __attribute__((target("neon"))) uint8x16_t vreinterpretq_u8_s8(int8x16_t __p0) {
60333	62307	uint8x16_t __ret;
60334		__ret = (uint8x16_t)(__p0);
	62308	__ret = __builtin_bit_cast(uint8x16_t, __p0);
60335	62309	return __ret;
60336	62310	}
60337	62311	__ai __attribute__((target("neon"))) uint8x16_t vreinterpretq_u8_f64(float64x2_t __p0) {
60338	62312	uint8x16_t __ret;
60339		__ret = (uint8x16_t)(__p0);
	62313	__ret = __builtin_bit_cast(uint8x16_t, __p0);
60340	62314	return __ret;
60341	62315	}
60342	62316	__ai __attribute__((target("neon"))) uint8x16_t vreinterpretq_u8_f32(float32x4_t __p0) {
60343	62317	uint8x16_t __ret;
60344		__ret = (uint8x16_t)(__p0);
	62318	__ret = __builtin_bit_cast(uint8x16_t, __p0);
60345	62319	return __ret;
60346	62320	}
60347	62321	__ai __attribute__((target("neon"))) uint8x16_t vreinterpretq_u8_f16(float16x8_t __p0) {
60348	62322	uint8x16_t __ret;
60349		__ret = (uint8x16_t)(__p0);
	62323	__ret = __builtin_bit_cast(uint8x16_t, __p0);
60350	62324	return __ret;
60351	62325	}
60352	62326	__ai __attribute__((target("neon"))) uint8x16_t vreinterpretq_u8_s32(int32x4_t __p0) {
60353	62327	uint8x16_t __ret;
60354		__ret = (uint8x16_t)(__p0);
	62328	__ret = __builtin_bit_cast(uint8x16_t, __p0);
60355	62329	return __ret;
60356	62330	}
60357	62331	__ai __attribute__((target("neon"))) uint8x16_t vreinterpretq_u8_s64(int64x2_t __p0) {
60358	62332	uint8x16_t __ret;
60359		__ret = (uint8x16_t)(__p0);
	62333	__ret = __builtin_bit_cast(uint8x16_t, __p0);
60360	62334	return __ret;
60361	62335	}
60362	62336	__ai __attribute__((target("neon"))) uint8x16_t vreinterpretq_u8_mf8(mfloat8x16_t __p0) {
60363	62337	uint8x16_t __ret;
60364		__ret = (uint8x16_t)(__p0);
	62338	__ret = __builtin_bit_cast(uint8x16_t, __p0);
60365	62339	return __ret;
60366	62340	}
60367	62341	__ai __attribute__((target("neon"))) uint8x16_t vreinterpretq_u8_s16(int16x8_t __p0) {
60368	62342	uint8x16_t __ret;
60369		__ret = (uint8x16_t)(__p0);
	62343	__ret = __builtin_bit_cast(uint8x16_t, __p0);
60370	62344	return __ret;
60371	62345	}
60372	62346	__ai __attribute__((target("neon"))) uint32x4_t vreinterpretq_u32_p8(poly8x16_t __p0) {
60373	62347	uint32x4_t __ret;
60374		__ret = (uint32x4_t)(__p0);
	62348	__ret = __builtin_bit_cast(uint32x4_t, __p0);
60375	62349	return __ret;
60376	62350	}
60377	62351	__ai __attribute__((target("neon"))) uint32x4_t vreinterpretq_u32_p128(poly128_t __p0) {
60378	62352	uint32x4_t __ret;
60379		__ret = (uint32x4_t)(__p0);
	62353	__ret = __builtin_bit_cast(uint32x4_t, __p0);
60380	62354	return __ret;
60381	62355	}
60382	62356	__ai __attribute__((target("neon"))) uint32x4_t vreinterpretq_u32_p64(poly64x2_t __p0) {
60383	62357	uint32x4_t __ret;
60384		__ret = (uint32x4_t)(__p0);
	62358	__ret = __builtin_bit_cast(uint32x4_t, __p0);
60385	62359	return __ret;
60386	62360	}
60387	62361	__ai __attribute__((target("neon"))) uint32x4_t vreinterpretq_u32_p16(poly16x8_t __p0) {
60388	62362	uint32x4_t __ret;
60389		__ret = (uint32x4_t)(__p0);
	62363	__ret = __builtin_bit_cast(uint32x4_t, __p0);
60390	62364	return __ret;
60391	62365	}
60392	62366	__ai __attribute__((target("neon"))) uint32x4_t vreinterpretq_u32_u8(uint8x16_t __p0) {
60393	62367	uint32x4_t __ret;
60394		__ret = (uint32x4_t)(__p0);
	62368	__ret = __builtin_bit_cast(uint32x4_t, __p0);
60395	62369	return __ret;
60396	62370	}
60397	62371	__ai __attribute__((target("neon"))) uint32x4_t vreinterpretq_u32_u64(uint64x2_t __p0) {
60398	62372	uint32x4_t __ret;
60399		__ret = (uint32x4_t)(__p0);
	62373	__ret = __builtin_bit_cast(uint32x4_t, __p0);
60400	62374	return __ret;
60401	62375	}
60402	62376	__ai __attribute__((target("neon"))) uint32x4_t vreinterpretq_u32_u16(uint16x8_t __p0) {
60403	62377	uint32x4_t __ret;
60404		__ret = (uint32x4_t)(__p0);
	62378	__ret = __builtin_bit_cast(uint32x4_t, __p0);
60405	62379	return __ret;
60406	62380	}
60407	62381	__ai __attribute__((target("neon"))) uint32x4_t vreinterpretq_u32_s8(int8x16_t __p0) {
60408	62382	uint32x4_t __ret;
60409		__ret = (uint32x4_t)(__p0);
	62383	__ret = __builtin_bit_cast(uint32x4_t, __p0);
60410	62384	return __ret;
60411	62385	}
60412	62386	__ai __attribute__((target("neon"))) uint32x4_t vreinterpretq_u32_f64(float64x2_t __p0) {
60413	62387	uint32x4_t __ret;
60414		__ret = (uint32x4_t)(__p0);
	62388	__ret = __builtin_bit_cast(uint32x4_t, __p0);
60415	62389	return __ret;
60416	62390	}
60417	62391	__ai __attribute__((target("neon"))) uint32x4_t vreinterpretq_u32_f32(float32x4_t __p0) {
60418	62392	uint32x4_t __ret;
60419		__ret = (uint32x4_t)(__p0);
	62393	__ret = __builtin_bit_cast(uint32x4_t, __p0);
60420	62394	return __ret;
60421	62395	}
60422	62396	__ai __attribute__((target("neon"))) uint32x4_t vreinterpretq_u32_f16(float16x8_t __p0) {
60423	62397	uint32x4_t __ret;
60424		__ret = (uint32x4_t)(__p0);
	62398	__ret = __builtin_bit_cast(uint32x4_t, __p0);
60425	62399	return __ret;
60426	62400	}
60427	62401	__ai __attribute__((target("neon"))) uint32x4_t vreinterpretq_u32_s32(int32x4_t __p0) {
60428	62402	uint32x4_t __ret;
60429		__ret = (uint32x4_t)(__p0);
	62403	__ret = __builtin_bit_cast(uint32x4_t, __p0);
60430	62404	return __ret;
60431	62405	}
60432	62406	__ai __attribute__((target("neon"))) uint32x4_t vreinterpretq_u32_s64(int64x2_t __p0) {
60433	62407	uint32x4_t __ret;
60434		__ret = (uint32x4_t)(__p0);
	62408	__ret = __builtin_bit_cast(uint32x4_t, __p0);
60435	62409	return __ret;
60436	62410	}
60437	62411	__ai __attribute__((target("neon"))) uint32x4_t vreinterpretq_u32_mf8(mfloat8x16_t __p0) {
60438	62412	uint32x4_t __ret;
60439		__ret = (uint32x4_t)(__p0);
	62413	__ret = __builtin_bit_cast(uint32x4_t, __p0);
60440	62414	return __ret;
60441	62415	}
60442	62416	__ai __attribute__((target("neon"))) uint32x4_t vreinterpretq_u32_s16(int16x8_t __p0) {
60443	62417	uint32x4_t __ret;
60444		__ret = (uint32x4_t)(__p0);
	62418	__ret = __builtin_bit_cast(uint32x4_t, __p0);
60445	62419	return __ret;
60446	62420	}
60447	62421	__ai __attribute__((target("neon"))) uint64x2_t vreinterpretq_u64_p8(poly8x16_t __p0) {
60448	62422	uint64x2_t __ret;
60449		__ret = (uint64x2_t)(__p0);
	62423	__ret = __builtin_bit_cast(uint64x2_t, __p0);
60450	62424	return __ret;
60451	62425	}
60452	62426	__ai __attribute__((target("neon"))) uint64x2_t vreinterpretq_u64_p128(poly128_t __p0) {
60453	62427	uint64x2_t __ret;
60454		__ret = (uint64x2_t)(__p0);
	62428	__ret = __builtin_bit_cast(uint64x2_t, __p0);
60455	62429	return __ret;
60456	62430	}
60457	62431	__ai __attribute__((target("neon"))) uint64x2_t vreinterpretq_u64_p64(poly64x2_t __p0) {
60458	62432	uint64x2_t __ret;
60459		__ret = (uint64x2_t)(__p0);
	62433	__ret = __builtin_bit_cast(uint64x2_t, __p0);
60460	62434	return __ret;
60461	62435	}
60462	62436	__ai __attribute__((target("neon"))) uint64x2_t vreinterpretq_u64_p16(poly16x8_t __p0) {
60463	62437	uint64x2_t __ret;
60464		__ret = (uint64x2_t)(__p0);
	62438	__ret = __builtin_bit_cast(uint64x2_t, __p0);
60465	62439	return __ret;
60466	62440	}
60467	62441	__ai __attribute__((target("neon"))) uint64x2_t vreinterpretq_u64_u8(uint8x16_t __p0) {
60468	62442	uint64x2_t __ret;
60469		__ret = (uint64x2_t)(__p0);
	62443	__ret = __builtin_bit_cast(uint64x2_t, __p0);
60470	62444	return __ret;
60471	62445	}
60472	62446	__ai __attribute__((target("neon"))) uint64x2_t vreinterpretq_u64_u32(uint32x4_t __p0) {
60473	62447	uint64x2_t __ret;
60474		__ret = (uint64x2_t)(__p0);
	62448	__ret = __builtin_bit_cast(uint64x2_t, __p0);
60475	62449	return __ret;
60476	62450	}
60477	62451	__ai __attribute__((target("neon"))) uint64x2_t vreinterpretq_u64_u16(uint16x8_t __p0) {
60478	62452	uint64x2_t __ret;
60479		__ret = (uint64x2_t)(__p0);
	62453	__ret = __builtin_bit_cast(uint64x2_t, __p0);
60480	62454	return __ret;
60481	62455	}
60482	62456	__ai __attribute__((target("neon"))) uint64x2_t vreinterpretq_u64_s8(int8x16_t __p0) {
60483	62457	uint64x2_t __ret;
60484		__ret = (uint64x2_t)(__p0);
	62458	__ret = __builtin_bit_cast(uint64x2_t, __p0);
60485	62459	return __ret;
60486	62460	}
60487	62461	__ai __attribute__((target("neon"))) uint64x2_t vreinterpretq_u64_f64(float64x2_t __p0) {
60488	62462	uint64x2_t __ret;
60489		__ret = (uint64x2_t)(__p0);
	62463	__ret = __builtin_bit_cast(uint64x2_t, __p0);
60490	62464	return __ret;
60491	62465	}
60492	62466	__ai __attribute__((target("neon"))) uint64x2_t vreinterpretq_u64_f32(float32x4_t __p0) {
60493	62467	uint64x2_t __ret;
60494		__ret = (uint64x2_t)(__p0);
	62468	__ret = __builtin_bit_cast(uint64x2_t, __p0);
60495	62469	return __ret;
60496	62470	}
60497	62471	__ai __attribute__((target("neon"))) uint64x2_t vreinterpretq_u64_f16(float16x8_t __p0) {
60498	62472	uint64x2_t __ret;
60499		__ret = (uint64x2_t)(__p0);
	62473	__ret = __builtin_bit_cast(uint64x2_t, __p0);
60500	62474	return __ret;
60501	62475	}
60502	62476	__ai __attribute__((target("neon"))) uint64x2_t vreinterpretq_u64_s32(int32x4_t __p0) {
60503	62477	uint64x2_t __ret;
60504		__ret = (uint64x2_t)(__p0);
	62478	__ret = __builtin_bit_cast(uint64x2_t, __p0);
60505	62479	return __ret;
60506	62480	}
60507	62481	__ai __attribute__((target("neon"))) uint64x2_t vreinterpretq_u64_s64(int64x2_t __p0) {
60508	62482	uint64x2_t __ret;
60509		__ret = (uint64x2_t)(__p0);
	62483	__ret = __builtin_bit_cast(uint64x2_t, __p0);
60510	62484	return __ret;
60511	62485	}
60512	62486	__ai __attribute__((target("neon"))) uint64x2_t vreinterpretq_u64_mf8(mfloat8x16_t __p0) {
60513	62487	uint64x2_t __ret;
60514		__ret = (uint64x2_t)(__p0);
	62488	__ret = __builtin_bit_cast(uint64x2_t, __p0);
60515	62489	return __ret;
60516	62490	}
60517	62491	__ai __attribute__((target("neon"))) uint64x2_t vreinterpretq_u64_s16(int16x8_t __p0) {
60518	62492	uint64x2_t __ret;
60519		__ret = (uint64x2_t)(__p0);
	62493	__ret = __builtin_bit_cast(uint64x2_t, __p0);
60520	62494	return __ret;
60521	62495	}
60522	62496	__ai __attribute__((target("neon"))) uint16x8_t vreinterpretq_u16_p8(poly8x16_t __p0) {
60523	62497	uint16x8_t __ret;
60524		__ret = (uint16x8_t)(__p0);
	62498	__ret = __builtin_bit_cast(uint16x8_t, __p0);
60525	62499	return __ret;
60526	62500	}
60527	62501	__ai __attribute__((target("neon"))) uint16x8_t vreinterpretq_u16_p128(poly128_t __p0) {
60528	62502	uint16x8_t __ret;
60529		__ret = (uint16x8_t)(__p0);
	62503	__ret = __builtin_bit_cast(uint16x8_t, __p0);
60530	62504	return __ret;
60531	62505	}
60532	62506	__ai __attribute__((target("neon"))) uint16x8_t vreinterpretq_u16_p64(poly64x2_t __p0) {
60533	62507	uint16x8_t __ret;
60534		__ret = (uint16x8_t)(__p0);
	62508	__ret = __builtin_bit_cast(uint16x8_t, __p0);
60535	62509	return __ret;
60536	62510	}
60537	62511	__ai __attribute__((target("neon"))) uint16x8_t vreinterpretq_u16_p16(poly16x8_t __p0) {
60538	62512	uint16x8_t __ret;
60539		__ret = (uint16x8_t)(__p0);
	62513	__ret = __builtin_bit_cast(uint16x8_t, __p0);
60540	62514	return __ret;
60541	62515	}
60542	62516	__ai __attribute__((target("neon"))) uint16x8_t vreinterpretq_u16_u8(uint8x16_t __p0) {
60543	62517	uint16x8_t __ret;
60544		__ret = (uint16x8_t)(__p0);
	62518	__ret = __builtin_bit_cast(uint16x8_t, __p0);
60545	62519	return __ret;
60546	62520	}
60547	62521	__ai __attribute__((target("neon"))) uint16x8_t vreinterpretq_u16_u32(uint32x4_t __p0) {
60548	62522	uint16x8_t __ret;
60549		__ret = (uint16x8_t)(__p0);
	62523	__ret = __builtin_bit_cast(uint16x8_t, __p0);
60550	62524	return __ret;
60551	62525	}
60552	62526	__ai __attribute__((target("neon"))) uint16x8_t vreinterpretq_u16_u64(uint64x2_t __p0) {
60553	62527	uint16x8_t __ret;
60554		__ret = (uint16x8_t)(__p0);
	62528	__ret = __builtin_bit_cast(uint16x8_t, __p0);
60555	62529	return __ret;
60556	62530	}
60557	62531	__ai __attribute__((target("neon"))) uint16x8_t vreinterpretq_u16_s8(int8x16_t __p0) {
60558	62532	uint16x8_t __ret;
60559		__ret = (uint16x8_t)(__p0);
	62533	__ret = __builtin_bit_cast(uint16x8_t, __p0);
60560	62534	return __ret;
60561	62535	}
60562	62536	__ai __attribute__((target("neon"))) uint16x8_t vreinterpretq_u16_f64(float64x2_t __p0) {
60563	62537	uint16x8_t __ret;
60564		__ret = (uint16x8_t)(__p0);
	62538	__ret = __builtin_bit_cast(uint16x8_t, __p0);
60565	62539	return __ret;
60566	62540	}
60567	62541	__ai __attribute__((target("neon"))) uint16x8_t vreinterpretq_u16_f32(float32x4_t __p0) {
60568	62542	uint16x8_t __ret;
60569		__ret = (uint16x8_t)(__p0);
	62543	__ret = __builtin_bit_cast(uint16x8_t, __p0);
60570	62544	return __ret;
60571	62545	}
60572	62546	__ai __attribute__((target("neon"))) uint16x8_t vreinterpretq_u16_f16(float16x8_t __p0) {
60573	62547	uint16x8_t __ret;
60574		__ret = (uint16x8_t)(__p0);
	62548	__ret = __builtin_bit_cast(uint16x8_t, __p0);
60575	62549	return __ret;
60576	62550	}
60577	62551	__ai __attribute__((target("neon"))) uint16x8_t vreinterpretq_u16_s32(int32x4_t __p0) {
60578	62552	uint16x8_t __ret;
60579		__ret = (uint16x8_t)(__p0);
	62553	__ret = __builtin_bit_cast(uint16x8_t, __p0);
60580	62554	return __ret;
60581	62555	}
60582	62556	__ai __attribute__((target("neon"))) uint16x8_t vreinterpretq_u16_s64(int64x2_t __p0) {
60583	62557	uint16x8_t __ret;
60584		__ret = (uint16x8_t)(__p0);
	62558	__ret = __builtin_bit_cast(uint16x8_t, __p0);
60585	62559	return __ret;
60586	62560	}
60587	62561	__ai __attribute__((target("neon"))) uint16x8_t vreinterpretq_u16_mf8(mfloat8x16_t __p0) {
60588	62562	uint16x8_t __ret;
60589		__ret = (uint16x8_t)(__p0);
	62563	__ret = __builtin_bit_cast(uint16x8_t, __p0);
60590	62564	return __ret;
60591	62565	}
60592	62566	__ai __attribute__((target("neon"))) uint16x8_t vreinterpretq_u16_s16(int16x8_t __p0) {
60593	62567	uint16x8_t __ret;
60594		__ret = (uint16x8_t)(__p0);
	62568	__ret = __builtin_bit_cast(uint16x8_t, __p0);
60595	62569	return __ret;
60596	62570	}
60597	62571	__ai __attribute__((target("neon"))) int8x16_t vreinterpretq_s8_p8(poly8x16_t __p0) {
60598	62572	int8x16_t __ret;
60599		__ret = (int8x16_t)(__p0);
	62573	__ret = __builtin_bit_cast(int8x16_t, __p0);
60600	62574	return __ret;
60601	62575	}
60602	62576	__ai __attribute__((target("neon"))) int8x16_t vreinterpretq_s8_p128(poly128_t __p0) {
60603	62577	int8x16_t __ret;
60604		__ret = (int8x16_t)(__p0);
	62578	__ret = __builtin_bit_cast(int8x16_t, __p0);
60605	62579	return __ret;
60606	62580	}
60607	62581	__ai __attribute__((target("neon"))) int8x16_t vreinterpretq_s8_p64(poly64x2_t __p0) {
60608	62582	int8x16_t __ret;
60609		__ret = (int8x16_t)(__p0);
	62583	__ret = __builtin_bit_cast(int8x16_t, __p0);
60610	62584	return __ret;
60611	62585	}
60612	62586	__ai __attribute__((target("neon"))) int8x16_t vreinterpretq_s8_p16(poly16x8_t __p0) {
60613	62587	int8x16_t __ret;
60614		__ret = (int8x16_t)(__p0);
	62588	__ret = __builtin_bit_cast(int8x16_t, __p0);
60615	62589	return __ret;
60616	62590	}
60617	62591	__ai __attribute__((target("neon"))) int8x16_t vreinterpretq_s8_u8(uint8x16_t __p0) {
60618	62592	int8x16_t __ret;
60619		__ret = (int8x16_t)(__p0);
	62593	__ret = __builtin_bit_cast(int8x16_t, __p0);
60620	62594	return __ret;
60621	62595	}
60622	62596	__ai __attribute__((target("neon"))) int8x16_t vreinterpretq_s8_u32(uint32x4_t __p0) {
60623	62597	int8x16_t __ret;
60624		__ret = (int8x16_t)(__p0);
	62598	__ret = __builtin_bit_cast(int8x16_t, __p0);
60625	62599	return __ret;
60626	62600	}
60627	62601	__ai __attribute__((target("neon"))) int8x16_t vreinterpretq_s8_u64(uint64x2_t __p0) {
60628	62602	int8x16_t __ret;
60629		__ret = (int8x16_t)(__p0);
	62603	__ret = __builtin_bit_cast(int8x16_t, __p0);
60630	62604	return __ret;
60631	62605	}
60632	62606	__ai __attribute__((target("neon"))) int8x16_t vreinterpretq_s8_u16(uint16x8_t __p0) {
60633	62607	int8x16_t __ret;
60634		__ret = (int8x16_t)(__p0);
	62608	__ret = __builtin_bit_cast(int8x16_t, __p0);
60635	62609	return __ret;
60636	62610	}
60637	62611	__ai __attribute__((target("neon"))) int8x16_t vreinterpretq_s8_f64(float64x2_t __p0) {
60638	62612	int8x16_t __ret;
60639		__ret = (int8x16_t)(__p0);
	62613	__ret = __builtin_bit_cast(int8x16_t, __p0);
60640	62614	return __ret;
60641	62615	}
60642	62616	__ai __attribute__((target("neon"))) int8x16_t vreinterpretq_s8_f32(float32x4_t __p0) {
60643	62617	int8x16_t __ret;
60644		__ret = (int8x16_t)(__p0);
	62618	__ret = __builtin_bit_cast(int8x16_t, __p0);
60645	62619	return __ret;
60646	62620	}
60647	62621	__ai __attribute__((target("neon"))) int8x16_t vreinterpretq_s8_f16(float16x8_t __p0) {
60648	62622	int8x16_t __ret;
60649		__ret = (int8x16_t)(__p0);
	62623	__ret = __builtin_bit_cast(int8x16_t, __p0);
60650	62624	return __ret;
60651	62625	}
60652	62626	__ai __attribute__((target("neon"))) int8x16_t vreinterpretq_s8_s32(int32x4_t __p0) {
60653	62627	int8x16_t __ret;
60654		__ret = (int8x16_t)(__p0);
	62628	__ret = __builtin_bit_cast(int8x16_t, __p0);
60655	62629	return __ret;
60656	62630	}
60657	62631	__ai __attribute__((target("neon"))) int8x16_t vreinterpretq_s8_s64(int64x2_t __p0) {
60658	62632	int8x16_t __ret;
60659		__ret = (int8x16_t)(__p0);
	62633	__ret = __builtin_bit_cast(int8x16_t, __p0);
60660	62634	return __ret;
60661	62635	}
60662	62636	__ai __attribute__((target("neon"))) int8x16_t vreinterpretq_s8_mf8(mfloat8x16_t __p0) {
60663	62637	int8x16_t __ret;
60664		__ret = (int8x16_t)(__p0);
	62638	__ret = __builtin_bit_cast(int8x16_t, __p0);
60665	62639	return __ret;
60666	62640	}
60667	62641	__ai __attribute__((target("neon"))) int8x16_t vreinterpretq_s8_s16(int16x8_t __p0) {
60668	62642	int8x16_t __ret;
60669		__ret = (int8x16_t)(__p0);
	62643	__ret = __builtin_bit_cast(int8x16_t, __p0);
60670	62644	return __ret;
60671	62645	}
60672	62646	__ai __attribute__((target("neon"))) float64x2_t vreinterpretq_f64_p8(poly8x16_t __p0) {
60673	62647	float64x2_t __ret;
60674		__ret = (float64x2_t)(__p0);
	62648	__ret = __builtin_bit_cast(float64x2_t, __p0);
60675	62649	return __ret;
60676	62650	}
60677	62651	__ai __attribute__((target("neon"))) float64x2_t vreinterpretq_f64_p128(poly128_t __p0) {
60678	62652	float64x2_t __ret;
60679		__ret = (float64x2_t)(__p0);
	62653	__ret = __builtin_bit_cast(float64x2_t, __p0);
60680	62654	return __ret;
60681	62655	}
60682	62656	__ai __attribute__((target("neon"))) float64x2_t vreinterpretq_f64_p64(poly64x2_t __p0) {
60683	62657	float64x2_t __ret;
60684		__ret = (float64x2_t)(__p0);
	62658	__ret = __builtin_bit_cast(float64x2_t, __p0);
60685	62659	return __ret;
60686	62660	}
60687	62661	__ai __attribute__((target("neon"))) float64x2_t vreinterpretq_f64_p16(poly16x8_t __p0) {
60688	62662	float64x2_t __ret;
60689		__ret = (float64x2_t)(__p0);
	62663	__ret = __builtin_bit_cast(float64x2_t, __p0);
60690	62664	return __ret;
60691	62665	}
60692	62666	__ai __attribute__((target("neon"))) float64x2_t vreinterpretq_f64_u8(uint8x16_t __p0) {
60693	62667	float64x2_t __ret;
60694		__ret = (float64x2_t)(__p0);
	62668	__ret = __builtin_bit_cast(float64x2_t, __p0);
60695	62669	return __ret;
60696	62670	}
60697	62671	__ai __attribute__((target("neon"))) float64x2_t vreinterpretq_f64_u32(uint32x4_t __p0) {
60698	62672	float64x2_t __ret;
60699		__ret = (float64x2_t)(__p0);
	62673	__ret = __builtin_bit_cast(float64x2_t, __p0);
60700	62674	return __ret;
60701	62675	}
60702	62676	__ai __attribute__((target("neon"))) float64x2_t vreinterpretq_f64_u64(uint64x2_t __p0) {
60703	62677	float64x2_t __ret;
60704		__ret = (float64x2_t)(__p0);
	62678	__ret = __builtin_bit_cast(float64x2_t, __p0);
60705	62679	return __ret;
60706	62680	}
60707	62681	__ai __attribute__((target("neon"))) float64x2_t vreinterpretq_f64_u16(uint16x8_t __p0) {
60708	62682	float64x2_t __ret;
60709		__ret = (float64x2_t)(__p0);
	62683	__ret = __builtin_bit_cast(float64x2_t, __p0);
60710	62684	return __ret;
60711	62685	}
60712	62686	__ai __attribute__((target("neon"))) float64x2_t vreinterpretq_f64_s8(int8x16_t __p0) {
60713	62687	float64x2_t __ret;
60714		__ret = (float64x2_t)(__p0);
	62688	__ret = __builtin_bit_cast(float64x2_t, __p0);
60715	62689	return __ret;
60716	62690	}
60717	62691	__ai __attribute__((target("neon"))) float64x2_t vreinterpretq_f64_f32(float32x4_t __p0) {
60718	62692	float64x2_t __ret;
60719		__ret = (float64x2_t)(__p0);
	62693	__ret = __builtin_bit_cast(float64x2_t, __p0);
60720	62694	return __ret;
60721	62695	}
60722	62696	__ai __attribute__((target("neon"))) float64x2_t vreinterpretq_f64_f16(float16x8_t __p0) {
60723	62697	float64x2_t __ret;
60724		__ret = (float64x2_t)(__p0);
	62698	__ret = __builtin_bit_cast(float64x2_t, __p0);
60725	62699	return __ret;
60726	62700	}
60727	62701	__ai __attribute__((target("neon"))) float64x2_t vreinterpretq_f64_s32(int32x4_t __p0) {
60728	62702	float64x2_t __ret;
60729		__ret = (float64x2_t)(__p0);
	62703	__ret = __builtin_bit_cast(float64x2_t, __p0);
60730	62704	return __ret;
60731	62705	}
60732	62706	__ai __attribute__((target("neon"))) float64x2_t vreinterpretq_f64_s64(int64x2_t __p0) {
60733	62707	float64x2_t __ret;
60734		__ret = (float64x2_t)(__p0);
	62708	__ret = __builtin_bit_cast(float64x2_t, __p0);
60735	62709	return __ret;
60736	62710	}
60737	62711	__ai __attribute__((target("neon"))) float64x2_t vreinterpretq_f64_mf8(mfloat8x16_t __p0) {
60738	62712	float64x2_t __ret;
60739		__ret = (float64x2_t)(__p0);
	62713	__ret = __builtin_bit_cast(float64x2_t, __p0);
60740	62714	return __ret;
60741	62715	}
60742	62716	__ai __attribute__((target("neon"))) float64x2_t vreinterpretq_f64_s16(int16x8_t __p0) {
60743	62717	float64x2_t __ret;
60744		__ret = (float64x2_t)(__p0);
	62718	__ret = __builtin_bit_cast(float64x2_t, __p0);
60745	62719	return __ret;
60746	62720	}
60747	62721	__ai __attribute__((target("neon"))) float32x4_t vreinterpretq_f32_p8(poly8x16_t __p0) {
60748	62722	float32x4_t __ret;
60749		__ret = (float32x4_t)(__p0);
	62723	__ret = __builtin_bit_cast(float32x4_t, __p0);
60750	62724	return __ret;
60751	62725	}
60752	62726	__ai __attribute__((target("neon"))) float32x4_t vreinterpretq_f32_p128(poly128_t __p0) {
60753	62727	float32x4_t __ret;
60754		__ret = (float32x4_t)(__p0);
	62728	__ret = __builtin_bit_cast(float32x4_t, __p0);
60755	62729	return __ret;
60756	62730	}
60757	62731	__ai __attribute__((target("neon"))) float32x4_t vreinterpretq_f32_p64(poly64x2_t __p0) {
60758	62732	float32x4_t __ret;
60759		__ret = (float32x4_t)(__p0);
	62733	__ret = __builtin_bit_cast(float32x4_t, __p0);
60760	62734	return __ret;
60761	62735	}
60762	62736	__ai __attribute__((target("neon"))) float32x4_t vreinterpretq_f32_p16(poly16x8_t __p0) {
60763	62737	float32x4_t __ret;
60764		__ret = (float32x4_t)(__p0);
	62738	__ret = __builtin_bit_cast(float32x4_t, __p0);
60765	62739	return __ret;
60766	62740	}
60767	62741	__ai __attribute__((target("neon"))) float32x4_t vreinterpretq_f32_u8(uint8x16_t __p0) {
60768	62742	float32x4_t __ret;
60769		__ret = (float32x4_t)(__p0);
	62743	__ret = __builtin_bit_cast(float32x4_t, __p0);
60770	62744	return __ret;
60771	62745	}
60772	62746	__ai __attribute__((target("neon"))) float32x4_t vreinterpretq_f32_u32(uint32x4_t __p0) {
60773	62747	float32x4_t __ret;
60774		__ret = (float32x4_t)(__p0);
	62748	__ret = __builtin_bit_cast(float32x4_t, __p0);
60775	62749	return __ret;
60776	62750	}
60777	62751	__ai __attribute__((target("neon"))) float32x4_t vreinterpretq_f32_u64(uint64x2_t __p0) {
60778	62752	float32x4_t __ret;
60779		__ret = (float32x4_t)(__p0);
	62753	__ret = __builtin_bit_cast(float32x4_t, __p0);
60780	62754	return __ret;
60781	62755	}
60782	62756	__ai __attribute__((target("neon"))) float32x4_t vreinterpretq_f32_u16(uint16x8_t __p0) {
60783	62757	float32x4_t __ret;
60784		__ret = (float32x4_t)(__p0);
	62758	__ret = __builtin_bit_cast(float32x4_t, __p0);
60785	62759	return __ret;
60786	62760	}
60787	62761	__ai __attribute__((target("neon"))) float32x4_t vreinterpretq_f32_s8(int8x16_t __p0) {
60788	62762	float32x4_t __ret;
60789		__ret = (float32x4_t)(__p0);
	62763	__ret = __builtin_bit_cast(float32x4_t, __p0);
60790	62764	return __ret;
60791	62765	}
60792	62766	__ai __attribute__((target("neon"))) float32x4_t vreinterpretq_f32_f64(float64x2_t __p0) {
60793	62767	float32x4_t __ret;
60794		__ret = (float32x4_t)(__p0);
	62768	__ret = __builtin_bit_cast(float32x4_t, __p0);
60795	62769	return __ret;
60796	62770	}
60797	62771	__ai __attribute__((target("neon"))) float32x4_t vreinterpretq_f32_f16(float16x8_t __p0) {
60798	62772	float32x4_t __ret;
60799		__ret = (float32x4_t)(__p0);
	62773	__ret = __builtin_bit_cast(float32x4_t, __p0);
60800	62774	return __ret;
60801	62775	}
60802	62776	__ai __attribute__((target("neon"))) float32x4_t vreinterpretq_f32_s32(int32x4_t __p0) {
60803	62777	float32x4_t __ret;
60804		__ret = (float32x4_t)(__p0);
	62778	__ret = __builtin_bit_cast(float32x4_t, __p0);
60805	62779	return __ret;
60806	62780	}
60807	62781	__ai __attribute__((target("neon"))) float32x4_t vreinterpretq_f32_s64(int64x2_t __p0) {
60808	62782	float32x4_t __ret;
60809		__ret = (float32x4_t)(__p0);
	62783	__ret = __builtin_bit_cast(float32x4_t, __p0);
60810	62784	return __ret;
60811	62785	}
60812	62786	__ai __attribute__((target("neon"))) float32x4_t vreinterpretq_f32_mf8(mfloat8x16_t __p0) {
60813	62787	float32x4_t __ret;
60814		__ret = (float32x4_t)(__p0);
	62788	__ret = __builtin_bit_cast(float32x4_t, __p0);
60815	62789	return __ret;
60816	62790	}
60817	62791	__ai __attribute__((target("neon"))) float32x4_t vreinterpretq_f32_s16(int16x8_t __p0) {
60818	62792	float32x4_t __ret;
60819		__ret = (float32x4_t)(__p0);
	62793	__ret = __builtin_bit_cast(float32x4_t, __p0);
60820	62794	return __ret;
60821	62795	}
60822	62796	__ai __attribute__((target("neon"))) float16x8_t vreinterpretq_f16_p8(poly8x16_t __p0) {
60823	62797	float16x8_t __ret;
60824		__ret = (float16x8_t)(__p0);
	62798	__ret = __builtin_bit_cast(float16x8_t, __p0);
60825	62799	return __ret;
60826	62800	}
60827	62801	__ai __attribute__((target("neon"))) float16x8_t vreinterpretq_f16_p128(poly128_t __p0) {
60828	62802	float16x8_t __ret;
60829		__ret = (float16x8_t)(__p0);
	62803	__ret = __builtin_bit_cast(float16x8_t, __p0);
60830	62804	return __ret;
60831	62805	}
60832	62806	__ai __attribute__((target("neon"))) float16x8_t vreinterpretq_f16_p64(poly64x2_t __p0) {
60833	62807	float16x8_t __ret;
60834		__ret = (float16x8_t)(__p0);
	62808	__ret = __builtin_bit_cast(float16x8_t, __p0);
60835	62809	return __ret;
60836	62810	}
60837	62811	__ai __attribute__((target("neon"))) float16x8_t vreinterpretq_f16_p16(poly16x8_t __p0) {
60838	62812	float16x8_t __ret;
60839		__ret = (float16x8_t)(__p0);
	62813	__ret = __builtin_bit_cast(float16x8_t, __p0);
60840	62814	return __ret;
60841	62815	}
60842	62816	__ai __attribute__((target("neon"))) float16x8_t vreinterpretq_f16_u8(uint8x16_t __p0) {
60843	62817	float16x8_t __ret;
60844		__ret = (float16x8_t)(__p0);
	62818	__ret = __builtin_bit_cast(float16x8_t, __p0);
60845	62819	return __ret;
60846	62820	}
60847	62821	__ai __attribute__((target("neon"))) float16x8_t vreinterpretq_f16_u32(uint32x4_t __p0) {
60848	62822	float16x8_t __ret;
60849		__ret = (float16x8_t)(__p0);
	62823	__ret = __builtin_bit_cast(float16x8_t, __p0);
60850	62824	return __ret;
60851	62825	}
60852	62826	__ai __attribute__((target("neon"))) float16x8_t vreinterpretq_f16_u64(uint64x2_t __p0) {
60853	62827	float16x8_t __ret;
60854		__ret = (float16x8_t)(__p0);
	62828	__ret = __builtin_bit_cast(float16x8_t, __p0);
60855	62829	return __ret;
60856	62830	}
60857	62831	__ai __attribute__((target("neon"))) float16x8_t vreinterpretq_f16_u16(uint16x8_t __p0) {
60858	62832	float16x8_t __ret;
60859		__ret = (float16x8_t)(__p0);
	62833	__ret = __builtin_bit_cast(float16x8_t, __p0);
60860	62834	return __ret;
60861	62835	}
60862	62836	__ai __attribute__((target("neon"))) float16x8_t vreinterpretq_f16_s8(int8x16_t __p0) {
60863	62837	float16x8_t __ret;
60864		__ret = (float16x8_t)(__p0);
	62838	__ret = __builtin_bit_cast(float16x8_t, __p0);
60865	62839	return __ret;
60866	62840	}
60867	62841	__ai __attribute__((target("neon"))) float16x8_t vreinterpretq_f16_f64(float64x2_t __p0) {
60868	62842	float16x8_t __ret;
60869		__ret = (float16x8_t)(__p0);
	62843	__ret = __builtin_bit_cast(float16x8_t, __p0);
60870	62844	return __ret;
60871	62845	}
60872	62846	__ai __attribute__((target("neon"))) float16x8_t vreinterpretq_f16_f32(float32x4_t __p0) {
60873	62847	float16x8_t __ret;
60874		__ret = (float16x8_t)(__p0);
	62848	__ret = __builtin_bit_cast(float16x8_t, __p0);
60875	62849	return __ret;
60876	62850	}
60877	62851	__ai __attribute__((target("neon"))) float16x8_t vreinterpretq_f16_s32(int32x4_t __p0) {
60878	62852	float16x8_t __ret;
60879		__ret = (float16x8_t)(__p0);
	62853	__ret = __builtin_bit_cast(float16x8_t, __p0);
60880	62854	return __ret;
60881	62855	}
60882	62856	__ai __attribute__((target("neon"))) float16x8_t vreinterpretq_f16_s64(int64x2_t __p0) {
60883	62857	float16x8_t __ret;
60884		__ret = (float16x8_t)(__p0);
	62858	__ret = __builtin_bit_cast(float16x8_t, __p0);
60885	62859	return __ret;
60886	62860	}
60887	62861	__ai __attribute__((target("neon"))) float16x8_t vreinterpretq_f16_mf8(mfloat8x16_t __p0) {
60888	62862	float16x8_t __ret;
60889		__ret = (float16x8_t)(__p0);
	62863	__ret = __builtin_bit_cast(float16x8_t, __p0);
60890	62864	return __ret;
60891	62865	}
60892	62866	__ai __attribute__((target("neon"))) float16x8_t vreinterpretq_f16_s16(int16x8_t __p0) {
60893	62867	float16x8_t __ret;
60894		__ret = (float16x8_t)(__p0);
	62868	__ret = __builtin_bit_cast(float16x8_t, __p0);
60895	62869	return __ret;
60896	62870	}
60897	62871	__ai __attribute__((target("neon"))) int32x4_t vreinterpretq_s32_p8(poly8x16_t __p0) {
60898	62872	int32x4_t __ret;
60899		__ret = (int32x4_t)(__p0);
	62873	__ret = __builtin_bit_cast(int32x4_t, __p0);
60900	62874	return __ret;
60901	62875	}
60902	62876	__ai __attribute__((target("neon"))) int32x4_t vreinterpretq_s32_p128(poly128_t __p0) {
60903	62877	int32x4_t __ret;
60904		__ret = (int32x4_t)(__p0);
	62878	__ret = __builtin_bit_cast(int32x4_t, __p0);
60905	62879	return __ret;
60906	62880	}
60907	62881	__ai __attribute__((target("neon"))) int32x4_t vreinterpretq_s32_p64(poly64x2_t __p0) {
60908	62882	int32x4_t __ret;
60909		__ret = (int32x4_t)(__p0);
	62883	__ret = __builtin_bit_cast(int32x4_t, __p0);
60910	62884	return __ret;
60911	62885	}
60912	62886	__ai __attribute__((target("neon"))) int32x4_t vreinterpretq_s32_p16(poly16x8_t __p0) {
60913	62887	int32x4_t __ret;
60914		__ret = (int32x4_t)(__p0);
	62888	__ret = __builtin_bit_cast(int32x4_t, __p0);
60915	62889	return __ret;
60916	62890	}
60917	62891	__ai __attribute__((target("neon"))) int32x4_t vreinterpretq_s32_u8(uint8x16_t __p0) {
60918	62892	int32x4_t __ret;
60919		__ret = (int32x4_t)(__p0);
	62893	__ret = __builtin_bit_cast(int32x4_t, __p0);
60920	62894	return __ret;
60921	62895	}
60922	62896	__ai __attribute__((target("neon"))) int32x4_t vreinterpretq_s32_u32(uint32x4_t __p0) {
60923	62897	int32x4_t __ret;
60924		__ret = (int32x4_t)(__p0);
	62898	__ret = __builtin_bit_cast(int32x4_t, __p0);
60925	62899	return __ret;
60926	62900	}
60927	62901	__ai __attribute__((target("neon"))) int32x4_t vreinterpretq_s32_u64(uint64x2_t __p0) {
60928	62902	int32x4_t __ret;
60929		__ret = (int32x4_t)(__p0);
	62903	__ret = __builtin_bit_cast(int32x4_t, __p0);
60930	62904	return __ret;
60931	62905	}
60932	62906	__ai __attribute__((target("neon"))) int32x4_t vreinterpretq_s32_u16(uint16x8_t __p0) {
60933	62907	int32x4_t __ret;
60934		__ret = (int32x4_t)(__p0);
	62908	__ret = __builtin_bit_cast(int32x4_t, __p0);
60935	62909	return __ret;
60936	62910	}
60937	62911	__ai __attribute__((target("neon"))) int32x4_t vreinterpretq_s32_s8(int8x16_t __p0) {
60938	62912	int32x4_t __ret;
60939		__ret = (int32x4_t)(__p0);
	62913	__ret = __builtin_bit_cast(int32x4_t, __p0);
60940	62914	return __ret;
60941	62915	}
60942	62916	__ai __attribute__((target("neon"))) int32x4_t vreinterpretq_s32_f64(float64x2_t __p0) {
60943	62917	int32x4_t __ret;
60944		__ret = (int32x4_t)(__p0);
	62918	__ret = __builtin_bit_cast(int32x4_t, __p0);
60945	62919	return __ret;
60946	62920	}
60947	62921	__ai __attribute__((target("neon"))) int32x4_t vreinterpretq_s32_f32(float32x4_t __p0) {
60948	62922	int32x4_t __ret;
60949		__ret = (int32x4_t)(__p0);
	62923	__ret = __builtin_bit_cast(int32x4_t, __p0);
60950	62924	return __ret;
60951	62925	}
60952	62926	__ai __attribute__((target("neon"))) int32x4_t vreinterpretq_s32_f16(float16x8_t __p0) {
60953	62927	int32x4_t __ret;
60954		__ret = (int32x4_t)(__p0);
	62928	__ret = __builtin_bit_cast(int32x4_t, __p0);
60955	62929	return __ret;
60956	62930	}
60957	62931	__ai __attribute__((target("neon"))) int32x4_t vreinterpretq_s32_s64(int64x2_t __p0) {
60958	62932	int32x4_t __ret;
60959		__ret = (int32x4_t)(__p0);
	62933	__ret = __builtin_bit_cast(int32x4_t, __p0);
60960	62934	return __ret;
60961	62935	}
60962	62936	__ai __attribute__((target("neon"))) int32x4_t vreinterpretq_s32_mf8(mfloat8x16_t __p0) {
60963	62937	int32x4_t __ret;
60964		__ret = (int32x4_t)(__p0);
	62938	__ret = __builtin_bit_cast(int32x4_t, __p0);
60965	62939	return __ret;
60966	62940	}
60967	62941	__ai __attribute__((target("neon"))) int32x4_t vreinterpretq_s32_s16(int16x8_t __p0) {
60968	62942	int32x4_t __ret;
60969		__ret = (int32x4_t)(__p0);
	62943	__ret = __builtin_bit_cast(int32x4_t, __p0);
60970	62944	return __ret;
60971	62945	}
60972	62946	__ai __attribute__((target("neon"))) int64x2_t vreinterpretq_s64_p8(poly8x16_t __p0) {
60973	62947	int64x2_t __ret;
60974		__ret = (int64x2_t)(__p0);
	62948	__ret = __builtin_bit_cast(int64x2_t, __p0);
60975	62949	return __ret;
60976	62950	}
60977	62951	__ai __attribute__((target("neon"))) int64x2_t vreinterpretq_s64_p128(poly128_t __p0) {
60978	62952	int64x2_t __ret;
60979		__ret = (int64x2_t)(__p0);
	62953	__ret = __builtin_bit_cast(int64x2_t, __p0);
60980	62954	return __ret;
60981	62955	}
60982	62956	__ai __attribute__((target("neon"))) int64x2_t vreinterpretq_s64_p64(poly64x2_t __p0) {
60983	62957	int64x2_t __ret;
60984		__ret = (int64x2_t)(__p0);
	62958	__ret = __builtin_bit_cast(int64x2_t, __p0);
60985	62959	return __ret;
60986	62960	}
60987	62961	__ai __attribute__((target("neon"))) int64x2_t vreinterpretq_s64_p16(poly16x8_t __p0) {
60988	62962	int64x2_t __ret;
60989		__ret = (int64x2_t)(__p0);
	62963	__ret = __builtin_bit_cast(int64x2_t, __p0);
60990	62964	return __ret;
60991	62965	}
60992	62966	__ai __attribute__((target("neon"))) int64x2_t vreinterpretq_s64_u8(uint8x16_t __p0) {
60993	62967	int64x2_t __ret;
60994		__ret = (int64x2_t)(__p0);
	62968	__ret = __builtin_bit_cast(int64x2_t, __p0);
60995	62969	return __ret;
60996	62970	}
60997	62971	__ai __attribute__((target("neon"))) int64x2_t vreinterpretq_s64_u32(uint32x4_t __p0) {
60998	62972	int64x2_t __ret;
60999		__ret = (int64x2_t)(__p0);
	62973	__ret = __builtin_bit_cast(int64x2_t, __p0);
61000	62974	return __ret;
61001	62975	}
61002	62976	__ai __attribute__((target("neon"))) int64x2_t vreinterpretq_s64_u64(uint64x2_t __p0) {
61003	62977	int64x2_t __ret;
61004		__ret = (int64x2_t)(__p0);
	62978	__ret = __builtin_bit_cast(int64x2_t, __p0);
61005	62979	return __ret;
61006	62980	}
61007	62981	__ai __attribute__((target("neon"))) int64x2_t vreinterpretq_s64_u16(uint16x8_t __p0) {
61008	62982	int64x2_t __ret;
61009		__ret = (int64x2_t)(__p0);
	62983	__ret = __builtin_bit_cast(int64x2_t, __p0);
61010	62984	return __ret;
61011	62985	}
61012	62986	__ai __attribute__((target("neon"))) int64x2_t vreinterpretq_s64_s8(int8x16_t __p0) {
61013	62987	int64x2_t __ret;
61014		__ret = (int64x2_t)(__p0);
	62988	__ret = __builtin_bit_cast(int64x2_t, __p0);
61015	62989	return __ret;
61016	62990	}
61017	62991	__ai __attribute__((target("neon"))) int64x2_t vreinterpretq_s64_f64(float64x2_t __p0) {
61018	62992	int64x2_t __ret;
61019		__ret = (int64x2_t)(__p0);
	62993	__ret = __builtin_bit_cast(int64x2_t, __p0);
61020	62994	return __ret;
61021	62995	}
61022	62996	__ai __attribute__((target("neon"))) int64x2_t vreinterpretq_s64_f32(float32x4_t __p0) {
61023	62997	int64x2_t __ret;
61024		__ret = (int64x2_t)(__p0);
	62998	__ret = __builtin_bit_cast(int64x2_t, __p0);
61025	62999	return __ret;
61026	63000	}
61027	63001	__ai __attribute__((target("neon"))) int64x2_t vreinterpretq_s64_f16(float16x8_t __p0) {
61028	63002	int64x2_t __ret;
61029		__ret = (int64x2_t)(__p0);
	63003	__ret = __builtin_bit_cast(int64x2_t, __p0);
61030	63004	return __ret;
61031	63005	}
61032	63006	__ai __attribute__((target("neon"))) int64x2_t vreinterpretq_s64_s32(int32x4_t __p0) {
61033	63007	int64x2_t __ret;
61034		__ret = (int64x2_t)(__p0);
	63008	__ret = __builtin_bit_cast(int64x2_t, __p0);
61035	63009	return __ret;
61036	63010	}
61037	63011	__ai __attribute__((target("neon"))) int64x2_t vreinterpretq_s64_mf8(mfloat8x16_t __p0) {
61038	63012	int64x2_t __ret;
61039		__ret = (int64x2_t)(__p0);
	63013	__ret = __builtin_bit_cast(int64x2_t, __p0);
61040	63014	return __ret;
61041	63015	}
61042	63016	__ai __attribute__((target("neon"))) int64x2_t vreinterpretq_s64_s16(int16x8_t __p0) {
61043	63017	int64x2_t __ret;
61044		__ret = (int64x2_t)(__p0);
	63018	__ret = __builtin_bit_cast(int64x2_t, __p0);
61045	63019	return __ret;
61046	63020	}
61047	63021	__ai __attribute__((target("neon"))) mfloat8x16_t vreinterpretq_mf8_p8(poly8x16_t __p0) {
61048	63022	mfloat8x16_t __ret;
61049		__ret = (mfloat8x16_t)(__p0);
	63023	__ret = __builtin_bit_cast(mfloat8x16_t, __p0);
61050	63024	return __ret;
61051	63025	}
61052	63026	__ai __attribute__((target("neon"))) mfloat8x16_t vreinterpretq_mf8_p128(poly128_t __p0) {
61053	63027	mfloat8x16_t __ret;
61054		__ret = (mfloat8x16_t)(__p0);
	63028	__ret = __builtin_bit_cast(mfloat8x16_t, __p0);
61055	63029	return __ret;
61056	63030	}
61057	63031	__ai __attribute__((target("neon"))) mfloat8x16_t vreinterpretq_mf8_p64(poly64x2_t __p0) {
61058	63032	mfloat8x16_t __ret;
61059		__ret = (mfloat8x16_t)(__p0);
	63033	__ret = __builtin_bit_cast(mfloat8x16_t, __p0);
61060	63034	return __ret;
61061	63035	}
61062	63036	__ai __attribute__((target("neon"))) mfloat8x16_t vreinterpretq_mf8_p16(poly16x8_t __p0) {
61063	63037	mfloat8x16_t __ret;
61064		__ret = (mfloat8x16_t)(__p0);
	63038	__ret = __builtin_bit_cast(mfloat8x16_t, __p0);
61065	63039	return __ret;
61066	63040	}
61067	63041	__ai __attribute__((target("neon"))) mfloat8x16_t vreinterpretq_mf8_u8(uint8x16_t __p0) {
61068	63042	mfloat8x16_t __ret;
61069		__ret = (mfloat8x16_t)(__p0);
	63043	__ret = __builtin_bit_cast(mfloat8x16_t, __p0);
61070	63044	return __ret;
61071	63045	}
61072	63046	__ai __attribute__((target("neon"))) mfloat8x16_t vreinterpretq_mf8_u32(uint32x4_t __p0) {
61073	63047	mfloat8x16_t __ret;
61074		__ret = (mfloat8x16_t)(__p0);
	63048	__ret = __builtin_bit_cast(mfloat8x16_t, __p0);
61075	63049	return __ret;
61076	63050	}
61077	63051	__ai __attribute__((target("neon"))) mfloat8x16_t vreinterpretq_mf8_u64(uint64x2_t __p0) {
61078	63052	mfloat8x16_t __ret;
61079		__ret = (mfloat8x16_t)(__p0);
	63053	__ret = __builtin_bit_cast(mfloat8x16_t, __p0);
61080	63054	return __ret;
61081	63055	}
61082	63056	__ai __attribute__((target("neon"))) mfloat8x16_t vreinterpretq_mf8_u16(uint16x8_t __p0) {
61083	63057	mfloat8x16_t __ret;
61084		__ret = (mfloat8x16_t)(__p0);
	63058	__ret = __builtin_bit_cast(mfloat8x16_t, __p0);
61085	63059	return __ret;
61086	63060	}
61087	63061	__ai __attribute__((target("neon"))) mfloat8x16_t vreinterpretq_mf8_s8(int8x16_t __p0) {
61088	63062	mfloat8x16_t __ret;
61089		__ret = (mfloat8x16_t)(__p0);
	63063	__ret = __builtin_bit_cast(mfloat8x16_t, __p0);
61090	63064	return __ret;
61091	63065	}
61092	63066	__ai __attribute__((target("neon"))) mfloat8x16_t vreinterpretq_mf8_f64(float64x2_t __p0) {
61093	63067	mfloat8x16_t __ret;
61094		__ret = (mfloat8x16_t)(__p0);
	63068	__ret = __builtin_bit_cast(mfloat8x16_t, __p0);
61095	63069	return __ret;
61096	63070	}
61097	63071	__ai __attribute__((target("neon"))) mfloat8x16_t vreinterpretq_mf8_f32(float32x4_t __p0) {
61098	63072	mfloat8x16_t __ret;
61099		__ret = (mfloat8x16_t)(__p0);
	63073	__ret = __builtin_bit_cast(mfloat8x16_t, __p0);
61100	63074	return __ret;
61101	63075	}
61102	63076	__ai __attribute__((target("neon"))) mfloat8x16_t vreinterpretq_mf8_f16(float16x8_t __p0) {
61103	63077	mfloat8x16_t __ret;
61104		__ret = (mfloat8x16_t)(__p0);
	63078	__ret = __builtin_bit_cast(mfloat8x16_t, __p0);
61105	63079	return __ret;
61106	63080	}
61107	63081	__ai __attribute__((target("neon"))) mfloat8x16_t vreinterpretq_mf8_s32(int32x4_t __p0) {
61108	63082	mfloat8x16_t __ret;
61109		__ret = (mfloat8x16_t)(__p0);
	63083	__ret = __builtin_bit_cast(mfloat8x16_t, __p0);
61110	63084	return __ret;
61111	63085	}
61112	63086	__ai __attribute__((target("neon"))) mfloat8x16_t vreinterpretq_mf8_s64(int64x2_t __p0) {
61113	63087	mfloat8x16_t __ret;
61114		__ret = (mfloat8x16_t)(__p0);
	63088	__ret = __builtin_bit_cast(mfloat8x16_t, __p0);
61115	63089	return __ret;
61116	63090	}
61117	63091	__ai __attribute__((target("neon"))) mfloat8x16_t vreinterpretq_mf8_s16(int16x8_t __p0) {
61118	63092	mfloat8x16_t __ret;
61119		__ret = (mfloat8x16_t)(__p0);
	63093	__ret = __builtin_bit_cast(mfloat8x16_t, __p0);
61120	63094	return __ret;
61121	63095	}
61122	63096	__ai __attribute__((target("neon"))) int16x8_t vreinterpretq_s16_p8(poly8x16_t __p0) {
61123	63097	int16x8_t __ret;
61124		__ret = (int16x8_t)(__p0);
	63098	__ret = __builtin_bit_cast(int16x8_t, __p0);
61125	63099	return __ret;
61126	63100	}
61127	63101	__ai __attribute__((target("neon"))) int16x8_t vreinterpretq_s16_p128(poly128_t __p0) {
61128	63102	int16x8_t __ret;
61129		__ret = (int16x8_t)(__p0);
	63103	__ret = __builtin_bit_cast(int16x8_t, __p0);
61130	63104	return __ret;
61131	63105	}
61132	63106	__ai __attribute__((target("neon"))) int16x8_t vreinterpretq_s16_p64(poly64x2_t __p0) {
61133	63107	int16x8_t __ret;
61134		__ret = (int16x8_t)(__p0);
	63108	__ret = __builtin_bit_cast(int16x8_t, __p0);
61135	63109	return __ret;
61136	63110	}
61137	63111	__ai __attribute__((target("neon"))) int16x8_t vreinterpretq_s16_p16(poly16x8_t __p0) {
61138	63112	int16x8_t __ret;
61139		__ret = (int16x8_t)(__p0);
	63113	__ret = __builtin_bit_cast(int16x8_t, __p0);
61140	63114	return __ret;
61141	63115	}
61142	63116	__ai __attribute__((target("neon"))) int16x8_t vreinterpretq_s16_u8(uint8x16_t __p0) {
61143	63117	int16x8_t __ret;
61144		__ret = (int16x8_t)(__p0);
	63118	__ret = __builtin_bit_cast(int16x8_t, __p0);
61145	63119	return __ret;
61146	63120	}
61147	63121	__ai __attribute__((target("neon"))) int16x8_t vreinterpretq_s16_u32(uint32x4_t __p0) {
61148	63122	int16x8_t __ret;
61149		__ret = (int16x8_t)(__p0);
	63123	__ret = __builtin_bit_cast(int16x8_t, __p0);
61150	63124	return __ret;
61151	63125	}
61152	63126	__ai __attribute__((target("neon"))) int16x8_t vreinterpretq_s16_u64(uint64x2_t __p0) {
61153	63127	int16x8_t __ret;
61154		__ret = (int16x8_t)(__p0);
	63128	__ret = __builtin_bit_cast(int16x8_t, __p0);
61155	63129	return __ret;
61156	63130	}
61157	63131	__ai __attribute__((target("neon"))) int16x8_t vreinterpretq_s16_u16(uint16x8_t __p0) {
61158	63132	int16x8_t __ret;
61159		__ret = (int16x8_t)(__p0);
	63133	__ret = __builtin_bit_cast(int16x8_t, __p0);
61160	63134	return __ret;
61161	63135	}
61162	63136	__ai __attribute__((target("neon"))) int16x8_t vreinterpretq_s16_s8(int8x16_t __p0) {
61163	63137	int16x8_t __ret;
61164		__ret = (int16x8_t)(__p0);
	63138	__ret = __builtin_bit_cast(int16x8_t, __p0);
61165	63139	return __ret;
61166	63140	}
61167	63141	__ai __attribute__((target("neon"))) int16x8_t vreinterpretq_s16_f64(float64x2_t __p0) {
61168	63142	int16x8_t __ret;
61169		__ret = (int16x8_t)(__p0);
	63143	__ret = __builtin_bit_cast(int16x8_t, __p0);
61170	63144	return __ret;
61171	63145	}
61172	63146	__ai __attribute__((target("neon"))) int16x8_t vreinterpretq_s16_f32(float32x4_t __p0) {
61173	63147	int16x8_t __ret;
61174		__ret = (int16x8_t)(__p0);
	63148	__ret = __builtin_bit_cast(int16x8_t, __p0);
61175	63149	return __ret;
61176	63150	}
61177	63151	__ai __attribute__((target("neon"))) int16x8_t vreinterpretq_s16_f16(float16x8_t __p0) {
61178	63152	int16x8_t __ret;
61179		__ret = (int16x8_t)(__p0);
	63153	__ret = __builtin_bit_cast(int16x8_t, __p0);
61180	63154	return __ret;
61181	63155	}
61182	63156	__ai __attribute__((target("neon"))) int16x8_t vreinterpretq_s16_s32(int32x4_t __p0) {
61183	63157	int16x8_t __ret;
61184		__ret = (int16x8_t)(__p0);
	63158	__ret = __builtin_bit_cast(int16x8_t, __p0);
61185	63159	return __ret;
61186	63160	}
61187	63161	__ai __attribute__((target("neon"))) int16x8_t vreinterpretq_s16_s64(int64x2_t __p0) {
61188	63162	int16x8_t __ret;
61189		__ret = (int16x8_t)(__p0);
	63163	__ret = __builtin_bit_cast(int16x8_t, __p0);
61190	63164	return __ret;
61191	63165	}
61192	63166	__ai __attribute__((target("neon"))) int16x8_t vreinterpretq_s16_mf8(mfloat8x16_t __p0) {
61193	63167	int16x8_t __ret;
61194		__ret = (int16x8_t)(__p0);
	63168	__ret = __builtin_bit_cast(int16x8_t, __p0);
61195	63169	return __ret;
61196	63170	}
61197	63171	__ai __attribute__((target("neon"))) uint8x8_t vreinterpret_u8_p8(poly8x8_t __p0) {
61198	63172	uint8x8_t __ret;
61199		__ret = (uint8x8_t)(__p0);
	63173	__ret = __builtin_bit_cast(uint8x8_t, __p0);
61200	63174	return __ret;
61201	63175	}
61202	63176	__ai __attribute__((target("neon"))) uint8x8_t vreinterpret_u8_p64(poly64x1_t __p0) {
61203	63177	uint8x8_t __ret;
61204		__ret = (uint8x8_t)(__p0);
	63178	__ret = __builtin_bit_cast(uint8x8_t, __p0);
61205	63179	return __ret;
61206	63180	}
61207	63181	__ai __attribute__((target("neon"))) uint8x8_t vreinterpret_u8_p16(poly16x4_t __p0) {
61208	63182	uint8x8_t __ret;
61209		__ret = (uint8x8_t)(__p0);
	63183	__ret = __builtin_bit_cast(uint8x8_t, __p0);
61210	63184	return __ret;
61211	63185	}
61212	63186	__ai __attribute__((target("neon"))) uint8x8_t vreinterpret_u8_u32(uint32x2_t __p0) {
61213	63187	uint8x8_t __ret;
61214		__ret = (uint8x8_t)(__p0);
	63188	__ret = __builtin_bit_cast(uint8x8_t, __p0);
61215	63189	return __ret;
61216	63190	}
61217	63191	__ai __attribute__((target("neon"))) uint8x8_t vreinterpret_u8_u64(uint64x1_t __p0) {
61218	63192	uint8x8_t __ret;
61219		__ret = (uint8x8_t)(__p0);
	63193	__ret = __builtin_bit_cast(uint8x8_t, __p0);
61220	63194	return __ret;
61221	63195	}
61222	63196	__ai __attribute__((target("neon"))) uint8x8_t vreinterpret_u8_u16(uint16x4_t __p0) {
61223	63197	uint8x8_t __ret;
61224		__ret = (uint8x8_t)(__p0);
	63198	__ret = __builtin_bit_cast(uint8x8_t, __p0);
61225	63199	return __ret;
61226	63200	}
61227	63201	__ai __attribute__((target("neon"))) uint8x8_t vreinterpret_u8_s8(int8x8_t __p0) {
61228	63202	uint8x8_t __ret;
61229		__ret = (uint8x8_t)(__p0);
	63203	__ret = __builtin_bit_cast(uint8x8_t, __p0);
61230	63204	return __ret;
61231	63205	}
61232	63206	__ai __attribute__((target("neon"))) uint8x8_t vreinterpret_u8_f64(float64x1_t __p0) {
61233	63207	uint8x8_t __ret;
61234		__ret = (uint8x8_t)(__p0);
	63208	__ret = __builtin_bit_cast(uint8x8_t, __p0);
61235	63209	return __ret;
61236	63210	}
61237	63211	__ai __attribute__((target("neon"))) uint8x8_t vreinterpret_u8_f32(float32x2_t __p0) {
61238	63212	uint8x8_t __ret;
61239		__ret = (uint8x8_t)(__p0);
	63213	__ret = __builtin_bit_cast(uint8x8_t, __p0);
61240	63214	return __ret;
61241	63215	}
61242	63216	__ai __attribute__((target("neon"))) uint8x8_t vreinterpret_u8_f16(float16x4_t __p0) {
61243	63217	uint8x8_t __ret;
61244		__ret = (uint8x8_t)(__p0);
	63218	__ret = __builtin_bit_cast(uint8x8_t, __p0);
61245	63219	return __ret;
61246	63220	}
61247	63221	__ai __attribute__((target("neon"))) uint8x8_t vreinterpret_u8_s32(int32x2_t __p0) {
61248	63222	uint8x8_t __ret;
61249		__ret = (uint8x8_t)(__p0);
	63223	__ret = __builtin_bit_cast(uint8x8_t, __p0);
61250	63224	return __ret;
61251	63225	}
61252	63226	__ai __attribute__((target("neon"))) uint8x8_t vreinterpret_u8_s64(int64x1_t __p0) {
61253	63227	uint8x8_t __ret;
61254		__ret = (uint8x8_t)(__p0);
	63228	__ret = __builtin_bit_cast(uint8x8_t, __p0);
61255	63229	return __ret;
61256	63230	}
61257	63231	__ai __attribute__((target("neon"))) uint8x8_t vreinterpret_u8_mf8(mfloat8x8_t __p0) {
61258	63232	uint8x8_t __ret;
61259		__ret = (uint8x8_t)(__p0);
	63233	__ret = __builtin_bit_cast(uint8x8_t, __p0);
61260	63234	return __ret;
61261	63235	}
61262	63236	__ai __attribute__((target("neon"))) uint8x8_t vreinterpret_u8_s16(int16x4_t __p0) {
61263	63237	uint8x8_t __ret;
61264		__ret = (uint8x8_t)(__p0);
	63238	__ret = __builtin_bit_cast(uint8x8_t, __p0);
61265	63239	return __ret;
61266	63240	}
61267	63241	__ai __attribute__((target("neon"))) uint32x2_t vreinterpret_u32_p8(poly8x8_t __p0) {
61268	63242	uint32x2_t __ret;
61269		__ret = (uint32x2_t)(__p0);
	63243	__ret = __builtin_bit_cast(uint32x2_t, __p0);
61270	63244	return __ret;
61271	63245	}
61272	63246	__ai __attribute__((target("neon"))) uint32x2_t vreinterpret_u32_p64(poly64x1_t __p0) {
61273	63247	uint32x2_t __ret;
61274		__ret = (uint32x2_t)(__p0);
	63248	__ret = __builtin_bit_cast(uint32x2_t, __p0);
61275	63249	return __ret;
61276	63250	}
61277	63251	__ai __attribute__((target("neon"))) uint32x2_t vreinterpret_u32_p16(poly16x4_t __p0) {
61278	63252	uint32x2_t __ret;
61279		__ret = (uint32x2_t)(__p0);
	63253	__ret = __builtin_bit_cast(uint32x2_t, __p0);
61280	63254	return __ret;
61281	63255	}
61282	63256	__ai __attribute__((target("neon"))) uint32x2_t vreinterpret_u32_u8(uint8x8_t __p0) {
61283	63257	uint32x2_t __ret;
61284		__ret = (uint32x2_t)(__p0);
	63258	__ret = __builtin_bit_cast(uint32x2_t, __p0);
61285	63259	return __ret;
61286	63260	}
61287	63261	__ai __attribute__((target("neon"))) uint32x2_t vreinterpret_u32_u64(uint64x1_t __p0) {
61288	63262	uint32x2_t __ret;
61289		__ret = (uint32x2_t)(__p0);
	63263	__ret = __builtin_bit_cast(uint32x2_t, __p0);
61290	63264	return __ret;
61291	63265	}
61292	63266	__ai __attribute__((target("neon"))) uint32x2_t vreinterpret_u32_u16(uint16x4_t __p0) {
61293	63267	uint32x2_t __ret;
61294		__ret = (uint32x2_t)(__p0);
	63268	__ret = __builtin_bit_cast(uint32x2_t, __p0);
61295	63269	return __ret;
61296	63270	}
61297	63271	__ai __attribute__((target("neon"))) uint32x2_t vreinterpret_u32_s8(int8x8_t __p0) {
61298	63272	uint32x2_t __ret;
61299		__ret = (uint32x2_t)(__p0);
	63273	__ret = __builtin_bit_cast(uint32x2_t, __p0);
61300	63274	return __ret;
61301	63275	}
61302	63276	__ai __attribute__((target("neon"))) uint32x2_t vreinterpret_u32_f64(float64x1_t __p0) {
61303	63277	uint32x2_t __ret;
61304		__ret = (uint32x2_t)(__p0);
	63278	__ret = __builtin_bit_cast(uint32x2_t, __p0);
61305	63279	return __ret;
61306	63280	}
61307	63281	__ai __attribute__((target("neon"))) uint32x2_t vreinterpret_u32_f32(float32x2_t __p0) {
61308	63282	uint32x2_t __ret;
61309		__ret = (uint32x2_t)(__p0);
	63283	__ret = __builtin_bit_cast(uint32x2_t, __p0);
61310	63284	return __ret;
61311	63285	}
61312	63286	__ai __attribute__((target("neon"))) uint32x2_t vreinterpret_u32_f16(float16x4_t __p0) {
61313	63287	uint32x2_t __ret;
61314		__ret = (uint32x2_t)(__p0);
	63288	__ret = __builtin_bit_cast(uint32x2_t, __p0);
61315	63289	return __ret;
61316	63290	}
61317	63291	__ai __attribute__((target("neon"))) uint32x2_t vreinterpret_u32_s32(int32x2_t __p0) {
61318	63292	uint32x2_t __ret;
61319		__ret = (uint32x2_t)(__p0);
	63293	__ret = __builtin_bit_cast(uint32x2_t, __p0);
61320	63294	return __ret;
61321	63295	}
61322	63296	__ai __attribute__((target("neon"))) uint32x2_t vreinterpret_u32_s64(int64x1_t __p0) {
61323	63297	uint32x2_t __ret;
61324		__ret = (uint32x2_t)(__p0);
	63298	__ret = __builtin_bit_cast(uint32x2_t, __p0);
61325	63299	return __ret;
61326	63300	}
61327	63301	__ai __attribute__((target("neon"))) uint32x2_t vreinterpret_u32_mf8(mfloat8x8_t __p0) {
61328	63302	uint32x2_t __ret;
61329		__ret = (uint32x2_t)(__p0);
	63303	__ret = __builtin_bit_cast(uint32x2_t, __p0);
61330	63304	return __ret;
61331	63305	}
61332	63306	__ai __attribute__((target("neon"))) uint32x2_t vreinterpret_u32_s16(int16x4_t __p0) {
61333	63307	uint32x2_t __ret;
61334		__ret = (uint32x2_t)(__p0);
	63308	__ret = __builtin_bit_cast(uint32x2_t, __p0);
61335	63309	return __ret;
61336	63310	}
61337	63311	__ai __attribute__((target("neon"))) uint64x1_t vreinterpret_u64_p8(poly8x8_t __p0) {
61338	63312	uint64x1_t __ret;
61339		__ret = (uint64x1_t)(__p0);
	63313	__ret = __builtin_bit_cast(uint64x1_t, __p0);
61340	63314	return __ret;
61341	63315	}
61342	63316	__ai __attribute__((target("neon"))) uint64x1_t vreinterpret_u64_p64(poly64x1_t __p0) {
61343	63317	uint64x1_t __ret;
61344		__ret = (uint64x1_t)(__p0);
	63318	__ret = __builtin_bit_cast(uint64x1_t, __p0);
61345	63319	return __ret;
61346	63320	}
61347	63321	__ai __attribute__((target("neon"))) uint64x1_t vreinterpret_u64_p16(poly16x4_t __p0) {
61348	63322	uint64x1_t __ret;
61349		__ret = (uint64x1_t)(__p0);
	63323	__ret = __builtin_bit_cast(uint64x1_t, __p0);
61350	63324	return __ret;
61351	63325	}
61352	63326	__ai __attribute__((target("neon"))) uint64x1_t vreinterpret_u64_u8(uint8x8_t __p0) {
61353	63327	uint64x1_t __ret;
61354		__ret = (uint64x1_t)(__p0);
	63328	__ret = __builtin_bit_cast(uint64x1_t, __p0);
61355	63329	return __ret;
61356	63330	}
61357	63331	__ai __attribute__((target("neon"))) uint64x1_t vreinterpret_u64_u32(uint32x2_t __p0) {
61358	63332	uint64x1_t __ret;
61359		__ret = (uint64x1_t)(__p0);
	63333	__ret = __builtin_bit_cast(uint64x1_t, __p0);
61360	63334	return __ret;
61361	63335	}
61362	63336	__ai __attribute__((target("neon"))) uint64x1_t vreinterpret_u64_u16(uint16x4_t __p0) {
61363	63337	uint64x1_t __ret;
61364		__ret = (uint64x1_t)(__p0);
	63338	__ret = __builtin_bit_cast(uint64x1_t, __p0);
61365	63339	return __ret;
61366	63340	}
61367	63341	__ai __attribute__((target("neon"))) uint64x1_t vreinterpret_u64_s8(int8x8_t __p0) {
61368	63342	uint64x1_t __ret;
61369		__ret = (uint64x1_t)(__p0);
	63343	__ret = __builtin_bit_cast(uint64x1_t, __p0);
61370	63344	return __ret;
61371	63345	}
61372	63346	__ai __attribute__((target("neon"))) uint64x1_t vreinterpret_u64_f64(float64x1_t __p0) {
61373	63347	uint64x1_t __ret;
61374		__ret = (uint64x1_t)(__p0);
	63348	__ret = __builtin_bit_cast(uint64x1_t, __p0);
61375	63349	return __ret;
61376	63350	}
61377	63351	__ai __attribute__((target("neon"))) uint64x1_t vreinterpret_u64_f32(float32x2_t __p0) {
61378	63352	uint64x1_t __ret;
61379		__ret = (uint64x1_t)(__p0);
	63353	__ret = __builtin_bit_cast(uint64x1_t, __p0);
61380	63354	return __ret;
61381	63355	}
61382	63356	__ai __attribute__((target("neon"))) uint64x1_t vreinterpret_u64_f16(float16x4_t __p0) {
61383	63357	uint64x1_t __ret;
61384		__ret = (uint64x1_t)(__p0);
	63358	__ret = __builtin_bit_cast(uint64x1_t, __p0);
61385	63359	return __ret;
61386	63360	}
61387	63361	__ai __attribute__((target("neon"))) uint64x1_t vreinterpret_u64_s32(int32x2_t __p0) {
61388	63362	uint64x1_t __ret;
61389		__ret = (uint64x1_t)(__p0);
	63363	__ret = __builtin_bit_cast(uint64x1_t, __p0);
61390	63364	return __ret;
61391	63365	}
61392	63366	__ai __attribute__((target("neon"))) uint64x1_t vreinterpret_u64_s64(int64x1_t __p0) {
61393	63367	uint64x1_t __ret;
61394		__ret = (uint64x1_t)(__p0);
	63368	__ret = __builtin_bit_cast(uint64x1_t, __p0);
61395	63369	return __ret;
61396	63370	}
61397	63371	__ai __attribute__((target("neon"))) uint64x1_t vreinterpret_u64_mf8(mfloat8x8_t __p0) {
61398	63372	uint64x1_t __ret;
61399		__ret = (uint64x1_t)(__p0);
	63373	__ret = __builtin_bit_cast(uint64x1_t, __p0);
61400	63374	return __ret;
61401	63375	}
61402	63376	__ai __attribute__((target("neon"))) uint64x1_t vreinterpret_u64_s16(int16x4_t __p0) {
61403	63377	uint64x1_t __ret;
61404		__ret = (uint64x1_t)(__p0);
	63378	__ret = __builtin_bit_cast(uint64x1_t, __p0);
61405	63379	return __ret;
61406	63380	}
61407	63381	__ai __attribute__((target("neon"))) uint16x4_t vreinterpret_u16_p8(poly8x8_t __p0) {
61408	63382	uint16x4_t __ret;
61409		__ret = (uint16x4_t)(__p0);
	63383	__ret = __builtin_bit_cast(uint16x4_t, __p0);
61410	63384	return __ret;
61411	63385	}
61412	63386	__ai __attribute__((target("neon"))) uint16x4_t vreinterpret_u16_p64(poly64x1_t __p0) {
61413	63387	uint16x4_t __ret;
61414		__ret = (uint16x4_t)(__p0);
	63388	__ret = __builtin_bit_cast(uint16x4_t, __p0);
61415	63389	return __ret;
61416	63390	}
61417	63391	__ai __attribute__((target("neon"))) uint16x4_t vreinterpret_u16_p16(poly16x4_t __p0) {
61418	63392	uint16x4_t __ret;
61419		__ret = (uint16x4_t)(__p0);
	63393	__ret = __builtin_bit_cast(uint16x4_t, __p0);
61420	63394	return __ret;
61421	63395	}
61422	63396	__ai __attribute__((target("neon"))) uint16x4_t vreinterpret_u16_u8(uint8x8_t __p0) {
61423	63397	uint16x4_t __ret;
61424		__ret = (uint16x4_t)(__p0);
	63398	__ret = __builtin_bit_cast(uint16x4_t, __p0);
61425	63399	return __ret;
61426	63400	}
61427	63401	__ai __attribute__((target("neon"))) uint16x4_t vreinterpret_u16_u32(uint32x2_t __p0) {
61428	63402	uint16x4_t __ret;
61429		__ret = (uint16x4_t)(__p0);
	63403	__ret = __builtin_bit_cast(uint16x4_t, __p0);
61430	63404	return __ret;
61431	63405	}
61432	63406	__ai __attribute__((target("neon"))) uint16x4_t vreinterpret_u16_u64(uint64x1_t __p0) {
61433	63407	uint16x4_t __ret;
61434		__ret = (uint16x4_t)(__p0);
	63408	__ret = __builtin_bit_cast(uint16x4_t, __p0);
61435	63409	return __ret;
61436	63410	}
61437	63411	__ai __attribute__((target("neon"))) uint16x4_t vreinterpret_u16_s8(int8x8_t __p0) {
61438	63412	uint16x4_t __ret;
61439		__ret = (uint16x4_t)(__p0);
	63413	__ret = __builtin_bit_cast(uint16x4_t, __p0);
61440	63414	return __ret;
61441	63415	}
61442	63416	__ai __attribute__((target("neon"))) uint16x4_t vreinterpret_u16_f64(float64x1_t __p0) {
61443	63417	uint16x4_t __ret;
61444		__ret = (uint16x4_t)(__p0);
	63418	__ret = __builtin_bit_cast(uint16x4_t, __p0);
61445	63419	return __ret;
61446	63420	}
61447	63421	__ai __attribute__((target("neon"))) uint16x4_t vreinterpret_u16_f32(float32x2_t __p0) {
61448	63422	uint16x4_t __ret;
61449		__ret = (uint16x4_t)(__p0);
	63423	__ret = __builtin_bit_cast(uint16x4_t, __p0);
61450	63424	return __ret;
61451	63425	}
61452	63426	__ai __attribute__((target("neon"))) uint16x4_t vreinterpret_u16_f16(float16x4_t __p0) {
61453	63427	uint16x4_t __ret;
61454		__ret = (uint16x4_t)(__p0);
	63428	__ret = __builtin_bit_cast(uint16x4_t, __p0);
61455	63429	return __ret;
61456	63430	}
61457	63431	__ai __attribute__((target("neon"))) uint16x4_t vreinterpret_u16_s32(int32x2_t __p0) {
61458	63432	uint16x4_t __ret;
61459		__ret = (uint16x4_t)(__p0);
	63433	__ret = __builtin_bit_cast(uint16x4_t, __p0);
61460	63434	return __ret;
61461	63435	}
61462	63436	__ai __attribute__((target("neon"))) uint16x4_t vreinterpret_u16_s64(int64x1_t __p0) {
61463	63437	uint16x4_t __ret;
61464		__ret = (uint16x4_t)(__p0);
	63438	__ret = __builtin_bit_cast(uint16x4_t, __p0);
61465	63439	return __ret;
61466	63440	}
61467	63441	__ai __attribute__((target("neon"))) uint16x4_t vreinterpret_u16_mf8(mfloat8x8_t __p0) {
61468	63442	uint16x4_t __ret;
61469		__ret = (uint16x4_t)(__p0);
	63443	__ret = __builtin_bit_cast(uint16x4_t, __p0);
61470	63444	return __ret;
61471	63445	}
61472	63446	__ai __attribute__((target("neon"))) uint16x4_t vreinterpret_u16_s16(int16x4_t __p0) {
61473	63447	uint16x4_t __ret;
61474		__ret = (uint16x4_t)(__p0);
	63448	__ret = __builtin_bit_cast(uint16x4_t, __p0);
61475	63449	return __ret;
61476	63450	}
61477	63451	__ai __attribute__((target("neon"))) int8x8_t vreinterpret_s8_p8(poly8x8_t __p0) {
61478	63452	int8x8_t __ret;
61479		__ret = (int8x8_t)(__p0);
	63453	__ret = __builtin_bit_cast(int8x8_t, __p0);
61480	63454	return __ret;
61481	63455	}
61482	63456	__ai __attribute__((target("neon"))) int8x8_t vreinterpret_s8_p64(poly64x1_t __p0) {
61483	63457	int8x8_t __ret;
61484		__ret = (int8x8_t)(__p0);
	63458	__ret = __builtin_bit_cast(int8x8_t, __p0);
61485	63459	return __ret;
61486	63460	}
61487	63461	__ai __attribute__((target("neon"))) int8x8_t vreinterpret_s8_p16(poly16x4_t __p0) {
61488	63462	int8x8_t __ret;
61489		__ret = (int8x8_t)(__p0);
	63463	__ret = __builtin_bit_cast(int8x8_t, __p0);
61490	63464	return __ret;
61491	63465	}
61492	63466	__ai __attribute__((target("neon"))) int8x8_t vreinterpret_s8_u8(uint8x8_t __p0) {
61493	63467	int8x8_t __ret;
61494		__ret = (int8x8_t)(__p0);
	63468	__ret = __builtin_bit_cast(int8x8_t, __p0);
61495	63469	return __ret;
61496	63470	}
61497	63471	__ai __attribute__((target("neon"))) int8x8_t vreinterpret_s8_u32(uint32x2_t __p0) {
61498	63472	int8x8_t __ret;
61499		__ret = (int8x8_t)(__p0);
	63473	__ret = __builtin_bit_cast(int8x8_t, __p0);
61500	63474	return __ret;
61501	63475	}
61502	63476	__ai __attribute__((target("neon"))) int8x8_t vreinterpret_s8_u64(uint64x1_t __p0) {
61503	63477	int8x8_t __ret;
61504		__ret = (int8x8_t)(__p0);
	63478	__ret = __builtin_bit_cast(int8x8_t, __p0);
61505	63479	return __ret;
61506	63480	}
61507	63481	__ai __attribute__((target("neon"))) int8x8_t vreinterpret_s8_u16(uint16x4_t __p0) {
61508	63482	int8x8_t __ret;
61509		__ret = (int8x8_t)(__p0);
	63483	__ret = __builtin_bit_cast(int8x8_t, __p0);
61510	63484	return __ret;
61511	63485	}
61512	63486	__ai __attribute__((target("neon"))) int8x8_t vreinterpret_s8_f64(float64x1_t __p0) {
61513	63487	int8x8_t __ret;
61514		__ret = (int8x8_t)(__p0);
	63488	__ret = __builtin_bit_cast(int8x8_t, __p0);
61515	63489	return __ret;
61516	63490	}
61517	63491	__ai __attribute__((target("neon"))) int8x8_t vreinterpret_s8_f32(float32x2_t __p0) {
61518	63492	int8x8_t __ret;
61519		__ret = (int8x8_t)(__p0);
	63493	__ret = __builtin_bit_cast(int8x8_t, __p0);
61520	63494	return __ret;
61521	63495	}
61522	63496	__ai __attribute__((target("neon"))) int8x8_t vreinterpret_s8_f16(float16x4_t __p0) {
61523	63497	int8x8_t __ret;
61524		__ret = (int8x8_t)(__p0);
	63498	__ret = __builtin_bit_cast(int8x8_t, __p0);
61525	63499	return __ret;
61526	63500	}
61527	63501	__ai __attribute__((target("neon"))) int8x8_t vreinterpret_s8_s32(int32x2_t __p0) {
61528	63502	int8x8_t __ret;
61529		__ret = (int8x8_t)(__p0);
	63503	__ret = __builtin_bit_cast(int8x8_t, __p0);
61530	63504	return __ret;
61531	63505	}
61532	63506	__ai __attribute__((target("neon"))) int8x8_t vreinterpret_s8_s64(int64x1_t __p0) {
61533	63507	int8x8_t __ret;
61534		__ret = (int8x8_t)(__p0);
	63508	__ret = __builtin_bit_cast(int8x8_t, __p0);
61535	63509	return __ret;
61536	63510	}
61537	63511	__ai __attribute__((target("neon"))) int8x8_t vreinterpret_s8_mf8(mfloat8x8_t __p0) {
61538	63512	int8x8_t __ret;
61539		__ret = (int8x8_t)(__p0);
	63513	__ret = __builtin_bit_cast(int8x8_t, __p0);
61540	63514	return __ret;
61541	63515	}
61542	63516	__ai __attribute__((target("neon"))) int8x8_t vreinterpret_s8_s16(int16x4_t __p0) {
61543	63517	int8x8_t __ret;
61544		__ret = (int8x8_t)(__p0);
	63518	__ret = __builtin_bit_cast(int8x8_t, __p0);
61545	63519	return __ret;
61546	63520	}
61547	63521	__ai __attribute__((target("neon"))) float64x1_t vreinterpret_f64_p8(poly8x8_t __p0) {
61548	63522	float64x1_t __ret;
61549		__ret = (float64x1_t)(__p0);
	63523	__ret = __builtin_bit_cast(float64x1_t, __p0);
61550	63524	return __ret;
61551	63525	}
61552	63526	__ai __attribute__((target("neon"))) float64x1_t vreinterpret_f64_p64(poly64x1_t __p0) {
61553	63527	float64x1_t __ret;
61554		__ret = (float64x1_t)(__p0);
	63528	__ret = __builtin_bit_cast(float64x1_t, __p0);
61555	63529	return __ret;
61556	63530	}
61557	63531	__ai __attribute__((target("neon"))) float64x1_t vreinterpret_f64_p16(poly16x4_t __p0) {
61558	63532	float64x1_t __ret;
61559		__ret = (float64x1_t)(__p0);
	63533	__ret = __builtin_bit_cast(float64x1_t, __p0);
61560	63534	return __ret;
61561	63535	}
61562	63536	__ai __attribute__((target("neon"))) float64x1_t vreinterpret_f64_u8(uint8x8_t __p0) {
61563	63537	float64x1_t __ret;
61564		__ret = (float64x1_t)(__p0);
	63538	__ret = __builtin_bit_cast(float64x1_t, __p0);
61565	63539	return __ret;
61566	63540	}
61567	63541	__ai __attribute__((target("neon"))) float64x1_t vreinterpret_f64_u32(uint32x2_t __p0) {
61568	63542	float64x1_t __ret;
61569		__ret = (float64x1_t)(__p0);
	63543	__ret = __builtin_bit_cast(float64x1_t, __p0);
61570	63544	return __ret;
61571	63545	}
61572	63546	__ai __attribute__((target("neon"))) float64x1_t vreinterpret_f64_u64(uint64x1_t __p0) {
61573	63547	float64x1_t __ret;
61574		__ret = (float64x1_t)(__p0);
	63548	__ret = __builtin_bit_cast(float64x1_t, __p0);
61575	63549	return __ret;
61576	63550	}
61577	63551	__ai __attribute__((target("neon"))) float64x1_t vreinterpret_f64_u16(uint16x4_t __p0) {
61578	63552	float64x1_t __ret;
61579		__ret = (float64x1_t)(__p0);
	63553	__ret = __builtin_bit_cast(float64x1_t, __p0);
61580	63554	return __ret;
61581	63555	}
61582	63556	__ai __attribute__((target("neon"))) float64x1_t vreinterpret_f64_s8(int8x8_t __p0) {
61583	63557	float64x1_t __ret;
61584		__ret = (float64x1_t)(__p0);
	63558	__ret = __builtin_bit_cast(float64x1_t, __p0);
61585	63559	return __ret;
61586	63560	}
61587	63561	__ai __attribute__((target("neon"))) float64x1_t vreinterpret_f64_f32(float32x2_t __p0) {
61588	63562	float64x1_t __ret;
61589		__ret = (float64x1_t)(__p0);
	63563	__ret = __builtin_bit_cast(float64x1_t, __p0);
61590	63564	return __ret;
61591	63565	}
61592	63566	__ai __attribute__((target("neon"))) float64x1_t vreinterpret_f64_f16(float16x4_t __p0) {
61593	63567	float64x1_t __ret;
61594		__ret = (float64x1_t)(__p0);
	63568	__ret = __builtin_bit_cast(float64x1_t, __p0);
61595	63569	return __ret;
61596	63570	}
61597	63571	__ai __attribute__((target("neon"))) float64x1_t vreinterpret_f64_s32(int32x2_t __p0) {
61598	63572	float64x1_t __ret;
61599		__ret = (float64x1_t)(__p0);
	63573	__ret = __builtin_bit_cast(float64x1_t, __p0);
61600	63574	return __ret;
61601	63575	}
61602	63576	__ai __attribute__((target("neon"))) float64x1_t vreinterpret_f64_s64(int64x1_t __p0) {
61603	63577	float64x1_t __ret;
61604		__ret = (float64x1_t)(__p0);
	63578	__ret = __builtin_bit_cast(float64x1_t, __p0);
61605	63579	return __ret;
61606	63580	}
61607	63581	__ai __attribute__((target("neon"))) float64x1_t vreinterpret_f64_mf8(mfloat8x8_t __p0) {
61608	63582	float64x1_t __ret;
61609		__ret = (float64x1_t)(__p0);
	63583	__ret = __builtin_bit_cast(float64x1_t, __p0);
61610	63584	return __ret;
61611	63585	}
61612	63586	__ai __attribute__((target("neon"))) float64x1_t vreinterpret_f64_s16(int16x4_t __p0) {
61613	63587	float64x1_t __ret;
61614		__ret = (float64x1_t)(__p0);
	63588	__ret = __builtin_bit_cast(float64x1_t, __p0);
61615	63589	return __ret;
61616	63590	}
61617	63591	__ai __attribute__((target("neon"))) float32x2_t vreinterpret_f32_p8(poly8x8_t __p0) {
61618	63592	float32x2_t __ret;
61619		__ret = (float32x2_t)(__p0);
	63593	__ret = __builtin_bit_cast(float32x2_t, __p0);
61620	63594	return __ret;
61621	63595	}
61622	63596	__ai __attribute__((target("neon"))) float32x2_t vreinterpret_f32_p64(poly64x1_t __p0) {
61623	63597	float32x2_t __ret;
61624		__ret = (float32x2_t)(__p0);
	63598	__ret = __builtin_bit_cast(float32x2_t, __p0);
61625	63599	return __ret;
61626	63600	}
61627	63601	__ai __attribute__((target("neon"))) float32x2_t vreinterpret_f32_p16(poly16x4_t __p0) {
61628	63602	float32x2_t __ret;
61629		__ret = (float32x2_t)(__p0);
	63603	__ret = __builtin_bit_cast(float32x2_t, __p0);
61630	63604	return __ret;
61631	63605	}
61632	63606	__ai __attribute__((target("neon"))) float32x2_t vreinterpret_f32_u8(uint8x8_t __p0) {
61633	63607	float32x2_t __ret;
61634		__ret = (float32x2_t)(__p0);
	63608	__ret = __builtin_bit_cast(float32x2_t, __p0);
61635	63609	return __ret;
61636	63610	}
61637	63611	__ai __attribute__((target("neon"))) float32x2_t vreinterpret_f32_u32(uint32x2_t __p0) {
61638	63612	float32x2_t __ret;
61639		__ret = (float32x2_t)(__p0);
	63613	__ret = __builtin_bit_cast(float32x2_t, __p0);
61640	63614	return __ret;
61641	63615	}
61642	63616	__ai __attribute__((target("neon"))) float32x2_t vreinterpret_f32_u64(uint64x1_t __p0) {
61643	63617	float32x2_t __ret;
61644		__ret = (float32x2_t)(__p0);
	63618	__ret = __builtin_bit_cast(float32x2_t, __p0);
61645	63619	return __ret;
61646	63620	}
61647	63621	__ai __attribute__((target("neon"))) float32x2_t vreinterpret_f32_u16(uint16x4_t __p0) {
61648	63622	float32x2_t __ret;
61649		__ret = (float32x2_t)(__p0);
	63623	__ret = __builtin_bit_cast(float32x2_t, __p0);
61650	63624	return __ret;
61651	63625	}
61652	63626	__ai __attribute__((target("neon"))) float32x2_t vreinterpret_f32_s8(int8x8_t __p0) {
61653	63627	float32x2_t __ret;
61654		__ret = (float32x2_t)(__p0);
	63628	__ret = __builtin_bit_cast(float32x2_t, __p0);
61655	63629	return __ret;
61656	63630	}
61657	63631	__ai __attribute__((target("neon"))) float32x2_t vreinterpret_f32_f64(float64x1_t __p0) {
61658	63632	float32x2_t __ret;
61659		__ret = (float32x2_t)(__p0);
	63633	__ret = __builtin_bit_cast(float32x2_t, __p0);
61660	63634	return __ret;
61661	63635	}
61662	63636	__ai __attribute__((target("neon"))) float32x2_t vreinterpret_f32_f16(float16x4_t __p0) {
61663	63637	float32x2_t __ret;
61664		__ret = (float32x2_t)(__p0);
	63638	__ret = __builtin_bit_cast(float32x2_t, __p0);
61665	63639	return __ret;
61666	63640	}
61667	63641	__ai __attribute__((target("neon"))) float32x2_t vreinterpret_f32_s32(int32x2_t __p0) {
61668	63642	float32x2_t __ret;
61669		__ret = (float32x2_t)(__p0);
	63643	__ret = __builtin_bit_cast(float32x2_t, __p0);
61670	63644	return __ret;
61671	63645	}
61672	63646	__ai __attribute__((target("neon"))) float32x2_t vreinterpret_f32_s64(int64x1_t __p0) {
61673	63647	float32x2_t __ret;
61674		__ret = (float32x2_t)(__p0);
	63648	__ret = __builtin_bit_cast(float32x2_t, __p0);
61675	63649	return __ret;
61676	63650	}
61677	63651	__ai __attribute__((target("neon"))) float32x2_t vreinterpret_f32_mf8(mfloat8x8_t __p0) {
61678	63652	float32x2_t __ret;
61679		__ret = (float32x2_t)(__p0);
	63653	__ret = __builtin_bit_cast(float32x2_t, __p0);
61680	63654	return __ret;
61681	63655	}
61682	63656	__ai __attribute__((target("neon"))) float32x2_t vreinterpret_f32_s16(int16x4_t __p0) {
61683	63657	float32x2_t __ret;
61684		__ret = (float32x2_t)(__p0);
	63658	__ret = __builtin_bit_cast(float32x2_t, __p0);
61685	63659	return __ret;
61686	63660	}
61687	63661	__ai __attribute__((target("neon"))) float16x4_t vreinterpret_f16_p8(poly8x8_t __p0) {
61688	63662	float16x4_t __ret;
61689		__ret = (float16x4_t)(__p0);
	63663	__ret = __builtin_bit_cast(float16x4_t, __p0);
61690	63664	return __ret;
61691	63665	}
61692	63666	__ai __attribute__((target("neon"))) float16x4_t vreinterpret_f16_p64(poly64x1_t __p0) {
61693	63667	float16x4_t __ret;
61694		__ret = (float16x4_t)(__p0);
	63668	__ret = __builtin_bit_cast(float16x4_t, __p0);
61695	63669	return __ret;
61696	63670	}
61697	63671	__ai __attribute__((target("neon"))) float16x4_t vreinterpret_f16_p16(poly16x4_t __p0) {
61698	63672	float16x4_t __ret;
61699		__ret = (float16x4_t)(__p0);
	63673	__ret = __builtin_bit_cast(float16x4_t, __p0);
61700	63674	return __ret;
61701	63675	}
61702	63676	__ai __attribute__((target("neon"))) float16x4_t vreinterpret_f16_u8(uint8x8_t __p0) {
61703	63677	float16x4_t __ret;
61704		__ret = (float16x4_t)(__p0);
	63678	__ret = __builtin_bit_cast(float16x4_t, __p0);
61705	63679	return __ret;
61706	63680	}
61707	63681	__ai __attribute__((target("neon"))) float16x4_t vreinterpret_f16_u32(uint32x2_t __p0) {
61708	63682	float16x4_t __ret;
61709		__ret = (float16x4_t)(__p0);
	63683	__ret = __builtin_bit_cast(float16x4_t, __p0);
61710	63684	return __ret;
61711	63685	}
61712	63686	__ai __attribute__((target("neon"))) float16x4_t vreinterpret_f16_u64(uint64x1_t __p0) {
61713	63687	float16x4_t __ret;
61714		__ret = (float16x4_t)(__p0);
	63688	__ret = __builtin_bit_cast(float16x4_t, __p0);
61715	63689	return __ret;
61716	63690	}
61717	63691	__ai __attribute__((target("neon"))) float16x4_t vreinterpret_f16_u16(uint16x4_t __p0) {
61718	63692	float16x4_t __ret;
61719		__ret = (float16x4_t)(__p0);
	63693	__ret = __builtin_bit_cast(float16x4_t, __p0);
61720	63694	return __ret;
61721	63695	}
61722	63696	__ai __attribute__((target("neon"))) float16x4_t vreinterpret_f16_s8(int8x8_t __p0) {
61723	63697	float16x4_t __ret;
61724		__ret = (float16x4_t)(__p0);
	63698	__ret = __builtin_bit_cast(float16x4_t, __p0);
61725	63699	return __ret;
61726	63700	}
61727	63701	__ai __attribute__((target("neon"))) float16x4_t vreinterpret_f16_f64(float64x1_t __p0) {
61728	63702	float16x4_t __ret;
61729		__ret = (float16x4_t)(__p0);
	63703	__ret = __builtin_bit_cast(float16x4_t, __p0);
61730	63704	return __ret;
61731	63705	}
61732	63706	__ai __attribute__((target("neon"))) float16x4_t vreinterpret_f16_f32(float32x2_t __p0) {
61733	63707	float16x4_t __ret;
61734		__ret = (float16x4_t)(__p0);
	63708	__ret = __builtin_bit_cast(float16x4_t, __p0);
61735	63709	return __ret;
61736	63710	}
61737	63711	__ai __attribute__((target("neon"))) float16x4_t vreinterpret_f16_s32(int32x2_t __p0) {
61738	63712	float16x4_t __ret;
61739		__ret = (float16x4_t)(__p0);
	63713	__ret = __builtin_bit_cast(float16x4_t, __p0);
61740	63714	return __ret;
61741	63715	}
61742	63716	__ai __attribute__((target("neon"))) float16x4_t vreinterpret_f16_s64(int64x1_t __p0) {
61743	63717	float16x4_t __ret;
61744		__ret = (float16x4_t)(__p0);
	63718	__ret = __builtin_bit_cast(float16x4_t, __p0);
61745	63719	return __ret;
61746	63720	}
61747	63721	__ai __attribute__((target("neon"))) float16x4_t vreinterpret_f16_mf8(mfloat8x8_t __p0) {
61748	63722	float16x4_t __ret;
61749		__ret = (float16x4_t)(__p0);
	63723	__ret = __builtin_bit_cast(float16x4_t, __p0);
61750	63724	return __ret;
61751	63725	}
61752	63726	__ai __attribute__((target("neon"))) float16x4_t vreinterpret_f16_s16(int16x4_t __p0) {
61753	63727	float16x4_t __ret;
61754		__ret = (float16x4_t)(__p0);
	63728	__ret = __builtin_bit_cast(float16x4_t, __p0);
61755	63729	return __ret;
61756	63730	}
61757	63731	__ai __attribute__((target("neon"))) int32x2_t vreinterpret_s32_p8(poly8x8_t __p0) {
61758	63732	int32x2_t __ret;
61759		__ret = (int32x2_t)(__p0);
	63733	__ret = __builtin_bit_cast(int32x2_t, __p0);
61760	63734	return __ret;
61761	63735	}
61762	63736	__ai __attribute__((target("neon"))) int32x2_t vreinterpret_s32_p64(poly64x1_t __p0) {
61763	63737	int32x2_t __ret;
61764		__ret = (int32x2_t)(__p0);
	63738	__ret = __builtin_bit_cast(int32x2_t, __p0);
61765	63739	return __ret;
61766	63740	}
61767	63741	__ai __attribute__((target("neon"))) int32x2_t vreinterpret_s32_p16(poly16x4_t __p0) {
61768	63742	int32x2_t __ret;
61769		__ret = (int32x2_t)(__p0);
	63743	__ret = __builtin_bit_cast(int32x2_t, __p0);
61770	63744	return __ret;
61771	63745	}
61772	63746	__ai __attribute__((target("neon"))) int32x2_t vreinterpret_s32_u8(uint8x8_t __p0) {
61773	63747	int32x2_t __ret;
61774		__ret = (int32x2_t)(__p0);
	63748	__ret = __builtin_bit_cast(int32x2_t, __p0);
61775	63749	return __ret;
61776	63750	}
61777	63751	__ai __attribute__((target("neon"))) int32x2_t vreinterpret_s32_u32(uint32x2_t __p0) {
61778	63752	int32x2_t __ret;
61779		__ret = (int32x2_t)(__p0);
	63753	__ret = __builtin_bit_cast(int32x2_t, __p0);
61780	63754	return __ret;
61781	63755	}
61782	63756	__ai __attribute__((target("neon"))) int32x2_t vreinterpret_s32_u64(uint64x1_t __p0) {
61783	63757	int32x2_t __ret;
61784		__ret = (int32x2_t)(__p0);
	63758	__ret = __builtin_bit_cast(int32x2_t, __p0);
61785	63759	return __ret;
61786	63760	}
61787	63761	__ai __attribute__((target("neon"))) int32x2_t vreinterpret_s32_u16(uint16x4_t __p0) {
61788	63762	int32x2_t __ret;
61789		__ret = (int32x2_t)(__p0);
	63763	__ret = __builtin_bit_cast(int32x2_t, __p0);
61790	63764	return __ret;
61791	63765	}
61792	63766	__ai __attribute__((target("neon"))) int32x2_t vreinterpret_s32_s8(int8x8_t __p0) {
61793	63767	int32x2_t __ret;
61794		__ret = (int32x2_t)(__p0);
	63768	__ret = __builtin_bit_cast(int32x2_t, __p0);
61795	63769	return __ret;
61796	63770	}
61797	63771	__ai __attribute__((target("neon"))) int32x2_t vreinterpret_s32_f64(float64x1_t __p0) {
61798	63772	int32x2_t __ret;
61799		__ret = (int32x2_t)(__p0);
	63773	__ret = __builtin_bit_cast(int32x2_t, __p0);
61800	63774	return __ret;
61801	63775	}
61802	63776	__ai __attribute__((target("neon"))) int32x2_t vreinterpret_s32_f32(float32x2_t __p0) {
61803	63777	int32x2_t __ret;
61804		__ret = (int32x2_t)(__p0);
	63778	__ret = __builtin_bit_cast(int32x2_t, __p0);
61805	63779	return __ret;
61806	63780	}
61807	63781	__ai __attribute__((target("neon"))) int32x2_t vreinterpret_s32_f16(float16x4_t __p0) {
61808	63782	int32x2_t __ret;
61809		__ret = (int32x2_t)(__p0);
	63783	__ret = __builtin_bit_cast(int32x2_t, __p0);
61810	63784	return __ret;
61811	63785	}
61812	63786	__ai __attribute__((target("neon"))) int32x2_t vreinterpret_s32_s64(int64x1_t __p0) {
61813	63787	int32x2_t __ret;
61814		__ret = (int32x2_t)(__p0);
	63788	__ret = __builtin_bit_cast(int32x2_t, __p0);
61815	63789	return __ret;
61816	63790	}
61817	63791	__ai __attribute__((target("neon"))) int32x2_t vreinterpret_s32_mf8(mfloat8x8_t __p0) {
61818	63792	int32x2_t __ret;
61819		__ret = (int32x2_t)(__p0);
	63793	__ret = __builtin_bit_cast(int32x2_t, __p0);
61820	63794	return __ret;
61821	63795	}
61822	63796	__ai __attribute__((target("neon"))) int32x2_t vreinterpret_s32_s16(int16x4_t __p0) {
61823	63797	int32x2_t __ret;
61824		__ret = (int32x2_t)(__p0);
	63798	__ret = __builtin_bit_cast(int32x2_t, __p0);
61825	63799	return __ret;
61826	63800	}
61827	63801	__ai __attribute__((target("neon"))) int64x1_t vreinterpret_s64_p8(poly8x8_t __p0) {
61828	63802	int64x1_t __ret;
61829		__ret = (int64x1_t)(__p0);
	63803	__ret = __builtin_bit_cast(int64x1_t, __p0);
61830	63804	return __ret;
61831	63805	}
61832	63806	__ai __attribute__((target("neon"))) int64x1_t vreinterpret_s64_p64(poly64x1_t __p0) {
61833	63807	int64x1_t __ret;
61834		__ret = (int64x1_t)(__p0);
	63808	__ret = __builtin_bit_cast(int64x1_t, __p0);
61835	63809	return __ret;
61836	63810	}
61837	63811	__ai __attribute__((target("neon"))) int64x1_t vreinterpret_s64_p16(poly16x4_t __p0) {
61838	63812	int64x1_t __ret;
61839		__ret = (int64x1_t)(__p0);
	63813	__ret = __builtin_bit_cast(int64x1_t, __p0);
61840	63814	return __ret;
61841	63815	}
61842	63816	__ai __attribute__((target("neon"))) int64x1_t vreinterpret_s64_u8(uint8x8_t __p0) {
61843	63817	int64x1_t __ret;
61844		__ret = (int64x1_t)(__p0);
	63818	__ret = __builtin_bit_cast(int64x1_t, __p0);
61845	63819	return __ret;
61846	63820	}
61847	63821	__ai __attribute__((target("neon"))) int64x1_t vreinterpret_s64_u32(uint32x2_t __p0) {
61848	63822	int64x1_t __ret;
61849		__ret = (int64x1_t)(__p0);
	63823	__ret = __builtin_bit_cast(int64x1_t, __p0);
61850	63824	return __ret;
61851	63825	}
61852	63826	__ai __attribute__((target("neon"))) int64x1_t vreinterpret_s64_u64(uint64x1_t __p0) {
61853	63827	int64x1_t __ret;
61854		__ret = (int64x1_t)(__p0);
	63828	__ret = __builtin_bit_cast(int64x1_t, __p0);
61855	63829	return __ret;
61856	63830	}
61857	63831	__ai __attribute__((target("neon"))) int64x1_t vreinterpret_s64_u16(uint16x4_t __p0) {
61858	63832	int64x1_t __ret;
61859		__ret = (int64x1_t)(__p0);
	63833	__ret = __builtin_bit_cast(int64x1_t, __p0);
61860	63834	return __ret;
61861	63835	}
61862	63836	__ai __attribute__((target("neon"))) int64x1_t vreinterpret_s64_s8(int8x8_t __p0) {
61863	63837	int64x1_t __ret;
61864		__ret = (int64x1_t)(__p0);
	63838	__ret = __builtin_bit_cast(int64x1_t, __p0);
61865	63839	return __ret;
61866	63840	}
61867	63841	__ai __attribute__((target("neon"))) int64x1_t vreinterpret_s64_f64(float64x1_t __p0) {
61868	63842	int64x1_t __ret;
61869		__ret = (int64x1_t)(__p0);
	63843	__ret = __builtin_bit_cast(int64x1_t, __p0);
61870	63844	return __ret;
61871	63845	}
61872	63846	__ai __attribute__((target("neon"))) int64x1_t vreinterpret_s64_f32(float32x2_t __p0) {
61873	63847	int64x1_t __ret;
61874		__ret = (int64x1_t)(__p0);
	63848	__ret = __builtin_bit_cast(int64x1_t, __p0);
61875	63849	return __ret;
61876	63850	}
61877	63851	__ai __attribute__((target("neon"))) int64x1_t vreinterpret_s64_f16(float16x4_t __p0) {
61878	63852	int64x1_t __ret;
61879		__ret = (int64x1_t)(__p0);
	63853	__ret = __builtin_bit_cast(int64x1_t, __p0);
61880	63854	return __ret;
61881	63855	}
61882	63856	__ai __attribute__((target("neon"))) int64x1_t vreinterpret_s64_s32(int32x2_t __p0) {
61883	63857	int64x1_t __ret;
61884		__ret = (int64x1_t)(__p0);
	63858	__ret = __builtin_bit_cast(int64x1_t, __p0);
61885	63859	return __ret;
61886	63860	}
61887	63861	__ai __attribute__((target("neon"))) int64x1_t vreinterpret_s64_mf8(mfloat8x8_t __p0) {
61888	63862	int64x1_t __ret;
61889		__ret = (int64x1_t)(__p0);
	63863	__ret = __builtin_bit_cast(int64x1_t, __p0);
61890	63864	return __ret;
61891	63865	}
61892	63866	__ai __attribute__((target("neon"))) int64x1_t vreinterpret_s64_s16(int16x4_t __p0) {
61893	63867	int64x1_t __ret;
61894		__ret = (int64x1_t)(__p0);
	63868	__ret = __builtin_bit_cast(int64x1_t, __p0);
61895	63869	return __ret;
61896	63870	}
61897	63871	__ai __attribute__((target("neon"))) mfloat8x8_t vreinterpret_mf8_p8(poly8x8_t __p0) {
61898	63872	mfloat8x8_t __ret;
61899		__ret = (mfloat8x8_t)(__p0);
	63873	__ret = __builtin_bit_cast(mfloat8x8_t, __p0);
61900	63874	return __ret;
61901	63875	}
61902	63876	__ai __attribute__((target("neon"))) mfloat8x8_t vreinterpret_mf8_p64(poly64x1_t __p0) {
61903	63877	mfloat8x8_t __ret;
61904		__ret = (mfloat8x8_t)(__p0);
	63878	__ret = __builtin_bit_cast(mfloat8x8_t, __p0);
61905	63879	return __ret;
61906	63880	}
61907	63881	__ai __attribute__((target("neon"))) mfloat8x8_t vreinterpret_mf8_p16(poly16x4_t __p0) {
61908	63882	mfloat8x8_t __ret;
61909		__ret = (mfloat8x8_t)(__p0);
	63883	__ret = __builtin_bit_cast(mfloat8x8_t, __p0);
61910	63884	return __ret;
61911	63885	}
61912	63886	__ai __attribute__((target("neon"))) mfloat8x8_t vreinterpret_mf8_u8(uint8x8_t __p0) {
61913	63887	mfloat8x8_t __ret;
61914		__ret = (mfloat8x8_t)(__p0);
	63888	__ret = __builtin_bit_cast(mfloat8x8_t, __p0);
61915	63889	return __ret;
61916	63890	}
61917	63891	__ai __attribute__((target("neon"))) mfloat8x8_t vreinterpret_mf8_u32(uint32x2_t __p0) {
61918	63892	mfloat8x8_t __ret;
61919		__ret = (mfloat8x8_t)(__p0);
	63893	__ret = __builtin_bit_cast(mfloat8x8_t, __p0);
61920	63894	return __ret;
61921	63895	}
61922	63896	__ai __attribute__((target("neon"))) mfloat8x8_t vreinterpret_mf8_u64(uint64x1_t __p0) {
61923	63897	mfloat8x8_t __ret;
61924		__ret = (mfloat8x8_t)(__p0);
	63898	__ret = __builtin_bit_cast(mfloat8x8_t, __p0);
61925	63899	return __ret;
61926	63900	}
61927	63901	__ai __attribute__((target("neon"))) mfloat8x8_t vreinterpret_mf8_u16(uint16x4_t __p0) {
61928	63902	mfloat8x8_t __ret;
61929		__ret = (mfloat8x8_t)(__p0);
	63903	__ret = __builtin_bit_cast(mfloat8x8_t, __p0);
61930	63904	return __ret;
61931	63905	}
61932	63906	__ai __attribute__((target("neon"))) mfloat8x8_t vreinterpret_mf8_s8(int8x8_t __p0) {
61933	63907	mfloat8x8_t __ret;
61934		__ret = (mfloat8x8_t)(__p0);
	63908	__ret = __builtin_bit_cast(mfloat8x8_t, __p0);
61935	63909	return __ret;
61936	63910	}
61937	63911	__ai __attribute__((target("neon"))) mfloat8x8_t vreinterpret_mf8_f64(float64x1_t __p0) {
61938	63912	mfloat8x8_t __ret;
61939		__ret = (mfloat8x8_t)(__p0);
	63913	__ret = __builtin_bit_cast(mfloat8x8_t, __p0);
61940	63914	return __ret;
61941	63915	}
61942	63916	__ai __attribute__((target("neon"))) mfloat8x8_t vreinterpret_mf8_f32(float32x2_t __p0) {
61943	63917	mfloat8x8_t __ret;
61944		__ret = (mfloat8x8_t)(__p0);
	63918	__ret = __builtin_bit_cast(mfloat8x8_t, __p0);
61945	63919	return __ret;
61946	63920	}
61947	63921	__ai __attribute__((target("neon"))) mfloat8x8_t vreinterpret_mf8_f16(float16x4_t __p0) {
61948	63922	mfloat8x8_t __ret;
61949		__ret = (mfloat8x8_t)(__p0);
	63923	__ret = __builtin_bit_cast(mfloat8x8_t, __p0);
61950	63924	return __ret;
61951	63925	}
61952	63926	__ai __attribute__((target("neon"))) mfloat8x8_t vreinterpret_mf8_s32(int32x2_t __p0) {
61953	63927	mfloat8x8_t __ret;
61954		__ret = (mfloat8x8_t)(__p0);
	63928	__ret = __builtin_bit_cast(mfloat8x8_t, __p0);
61955	63929	return __ret;
61956	63930	}
61957	63931	__ai __attribute__((target("neon"))) mfloat8x8_t vreinterpret_mf8_s64(int64x1_t __p0) {
61958	63932	mfloat8x8_t __ret;
61959		__ret = (mfloat8x8_t)(__p0);
	63933	__ret = __builtin_bit_cast(mfloat8x8_t, __p0);
61960	63934	return __ret;
61961	63935	}
61962	63936	__ai __attribute__((target("neon"))) mfloat8x8_t vreinterpret_mf8_s16(int16x4_t __p0) {
61963	63937	mfloat8x8_t __ret;
61964		__ret = (mfloat8x8_t)(__p0);
	63938	__ret = __builtin_bit_cast(mfloat8x8_t, __p0);
61965	63939	return __ret;
61966	63940	}
61967	63941	__ai __attribute__((target("neon"))) int16x4_t vreinterpret_s16_p8(poly8x8_t __p0) {
61968	63942	int16x4_t __ret;
61969		__ret = (int16x4_t)(__p0);
	63943	__ret = __builtin_bit_cast(int16x4_t, __p0);
61970	63944	return __ret;
61971	63945	}
61972	63946	__ai __attribute__((target("neon"))) int16x4_t vreinterpret_s16_p64(poly64x1_t __p0) {
61973	63947	int16x4_t __ret;
61974		__ret = (int16x4_t)(__p0);
	63948	__ret = __builtin_bit_cast(int16x4_t, __p0);
61975	63949	return __ret;
61976	63950	}
61977	63951	__ai __attribute__((target("neon"))) int16x4_t vreinterpret_s16_p16(poly16x4_t __p0) {
61978	63952	int16x4_t __ret;
61979		__ret = (int16x4_t)(__p0);
	63953	__ret = __builtin_bit_cast(int16x4_t, __p0);
61980	63954	return __ret;
61981	63955	}
61982	63956	__ai __attribute__((target("neon"))) int16x4_t vreinterpret_s16_u8(uint8x8_t __p0) {
61983	63957	int16x4_t __ret;
61984		__ret = (int16x4_t)(__p0);
	63958	__ret = __builtin_bit_cast(int16x4_t, __p0);
61985	63959	return __ret;
61986	63960	}
61987	63961	__ai __attribute__((target("neon"))) int16x4_t vreinterpret_s16_u32(uint32x2_t __p0) {
61988	63962	int16x4_t __ret;
61989		__ret = (int16x4_t)(__p0);
	63963	__ret = __builtin_bit_cast(int16x4_t, __p0);
61990	63964	return __ret;
61991	63965	}
61992	63966	__ai __attribute__((target("neon"))) int16x4_t vreinterpret_s16_u64(uint64x1_t __p0) {
61993	63967	int16x4_t __ret;
61994		__ret = (int16x4_t)(__p0);
	63968	__ret = __builtin_bit_cast(int16x4_t, __p0);
61995	63969	return __ret;
61996	63970	}
61997	63971	__ai __attribute__((target("neon"))) int16x4_t vreinterpret_s16_u16(uint16x4_t __p0) {
61998	63972	int16x4_t __ret;
61999		__ret = (int16x4_t)(__p0);
	63973	__ret = __builtin_bit_cast(int16x4_t, __p0);
62000	63974	return __ret;
62001	63975	}
62002	63976	__ai __attribute__((target("neon"))) int16x4_t vreinterpret_s16_s8(int8x8_t __p0) {
62003	63977	int16x4_t __ret;
62004		__ret = (int16x4_t)(__p0);
	63978	__ret = __builtin_bit_cast(int16x4_t, __p0);
62005	63979	return __ret;
62006	63980	}
62007	63981	__ai __attribute__((target("neon"))) int16x4_t vreinterpret_s16_f64(float64x1_t __p0) {
62008	63982	int16x4_t __ret;
62009		__ret = (int16x4_t)(__p0);
	63983	__ret = __builtin_bit_cast(int16x4_t, __p0);
62010	63984	return __ret;
62011	63985	}
62012	63986	__ai __attribute__((target("neon"))) int16x4_t vreinterpret_s16_f32(float32x2_t __p0) {
62013	63987	int16x4_t __ret;
62014		__ret = (int16x4_t)(__p0);
	63988	__ret = __builtin_bit_cast(int16x4_t, __p0);
62015	63989	return __ret;
62016	63990	}
62017	63991	__ai __attribute__((target("neon"))) int16x4_t vreinterpret_s16_f16(float16x4_t __p0) {
62018	63992	int16x4_t __ret;
62019		__ret = (int16x4_t)(__p0);
	63993	__ret = __builtin_bit_cast(int16x4_t, __p0);
62020	63994	return __ret;
62021	63995	}
62022	63996	__ai __attribute__((target("neon"))) int16x4_t vreinterpret_s16_s32(int32x2_t __p0) {
62023	63997	int16x4_t __ret;
62024		__ret = (int16x4_t)(__p0);
	63998	__ret = __builtin_bit_cast(int16x4_t, __p0);
62025	63999	return __ret;
62026	64000	}
62027	64001	__ai __attribute__((target("neon"))) int16x4_t vreinterpret_s16_s64(int64x1_t __p0) {
62028	64002	int16x4_t __ret;
62029		__ret = (int16x4_t)(__p0);
	64003	__ret = __builtin_bit_cast(int16x4_t, __p0);
62030	64004	return __ret;
62031	64005	}
62032	64006	__ai __attribute__((target("neon"))) int16x4_t vreinterpret_s16_mf8(mfloat8x8_t __p0) {
62033	64007	int16x4_t __ret;
62034		__ret = (int16x4_t)(__p0);
	64008	__ret = __builtin_bit_cast(int16x4_t, __p0);
62035	64009	return __ret;
62036	64010	}
62037	64011	__ai __attribute__((target("neon"))) uint64_t vrshld_u64(uint64_t __p0, int64_t __p1) {
62038	64012	uint64_t __ret;
62039		__ret = (uint64_t) __builtin_neon_vrshld_u64(__p0, __p1);
	64013	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vrshld_u64(__p0, __p1));
62040	64014	return __ret;
62041	64015	}
62042	64016	__ai __attribute__((target("neon"))) int64_t vrshld_s64(int64_t __p0, int64_t __p1) {
62043	64017	int64_t __ret;
62044		__ret = (int64_t) __builtin_neon_vrshld_s64(__p0, __p1);
	64018	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vrshld_s64(__p0, __p1));
62045	64019	return __ret;
62046	64020	}
62047	64021	#define vrshrd_n_u64(__p0, __p1) __extension__ ({ \
62048	64022	uint64_t __ret; \
62049	64023	uint64_t __s0 = __p0; \
62050		__ret = (uint64_t) __builtin_neon_vrshrd_n_u64(__s0, __p1); \
	64024	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vrshrd_n_u64(__s0, __p1)); \
62051	64025	__ret; \
62052	64026	})
62053	64027	#define vrshrd_n_s64(__p0, __p1) __extension__ ({ \
62054	64028	int64_t __ret; \
62055	64029	int64_t __s0 = __p0; \
62056		__ret = (int64_t) __builtin_neon_vrshrd_n_s64(__s0, __p1); \
	64030	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vrshrd_n_s64(__s0, __p1)); \
62057	64031	__ret; \
62058	64032	})
62059	64033	#ifdef __LITTLE_ENDIAN__
62060		#define vrshrn_high_n_u32(__p0_724, __p1_724, __p2_724) __extension__ ({ \
62061		uint16x8_t __ret_724; \
62062		uint16x4_t __s0_724 = __p0_724; \
62063		uint32x4_t __s1_724 = __p1_724; \
62064		__ret_724 = (uint16x8_t)(vcombine_u16((uint16x4_t)(__s0_724), (uint16x4_t)(vrshrn_n_u32(__s1_724, __p2_724)))); \
62065		__ret_724; \
	64034	#define vrshrn_high_n_u32(__p0_740, __p1_740, __p2_740) __extension__ ({ \
	64035	uint16x8_t __ret_740; \
	64036	uint16x4_t __s0_740 = __p0_740; \
	64037	uint32x4_t __s1_740 = __p1_740; \
	64038	__ret_740 = __builtin_bit_cast(uint16x8_t, vcombine_u16(__builtin_bit_cast(uint16x4_t, __s0_740), __builtin_bit_cast(uint16x4_t, vrshrn_n_u32(__s1_740, __p2_740)))); \
	64039	__ret_740; \
62066	64040	})
62067	64041	#else
62068		#define vrshrn_high_n_u32(__p0_725, __p1_725, __p2_725) __extension__ ({ \
62069		uint16x8_t __ret_725; \
62070		uint16x4_t __s0_725 = __p0_725; \
62071		uint32x4_t __s1_725 = __p1_725; \
62072		uint16x4_t __rev0_725; __rev0_725 = __builtin_shufflevector(__s0_725, __s0_725, 3, 2, 1, 0); \
62073		uint32x4_t __rev1_725; __rev1_725 = __builtin_shufflevector(__s1_725, __s1_725, 3, 2, 1, 0); \
62074		__ret_725 = (uint16x8_t)(__noswap_vcombine_u16((uint16x4_t)(__rev0_725), (uint16x4_t)(__noswap_vrshrn_n_u32(__rev1_725, __p2_725)))); \
62075		__ret_725 = __builtin_shufflevector(__ret_725, __ret_725, 7, 6, 5, 4, 3, 2, 1, 0); \
62076		__ret_725; \
	64042	#define vrshrn_high_n_u32(__p0_741, __p1_741, __p2_741) __extension__ ({ \
	64043	uint16x8_t __ret_741; \
	64044	uint16x4_t __s0_741 = __p0_741; \
	64045	uint32x4_t __s1_741 = __p1_741; \
	64046	uint16x4_t __rev0_741; __rev0_741 = __builtin_shufflevector(__s0_741, __s0_741, __lane_reverse_64_16); \
	64047	uint32x4_t __rev1_741; __rev1_741 = __builtin_shufflevector(__s1_741, __s1_741, __lane_reverse_128_32); \
	64048	__ret_741 = __builtin_bit_cast(uint16x8_t, __noswap_vcombine_u16(__builtin_bit_cast(uint16x4_t, __rev0_741), __builtin_bit_cast(uint16x4_t, __noswap_vrshrn_n_u32(__rev1_741, __p2_741)))); \
	64049	__ret_741 = __builtin_shufflevector(__ret_741, __ret_741, __lane_reverse_128_16); \
	64050	__ret_741; \
62077	64051	})
62078	64052	#endif
62079	64053
62080	64054	#ifdef __LITTLE_ENDIAN__
62081		#define vrshrn_high_n_u64(__p0_726, __p1_726, __p2_726) __extension__ ({ \
62082		uint32x4_t __ret_726; \
62083		uint32x2_t __s0_726 = __p0_726; \
62084		uint64x2_t __s1_726 = __p1_726; \
62085		__ret_726 = (uint32x4_t)(vcombine_u32((uint32x2_t)(__s0_726), (uint32x2_t)(vrshrn_n_u64(__s1_726, __p2_726)))); \
62086		__ret_726; \
	64055	#define vrshrn_high_n_u64(__p0_742, __p1_742, __p2_742) __extension__ ({ \
	64056	uint32x4_t __ret_742; \
	64057	uint32x2_t __s0_742 = __p0_742; \
	64058	uint64x2_t __s1_742 = __p1_742; \
	64059	__ret_742 = __builtin_bit_cast(uint32x4_t, vcombine_u32(__builtin_bit_cast(uint32x2_t, __s0_742), __builtin_bit_cast(uint32x2_t, vrshrn_n_u64(__s1_742, __p2_742)))); \
	64060	__ret_742; \
62087	64061	})
62088	64062	#else
62089		#define vrshrn_high_n_u64(__p0_727, __p1_727, __p2_727) __extension__ ({ \
62090		uint32x4_t __ret_727; \
62091		uint32x2_t __s0_727 = __p0_727; \
62092		uint64x2_t __s1_727 = __p1_727; \
62093		uint32x2_t __rev0_727; __rev0_727 = __builtin_shufflevector(__s0_727, __s0_727, 1, 0); \
62094		uint64x2_t __rev1_727; __rev1_727 = __builtin_shufflevector(__s1_727, __s1_727, 1, 0); \
62095		__ret_727 = (uint32x4_t)(__noswap_vcombine_u32((uint32x2_t)(__rev0_727), (uint32x2_t)(__noswap_vrshrn_n_u64(__rev1_727, __p2_727)))); \
62096		__ret_727 = __builtin_shufflevector(__ret_727, __ret_727, 3, 2, 1, 0); \
62097		__ret_727; \
	64063	#define vrshrn_high_n_u64(__p0_743, __p1_743, __p2_743) __extension__ ({ \
	64064	uint32x4_t __ret_743; \
	64065	uint32x2_t __s0_743 = __p0_743; \
	64066	uint64x2_t __s1_743 = __p1_743; \
	64067	uint32x2_t __rev0_743; __rev0_743 = __builtin_shufflevector(__s0_743, __s0_743, __lane_reverse_64_32); \
	64068	uint64x2_t __rev1_743; __rev1_743 = __builtin_shufflevector(__s1_743, __s1_743, __lane_reverse_128_64); \
	64069	__ret_743 = __builtin_bit_cast(uint32x4_t, __noswap_vcombine_u32(__builtin_bit_cast(uint32x2_t, __rev0_743), __builtin_bit_cast(uint32x2_t, __noswap_vrshrn_n_u64(__rev1_743, __p2_743)))); \
	64070	__ret_743 = __builtin_shufflevector(__ret_743, __ret_743, __lane_reverse_128_32); \
	64071	__ret_743; \
62098	64072	})
62099	64073	#endif
62100	64074
62101	64075	#ifdef __LITTLE_ENDIAN__
62102		#define vrshrn_high_n_u16(__p0_728, __p1_728, __p2_728) __extension__ ({ \
62103		uint8x16_t __ret_728; \
62104		uint8x8_t __s0_728 = __p0_728; \
62105		uint16x8_t __s1_728 = __p1_728; \
62106		__ret_728 = (uint8x16_t)(vcombine_u8((uint8x8_t)(__s0_728), (uint8x8_t)(vrshrn_n_u16(__s1_728, __p2_728)))); \
62107		__ret_728; \
	64076	#define vrshrn_high_n_u16(__p0_744, __p1_744, __p2_744) __extension__ ({ \
	64077	uint8x16_t __ret_744; \
	64078	uint8x8_t __s0_744 = __p0_744; \
	64079	uint16x8_t __s1_744 = __p1_744; \
	64080	__ret_744 = __builtin_bit_cast(uint8x16_t, vcombine_u8(__builtin_bit_cast(uint8x8_t, __s0_744), __builtin_bit_cast(uint8x8_t, vrshrn_n_u16(__s1_744, __p2_744)))); \
	64081	__ret_744; \
62108	64082	})
62109	64083	#else
62110		#define vrshrn_high_n_u16(__p0_729, __p1_729, __p2_729) __extension__ ({ \
62111		uint8x16_t __ret_729; \
62112		uint8x8_t __s0_729 = __p0_729; \
62113		uint16x8_t __s1_729 = __p1_729; \
62114		uint8x8_t __rev0_729; __rev0_729 = __builtin_shufflevector(__s0_729, __s0_729, 7, 6, 5, 4, 3, 2, 1, 0); \
62115		uint16x8_t __rev1_729; __rev1_729 = __builtin_shufflevector(__s1_729, __s1_729, 7, 6, 5, 4, 3, 2, 1, 0); \
62116		__ret_729 = (uint8x16_t)(__noswap_vcombine_u8((uint8x8_t)(__rev0_729), (uint8x8_t)(__noswap_vrshrn_n_u16(__rev1_729, __p2_729)))); \
62117		__ret_729 = __builtin_shufflevector(__ret_729, __ret_729, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
62118		__ret_729; \
	64084	#define vrshrn_high_n_u16(__p0_745, __p1_745, __p2_745) __extension__ ({ \
	64085	uint8x16_t __ret_745; \
	64086	uint8x8_t __s0_745 = __p0_745; \
	64087	uint16x8_t __s1_745 = __p1_745; \
	64088	uint8x8_t __rev0_745; __rev0_745 = __builtin_shufflevector(__s0_745, __s0_745, __lane_reverse_64_8); \
	64089	uint16x8_t __rev1_745; __rev1_745 = __builtin_shufflevector(__s1_745, __s1_745, __lane_reverse_128_16); \
	64090	__ret_745 = __builtin_bit_cast(uint8x16_t, __noswap_vcombine_u8(__builtin_bit_cast(uint8x8_t, __rev0_745), __builtin_bit_cast(uint8x8_t, __noswap_vrshrn_n_u16(__rev1_745, __p2_745)))); \
	64091	__ret_745 = __builtin_shufflevector(__ret_745, __ret_745, __lane_reverse_128_8); \
	64092	__ret_745; \
62119	64093	})
62120	64094	#endif
62121	64095
62122	64096	#ifdef __LITTLE_ENDIAN__
62123		#define vrshrn_high_n_s32(__p0_730, __p1_730, __p2_730) __extension__ ({ \
62124		int16x8_t __ret_730; \
62125		int16x4_t __s0_730 = __p0_730; \
62126		int32x4_t __s1_730 = __p1_730; \
62127		__ret_730 = (int16x8_t)(vcombine_s16((int16x4_t)(__s0_730), (int16x4_t)(vrshrn_n_s32(__s1_730, __p2_730)))); \
62128		__ret_730; \
	64097	#define vrshrn_high_n_s32(__p0_746, __p1_746, __p2_746) __extension__ ({ \
	64098	int16x8_t __ret_746; \
	64099	int16x4_t __s0_746 = __p0_746; \
	64100	int32x4_t __s1_746 = __p1_746; \
	64101	__ret_746 = __builtin_bit_cast(int16x8_t, vcombine_s16(__builtin_bit_cast(int16x4_t, __s0_746), __builtin_bit_cast(int16x4_t, vrshrn_n_s32(__s1_746, __p2_746)))); \
	64102	__ret_746; \
62129	64103	})
62130	64104	#else
62131		#define vrshrn_high_n_s32(__p0_731, __p1_731, __p2_731) __extension__ ({ \
62132		int16x8_t __ret_731; \
62133		int16x4_t __s0_731 = __p0_731; \
62134		int32x4_t __s1_731 = __p1_731; \
62135		int16x4_t __rev0_731; __rev0_731 = __builtin_shufflevector(__s0_731, __s0_731, 3, 2, 1, 0); \
62136		int32x4_t __rev1_731; __rev1_731 = __builtin_shufflevector(__s1_731, __s1_731, 3, 2, 1, 0); \
62137		__ret_731 = (int16x8_t)(__noswap_vcombine_s16((int16x4_t)(__rev0_731), (int16x4_t)(__noswap_vrshrn_n_s32(__rev1_731, __p2_731)))); \
62138		__ret_731 = __builtin_shufflevector(__ret_731, __ret_731, 7, 6, 5, 4, 3, 2, 1, 0); \
62139		__ret_731; \
	64105	#define vrshrn_high_n_s32(__p0_747, __p1_747, __p2_747) __extension__ ({ \
	64106	int16x8_t __ret_747; \
	64107	int16x4_t __s0_747 = __p0_747; \
	64108	int32x4_t __s1_747 = __p1_747; \
	64109	int16x4_t __rev0_747; __rev0_747 = __builtin_shufflevector(__s0_747, __s0_747, __lane_reverse_64_16); \
	64110	int32x4_t __rev1_747; __rev1_747 = __builtin_shufflevector(__s1_747, __s1_747, __lane_reverse_128_32); \
	64111	__ret_747 = __builtin_bit_cast(int16x8_t, __noswap_vcombine_s16(__builtin_bit_cast(int16x4_t, __rev0_747), __builtin_bit_cast(int16x4_t, __noswap_vrshrn_n_s32(__rev1_747, __p2_747)))); \
	64112	__ret_747 = __builtin_shufflevector(__ret_747, __ret_747, __lane_reverse_128_16); \
	64113	__ret_747; \
62140	64114	})
62141	64115	#endif
62142	64116
62143	64117	#ifdef __LITTLE_ENDIAN__
62144		#define vrshrn_high_n_s64(__p0_732, __p1_732, __p2_732) __extension__ ({ \
62145		int32x4_t __ret_732; \
62146		int32x2_t __s0_732 = __p0_732; \
62147		int64x2_t __s1_732 = __p1_732; \
62148		__ret_732 = (int32x4_t)(vcombine_s32((int32x2_t)(__s0_732), (int32x2_t)(vrshrn_n_s64(__s1_732, __p2_732)))); \
62149		__ret_732; \
	64118	#define vrshrn_high_n_s64(__p0_748, __p1_748, __p2_748) __extension__ ({ \
	64119	int32x4_t __ret_748; \
	64120	int32x2_t __s0_748 = __p0_748; \
	64121	int64x2_t __s1_748 = __p1_748; \
	64122	__ret_748 = __builtin_bit_cast(int32x4_t, vcombine_s32(__builtin_bit_cast(int32x2_t, __s0_748), __builtin_bit_cast(int32x2_t, vrshrn_n_s64(__s1_748, __p2_748)))); \
	64123	__ret_748; \
62150	64124	})
62151	64125	#else
62152		#define vrshrn_high_n_s64(__p0_733, __p1_733, __p2_733) __extension__ ({ \
62153		int32x4_t __ret_733; \
62154		int32x2_t __s0_733 = __p0_733; \
62155		int64x2_t __s1_733 = __p1_733; \
62156		int32x2_t __rev0_733; __rev0_733 = __builtin_shufflevector(__s0_733, __s0_733, 1, 0); \
62157		int64x2_t __rev1_733; __rev1_733 = __builtin_shufflevector(__s1_733, __s1_733, 1, 0); \
62158		__ret_733 = (int32x4_t)(__noswap_vcombine_s32((int32x2_t)(__rev0_733), (int32x2_t)(__noswap_vrshrn_n_s64(__rev1_733, __p2_733)))); \
62159		__ret_733 = __builtin_shufflevector(__ret_733, __ret_733, 3, 2, 1, 0); \
62160		__ret_733; \
	64126	#define vrshrn_high_n_s64(__p0_749, __p1_749, __p2_749) __extension__ ({ \
	64127	int32x4_t __ret_749; \
	64128	int32x2_t __s0_749 = __p0_749; \
	64129	int64x2_t __s1_749 = __p1_749; \
	64130	int32x2_t __rev0_749; __rev0_749 = __builtin_shufflevector(__s0_749, __s0_749, __lane_reverse_64_32); \
	64131	int64x2_t __rev1_749; __rev1_749 = __builtin_shufflevector(__s1_749, __s1_749, __lane_reverse_128_64); \
	64132	__ret_749 = __builtin_bit_cast(int32x4_t, __noswap_vcombine_s32(__builtin_bit_cast(int32x2_t, __rev0_749), __builtin_bit_cast(int32x2_t, __noswap_vrshrn_n_s64(__rev1_749, __p2_749)))); \
	64133	__ret_749 = __builtin_shufflevector(__ret_749, __ret_749, __lane_reverse_128_32); \
	64134	__ret_749; \
62161	64135	})
62162	64136	#endif
62163	64137
62164	64138	#ifdef __LITTLE_ENDIAN__
62165		#define vrshrn_high_n_s16(__p0_734, __p1_734, __p2_734) __extension__ ({ \
62166		int8x16_t __ret_734; \
62167		int8x8_t __s0_734 = __p0_734; \
62168		int16x8_t __s1_734 = __p1_734; \
62169		__ret_734 = (int8x16_t)(vcombine_s8((int8x8_t)(__s0_734), (int8x8_t)(vrshrn_n_s16(__s1_734, __p2_734)))); \
62170		__ret_734; \
	64139	#define vrshrn_high_n_s16(__p0_750, __p1_750, __p2_750) __extension__ ({ \
	64140	int8x16_t __ret_750; \
	64141	int8x8_t __s0_750 = __p0_750; \
	64142	int16x8_t __s1_750 = __p1_750; \
	64143	__ret_750 = __builtin_bit_cast(int8x16_t, vcombine_s8(__builtin_bit_cast(int8x8_t, __s0_750), __builtin_bit_cast(int8x8_t, vrshrn_n_s16(__s1_750, __p2_750)))); \
	64144	__ret_750; \
62171	64145	})
62172	64146	#else
62173		#define vrshrn_high_n_s16(__p0_735, __p1_735, __p2_735) __extension__ ({ \
62174		int8x16_t __ret_735; \
62175		int8x8_t __s0_735 = __p0_735; \
62176		int16x8_t __s1_735 = __p1_735; \
62177		int8x8_t __rev0_735; __rev0_735 = __builtin_shufflevector(__s0_735, __s0_735, 7, 6, 5, 4, 3, 2, 1, 0); \
62178		int16x8_t __rev1_735; __rev1_735 = __builtin_shufflevector(__s1_735, __s1_735, 7, 6, 5, 4, 3, 2, 1, 0); \
62179		__ret_735 = (int8x16_t)(__noswap_vcombine_s8((int8x8_t)(__rev0_735), (int8x8_t)(__noswap_vrshrn_n_s16(__rev1_735, __p2_735)))); \
62180		__ret_735 = __builtin_shufflevector(__ret_735, __ret_735, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
62181		__ret_735; \
	64147	#define vrshrn_high_n_s16(__p0_751, __p1_751, __p2_751) __extension__ ({ \
	64148	int8x16_t __ret_751; \
	64149	int8x8_t __s0_751 = __p0_751; \
	64150	int16x8_t __s1_751 = __p1_751; \
	64151	int8x8_t __rev0_751; __rev0_751 = __builtin_shufflevector(__s0_751, __s0_751, __lane_reverse_64_8); \
	64152	int16x8_t __rev1_751; __rev1_751 = __builtin_shufflevector(__s1_751, __s1_751, __lane_reverse_128_16); \
	64153	__ret_751 = __builtin_bit_cast(int8x16_t, __noswap_vcombine_s8(__builtin_bit_cast(int8x8_t, __rev0_751), __builtin_bit_cast(int8x8_t, __noswap_vrshrn_n_s16(__rev1_751, __p2_751)))); \
	64154	__ret_751 = __builtin_shufflevector(__ret_751, __ret_751, __lane_reverse_128_8); \
	64155	__ret_751; \
62182	64156	})
62183	64157	#endif
62184	64158
62185	64159	#ifdef __LITTLE_ENDIAN__
62186	64160	__ai __attribute__((target("neon"))) float64x2_t vrsqrteq_f64(float64x2_t __p0) {
62187	64161	float64x2_t __ret;
62188		__ret = (float64x2_t) __builtin_neon_vrsqrteq_v((int8x16_t)__p0, 42);
	64162	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vrsqrteq_v(__builtin_bit_cast(int8x16_t, __p0), 42));
62189	64163	return __ret;
62190	64164	}
62191	64165	#else
62192	64166	__ai __attribute__((target("neon"))) float64x2_t vrsqrteq_f64(float64x2_t __p0) {
62193	64167	float64x2_t __ret;
62194		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
62195		__ret = (float64x2_t) __builtin_neon_vrsqrteq_v((int8x16_t)__rev0, 42);
62196		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	64168	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	64169	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vrsqrteq_v(__builtin_bit_cast(int8x16_t, __rev0), 42));
	64170	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
62197	64171	return __ret;
62198	64172	}
62199	64173	#endif
62200	64174
62201	64175	__ai __attribute__((target("neon"))) float64x1_t vrsqrte_f64(float64x1_t __p0) {
62202	64176	float64x1_t __ret;
62203		__ret = (float64x1_t) __builtin_neon_vrsqrte_v((int8x8_t)__p0, 10);
	64177	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vrsqrte_v(__builtin_bit_cast(int8x8_t, __p0), 10));
62204	64178	return __ret;
62205	64179	}
62206	64180	__ai __attribute__((target("neon"))) float64_t vrsqrted_f64(float64_t __p0) {
62207	64181	float64_t __ret;
62208		__ret = (float64_t) __builtin_neon_vrsqrted_f64(__p0);
	64182	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vrsqrted_f64(__p0));
62209	64183	return __ret;
62210	64184	}
62211	64185	__ai __attribute__((target("neon"))) float32_t vrsqrtes_f32(float32_t __p0) {
62212	64186	float32_t __ret;
62213		__ret = (float32_t) __builtin_neon_vrsqrtes_f32(__p0);
	64187	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vrsqrtes_f32(__p0));
62214	64188	return __ret;
62215	64189	}
62216	64190	#ifdef __LITTLE_ENDIAN__
62217	64191	__ai __attribute__((target("neon"))) float64x2_t vrsqrtsq_f64(float64x2_t __p0, float64x2_t __p1) {
62218	64192	float64x2_t __ret;
62219		__ret = (float64x2_t) __builtin_neon_vrsqrtsq_v((int8x16_t)__p0, (int8x16_t)__p1, 42);
	64193	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vrsqrtsq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 42));
62220	64194	return __ret;
62221	64195	}
62222	64196	#else
62223	64197	__ai __attribute__((target("neon"))) float64x2_t vrsqrtsq_f64(float64x2_t __p0, float64x2_t __p1) {
62224	64198	float64x2_t __ret;
62225		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
62226		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
62227		__ret = (float64x2_t) __builtin_neon_vrsqrtsq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 42);
62228		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	64199	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	64200	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	64201	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vrsqrtsq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 42));
	64202	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
62229	64203	return __ret;
62230	64204	}
62231	64205	#endif
62232	64206
62233	64207	__ai __attribute__((target("neon"))) float64x1_t vrsqrts_f64(float64x1_t __p0, float64x1_t __p1) {
62234	64208	float64x1_t __ret;
62235		__ret = (float64x1_t) __builtin_neon_vrsqrts_v((int8x8_t)__p0, (int8x8_t)__p1, 10);
	64209	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vrsqrts_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 10));
62236	64210	return __ret;
62237	64211	}
62238	64212	__ai __attribute__((target("neon"))) float64_t vrsqrtsd_f64(float64_t __p0, float64_t __p1) {
62239	64213	float64_t __ret;
62240		__ret = (float64_t) __builtin_neon_vrsqrtsd_f64(__p0, __p1);
	64214	__ret = __builtin_bit_cast(float64_t, __builtin_neon_vrsqrtsd_f64(__p0, __p1));
62241	64215	return __ret;
62242	64216	}
62243	64217	__ai __attribute__((target("neon"))) float32_t vrsqrtss_f32(float32_t __p0, float32_t __p1) {
62244	64218	float32_t __ret;
62245		__ret = (float32_t) __builtin_neon_vrsqrtss_f32(__p0, __p1);
	64219	__ret = __builtin_bit_cast(float32_t, __builtin_neon_vrsqrtss_f32(__p0, __p1));
62246	64220	return __ret;
62247	64221	}
62248	64222	#define vrsrad_n_u64(__p0, __p1, __p2) __extension__ ({ \
62249	64223	uint64_t __ret; \
62250	64224	uint64_t __s0 = __p0; \
62251	64225	uint64_t __s1 = __p1; \
62252		__ret = (uint64_t) __builtin_neon_vrsrad_n_u64(__s0, __s1, __p2); \
	64226	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vrsrad_n_u64(__s0, __s1, __p2)); \
62253	64227	__ret; \
62254	64228	})
62255	64229	#define vrsrad_n_s64(__p0, __p1, __p2) __extension__ ({ \
62256	64230	int64_t __ret; \
62257	64231	int64_t __s0 = __p0; \
62258	64232	int64_t __s1 = __p1; \
62259		__ret = (int64_t) __builtin_neon_vrsrad_n_s64(__s0, __s1, __p2); \
	64233	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vrsrad_n_s64(__s0, __s1, __p2)); \
62260	64234	__ret; \
62261	64235	})
62262	64236	#ifdef __LITTLE_ENDIAN__
...	...	@@ -62268,11 +64242,11 @@ __ai __attribute__((target("neon"))) uint16x8_t vrsubhn_high_u32(uint16x4_t __p0
62268	64242	#else
62269	64243	__ai __attribute__((target("neon"))) uint16x8_t vrsubhn_high_u32(uint16x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
62270	64244	uint16x8_t __ret;
62271		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
62272		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
62273		uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	64245	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	64246	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	64247	uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
62274	64248	__ret = __noswap_vcombine_u16(__rev0, __noswap_vrsubhn_u32(__rev1, __rev2));
62275		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	64249	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
62276	64250	return __ret;
62277	64251	}
62278	64252	#endif
...	...	@@ -62286,11 +64260,11 @@ __ai __attribute__((target("neon"))) uint32x4_t vrsubhn_high_u64(uint32x2_t __p0
62286	64260	#else
62287	64261	__ai __attribute__((target("neon"))) uint32x4_t vrsubhn_high_u64(uint32x2_t __p0, uint64x2_t __p1, uint64x2_t __p2) {
62288	64262	uint32x4_t __ret;
62289		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
62290		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
62291		uint64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
	64263	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	64264	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	64265	uint64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_64);
62292	64266	__ret = __noswap_vcombine_u32(__rev0, __noswap_vrsubhn_u64(__rev1, __rev2));
62293		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	64267	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
62294	64268	return __ret;
62295	64269	}
62296	64270	#endif
...	...	@@ -62304,11 +64278,11 @@ __ai __attribute__((target("neon"))) uint8x16_t vrsubhn_high_u16(uint8x8_t __p0,
62304	64278	#else
62305	64279	__ai __attribute__((target("neon"))) uint8x16_t vrsubhn_high_u16(uint8x8_t __p0, uint16x8_t __p1, uint16x8_t __p2) {
62306	64280	uint8x16_t __ret;
62307		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
62308		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
62309		uint16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	64281	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	64282	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	64283	uint16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
62310	64284	__ret = __noswap_vcombine_u8(__rev0, __noswap_vrsubhn_u16(__rev1, __rev2));
62311		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	64285	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
62312	64286	return __ret;
62313	64287	}
62314	64288	#endif
...	...	@@ -62322,11 +64296,11 @@ __ai __attribute__((target("neon"))) int16x8_t vrsubhn_high_s32(int16x4_t __p0,
62322	64296	#else
62323	64297	__ai __attribute__((target("neon"))) int16x8_t vrsubhn_high_s32(int16x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
62324	64298	int16x8_t __ret;
62325		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
62326		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
62327		int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	64299	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	64300	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	64301	int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
62328	64302	__ret = __noswap_vcombine_s16(__rev0, __noswap_vrsubhn_s32(__rev1, __rev2));
62329		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	64303	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
62330	64304	return __ret;
62331	64305	}
62332	64306	#endif
...	...	@@ -62340,11 +64314,11 @@ __ai __attribute__((target("neon"))) int32x4_t vrsubhn_high_s64(int32x2_t __p0,
62340	64314	#else
62341	64315	__ai __attribute__((target("neon"))) int32x4_t vrsubhn_high_s64(int32x2_t __p0, int64x2_t __p1, int64x2_t __p2) {
62342	64316	int32x4_t __ret;
62343		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
62344		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
62345		int64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
	64317	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	64318	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	64319	int64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_64);
62346	64320	__ret = __noswap_vcombine_s32(__rev0, __noswap_vrsubhn_s64(__rev1, __rev2));
62347		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	64321	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
62348	64322	return __ret;
62349	64323	}
62350	64324	#endif
...	...	@@ -62358,11 +64332,11 @@ __ai __attribute__((target("neon"))) int8x16_t vrsubhn_high_s16(int8x8_t __p0, i
62358	64332	#else
62359	64333	__ai __attribute__((target("neon"))) int8x16_t vrsubhn_high_s16(int8x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
62360	64334	int8x16_t __ret;
62361		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
62362		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
62363		int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	64335	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	64336	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	64337	int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
62364	64338	__ret = __noswap_vcombine_s8(__rev0, __noswap_vrsubhn_s16(__rev1, __rev2));
62365		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	64339	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
62366	64340	return __ret;
62367	64341	}
62368	64342	#endif
...	...	@@ -62371,7 +64345,7 @@ __ai __attribute__((target("neon"))) int8x16_t vrsubhn_high_s16(int8x8_t __p0, i
62371	64345	poly64x1_t __ret; \
62372	64346	poly64_t __s0 = __p0; \
62373	64347	poly64x1_t __s1 = __p1; \
62374		__ret = (poly64x1_t) __builtin_neon_vset_lane_i64(__s0, (poly64x1_t)__s1, __p2); \
	64348	__ret = __builtin_bit_cast(poly64x1_t, __builtin_neon_vset_lane_i64(__s0, __s1, __p2)); \
62375	64349	__ret; \
62376	64350	})
62377	64351	#ifdef __LITTLE_ENDIAN__
...	...	@@ -62379,7 +64353,7 @@ __ai __attribute__((target("neon"))) int8x16_t vrsubhn_high_s16(int8x8_t __p0, i
62379	64353	poly64x2_t __ret; \
62380	64354	poly64_t __s0 = __p0; \
62381	64355	poly64x2_t __s1 = __p1; \
62382		__ret = (poly64x2_t) __builtin_neon_vsetq_lane_i64(__s0, (poly64x2_t)__s1, __p2); \
	64356	__ret = __builtin_bit_cast(poly64x2_t, __builtin_neon_vsetq_lane_i64(__s0, __s1, __p2)); \
62383	64357	__ret; \
62384	64358	})
62385	64359	#else
...	...	@@ -62387,16 +64361,16 @@ __ai __attribute__((target("neon"))) int8x16_t vrsubhn_high_s16(int8x8_t __p0, i
62387	64361	poly64x2_t __ret; \
62388	64362	poly64_t __s0 = __p0; \
62389	64363	poly64x2_t __s1 = __p1; \
62390		poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
62391		__ret = (poly64x2_t) __builtin_neon_vsetq_lane_i64(__s0, (poly64x2_t)__rev1, __p2); \
62392		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	64364	poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	64365	__ret = __builtin_bit_cast(poly64x2_t, __builtin_neon_vsetq_lane_i64(__s0, __rev1, __p2)); \
	64366	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
62393	64367	__ret; \
62394	64368	})
62395	64369	#define __noswap_vsetq_lane_p64(__p0, __p1, __p2) __extension__ ({ \
62396	64370	poly64x2_t __ret; \
62397	64371	poly64_t __s0 = __p0; \
62398	64372	poly64x2_t __s1 = __p1; \
62399		__ret = (poly64x2_t) __builtin_neon_vsetq_lane_i64(__s0, (poly64x2_t)__s1, __p2); \
	64373	__ret = __builtin_bit_cast(poly64x2_t, __builtin_neon_vsetq_lane_i64(__s0, __s1, __p2)); \
62400	64374	__ret; \
62401	64375	})
62402	64376	#endif
...	...	@@ -62406,7 +64380,7 @@ __ai __attribute__((target("neon"))) int8x16_t vrsubhn_high_s16(int8x8_t __p0, i
62406	64380	float64x2_t __ret; \
62407	64381	float64_t __s0 = __p0; \
62408	64382	float64x2_t __s1 = __p1; \
62409		__ret = (float64x2_t) __builtin_neon_vsetq_lane_f64(__s0, (float64x2_t)__s1, __p2); \
	64383	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vsetq_lane_f64(__s0, __s1, __p2)); \
62410	64384	__ret; \
62411	64385	})
62412	64386	#else
...	...	@@ -62414,16 +64388,16 @@ __ai __attribute__((target("neon"))) int8x16_t vrsubhn_high_s16(int8x8_t __p0, i
62414	64388	float64x2_t __ret; \
62415	64389	float64_t __s0 = __p0; \
62416	64390	float64x2_t __s1 = __p1; \
62417		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
62418		__ret = (float64x2_t) __builtin_neon_vsetq_lane_f64(__s0, (float64x2_t)__rev1, __p2); \
62419		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	64391	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	64392	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vsetq_lane_f64(__s0, __rev1, __p2)); \
	64393	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
62420	64394	__ret; \
62421	64395	})
62422	64396	#define __noswap_vsetq_lane_f64(__p0, __p1, __p2) __extension__ ({ \
62423	64397	float64x2_t __ret; \
62424	64398	float64_t __s0 = __p0; \
62425	64399	float64x2_t __s1 = __p1; \
62426		__ret = (float64x2_t) __builtin_neon_vsetq_lane_f64(__s0, (float64x2_t)__s1, __p2); \
	64400	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vsetq_lane_f64(__s0, __s1, __p2)); \
62427	64401	__ret; \
62428	64402	})
62429	64403	#endif
...	...	@@ -62432,274 +64406,274 @@ __ai __attribute__((target("neon"))) int8x16_t vrsubhn_high_s16(int8x8_t __p0, i
62432	64406	float64x1_t __ret; \
62433	64407	float64_t __s0 = __p0; \
62434	64408	float64x1_t __s1 = __p1; \
62435		__ret = (float64x1_t) __builtin_neon_vset_lane_f64(__s0, (float64x1_t)__s1, __p2); \
	64409	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vset_lane_f64(__s0, __s1, __p2)); \
62436	64410	__ret; \
62437	64411	})
62438	64412	__ai __attribute__((target("neon"))) uint64_t vshld_u64(uint64_t __p0, int64_t __p1) {
62439	64413	uint64_t __ret;
62440		__ret = (uint64_t) __builtin_neon_vshld_u64(__p0, __p1);
	64414	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vshld_u64(__p0, __p1));
62441	64415	return __ret;
62442	64416	}
62443	64417	__ai __attribute__((target("neon"))) int64_t vshld_s64(int64_t __p0, int64_t __p1) {
62444	64418	int64_t __ret;
62445		__ret = (int64_t) __builtin_neon_vshld_s64(__p0, __p1);
	64419	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vshld_s64(__p0, __p1));
62446	64420	return __ret;
62447	64421	}
62448	64422	#define vshld_n_u64(__p0, __p1) __extension__ ({ \
62449	64423	uint64_t __ret; \
62450	64424	uint64_t __s0 = __p0; \
62451		__ret = (uint64_t) __builtin_neon_vshld_n_u64(__s0, __p1); \
	64425	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vshld_n_u64(__s0, __p1)); \
62452	64426	__ret; \
62453	64427	})
62454	64428	#define vshld_n_s64(__p0, __p1) __extension__ ({ \
62455	64429	int64_t __ret; \
62456	64430	int64_t __s0 = __p0; \
62457		__ret = (int64_t) __builtin_neon_vshld_n_s64(__s0, __p1); \
	64431	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vshld_n_s64(__s0, __p1)); \
62458	64432	__ret; \
62459	64433	})
62460	64434	#ifdef __LITTLE_ENDIAN__
62461		#define vshll_high_n_u8(__p0_736, __p1_736) __extension__ ({ \
62462		uint16x8_t __ret_736; \
62463		uint8x16_t __s0_736 = __p0_736; \
62464		__ret_736 = (uint16x8_t)(vshll_n_u8(vget_high_u8(__s0_736), __p1_736)); \
62465		__ret_736; \
	64435	#define vshll_high_n_u8(__p0_752, __p1_752) __extension__ ({ \
	64436	uint16x8_t __ret_752; \
	64437	uint8x16_t __s0_752 = __p0_752; \
	64438	__ret_752 = __builtin_bit_cast(uint16x8_t, vshll_n_u8(vget_high_u8(__s0_752), __p1_752)); \
	64439	__ret_752; \
62466	64440	})
62467	64441	#else
62468		#define vshll_high_n_u8(__p0_737, __p1_737) __extension__ ({ \
62469		uint16x8_t __ret_737; \
62470		uint8x16_t __s0_737 = __p0_737; \
62471		uint8x16_t __rev0_737; __rev0_737 = __builtin_shufflevector(__s0_737, __s0_737, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
62472		__ret_737 = (uint16x8_t)(__noswap_vshll_n_u8(__noswap_vget_high_u8(__rev0_737), __p1_737)); \
62473		__ret_737 = __builtin_shufflevector(__ret_737, __ret_737, 7, 6, 5, 4, 3, 2, 1, 0); \
62474		__ret_737; \
	64442	#define vshll_high_n_u8(__p0_753, __p1_753) __extension__ ({ \
	64443	uint16x8_t __ret_753; \
	64444	uint8x16_t __s0_753 = __p0_753; \
	64445	uint8x16_t __rev0_753; __rev0_753 = __builtin_shufflevector(__s0_753, __s0_753, __lane_reverse_128_8); \
	64446	__ret_753 = __builtin_bit_cast(uint16x8_t, __noswap_vshll_n_u8(__noswap_vget_high_u8(__rev0_753), __p1_753)); \
	64447	__ret_753 = __builtin_shufflevector(__ret_753, __ret_753, __lane_reverse_128_16); \
	64448	__ret_753; \
62475	64449	})
62476	64450	#endif
62477	64451
62478	64452	#ifdef __LITTLE_ENDIAN__
62479		#define vshll_high_n_u32(__p0_738, __p1_738) __extension__ ({ \
62480		uint64x2_t __ret_738; \
62481		uint32x4_t __s0_738 = __p0_738; \
62482		__ret_738 = (uint64x2_t)(vshll_n_u32(vget_high_u32(__s0_738), __p1_738)); \
62483		__ret_738; \
	64453	#define vshll_high_n_u32(__p0_754, __p1_754) __extension__ ({ \
	64454	uint64x2_t __ret_754; \
	64455	uint32x4_t __s0_754 = __p0_754; \
	64456	__ret_754 = __builtin_bit_cast(uint64x2_t, vshll_n_u32(vget_high_u32(__s0_754), __p1_754)); \
	64457	__ret_754; \
62484	64458	})
62485	64459	#else
62486		#define vshll_high_n_u32(__p0_739, __p1_739) __extension__ ({ \
62487		uint64x2_t __ret_739; \
62488		uint32x4_t __s0_739 = __p0_739; \
62489		uint32x4_t __rev0_739; __rev0_739 = __builtin_shufflevector(__s0_739, __s0_739, 3, 2, 1, 0); \
62490		__ret_739 = (uint64x2_t)(__noswap_vshll_n_u32(__noswap_vget_high_u32(__rev0_739), __p1_739)); \
62491		__ret_739 = __builtin_shufflevector(__ret_739, __ret_739, 1, 0); \
62492		__ret_739; \
	64460	#define vshll_high_n_u32(__p0_755, __p1_755) __extension__ ({ \
	64461	uint64x2_t __ret_755; \
	64462	uint32x4_t __s0_755 = __p0_755; \
	64463	uint32x4_t __rev0_755; __rev0_755 = __builtin_shufflevector(__s0_755, __s0_755, __lane_reverse_128_32); \
	64464	__ret_755 = __builtin_bit_cast(uint64x2_t, __noswap_vshll_n_u32(__noswap_vget_high_u32(__rev0_755), __p1_755)); \
	64465	__ret_755 = __builtin_shufflevector(__ret_755, __ret_755, __lane_reverse_128_64); \
	64466	__ret_755; \
62493	64467	})
62494	64468	#endif
62495	64469
62496	64470	#ifdef __LITTLE_ENDIAN__
62497		#define vshll_high_n_u16(__p0_740, __p1_740) __extension__ ({ \
62498		uint32x4_t __ret_740; \
62499		uint16x8_t __s0_740 = __p0_740; \
62500		__ret_740 = (uint32x4_t)(vshll_n_u16(vget_high_u16(__s0_740), __p1_740)); \
62501		__ret_740; \
	64471	#define vshll_high_n_u16(__p0_756, __p1_756) __extension__ ({ \
	64472	uint32x4_t __ret_756; \
	64473	uint16x8_t __s0_756 = __p0_756; \
	64474	__ret_756 = __builtin_bit_cast(uint32x4_t, vshll_n_u16(vget_high_u16(__s0_756), __p1_756)); \
	64475	__ret_756; \
62502	64476	})
62503	64477	#else
62504		#define vshll_high_n_u16(__p0_741, __p1_741) __extension__ ({ \
62505		uint32x4_t __ret_741; \
62506		uint16x8_t __s0_741 = __p0_741; \
62507		uint16x8_t __rev0_741; __rev0_741 = __builtin_shufflevector(__s0_741, __s0_741, 7, 6, 5, 4, 3, 2, 1, 0); \
62508		__ret_741 = (uint32x4_t)(__noswap_vshll_n_u16(__noswap_vget_high_u16(__rev0_741), __p1_741)); \
62509		__ret_741 = __builtin_shufflevector(__ret_741, __ret_741, 3, 2, 1, 0); \
62510		__ret_741; \
	64478	#define vshll_high_n_u16(__p0_757, __p1_757) __extension__ ({ \
	64479	uint32x4_t __ret_757; \
	64480	uint16x8_t __s0_757 = __p0_757; \
	64481	uint16x8_t __rev0_757; __rev0_757 = __builtin_shufflevector(__s0_757, __s0_757, __lane_reverse_128_16); \
	64482	__ret_757 = __builtin_bit_cast(uint32x4_t, __noswap_vshll_n_u16(__noswap_vget_high_u16(__rev0_757), __p1_757)); \
	64483	__ret_757 = __builtin_shufflevector(__ret_757, __ret_757, __lane_reverse_128_32); \
	64484	__ret_757; \
62511	64485	})
62512	64486	#endif
62513	64487
62514	64488	#ifdef __LITTLE_ENDIAN__
62515		#define vshll_high_n_s8(__p0_742, __p1_742) __extension__ ({ \
62516		int16x8_t __ret_742; \
62517		int8x16_t __s0_742 = __p0_742; \
62518		__ret_742 = (int16x8_t)(vshll_n_s8(vget_high_s8(__s0_742), __p1_742)); \
62519		__ret_742; \
	64489	#define vshll_high_n_s8(__p0_758, __p1_758) __extension__ ({ \
	64490	int16x8_t __ret_758; \
	64491	int8x16_t __s0_758 = __p0_758; \
	64492	__ret_758 = __builtin_bit_cast(int16x8_t, vshll_n_s8(vget_high_s8(__s0_758), __p1_758)); \
	64493	__ret_758; \
62520	64494	})
62521	64495	#else
62522		#define vshll_high_n_s8(__p0_743, __p1_743) __extension__ ({ \
62523		int16x8_t __ret_743; \
62524		int8x16_t __s0_743 = __p0_743; \
62525		int8x16_t __rev0_743; __rev0_743 = __builtin_shufflevector(__s0_743, __s0_743, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
62526		__ret_743 = (int16x8_t)(__noswap_vshll_n_s8(__noswap_vget_high_s8(__rev0_743), __p1_743)); \
62527		__ret_743 = __builtin_shufflevector(__ret_743, __ret_743, 7, 6, 5, 4, 3, 2, 1, 0); \
62528		__ret_743; \
	64496	#define vshll_high_n_s8(__p0_759, __p1_759) __extension__ ({ \
	64497	int16x8_t __ret_759; \
	64498	int8x16_t __s0_759 = __p0_759; \
	64499	int8x16_t __rev0_759; __rev0_759 = __builtin_shufflevector(__s0_759, __s0_759, __lane_reverse_128_8); \
	64500	__ret_759 = __builtin_bit_cast(int16x8_t, __noswap_vshll_n_s8(__noswap_vget_high_s8(__rev0_759), __p1_759)); \
	64501	__ret_759 = __builtin_shufflevector(__ret_759, __ret_759, __lane_reverse_128_16); \
	64502	__ret_759; \
62529	64503	})
62530	64504	#endif
62531	64505
62532	64506	#ifdef __LITTLE_ENDIAN__
62533		#define vshll_high_n_s32(__p0_744, __p1_744) __extension__ ({ \
62534		int64x2_t __ret_744; \
62535		int32x4_t __s0_744 = __p0_744; \
62536		__ret_744 = (int64x2_t)(vshll_n_s32(vget_high_s32(__s0_744), __p1_744)); \
62537		__ret_744; \
	64507	#define vshll_high_n_s32(__p0_760, __p1_760) __extension__ ({ \
	64508	int64x2_t __ret_760; \
	64509	int32x4_t __s0_760 = __p0_760; \
	64510	__ret_760 = __builtin_bit_cast(int64x2_t, vshll_n_s32(vget_high_s32(__s0_760), __p1_760)); \
	64511	__ret_760; \
62538	64512	})
62539	64513	#else
62540		#define vshll_high_n_s32(__p0_745, __p1_745) __extension__ ({ \
62541		int64x2_t __ret_745; \
62542		int32x4_t __s0_745 = __p0_745; \
62543		int32x4_t __rev0_745; __rev0_745 = __builtin_shufflevector(__s0_745, __s0_745, 3, 2, 1, 0); \
62544		__ret_745 = (int64x2_t)(__noswap_vshll_n_s32(__noswap_vget_high_s32(__rev0_745), __p1_745)); \
62545		__ret_745 = __builtin_shufflevector(__ret_745, __ret_745, 1, 0); \
62546		__ret_745; \
	64514	#define vshll_high_n_s32(__p0_761, __p1_761) __extension__ ({ \
	64515	int64x2_t __ret_761; \
	64516	int32x4_t __s0_761 = __p0_761; \
	64517	int32x4_t __rev0_761; __rev0_761 = __builtin_shufflevector(__s0_761, __s0_761, __lane_reverse_128_32); \
	64518	__ret_761 = __builtin_bit_cast(int64x2_t, __noswap_vshll_n_s32(__noswap_vget_high_s32(__rev0_761), __p1_761)); \
	64519	__ret_761 = __builtin_shufflevector(__ret_761, __ret_761, __lane_reverse_128_64); \
	64520	__ret_761; \
62547	64521	})
62548	64522	#endif
62549	64523
62550	64524	#ifdef __LITTLE_ENDIAN__
62551		#define vshll_high_n_s16(__p0_746, __p1_746) __extension__ ({ \
62552		int32x4_t __ret_746; \
62553		int16x8_t __s0_746 = __p0_746; \
62554		__ret_746 = (int32x4_t)(vshll_n_s16(vget_high_s16(__s0_746), __p1_746)); \
62555		__ret_746; \
	64525	#define vshll_high_n_s16(__p0_762, __p1_762) __extension__ ({ \
	64526	int32x4_t __ret_762; \
	64527	int16x8_t __s0_762 = __p0_762; \
	64528	__ret_762 = __builtin_bit_cast(int32x4_t, vshll_n_s16(vget_high_s16(__s0_762), __p1_762)); \
	64529	__ret_762; \
62556	64530	})
62557	64531	#else
62558		#define vshll_high_n_s16(__p0_747, __p1_747) __extension__ ({ \
62559		int32x4_t __ret_747; \
62560		int16x8_t __s0_747 = __p0_747; \
62561		int16x8_t __rev0_747; __rev0_747 = __builtin_shufflevector(__s0_747, __s0_747, 7, 6, 5, 4, 3, 2, 1, 0); \
62562		__ret_747 = (int32x4_t)(__noswap_vshll_n_s16(__noswap_vget_high_s16(__rev0_747), __p1_747)); \
62563		__ret_747 = __builtin_shufflevector(__ret_747, __ret_747, 3, 2, 1, 0); \
62564		__ret_747; \
	64532	#define vshll_high_n_s16(__p0_763, __p1_763) __extension__ ({ \
	64533	int32x4_t __ret_763; \
	64534	int16x8_t __s0_763 = __p0_763; \
	64535	int16x8_t __rev0_763; __rev0_763 = __builtin_shufflevector(__s0_763, __s0_763, __lane_reverse_128_16); \
	64536	__ret_763 = __builtin_bit_cast(int32x4_t, __noswap_vshll_n_s16(__noswap_vget_high_s16(__rev0_763), __p1_763)); \
	64537	__ret_763 = __builtin_shufflevector(__ret_763, __ret_763, __lane_reverse_128_32); \
	64538	__ret_763; \
62565	64539	})
62566	64540	#endif
62567	64541
62568	64542	#define vshrd_n_u64(__p0, __p1) __extension__ ({ \
62569	64543	uint64_t __ret; \
62570	64544	uint64_t __s0 = __p0; \
62571		__ret = (uint64_t) __builtin_neon_vshrd_n_u64(__s0, __p1); \
	64545	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vshrd_n_u64(__s0, __p1)); \
62572	64546	__ret; \
62573	64547	})
62574	64548	#define vshrd_n_s64(__p0, __p1) __extension__ ({ \
62575	64549	int64_t __ret; \
62576	64550	int64_t __s0 = __p0; \
62577		__ret = (int64_t) __builtin_neon_vshrd_n_s64(__s0, __p1); \
	64551	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vshrd_n_s64(__s0, __p1)); \
62578	64552	__ret; \
62579	64553	})
62580	64554	#ifdef __LITTLE_ENDIAN__
62581		#define vshrn_high_n_u32(__p0_748, __p1_748, __p2_748) __extension__ ({ \
62582		uint16x8_t __ret_748; \
62583		uint16x4_t __s0_748 = __p0_748; \
62584		uint32x4_t __s1_748 = __p1_748; \
62585		__ret_748 = (uint16x8_t)(vcombine_u16((uint16x4_t)(__s0_748), (uint16x4_t)(vshrn_n_u32(__s1_748, __p2_748)))); \
62586		__ret_748; \
	64555	#define vshrn_high_n_u32(__p0_764, __p1_764, __p2_764) __extension__ ({ \
	64556	uint16x8_t __ret_764; \
	64557	uint16x4_t __s0_764 = __p0_764; \
	64558	uint32x4_t __s1_764 = __p1_764; \
	64559	__ret_764 = __builtin_bit_cast(uint16x8_t, vcombine_u16(__builtin_bit_cast(uint16x4_t, __s0_764), __builtin_bit_cast(uint16x4_t, vshrn_n_u32(__s1_764, __p2_764)))); \
	64560	__ret_764; \
62587	64561	})
62588	64562	#else
62589		#define vshrn_high_n_u32(__p0_749, __p1_749, __p2_749) __extension__ ({ \
62590		uint16x8_t __ret_749; \
62591		uint16x4_t __s0_749 = __p0_749; \
62592		uint32x4_t __s1_749 = __p1_749; \
62593		uint16x4_t __rev0_749; __rev0_749 = __builtin_shufflevector(__s0_749, __s0_749, 3, 2, 1, 0); \
62594		uint32x4_t __rev1_749; __rev1_749 = __builtin_shufflevector(__s1_749, __s1_749, 3, 2, 1, 0); \
62595		__ret_749 = (uint16x8_t)(__noswap_vcombine_u16((uint16x4_t)(__rev0_749), (uint16x4_t)(__noswap_vshrn_n_u32(__rev1_749, __p2_749)))); \
62596		__ret_749 = __builtin_shufflevector(__ret_749, __ret_749, 7, 6, 5, 4, 3, 2, 1, 0); \
62597		__ret_749; \
	64563	#define vshrn_high_n_u32(__p0_765, __p1_765, __p2_765) __extension__ ({ \
	64564	uint16x8_t __ret_765; \
	64565	uint16x4_t __s0_765 = __p0_765; \
	64566	uint32x4_t __s1_765 = __p1_765; \
	64567	uint16x4_t __rev0_765; __rev0_765 = __builtin_shufflevector(__s0_765, __s0_765, __lane_reverse_64_16); \
	64568	uint32x4_t __rev1_765; __rev1_765 = __builtin_shufflevector(__s1_765, __s1_765, __lane_reverse_128_32); \
	64569	__ret_765 = __builtin_bit_cast(uint16x8_t, __noswap_vcombine_u16(__builtin_bit_cast(uint16x4_t, __rev0_765), __builtin_bit_cast(uint16x4_t, __noswap_vshrn_n_u32(__rev1_765, __p2_765)))); \
	64570	__ret_765 = __builtin_shufflevector(__ret_765, __ret_765, __lane_reverse_128_16); \
	64571	__ret_765; \
62598	64572	})
62599	64573	#endif
62600	64574
62601	64575	#ifdef __LITTLE_ENDIAN__
62602		#define vshrn_high_n_u64(__p0_750, __p1_750, __p2_750) __extension__ ({ \
62603		uint32x4_t __ret_750; \
62604		uint32x2_t __s0_750 = __p0_750; \
62605		uint64x2_t __s1_750 = __p1_750; \
62606		__ret_750 = (uint32x4_t)(vcombine_u32((uint32x2_t)(__s0_750), (uint32x2_t)(vshrn_n_u64(__s1_750, __p2_750)))); \
62607		__ret_750; \
	64576	#define vshrn_high_n_u64(__p0_766, __p1_766, __p2_766) __extension__ ({ \
	64577	uint32x4_t __ret_766; \
	64578	uint32x2_t __s0_766 = __p0_766; \
	64579	uint64x2_t __s1_766 = __p1_766; \
	64580	__ret_766 = __builtin_bit_cast(uint32x4_t, vcombine_u32(__builtin_bit_cast(uint32x2_t, __s0_766), __builtin_bit_cast(uint32x2_t, vshrn_n_u64(__s1_766, __p2_766)))); \
	64581	__ret_766; \
62608	64582	})
62609	64583	#else
62610		#define vshrn_high_n_u64(__p0_751, __p1_751, __p2_751) __extension__ ({ \
62611		uint32x4_t __ret_751; \
62612		uint32x2_t __s0_751 = __p0_751; \
62613		uint64x2_t __s1_751 = __p1_751; \
62614		uint32x2_t __rev0_751; __rev0_751 = __builtin_shufflevector(__s0_751, __s0_751, 1, 0); \
62615		uint64x2_t __rev1_751; __rev1_751 = __builtin_shufflevector(__s1_751, __s1_751, 1, 0); \
62616		__ret_751 = (uint32x4_t)(__noswap_vcombine_u32((uint32x2_t)(__rev0_751), (uint32x2_t)(__noswap_vshrn_n_u64(__rev1_751, __p2_751)))); \
62617		__ret_751 = __builtin_shufflevector(__ret_751, __ret_751, 3, 2, 1, 0); \
62618		__ret_751; \
	64584	#define vshrn_high_n_u64(__p0_767, __p1_767, __p2_767) __extension__ ({ \
	64585	uint32x4_t __ret_767; \
	64586	uint32x2_t __s0_767 = __p0_767; \
	64587	uint64x2_t __s1_767 = __p1_767; \
	64588	uint32x2_t __rev0_767; __rev0_767 = __builtin_shufflevector(__s0_767, __s0_767, __lane_reverse_64_32); \
	64589	uint64x2_t __rev1_767; __rev1_767 = __builtin_shufflevector(__s1_767, __s1_767, __lane_reverse_128_64); \
	64590	__ret_767 = __builtin_bit_cast(uint32x4_t, __noswap_vcombine_u32(__builtin_bit_cast(uint32x2_t, __rev0_767), __builtin_bit_cast(uint32x2_t, __noswap_vshrn_n_u64(__rev1_767, __p2_767)))); \
	64591	__ret_767 = __builtin_shufflevector(__ret_767, __ret_767, __lane_reverse_128_32); \
	64592	__ret_767; \
62619	64593	})
62620	64594	#endif
62621	64595
62622	64596	#ifdef __LITTLE_ENDIAN__
62623		#define vshrn_high_n_u16(__p0_752, __p1_752, __p2_752) __extension__ ({ \
62624		uint8x16_t __ret_752; \
62625		uint8x8_t __s0_752 = __p0_752; \
62626		uint16x8_t __s1_752 = __p1_752; \
62627		__ret_752 = (uint8x16_t)(vcombine_u8((uint8x8_t)(__s0_752), (uint8x8_t)(vshrn_n_u16(__s1_752, __p2_752)))); \
62628		__ret_752; \
	64597	#define vshrn_high_n_u16(__p0_768, __p1_768, __p2_768) __extension__ ({ \
	64598	uint8x16_t __ret_768; \
	64599	uint8x8_t __s0_768 = __p0_768; \
	64600	uint16x8_t __s1_768 = __p1_768; \
	64601	__ret_768 = __builtin_bit_cast(uint8x16_t, vcombine_u8(__builtin_bit_cast(uint8x8_t, __s0_768), __builtin_bit_cast(uint8x8_t, vshrn_n_u16(__s1_768, __p2_768)))); \
	64602	__ret_768; \
62629	64603	})
62630	64604	#else
62631		#define vshrn_high_n_u16(__p0_753, __p1_753, __p2_753) __extension__ ({ \
62632		uint8x16_t __ret_753; \
62633		uint8x8_t __s0_753 = __p0_753; \
62634		uint16x8_t __s1_753 = __p1_753; \
62635		uint8x8_t __rev0_753; __rev0_753 = __builtin_shufflevector(__s0_753, __s0_753, 7, 6, 5, 4, 3, 2, 1, 0); \
62636		uint16x8_t __rev1_753; __rev1_753 = __builtin_shufflevector(__s1_753, __s1_753, 7, 6, 5, 4, 3, 2, 1, 0); \
62637		__ret_753 = (uint8x16_t)(__noswap_vcombine_u8((uint8x8_t)(__rev0_753), (uint8x8_t)(__noswap_vshrn_n_u16(__rev1_753, __p2_753)))); \
62638		__ret_753 = __builtin_shufflevector(__ret_753, __ret_753, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
62639		__ret_753; \
	64605	#define vshrn_high_n_u16(__p0_769, __p1_769, __p2_769) __extension__ ({ \
	64606	uint8x16_t __ret_769; \
	64607	uint8x8_t __s0_769 = __p0_769; \
	64608	uint16x8_t __s1_769 = __p1_769; \
	64609	uint8x8_t __rev0_769; __rev0_769 = __builtin_shufflevector(__s0_769, __s0_769, __lane_reverse_64_8); \
	64610	uint16x8_t __rev1_769; __rev1_769 = __builtin_shufflevector(__s1_769, __s1_769, __lane_reverse_128_16); \
	64611	__ret_769 = __builtin_bit_cast(uint8x16_t, __noswap_vcombine_u8(__builtin_bit_cast(uint8x8_t, __rev0_769), __builtin_bit_cast(uint8x8_t, __noswap_vshrn_n_u16(__rev1_769, __p2_769)))); \
	64612	__ret_769 = __builtin_shufflevector(__ret_769, __ret_769, __lane_reverse_128_8); \
	64613	__ret_769; \
62640	64614	})
62641	64615	#endif
62642	64616
62643	64617	#ifdef __LITTLE_ENDIAN__
62644		#define vshrn_high_n_s32(__p0_754, __p1_754, __p2_754) __extension__ ({ \
62645		int16x8_t __ret_754; \
62646		int16x4_t __s0_754 = __p0_754; \
62647		int32x4_t __s1_754 = __p1_754; \
62648		__ret_754 = (int16x8_t)(vcombine_s16((int16x4_t)(__s0_754), (int16x4_t)(vshrn_n_s32(__s1_754, __p2_754)))); \
62649		__ret_754; \
	64618	#define vshrn_high_n_s32(__p0_770, __p1_770, __p2_770) __extension__ ({ \
	64619	int16x8_t __ret_770; \
	64620	int16x4_t __s0_770 = __p0_770; \
	64621	int32x4_t __s1_770 = __p1_770; \
	64622	__ret_770 = __builtin_bit_cast(int16x8_t, vcombine_s16(__builtin_bit_cast(int16x4_t, __s0_770), __builtin_bit_cast(int16x4_t, vshrn_n_s32(__s1_770, __p2_770)))); \
	64623	__ret_770; \
62650	64624	})
62651	64625	#else
62652		#define vshrn_high_n_s32(__p0_755, __p1_755, __p2_755) __extension__ ({ \
62653		int16x8_t __ret_755; \
62654		int16x4_t __s0_755 = __p0_755; \
62655		int32x4_t __s1_755 = __p1_755; \
62656		int16x4_t __rev0_755; __rev0_755 = __builtin_shufflevector(__s0_755, __s0_755, 3, 2, 1, 0); \
62657		int32x4_t __rev1_755; __rev1_755 = __builtin_shufflevector(__s1_755, __s1_755, 3, 2, 1, 0); \
62658		__ret_755 = (int16x8_t)(__noswap_vcombine_s16((int16x4_t)(__rev0_755), (int16x4_t)(__noswap_vshrn_n_s32(__rev1_755, __p2_755)))); \
62659		__ret_755 = __builtin_shufflevector(__ret_755, __ret_755, 7, 6, 5, 4, 3, 2, 1, 0); \
62660		__ret_755; \
	64626	#define vshrn_high_n_s32(__p0_771, __p1_771, __p2_771) __extension__ ({ \
	64627	int16x8_t __ret_771; \
	64628	int16x4_t __s0_771 = __p0_771; \
	64629	int32x4_t __s1_771 = __p1_771; \
	64630	int16x4_t __rev0_771; __rev0_771 = __builtin_shufflevector(__s0_771, __s0_771, __lane_reverse_64_16); \
	64631	int32x4_t __rev1_771; __rev1_771 = __builtin_shufflevector(__s1_771, __s1_771, __lane_reverse_128_32); \
	64632	__ret_771 = __builtin_bit_cast(int16x8_t, __noswap_vcombine_s16(__builtin_bit_cast(int16x4_t, __rev0_771), __builtin_bit_cast(int16x4_t, __noswap_vshrn_n_s32(__rev1_771, __p2_771)))); \
	64633	__ret_771 = __builtin_shufflevector(__ret_771, __ret_771, __lane_reverse_128_16); \
	64634	__ret_771; \
62661	64635	})
62662	64636	#endif
62663	64637
62664	64638	#ifdef __LITTLE_ENDIAN__
62665		#define vshrn_high_n_s64(__p0_756, __p1_756, __p2_756) __extension__ ({ \
62666		int32x4_t __ret_756; \
62667		int32x2_t __s0_756 = __p0_756; \
62668		int64x2_t __s1_756 = __p1_756; \
62669		__ret_756 = (int32x4_t)(vcombine_s32((int32x2_t)(__s0_756), (int32x2_t)(vshrn_n_s64(__s1_756, __p2_756)))); \
62670		__ret_756; \
	64639	#define vshrn_high_n_s64(__p0_772, __p1_772, __p2_772) __extension__ ({ \
	64640	int32x4_t __ret_772; \
	64641	int32x2_t __s0_772 = __p0_772; \
	64642	int64x2_t __s1_772 = __p1_772; \
	64643	__ret_772 = __builtin_bit_cast(int32x4_t, vcombine_s32(__builtin_bit_cast(int32x2_t, __s0_772), __builtin_bit_cast(int32x2_t, vshrn_n_s64(__s1_772, __p2_772)))); \
	64644	__ret_772; \
62671	64645	})
62672	64646	#else
62673		#define vshrn_high_n_s64(__p0_757, __p1_757, __p2_757) __extension__ ({ \
62674		int32x4_t __ret_757; \
62675		int32x2_t __s0_757 = __p0_757; \
62676		int64x2_t __s1_757 = __p1_757; \
62677		int32x2_t __rev0_757; __rev0_757 = __builtin_shufflevector(__s0_757, __s0_757, 1, 0); \
62678		int64x2_t __rev1_757; __rev1_757 = __builtin_shufflevector(__s1_757, __s1_757, 1, 0); \
62679		__ret_757 = (int32x4_t)(__noswap_vcombine_s32((int32x2_t)(__rev0_757), (int32x2_t)(__noswap_vshrn_n_s64(__rev1_757, __p2_757)))); \
62680		__ret_757 = __builtin_shufflevector(__ret_757, __ret_757, 3, 2, 1, 0); \
62681		__ret_757; \
	64647	#define vshrn_high_n_s64(__p0_773, __p1_773, __p2_773) __extension__ ({ \
	64648	int32x4_t __ret_773; \
	64649	int32x2_t __s0_773 = __p0_773; \
	64650	int64x2_t __s1_773 = __p1_773; \
	64651	int32x2_t __rev0_773; __rev0_773 = __builtin_shufflevector(__s0_773, __s0_773, __lane_reverse_64_32); \
	64652	int64x2_t __rev1_773; __rev1_773 = __builtin_shufflevector(__s1_773, __s1_773, __lane_reverse_128_64); \
	64653	__ret_773 = __builtin_bit_cast(int32x4_t, __noswap_vcombine_s32(__builtin_bit_cast(int32x2_t, __rev0_773), __builtin_bit_cast(int32x2_t, __noswap_vshrn_n_s64(__rev1_773, __p2_773)))); \
	64654	__ret_773 = __builtin_shufflevector(__ret_773, __ret_773, __lane_reverse_128_32); \
	64655	__ret_773; \
62682	64656	})
62683	64657	#endif
62684	64658
62685	64659	#ifdef __LITTLE_ENDIAN__
62686		#define vshrn_high_n_s16(__p0_758, __p1_758, __p2_758) __extension__ ({ \
62687		int8x16_t __ret_758; \
62688		int8x8_t __s0_758 = __p0_758; \
62689		int16x8_t __s1_758 = __p1_758; \
62690		__ret_758 = (int8x16_t)(vcombine_s8((int8x8_t)(__s0_758), (int8x8_t)(vshrn_n_s16(__s1_758, __p2_758)))); \
62691		__ret_758; \
	64660	#define vshrn_high_n_s16(__p0_774, __p1_774, __p2_774) __extension__ ({ \
	64661	int8x16_t __ret_774; \
	64662	int8x8_t __s0_774 = __p0_774; \
	64663	int16x8_t __s1_774 = __p1_774; \
	64664	__ret_774 = __builtin_bit_cast(int8x16_t, vcombine_s8(__builtin_bit_cast(int8x8_t, __s0_774), __builtin_bit_cast(int8x8_t, vshrn_n_s16(__s1_774, __p2_774)))); \
	64665	__ret_774; \
62692	64666	})
62693	64667	#else
62694		#define vshrn_high_n_s16(__p0_759, __p1_759, __p2_759) __extension__ ({ \
62695		int8x16_t __ret_759; \
62696		int8x8_t __s0_759 = __p0_759; \
62697		int16x8_t __s1_759 = __p1_759; \
62698		int8x8_t __rev0_759; __rev0_759 = __builtin_shufflevector(__s0_759, __s0_759, 7, 6, 5, 4, 3, 2, 1, 0); \
62699		int16x8_t __rev1_759; __rev1_759 = __builtin_shufflevector(__s1_759, __s1_759, 7, 6, 5, 4, 3, 2, 1, 0); \
62700		__ret_759 = (int8x16_t)(__noswap_vcombine_s8((int8x8_t)(__rev0_759), (int8x8_t)(__noswap_vshrn_n_s16(__rev1_759, __p2_759)))); \
62701		__ret_759 = __builtin_shufflevector(__ret_759, __ret_759, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
62702		__ret_759; \
	64668	#define vshrn_high_n_s16(__p0_775, __p1_775, __p2_775) __extension__ ({ \
	64669	int8x16_t __ret_775; \
	64670	int8x8_t __s0_775 = __p0_775; \
	64671	int16x8_t __s1_775 = __p1_775; \
	64672	int8x8_t __rev0_775; __rev0_775 = __builtin_shufflevector(__s0_775, __s0_775, __lane_reverse_64_8); \
	64673	int16x8_t __rev1_775; __rev1_775 = __builtin_shufflevector(__s1_775, __s1_775, __lane_reverse_128_16); \
	64674	__ret_775 = __builtin_bit_cast(int8x16_t, __noswap_vcombine_s8(__builtin_bit_cast(int8x8_t, __rev0_775), __builtin_bit_cast(int8x8_t, __noswap_vshrn_n_s16(__rev1_775, __p2_775)))); \
	64675	__ret_775 = __builtin_shufflevector(__ret_775, __ret_775, __lane_reverse_128_8); \
	64676	__ret_775; \
62703	64677	})
62704	64678	#endif
62705	64679
...	...	@@ -62707,21 +64681,21 @@ __ai __attribute__((target("neon"))) int64_t vshld_s64(int64_t __p0, int64_t __p
62707	64681	uint64_t __ret; \
62708	64682	uint64_t __s0 = __p0; \
62709	64683	uint64_t __s1 = __p1; \
62710		__ret = (uint64_t) __builtin_neon_vslid_n_u64(__s0, __s1, __p2); \
	64684	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vslid_n_u64(__s0, __s1, __p2)); \
62711	64685	__ret; \
62712	64686	})
62713	64687	#define vslid_n_s64(__p0, __p1, __p2) __extension__ ({ \
62714	64688	int64_t __ret; \
62715	64689	int64_t __s0 = __p0; \
62716	64690	int64_t __s1 = __p1; \
62717		__ret = (int64_t) __builtin_neon_vslid_n_s64(__s0, __s1, __p2); \
	64691	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vslid_n_s64(__s0, __s1, __p2)); \
62718	64692	__ret; \
62719	64693	})
62720	64694	#define vsli_n_p64(__p0, __p1, __p2) __extension__ ({ \
62721	64695	poly64x1_t __ret; \
62722	64696	poly64x1_t __s0 = __p0; \
62723	64697	poly64x1_t __s1 = __p1; \
62724		__ret = (poly64x1_t) __builtin_neon_vsli_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 6); \
	64698	__ret = __builtin_bit_cast(poly64x1_t, __builtin_neon_vsli_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 6)); \
62725	64699	__ret; \
62726	64700	})
62727	64701	#ifdef __LITTLE_ENDIAN__
...	...	@@ -62729,7 +64703,7 @@ __ai __attribute__((target("neon"))) int64_t vshld_s64(int64_t __p0, int64_t __p
62729	64703	poly64x2_t __ret; \
62730	64704	poly64x2_t __s0 = __p0; \
62731	64705	poly64x2_t __s1 = __p1; \
62732		__ret = (poly64x2_t) __builtin_neon_vsliq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 38); \
	64706	__ret = __builtin_bit_cast(poly64x2_t, __builtin_neon_vsliq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 38)); \
62733	64707	__ret; \
62734	64708	})
62735	64709	#else
...	...	@@ -62737,47 +64711,47 @@ __ai __attribute__((target("neon"))) int64_t vshld_s64(int64_t __p0, int64_t __p
62737	64711	poly64x2_t __ret; \
62738	64712	poly64x2_t __s0 = __p0; \
62739	64713	poly64x2_t __s1 = __p1; \
62740		poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
62741		poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
62742		__ret = (poly64x2_t) __builtin_neon_vsliq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 38); \
62743		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	64714	poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	64715	poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	64716	__ret = __builtin_bit_cast(poly64x2_t, __builtin_neon_vsliq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 38)); \
	64717	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
62744	64718	__ret; \
62745	64719	})
62746	64720	#endif
62747	64721
62748	64722	__ai __attribute__((target("neon"))) uint8_t vsqaddb_u8(uint8_t __p0, int8_t __p1) {
62749	64723	uint8_t __ret;
62750		__ret = (uint8_t) __builtin_neon_vsqaddb_u8(__p0, __p1);
	64724	__ret = __builtin_bit_cast(uint8_t, __builtin_neon_vsqaddb_u8(__p0, __p1));
62751	64725	return __ret;
62752	64726	}
62753	64727	__ai __attribute__((target("neon"))) uint32_t vsqadds_u32(uint32_t __p0, int32_t __p1) {
62754	64728	uint32_t __ret;
62755		__ret = (uint32_t) __builtin_neon_vsqadds_u32(__p0, __p1);
	64729	__ret = __builtin_bit_cast(uint32_t, __builtin_neon_vsqadds_u32(__p0, __p1));
62756	64730	return __ret;
62757	64731	}
62758	64732	__ai __attribute__((target("neon"))) uint64_t vsqaddd_u64(uint64_t __p0, int64_t __p1) {
62759	64733	uint64_t __ret;
62760		__ret = (uint64_t) __builtin_neon_vsqaddd_u64(__p0, __p1);
	64734	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vsqaddd_u64(__p0, __p1));
62761	64735	return __ret;
62762	64736	}
62763	64737	__ai __attribute__((target("neon"))) uint16_t vsqaddh_u16(uint16_t __p0, int16_t __p1) {
62764	64738	uint16_t __ret;
62765		__ret = (uint16_t) __builtin_neon_vsqaddh_u16(__p0, __p1);
	64739	__ret = __builtin_bit_cast(uint16_t, __builtin_neon_vsqaddh_u16(__p0, __p1));
62766	64740	return __ret;
62767	64741	}
62768	64742	#ifdef __LITTLE_ENDIAN__
62769	64743	__ai __attribute__((target("neon"))) uint8x16_t vsqaddq_u8(uint8x16_t __p0, int8x16_t __p1) {
62770	64744	uint8x16_t __ret;
62771		__ret = (uint8x16_t) __builtin_neon_vsqaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
	64745	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vsqaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 48));
62772	64746	return __ret;
62773	64747	}
62774	64748	#else
62775	64749	__ai __attribute__((target("neon"))) uint8x16_t vsqaddq_u8(uint8x16_t __p0, int8x16_t __p1) {
62776	64750	uint8x16_t __ret;
62777		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
62778		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
62779		__ret = (uint8x16_t) __builtin_neon_vsqaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
62780		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	64751	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	64752	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	64753	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vsqaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 48));
	64754	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
62781	64755	return __ret;
62782	64756	}
62783	64757	#endif
...	...	@@ -62785,16 +64759,16 @@ __ai __attribute__((target("neon"))) uint8x16_t vsqaddq_u8(uint8x16_t __p0, int8
62785	64759	#ifdef __LITTLE_ENDIAN__
62786	64760	__ai __attribute__((target("neon"))) uint32x4_t vsqaddq_u32(uint32x4_t __p0, int32x4_t __p1) {
62787	64761	uint32x4_t __ret;
62788		__ret = (uint32x4_t) __builtin_neon_vsqaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
	64762	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsqaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50));
62789	64763	return __ret;
62790	64764	}
62791	64765	#else
62792	64766	__ai __attribute__((target("neon"))) uint32x4_t vsqaddq_u32(uint32x4_t __p0, int32x4_t __p1) {
62793	64767	uint32x4_t __ret;
62794		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
62795		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
62796		__ret = (uint32x4_t) __builtin_neon_vsqaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
62797		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	64768	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	64769	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	64770	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsqaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 50));
	64771	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
62798	64772	return __ret;
62799	64773	}
62800	64774	#endif
...	...	@@ -62802,16 +64776,16 @@ __ai __attribute__((target("neon"))) uint32x4_t vsqaddq_u32(uint32x4_t __p0, int
62802	64776	#ifdef __LITTLE_ENDIAN__
62803	64777	__ai __attribute__((target("neon"))) uint64x2_t vsqaddq_u64(uint64x2_t __p0, int64x2_t __p1) {
62804	64778	uint64x2_t __ret;
62805		__ret = (uint64x2_t) __builtin_neon_vsqaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 51);
	64779	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vsqaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 51));
62806	64780	return __ret;
62807	64781	}
62808	64782	#else
62809	64783	__ai __attribute__((target("neon"))) uint64x2_t vsqaddq_u64(uint64x2_t __p0, int64x2_t __p1) {
62810	64784	uint64x2_t __ret;
62811		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
62812		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
62813		__ret = (uint64x2_t) __builtin_neon_vsqaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 51);
62814		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	64785	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	64786	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	64787	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vsqaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 51));
	64788	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
62815	64789	return __ret;
62816	64790	}
62817	64791	#endif
...	...	@@ -62819,16 +64793,16 @@ __ai __attribute__((target("neon"))) uint64x2_t vsqaddq_u64(uint64x2_t __p0, int
62819	64793	#ifdef __LITTLE_ENDIAN__
62820	64794	__ai __attribute__((target("neon"))) uint16x8_t vsqaddq_u16(uint16x8_t __p0, int16x8_t __p1) {
62821	64795	uint16x8_t __ret;
62822		__ret = (uint16x8_t) __builtin_neon_vsqaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
	64796	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vsqaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 49));
62823	64797	return __ret;
62824	64798	}
62825	64799	#else
62826	64800	__ai __attribute__((target("neon"))) uint16x8_t vsqaddq_u16(uint16x8_t __p0, int16x8_t __p1) {
62827	64801	uint16x8_t __ret;
62828		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
62829		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
62830		__ret = (uint16x8_t) __builtin_neon_vsqaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
62831		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	64802	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	64803	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	64804	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vsqaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 49));
	64805	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
62832	64806	return __ret;
62833	64807	}
62834	64808	#endif
...	...	@@ -62836,16 +64810,16 @@ __ai __attribute__((target("neon"))) uint16x8_t vsqaddq_u16(uint16x8_t __p0, int
62836	64810	#ifdef __LITTLE_ENDIAN__
62837	64811	__ai __attribute__((target("neon"))) uint8x8_t vsqadd_u8(uint8x8_t __p0, int8x8_t __p1) {
62838	64812	uint8x8_t __ret;
62839		__ret = (uint8x8_t) __builtin_neon_vsqadd_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
	64813	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vsqadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 16));
62840	64814	return __ret;
62841	64815	}
62842	64816	#else
62843	64817	__ai __attribute__((target("neon"))) uint8x8_t vsqadd_u8(uint8x8_t __p0, int8x8_t __p1) {
62844	64818	uint8x8_t __ret;
62845		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
62846		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
62847		__ret = (uint8x8_t) __builtin_neon_vsqadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
62848		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	64819	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	64820	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	64821	__ret = __builtin_bit_cast(uint8x8_t, __builtin_neon_vsqadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 16));
	64822	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
62849	64823	return __ret;
62850	64824	}
62851	64825	#endif
...	...	@@ -62853,38 +64827,38 @@ __ai __attribute__((target("neon"))) uint8x8_t vsqadd_u8(uint8x8_t __p0, int8x8_
62853	64827	#ifdef __LITTLE_ENDIAN__
62854	64828	__ai __attribute__((target("neon"))) uint32x2_t vsqadd_u32(uint32x2_t __p0, int32x2_t __p1) {
62855	64829	uint32x2_t __ret;
62856		__ret = (uint32x2_t) __builtin_neon_vsqadd_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
	64830	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vsqadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 18));
62857	64831	return __ret;
62858	64832	}
62859	64833	#else
62860	64834	__ai __attribute__((target("neon"))) uint32x2_t vsqadd_u32(uint32x2_t __p0, int32x2_t __p1) {
62861	64835	uint32x2_t __ret;
62862		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
62863		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
62864		__ret = (uint32x2_t) __builtin_neon_vsqadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
62865		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	64836	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	64837	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	64838	__ret = __builtin_bit_cast(uint32x2_t, __builtin_neon_vsqadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 18));
	64839	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
62866	64840	return __ret;
62867	64841	}
62868	64842	#endif
62869	64843
62870	64844	__ai __attribute__((target("neon"))) uint64x1_t vsqadd_u64(uint64x1_t __p0, int64x1_t __p1) {
62871	64845	uint64x1_t __ret;
62872		__ret = (uint64x1_t) __builtin_neon_vsqadd_v((int8x8_t)__p0, (int8x8_t)__p1, 19);
	64846	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vsqadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 19));
62873	64847	return __ret;
62874	64848	}
62875	64849	#ifdef __LITTLE_ENDIAN__
62876	64850	__ai __attribute__((target("neon"))) uint16x4_t vsqadd_u16(uint16x4_t __p0, int16x4_t __p1) {
62877	64851	uint16x4_t __ret;
62878		__ret = (uint16x4_t) __builtin_neon_vsqadd_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
	64852	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vsqadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 17));
62879	64853	return __ret;
62880	64854	}
62881	64855	#else
62882	64856	__ai __attribute__((target("neon"))) uint16x4_t vsqadd_u16(uint16x4_t __p0, int16x4_t __p1) {
62883	64857	uint16x4_t __ret;
62884		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
62885		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
62886		__ret = (uint16x4_t) __builtin_neon_vsqadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
62887		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	64858	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	64859	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	64860	__ret = __builtin_bit_cast(uint16x4_t, __builtin_neon_vsqadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 17));
	64861	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
62888	64862	return __ret;
62889	64863	}
62890	64864	#endif
...	...	@@ -62892,15 +64866,15 @@ __ai __attribute__((target("neon"))) uint16x4_t vsqadd_u16(uint16x4_t __p0, int1
62892	64866	#ifdef __LITTLE_ENDIAN__
62893	64867	__ai __attribute__((target("neon"))) float64x2_t vsqrtq_f64(float64x2_t __p0) {
62894	64868	float64x2_t __ret;
62895		__ret = (float64x2_t) __builtin_neon_vsqrtq_v((int8x16_t)__p0, 42);
	64869	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vsqrtq_v(__builtin_bit_cast(int8x16_t, __p0), 42));
62896	64870	return __ret;
62897	64871	}
62898	64872	#else
62899	64873	__ai __attribute__((target("neon"))) float64x2_t vsqrtq_f64(float64x2_t __p0) {
62900	64874	float64x2_t __ret;
62901		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
62902		__ret = (float64x2_t) __builtin_neon_vsqrtq_v((int8x16_t)__rev0, 42);
62903		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	64875	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	64876	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vsqrtq_v(__builtin_bit_cast(int8x16_t, __rev0), 42));
	64877	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
62904	64878	return __ret;
62905	64879	}
62906	64880	#endif
...	...	@@ -62908,36 +64882,36 @@ __ai __attribute__((target("neon"))) float64x2_t vsqrtq_f64(float64x2_t __p0) {
62908	64882	#ifdef __LITTLE_ENDIAN__
62909	64883	__ai __attribute__((target("neon"))) float32x4_t vsqrtq_f32(float32x4_t __p0) {
62910	64884	float32x4_t __ret;
62911		__ret = (float32x4_t) __builtin_neon_vsqrtq_v((int8x16_t)__p0, 41);
	64885	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vsqrtq_v(__builtin_bit_cast(int8x16_t, __p0), 41));
62912	64886	return __ret;
62913	64887	}
62914	64888	#else
62915	64889	__ai __attribute__((target("neon"))) float32x4_t vsqrtq_f32(float32x4_t __p0) {
62916	64890	float32x4_t __ret;
62917		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
62918		__ret = (float32x4_t) __builtin_neon_vsqrtq_v((int8x16_t)__rev0, 41);
62919		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	64891	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	64892	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vsqrtq_v(__builtin_bit_cast(int8x16_t, __rev0), 41));
	64893	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
62920	64894	return __ret;
62921	64895	}
62922	64896	#endif
62923	64897
62924	64898	__ai __attribute__((target("neon"))) float64x1_t vsqrt_f64(float64x1_t __p0) {
62925	64899	float64x1_t __ret;
62926		__ret = (float64x1_t) __builtin_neon_vsqrt_v((int8x8_t)__p0, 10);
	64900	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vsqrt_v(__builtin_bit_cast(int8x8_t, __p0), 10));
62927	64901	return __ret;
62928	64902	}
62929	64903	#ifdef __LITTLE_ENDIAN__
62930	64904	__ai __attribute__((target("neon"))) float32x2_t vsqrt_f32(float32x2_t __p0) {
62931	64905	float32x2_t __ret;
62932		__ret = (float32x2_t) __builtin_neon_vsqrt_v((int8x8_t)__p0, 9);
	64906	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vsqrt_v(__builtin_bit_cast(int8x8_t, __p0), 9));
62933	64907	return __ret;
62934	64908	}
62935	64909	#else
62936	64910	__ai __attribute__((target("neon"))) float32x2_t vsqrt_f32(float32x2_t __p0) {
62937	64911	float32x2_t __ret;
62938		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
62939		__ret = (float32x2_t) __builtin_neon_vsqrt_v((int8x8_t)__rev0, 9);
62940		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	64912	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	64913	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vsqrt_v(__builtin_bit_cast(int8x8_t, __rev0), 9));
	64914	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
62941	64915	return __ret;
62942	64916	}
62943	64917	#endif
...	...	@@ -62946,35 +64920,35 @@ __ai __attribute__((target("neon"))) float32x2_t vsqrt_f32(float32x2_t __p0) {
62946	64920	uint64_t __ret; \
62947	64921	uint64_t __s0 = __p0; \
62948	64922	uint64_t __s1 = __p1; \
62949		__ret = (uint64_t) __builtin_neon_vsrad_n_u64(__s0, __s1, __p2); \
	64923	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vsrad_n_u64(__s0, __s1, __p2)); \
62950	64924	__ret; \
62951	64925	})
62952	64926	#define vsrad_n_s64(__p0, __p1, __p2) __extension__ ({ \
62953	64927	int64_t __ret; \
62954	64928	int64_t __s0 = __p0; \
62955	64929	int64_t __s1 = __p1; \
62956		__ret = (int64_t) __builtin_neon_vsrad_n_s64(__s0, __s1, __p2); \
	64930	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vsrad_n_s64(__s0, __s1, __p2)); \
62957	64931	__ret; \
62958	64932	})
62959	64933	#define vsrid_n_u64(__p0, __p1, __p2) __extension__ ({ \
62960	64934	uint64_t __ret; \
62961	64935	uint64_t __s0 = __p0; \
62962	64936	uint64_t __s1 = __p1; \
62963		__ret = (uint64_t) __builtin_neon_vsrid_n_u64(__s0, __s1, __p2); \
	64937	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vsrid_n_u64(__s0, __s1, __p2)); \
62964	64938	__ret; \
62965	64939	})
62966	64940	#define vsrid_n_s64(__p0, __p1, __p2) __extension__ ({ \
62967	64941	int64_t __ret; \
62968	64942	int64_t __s0 = __p0; \
62969	64943	int64_t __s1 = __p1; \
62970		__ret = (int64_t) __builtin_neon_vsrid_n_s64(__s0, __s1, __p2); \
	64944	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vsrid_n_s64(__s0, __s1, __p2)); \
62971	64945	__ret; \
62972	64946	})
62973	64947	#define vsri_n_p64(__p0, __p1, __p2) __extension__ ({ \
62974	64948	poly64x1_t __ret; \
62975	64949	poly64x1_t __s0 = __p0; \
62976	64950	poly64x1_t __s1 = __p1; \
62977		__ret = (poly64x1_t) __builtin_neon_vsri_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 6); \
	64951	__ret = __builtin_bit_cast(poly64x1_t, __builtin_neon_vsri_n_v(__builtin_bit_cast(int8x8_t, __s0), __builtin_bit_cast(int8x8_t, __s1), __p2, 6)); \
62978	64952	__ret; \
62979	64953	})
62980	64954	#ifdef __LITTLE_ENDIAN__
...	...	@@ -62982,7 +64956,7 @@ __ai __attribute__((target("neon"))) float32x2_t vsqrt_f32(float32x2_t __p0) {
62982	64956	poly64x2_t __ret; \
62983	64957	poly64x2_t __s0 = __p0; \
62984	64958	poly64x2_t __s1 = __p1; \
62985		__ret = (poly64x2_t) __builtin_neon_vsriq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 38); \
	64959	__ret = __builtin_bit_cast(poly64x2_t, __builtin_neon_vsriq_n_v(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 38)); \
62986	64960	__ret; \
62987	64961	})
62988	64962	#else
...	...	@@ -62990,814 +64964,1154 @@ __ai __attribute__((target("neon"))) float32x2_t vsqrt_f32(float32x2_t __p0) {
62990	64964	poly64x2_t __ret; \
62991	64965	poly64x2_t __s0 = __p0; \
62992	64966	poly64x2_t __s1 = __p1; \
62993		poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
62994		poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
62995		__ret = (poly64x2_t) __builtin_neon_vsriq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 38); \
62996		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	64967	poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	64968	poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	64969	__ret = __builtin_bit_cast(poly64x2_t, __builtin_neon_vsriq_n_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 38)); \
	64970	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
62997	64971	__ret; \
62998	64972	})
62999	64973	#endif
63000	64974
63001	64975	#define vst1_p64(__p0, __p1) __extension__ ({ \
63002	64976	poly64x1_t __s1 = __p1; \
63003		__builtin_neon_vst1_v(__p0, (int8x8_t)__s1, 6); \
	64977	__builtin_neon_vst1_v(__p0, __builtin_bit_cast(int8x8_t, __s1), 6); \
63004	64978	})
63005	64979	#ifdef __LITTLE_ENDIAN__
63006	64980	#define vst1q_p64(__p0, __p1) __extension__ ({ \
63007	64981	poly64x2_t __s1 = __p1; \
63008		__builtin_neon_vst1q_v(__p0, (int8x16_t)__s1, 38); \
	64982	__builtin_neon_vst1q_v(__p0, __builtin_bit_cast(int8x16_t, __s1), 38); \
63009	64983	})
63010	64984	#else
63011	64985	#define vst1q_p64(__p0, __p1) __extension__ ({ \
63012	64986	poly64x2_t __s1 = __p1; \
63013		poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
63014		__builtin_neon_vst1q_v(__p0, (int8x16_t)__rev1, 38); \
	64987	poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	64988	__builtin_neon_vst1q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), 38); \
63015	64989	})
63016	64990	#endif
63017	64991
63018	64992	#ifdef __LITTLE_ENDIAN__
63019	64993	#define vst1q_f64(__p0, __p1) __extension__ ({ \
63020	64994	float64x2_t __s1 = __p1; \
63021		__builtin_neon_vst1q_v(__p0, (int8x16_t)__s1, 42); \
	64995	__builtin_neon_vst1q_v(__p0, __builtin_bit_cast(int8x16_t, __s1), 42); \
63022	64996	})
63023	64997	#else
63024	64998	#define vst1q_f64(__p0, __p1) __extension__ ({ \
63025	64999	float64x2_t __s1 = __p1; \
63026		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
63027		__builtin_neon_vst1q_v(__p0, (int8x16_t)__rev1, 42); \
	65000	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	65001	__builtin_neon_vst1q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), 42); \
	65002	})
	65003	#endif
	65004
	65005	#ifdef __LITTLE_ENDIAN__
	65006	#define vst1q_mf8(__p0, __p1) __extension__ ({ \
	65007	mfloat8x16_t __s1 = __p1; \
	65008	__builtin_neon_vst1q_v(__p0, __builtin_bit_cast(int8x16_t, __s1), 44); \
	65009	})
	65010	#else
	65011	#define vst1q_mf8(__p0, __p1) __extension__ ({ \
	65012	mfloat8x16_t __s1 = __p1; \
	65013	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	65014	__builtin_neon_vst1q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), 44); \
63028	65015	})
63029	65016	#endif
63030	65017
63031	65018	#define vst1_f64(__p0, __p1) __extension__ ({ \
63032	65019	float64x1_t __s1 = __p1; \
63033		__builtin_neon_vst1_v(__p0, (int8x8_t)__s1, 10); \
	65020	__builtin_neon_vst1_v(__p0, __builtin_bit_cast(int8x8_t, __s1), 10); \
	65021	})
	65022	#ifdef __LITTLE_ENDIAN__
	65023	#define vst1_mf8(__p0, __p1) __extension__ ({ \
	65024	mfloat8x8_t __s1 = __p1; \
	65025	__builtin_neon_vst1_v(__p0, __builtin_bit_cast(int8x8_t, __s1), 12); \
	65026	})
	65027	#else
	65028	#define vst1_mf8(__p0, __p1) __extension__ ({ \
	65029	mfloat8x8_t __s1 = __p1; \
	65030	mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	65031	__builtin_neon_vst1_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), 12); \
63034	65032	})
	65033	#endif
	65034
63035	65035	#define vst1_lane_p64(__p0, __p1, __p2) __extension__ ({ \
63036	65036	poly64x1_t __s1 = __p1; \
63037		__builtin_neon_vst1_lane_v(__p0, (int8x8_t)__s1, __p2, 6); \
	65037	__builtin_neon_vst1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 6); \
63038	65038	})
63039	65039	#ifdef __LITTLE_ENDIAN__
63040	65040	#define vst1q_lane_p64(__p0, __p1, __p2) __extension__ ({ \
63041	65041	poly64x2_t __s1 = __p1; \
63042		__builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__s1, __p2, 38); \
	65042	__builtin_neon_vst1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 38); \
63043	65043	})
63044	65044	#else
63045	65045	#define vst1q_lane_p64(__p0, __p1, __p2) __extension__ ({ \
63046	65046	poly64x2_t __s1 = __p1; \
63047		poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
63048		__builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 38); \
	65047	poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	65048	__builtin_neon_vst1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 38); \
63049	65049	})
63050	65050	#endif
63051	65051
63052	65052	#ifdef __LITTLE_ENDIAN__
63053	65053	#define vst1q_lane_f64(__p0, __p1, __p2) __extension__ ({ \
63054	65054	float64x2_t __s1 = __p1; \
63055		__builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__s1, __p2, 42); \
	65055	__builtin_neon_vst1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 42); \
63056	65056	})
63057	65057	#else
63058	65058	#define vst1q_lane_f64(__p0, __p1, __p2) __extension__ ({ \
63059	65059	float64x2_t __s1 = __p1; \
63060		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
63061		__builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 42); \
	65060	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	65061	__builtin_neon_vst1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 42); \
	65062	})
	65063	#endif
	65064
	65065	#ifdef __LITTLE_ENDIAN__
	65066	#define vst1q_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	65067	mfloat8x16_t __s1 = __p1; \
	65068	__builtin_neon_vst1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 44); \
	65069	})
	65070	#else
	65071	#define vst1q_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	65072	mfloat8x16_t __s1 = __p1; \
	65073	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_8); \
	65074	__builtin_neon_vst1q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 44); \
63062	65075	})
63063	65076	#endif
63064	65077
63065	65078	#define vst1_lane_f64(__p0, __p1, __p2) __extension__ ({ \
63066	65079	float64x1_t __s1 = __p1; \
63067		__builtin_neon_vst1_lane_v(__p0, (int8x8_t)__s1, __p2, 10); \
	65080	__builtin_neon_vst1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 10); \
	65081	})
	65082	#ifdef __LITTLE_ENDIAN__
	65083	#define vst1_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	65084	mfloat8x8_t __s1 = __p1; \
	65085	__builtin_neon_vst1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 12); \
	65086	})
	65087	#else
	65088	#define vst1_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	65089	mfloat8x8_t __s1 = __p1; \
	65090	mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_64_8); \
	65091	__builtin_neon_vst1_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1), __p2, 12); \
63068	65092	})
	65093	#endif
	65094
63069	65095	#define vst1_p64_x2(__p0, __p1) __extension__ ({ \
63070	65096	poly64x1x2_t __s1 = __p1; \
63071		__builtin_neon_vst1_x2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 6); \
	65097	__builtin_neon_vst1_x2_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 6); \
63072	65098	})
63073	65099	#ifdef __LITTLE_ENDIAN__
63074	65100	#define vst1q_p64_x2(__p0, __p1) __extension__ ({ \
63075	65101	poly64x2x2_t __s1 = __p1; \
63076		__builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 38); \
	65102	__builtin_neon_vst1q_x2_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 38); \
63077	65103	})
63078	65104	#else
63079	65105	#define vst1q_p64_x2(__p0, __p1) __extension__ ({ \
63080	65106	poly64x2x2_t __s1 = __p1; \
63081	65107	poly64x2x2_t __rev1; \
63082		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
63083		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
63084		__builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 38); \
	65108	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	65109	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	65110	__builtin_neon_vst1q_x2_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 38); \
63085	65111	})
63086	65112	#endif
63087	65113
63088	65114	#ifdef __LITTLE_ENDIAN__
63089	65115	#define vst1q_f64_x2(__p0, __p1) __extension__ ({ \
63090	65116	float64x2x2_t __s1 = __p1; \
63091		__builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 42); \
	65117	__builtin_neon_vst1q_x2_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 42); \
63092	65118	})
63093	65119	#else
63094	65120	#define vst1q_f64_x2(__p0, __p1) __extension__ ({ \
63095	65121	float64x2x2_t __s1 = __p1; \
63096	65122	float64x2x2_t __rev1; \
63097		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
63098		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
63099		__builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 42); \
	65123	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	65124	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	65125	__builtin_neon_vst1q_x2_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 42); \
	65126	})
	65127	#endif
	65128
	65129	#ifdef __LITTLE_ENDIAN__
	65130	#define vst1q_mf8_x2(__p0, __p1) __extension__ ({ \
	65131	mfloat8x16x2_t __s1 = __p1; \
	65132	__builtin_neon_vst1q_x2_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 44); \
	65133	})
	65134	#else
	65135	#define vst1q_mf8_x2(__p0, __p1) __extension__ ({ \
	65136	mfloat8x16x2_t __s1 = __p1; \
	65137	mfloat8x16x2_t __rev1; \
	65138	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	65139	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	65140	__builtin_neon_vst1q_x2_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 44); \
63100	65141	})
63101	65142	#endif
63102	65143
63103	65144	#define vst1_f64_x2(__p0, __p1) __extension__ ({ \
63104	65145	float64x1x2_t __s1 = __p1; \
63105		__builtin_neon_vst1_x2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 10); \
	65146	__builtin_neon_vst1_x2_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 10); \
63106	65147	})
	65148	#ifdef __LITTLE_ENDIAN__
	65149	#define vst1_mf8_x2(__p0, __p1) __extension__ ({ \
	65150	mfloat8x8x2_t __s1 = __p1; \
	65151	__builtin_neon_vst1_x2_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 12); \
	65152	})
	65153	#else
	65154	#define vst1_mf8_x2(__p0, __p1) __extension__ ({ \
	65155	mfloat8x8x2_t __s1 = __p1; \
	65156	mfloat8x8x2_t __rev1; \
	65157	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	65158	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	65159	__builtin_neon_vst1_x2_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), 12); \
	65160	})
	65161	#endif
	65162
63107	65163	#define vst1_p64_x3(__p0, __p1) __extension__ ({ \
63108	65164	poly64x1x3_t __s1 = __p1; \
63109		__builtin_neon_vst1_x3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 6); \
	65165	__builtin_neon_vst1_x3_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 6); \
63110	65166	})
63111	65167	#ifdef __LITTLE_ENDIAN__
63112	65168	#define vst1q_p64_x3(__p0, __p1) __extension__ ({ \
63113	65169	poly64x2x3_t __s1 = __p1; \
63114		__builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 38); \
	65170	__builtin_neon_vst1q_x3_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 38); \
63115	65171	})
63116	65172	#else
63117	65173	#define vst1q_p64_x3(__p0, __p1) __extension__ ({ \
63118	65174	poly64x2x3_t __s1 = __p1; \
63119	65175	poly64x2x3_t __rev1; \
63120		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
63121		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
63122		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
63123		__builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 38); \
	65176	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	65177	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	65178	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	65179	__builtin_neon_vst1q_x3_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 38); \
63124	65180	})
63125	65181	#endif
63126	65182
63127	65183	#ifdef __LITTLE_ENDIAN__
63128	65184	#define vst1q_f64_x3(__p0, __p1) __extension__ ({ \
63129	65185	float64x2x3_t __s1 = __p1; \
63130		__builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 42); \
	65186	__builtin_neon_vst1q_x3_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 42); \
63131	65187	})
63132	65188	#else
63133	65189	#define vst1q_f64_x3(__p0, __p1) __extension__ ({ \
63134	65190	float64x2x3_t __s1 = __p1; \
63135	65191	float64x2x3_t __rev1; \
63136		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
63137		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
63138		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
63139		__builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 42); \
	65192	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	65193	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	65194	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	65195	__builtin_neon_vst1q_x3_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 42); \
	65196	})
	65197	#endif
	65198
	65199	#ifdef __LITTLE_ENDIAN__
	65200	#define vst1q_mf8_x3(__p0, __p1) __extension__ ({ \
	65201	mfloat8x16x3_t __s1 = __p1; \
	65202	__builtin_neon_vst1q_x3_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 44); \
	65203	})
	65204	#else
	65205	#define vst1q_mf8_x3(__p0, __p1) __extension__ ({ \
	65206	mfloat8x16x3_t __s1 = __p1; \
	65207	mfloat8x16x3_t __rev1; \
	65208	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	65209	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	65210	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	65211	__builtin_neon_vst1q_x3_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 44); \
63140	65212	})
63141	65213	#endif
63142	65214
63143	65215	#define vst1_f64_x3(__p0, __p1) __extension__ ({ \
63144	65216	float64x1x3_t __s1 = __p1; \
63145		__builtin_neon_vst1_x3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 10); \
	65217	__builtin_neon_vst1_x3_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 10); \
63146	65218	})
	65219	#ifdef __LITTLE_ENDIAN__
	65220	#define vst1_mf8_x3(__p0, __p1) __extension__ ({ \
	65221	mfloat8x8x3_t __s1 = __p1; \
	65222	__builtin_neon_vst1_x3_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 12); \
	65223	})
	65224	#else
	65225	#define vst1_mf8_x3(__p0, __p1) __extension__ ({ \
	65226	mfloat8x8x3_t __s1 = __p1; \
	65227	mfloat8x8x3_t __rev1; \
	65228	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	65229	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	65230	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	65231	__builtin_neon_vst1_x3_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), 12); \
	65232	})
	65233	#endif
	65234
63147	65235	#define vst1_p64_x4(__p0, __p1) __extension__ ({ \
63148	65236	poly64x1x4_t __s1 = __p1; \
63149		__builtin_neon_vst1_x4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 6); \
	65237	__builtin_neon_vst1_x4_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 6); \
63150	65238	})
63151	65239	#ifdef __LITTLE_ENDIAN__
63152	65240	#define vst1q_p64_x4(__p0, __p1) __extension__ ({ \
63153	65241	poly64x2x4_t __s1 = __p1; \
63154		__builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 38); \
	65242	__builtin_neon_vst1q_x4_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 38); \
63155	65243	})
63156	65244	#else
63157	65245	#define vst1q_p64_x4(__p0, __p1) __extension__ ({ \
63158	65246	poly64x2x4_t __s1 = __p1; \
63159	65247	poly64x2x4_t __rev1; \
63160		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
63161		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
63162		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
63163		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
63164		__builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 38); \
	65248	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	65249	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	65250	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	65251	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_64); \
	65252	__builtin_neon_vst1q_x4_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 38); \
63165	65253	})
63166	65254	#endif
63167	65255
63168	65256	#ifdef __LITTLE_ENDIAN__
63169	65257	#define vst1q_f64_x4(__p0, __p1) __extension__ ({ \
63170	65258	float64x2x4_t __s1 = __p1; \
63171		__builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 42); \
	65259	__builtin_neon_vst1q_x4_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 42); \
63172	65260	})
63173	65261	#else
63174	65262	#define vst1q_f64_x4(__p0, __p1) __extension__ ({ \
63175	65263	float64x2x4_t __s1 = __p1; \
63176	65264	float64x2x4_t __rev1; \
63177		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
63178		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
63179		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
63180		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
63181		__builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 42); \
	65265	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	65266	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	65267	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	65268	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_64); \
	65269	__builtin_neon_vst1q_x4_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 42); \
	65270	})
	65271	#endif
	65272
	65273	#ifdef __LITTLE_ENDIAN__
	65274	#define vst1q_mf8_x4(__p0, __p1) __extension__ ({ \
	65275	mfloat8x16x4_t __s1 = __p1; \
	65276	__builtin_neon_vst1q_x4_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 44); \
	65277	})
	65278	#else
	65279	#define vst1q_mf8_x4(__p0, __p1) __extension__ ({ \
	65280	mfloat8x16x4_t __s1 = __p1; \
	65281	mfloat8x16x4_t __rev1; \
	65282	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	65283	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	65284	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	65285	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_8); \
	65286	__builtin_neon_vst1q_x4_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 44); \
63182	65287	})
63183	65288	#endif
63184	65289
63185	65290	#define vst1_f64_x4(__p0, __p1) __extension__ ({ \
63186	65291	float64x1x4_t __s1 = __p1; \
63187		__builtin_neon_vst1_x4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 10); \
	65292	__builtin_neon_vst1_x4_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 10); \
	65293	})
	65294	#ifdef __LITTLE_ENDIAN__
	65295	#define vst1_mf8_x4(__p0, __p1) __extension__ ({ \
	65296	mfloat8x8x4_t __s1 = __p1; \
	65297	__builtin_neon_vst1_x4_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 12); \
	65298	})
	65299	#else
	65300	#define vst1_mf8_x4(__p0, __p1) __extension__ ({ \
	65301	mfloat8x8x4_t __s1 = __p1; \
	65302	mfloat8x8x4_t __rev1; \
	65303	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	65304	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	65305	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	65306	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_8); \
	65307	__builtin_neon_vst1_x4_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), 12); \
63188	65308	})
	65309	#endif
	65310
63189	65311	#define vst2_p64(__p0, __p1) __extension__ ({ \
63190	65312	poly64x1x2_t __s1 = __p1; \
63191		__builtin_neon_vst2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 6); \
	65313	__builtin_neon_vst2_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 6); \
63192	65314	})
63193	65315	#ifdef __LITTLE_ENDIAN__
63194	65316	#define vst2q_p64(__p0, __p1) __extension__ ({ \
63195	65317	poly64x2x2_t __s1 = __p1; \
63196		__builtin_neon_vst2q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 38); \
	65318	__builtin_neon_vst2q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 38); \
63197	65319	})
63198	65320	#else
63199	65321	#define vst2q_p64(__p0, __p1) __extension__ ({ \
63200	65322	poly64x2x2_t __s1 = __p1; \
63201	65323	poly64x2x2_t __rev1; \
63202		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
63203		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
63204		__builtin_neon_vst2q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 38); \
	65324	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	65325	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	65326	__builtin_neon_vst2q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 38); \
63205	65327	})
63206	65328	#endif
63207	65329
63208	65330	#ifdef __LITTLE_ENDIAN__
63209	65331	#define vst2q_u64(__p0, __p1) __extension__ ({ \
63210	65332	uint64x2x2_t __s1 = __p1; \
63211		__builtin_neon_vst2q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 51); \
	65333	__builtin_neon_vst2q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 51); \
63212	65334	})
63213	65335	#else
63214	65336	#define vst2q_u64(__p0, __p1) __extension__ ({ \
63215	65337	uint64x2x2_t __s1 = __p1; \
63216	65338	uint64x2x2_t __rev1; \
63217		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
63218		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
63219		__builtin_neon_vst2q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 51); \
	65339	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	65340	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	65341	__builtin_neon_vst2q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 51); \
63220	65342	})
63221	65343	#endif
63222	65344
63223	65345	#ifdef __LITTLE_ENDIAN__
63224	65346	#define vst2q_f64(__p0, __p1) __extension__ ({ \
63225	65347	float64x2x2_t __s1 = __p1; \
63226		__builtin_neon_vst2q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 42); \
	65348	__builtin_neon_vst2q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 42); \
63227	65349	})
63228	65350	#else
63229	65351	#define vst2q_f64(__p0, __p1) __extension__ ({ \
63230	65352	float64x2x2_t __s1 = __p1; \
63231	65353	float64x2x2_t __rev1; \
63232		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
63233		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
63234		__builtin_neon_vst2q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 42); \
	65354	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	65355	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	65356	__builtin_neon_vst2q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 42); \
63235	65357	})
63236	65358	#endif
63237	65359
63238	65360	#ifdef __LITTLE_ENDIAN__
63239	65361	#define vst2q_s64(__p0, __p1) __extension__ ({ \
63240	65362	int64x2x2_t __s1 = __p1; \
63241		__builtin_neon_vst2q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 35); \
	65363	__builtin_neon_vst2q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 35); \
63242	65364	})
63243	65365	#else
63244	65366	#define vst2q_s64(__p0, __p1) __extension__ ({ \
63245	65367	int64x2x2_t __s1 = __p1; \
63246	65368	int64x2x2_t __rev1; \
63247		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
63248		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
63249		__builtin_neon_vst2q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 35); \
	65369	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	65370	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	65371	__builtin_neon_vst2q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 35); \
	65372	})
	65373	#endif
	65374
	65375	#ifdef __LITTLE_ENDIAN__
	65376	#define vst2q_mf8(__p0, __p1) __extension__ ({ \
	65377	mfloat8x16x2_t __s1 = __p1; \
	65378	__builtin_neon_vst2q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), 44); \
	65379	})
	65380	#else
	65381	#define vst2q_mf8(__p0, __p1) __extension__ ({ \
	65382	mfloat8x16x2_t __s1 = __p1; \
	65383	mfloat8x16x2_t __rev1; \
	65384	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	65385	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	65386	__builtin_neon_vst2q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), 44); \
63250	65387	})
63251	65388	#endif
63252	65389
63253	65390	#define vst2_f64(__p0, __p1) __extension__ ({ \
63254	65391	float64x1x2_t __s1 = __p1; \
63255		__builtin_neon_vst2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 10); \
	65392	__builtin_neon_vst2_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 10); \
	65393	})
	65394	#ifdef __LITTLE_ENDIAN__
	65395	#define vst2_mf8(__p0, __p1) __extension__ ({ \
	65396	mfloat8x8x2_t __s1 = __p1; \
	65397	__builtin_neon_vst2_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), 12); \
	65398	})
	65399	#else
	65400	#define vst2_mf8(__p0, __p1) __extension__ ({ \
	65401	mfloat8x8x2_t __s1 = __p1; \
	65402	mfloat8x8x2_t __rev1; \
	65403	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	65404	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	65405	__builtin_neon_vst2_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), 12); \
63256	65406	})
	65407	#endif
	65408
63257	65409	#define vst2_lane_p64(__p0, __p1, __p2) __extension__ ({ \
63258	65410	poly64x1x2_t __s1 = __p1; \
63259		__builtin_neon_vst2_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 6); \
	65411	__builtin_neon_vst2_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 6); \
63260	65412	})
63261	65413	#ifdef __LITTLE_ENDIAN__
63262	65414	#define vst2q_lane_p8(__p0, __p1, __p2) __extension__ ({ \
63263	65415	poly8x16x2_t __s1 = __p1; \
63264		__builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 36); \
	65416	__builtin_neon_vst2q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 36); \
63265	65417	})
63266	65418	#else
63267	65419	#define vst2q_lane_p8(__p0, __p1, __p2) __extension__ ({ \
63268	65420	poly8x16x2_t __s1 = __p1; \
63269	65421	poly8x16x2_t __rev1; \
63270		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
63271		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
63272		__builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 36); \
	65422	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	65423	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	65424	__builtin_neon_vst2q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 36); \
63273	65425	})
63274	65426	#endif
63275	65427
63276	65428	#ifdef __LITTLE_ENDIAN__
63277	65429	#define vst2q_lane_p64(__p0, __p1, __p2) __extension__ ({ \
63278	65430	poly64x2x2_t __s1 = __p1; \
63279		__builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 38); \
	65431	__builtin_neon_vst2q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 38); \
63280	65432	})
63281	65433	#else
63282	65434	#define vst2q_lane_p64(__p0, __p1, __p2) __extension__ ({ \
63283	65435	poly64x2x2_t __s1 = __p1; \
63284	65436	poly64x2x2_t __rev1; \
63285		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
63286		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
63287		__builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 38); \
	65437	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	65438	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	65439	__builtin_neon_vst2q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 38); \
63288	65440	})
63289	65441	#endif
63290	65442
63291	65443	#ifdef __LITTLE_ENDIAN__
63292	65444	#define vst2q_lane_u8(__p0, __p1, __p2) __extension__ ({ \
63293	65445	uint8x16x2_t __s1 = __p1; \
63294		__builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 48); \
	65446	__builtin_neon_vst2q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 48); \
63295	65447	})
63296	65448	#else
63297	65449	#define vst2q_lane_u8(__p0, __p1, __p2) __extension__ ({ \
63298	65450	uint8x16x2_t __s1 = __p1; \
63299	65451	uint8x16x2_t __rev1; \
63300		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
63301		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
63302		__builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 48); \
	65452	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	65453	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	65454	__builtin_neon_vst2q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 48); \
63303	65455	})
63304	65456	#endif
63305	65457
63306	65458	#ifdef __LITTLE_ENDIAN__
63307	65459	#define vst2q_lane_u64(__p0, __p1, __p2) __extension__ ({ \
63308	65460	uint64x2x2_t __s1 = __p1; \
63309		__builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 51); \
	65461	__builtin_neon_vst2q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 51); \
63310	65462	})
63311	65463	#else
63312	65464	#define vst2q_lane_u64(__p0, __p1, __p2) __extension__ ({ \
63313	65465	uint64x2x2_t __s1 = __p1; \
63314	65466	uint64x2x2_t __rev1; \
63315		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
63316		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
63317		__builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 51); \
	65467	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	65468	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	65469	__builtin_neon_vst2q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 51); \
63318	65470	})
63319	65471	#endif
63320	65472
63321	65473	#ifdef __LITTLE_ENDIAN__
63322	65474	#define vst2q_lane_s8(__p0, __p1, __p2) __extension__ ({ \
63323	65475	int8x16x2_t __s1 = __p1; \
63324		__builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 32); \
	65476	__builtin_neon_vst2q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 32); \
63325	65477	})
63326	65478	#else
63327	65479	#define vst2q_lane_s8(__p0, __p1, __p2) __extension__ ({ \
63328	65480	int8x16x2_t __s1 = __p1; \
63329	65481	int8x16x2_t __rev1; \
63330		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
63331		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
63332		__builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 32); \
	65482	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	65483	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	65484	__builtin_neon_vst2q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 32); \
63333	65485	})
63334	65486	#endif
63335	65487
63336	65488	#ifdef __LITTLE_ENDIAN__
63337	65489	#define vst2q_lane_f64(__p0, __p1, __p2) __extension__ ({ \
63338	65490	float64x2x2_t __s1 = __p1; \
63339		__builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 42); \
	65491	__builtin_neon_vst2q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 42); \
63340	65492	})
63341	65493	#else
63342	65494	#define vst2q_lane_f64(__p0, __p1, __p2) __extension__ ({ \
63343	65495	float64x2x2_t __s1 = __p1; \
63344	65496	float64x2x2_t __rev1; \
63345		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
63346		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
63347		__builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 42); \
	65497	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	65498	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	65499	__builtin_neon_vst2q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 42); \
63348	65500	})
63349	65501	#endif
63350	65502
63351	65503	#ifdef __LITTLE_ENDIAN__
63352	65504	#define vst2q_lane_s64(__p0, __p1, __p2) __extension__ ({ \
63353	65505	int64x2x2_t __s1 = __p1; \
63354		__builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 35); \
	65506	__builtin_neon_vst2q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 35); \
63355	65507	})
63356	65508	#else
63357	65509	#define vst2q_lane_s64(__p0, __p1, __p2) __extension__ ({ \
63358	65510	int64x2x2_t __s1 = __p1; \
63359	65511	int64x2x2_t __rev1; \
63360		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
63361		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
63362		__builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 35); \
	65512	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	65513	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	65514	__builtin_neon_vst2q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 35); \
	65515	})
	65516	#endif
	65517
	65518	#ifdef __LITTLE_ENDIAN__
	65519	#define vst2q_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	65520	mfloat8x16x2_t __s1 = __p1; \
	65521	__builtin_neon_vst2q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __p2, 44); \
	65522	})
	65523	#else
	65524	#define vst2q_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	65525	mfloat8x16x2_t __s1 = __p1; \
	65526	mfloat8x16x2_t __rev1; \
	65527	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	65528	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	65529	__builtin_neon_vst2q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __p2, 44); \
63363	65530	})
63364	65531	#endif
63365	65532
63366	65533	#define vst2_lane_u64(__p0, __p1, __p2) __extension__ ({ \
63367	65534	uint64x1x2_t __s1 = __p1; \
63368		__builtin_neon_vst2_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 19); \
	65535	__builtin_neon_vst2_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 19); \
63369	65536	})
63370	65537	#define vst2_lane_f64(__p0, __p1, __p2) __extension__ ({ \
63371	65538	float64x1x2_t __s1 = __p1; \
63372		__builtin_neon_vst2_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 10); \
	65539	__builtin_neon_vst2_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 10); \
63373	65540	})
63374	65541	#define vst2_lane_s64(__p0, __p1, __p2) __extension__ ({ \
63375	65542	int64x1x2_t __s1 = __p1; \
63376		__builtin_neon_vst2_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 3); \
	65543	__builtin_neon_vst2_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 3); \
	65544	})
	65545	#ifdef __LITTLE_ENDIAN__
	65546	#define vst2_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	65547	mfloat8x8x2_t __s1 = __p1; \
	65548	__builtin_neon_vst2_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __p2, 12); \
	65549	})
	65550	#else
	65551	#define vst2_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	65552	mfloat8x8x2_t __s1 = __p1; \
	65553	mfloat8x8x2_t __rev1; \
	65554	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	65555	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	65556	__builtin_neon_vst2_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __p2, 12); \
63377	65557	})
	65558	#endif
	65559
63378	65560	#define vst3_p64(__p0, __p1) __extension__ ({ \
63379	65561	poly64x1x3_t __s1 = __p1; \
63380		__builtin_neon_vst3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 6); \
	65562	__builtin_neon_vst3_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 6); \
63381	65563	})
63382	65564	#ifdef __LITTLE_ENDIAN__
63383	65565	#define vst3q_p64(__p0, __p1) __extension__ ({ \
63384	65566	poly64x2x3_t __s1 = __p1; \
63385		__builtin_neon_vst3q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 38); \
	65567	__builtin_neon_vst3q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 38); \
63386	65568	})
63387	65569	#else
63388	65570	#define vst3q_p64(__p0, __p1) __extension__ ({ \
63389	65571	poly64x2x3_t __s1 = __p1; \
63390	65572	poly64x2x3_t __rev1; \
63391		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
63392		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
63393		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
63394		__builtin_neon_vst3q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 38); \
	65573	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	65574	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	65575	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	65576	__builtin_neon_vst3q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 38); \
63395	65577	})
63396	65578	#endif
63397	65579
63398	65580	#ifdef __LITTLE_ENDIAN__
63399	65581	#define vst3q_u64(__p0, __p1) __extension__ ({ \
63400	65582	uint64x2x3_t __s1 = __p1; \
63401		__builtin_neon_vst3q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 51); \
	65583	__builtin_neon_vst3q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 51); \
63402	65584	})
63403	65585	#else
63404	65586	#define vst3q_u64(__p0, __p1) __extension__ ({ \
63405	65587	uint64x2x3_t __s1 = __p1; \
63406	65588	uint64x2x3_t __rev1; \
63407		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
63408		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
63409		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
63410		__builtin_neon_vst3q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 51); \
	65589	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	65590	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	65591	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	65592	__builtin_neon_vst3q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 51); \
63411	65593	})
63412	65594	#endif
63413	65595
63414	65596	#ifdef __LITTLE_ENDIAN__
63415	65597	#define vst3q_f64(__p0, __p1) __extension__ ({ \
63416	65598	float64x2x3_t __s1 = __p1; \
63417		__builtin_neon_vst3q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 42); \
	65599	__builtin_neon_vst3q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 42); \
63418	65600	})
63419	65601	#else
63420	65602	#define vst3q_f64(__p0, __p1) __extension__ ({ \
63421	65603	float64x2x3_t __s1 = __p1; \
63422	65604	float64x2x3_t __rev1; \
63423		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
63424		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
63425		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
63426		__builtin_neon_vst3q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 42); \
	65605	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	65606	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	65607	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	65608	__builtin_neon_vst3q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 42); \
63427	65609	})
63428	65610	#endif
63429	65611
63430	65612	#ifdef __LITTLE_ENDIAN__
63431	65613	#define vst3q_s64(__p0, __p1) __extension__ ({ \
63432	65614	int64x2x3_t __s1 = __p1; \
63433		__builtin_neon_vst3q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 35); \
	65615	__builtin_neon_vst3q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 35); \
63434	65616	})
63435	65617	#else
63436	65618	#define vst3q_s64(__p0, __p1) __extension__ ({ \
63437	65619	int64x2x3_t __s1 = __p1; \
63438	65620	int64x2x3_t __rev1; \
63439		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
63440		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
63441		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
63442		__builtin_neon_vst3q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 35); \
	65621	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	65622	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	65623	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	65624	__builtin_neon_vst3q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 35); \
	65625	})
	65626	#endif
	65627
	65628	#ifdef __LITTLE_ENDIAN__
	65629	#define vst3q_mf8(__p0, __p1) __extension__ ({ \
	65630	mfloat8x16x3_t __s1 = __p1; \
	65631	__builtin_neon_vst3q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), 44); \
	65632	})
	65633	#else
	65634	#define vst3q_mf8(__p0, __p1) __extension__ ({ \
	65635	mfloat8x16x3_t __s1 = __p1; \
	65636	mfloat8x16x3_t __rev1; \
	65637	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	65638	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	65639	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	65640	__builtin_neon_vst3q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), 44); \
63443	65641	})
63444	65642	#endif
63445	65643
63446	65644	#define vst3_f64(__p0, __p1) __extension__ ({ \
63447	65645	float64x1x3_t __s1 = __p1; \
63448		__builtin_neon_vst3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 10); \
	65646	__builtin_neon_vst3_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 10); \
	65647	})
	65648	#ifdef __LITTLE_ENDIAN__
	65649	#define vst3_mf8(__p0, __p1) __extension__ ({ \
	65650	mfloat8x8x3_t __s1 = __p1; \
	65651	__builtin_neon_vst3_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), 12); \
	65652	})
	65653	#else
	65654	#define vst3_mf8(__p0, __p1) __extension__ ({ \
	65655	mfloat8x8x3_t __s1 = __p1; \
	65656	mfloat8x8x3_t __rev1; \
	65657	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	65658	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	65659	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	65660	__builtin_neon_vst3_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), 12); \
63449	65661	})
	65662	#endif
	65663
63450	65664	#define vst3_lane_p64(__p0, __p1, __p2) __extension__ ({ \
63451	65665	poly64x1x3_t __s1 = __p1; \
63452		__builtin_neon_vst3_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 6); \
	65666	__builtin_neon_vst3_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 6); \
63453	65667	})
63454	65668	#ifdef __LITTLE_ENDIAN__
63455	65669	#define vst3q_lane_p8(__p0, __p1, __p2) __extension__ ({ \
63456	65670	poly8x16x3_t __s1 = __p1; \
63457		__builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 36); \
	65671	__builtin_neon_vst3q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 36); \
63458	65672	})
63459	65673	#else
63460	65674	#define vst3q_lane_p8(__p0, __p1, __p2) __extension__ ({ \
63461	65675	poly8x16x3_t __s1 = __p1; \
63462	65676	poly8x16x3_t __rev1; \
63463		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
63464		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
63465		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
63466		__builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 36); \
	65677	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	65678	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	65679	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	65680	__builtin_neon_vst3q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 36); \
63467	65681	})
63468	65682	#endif
63469	65683
63470	65684	#ifdef __LITTLE_ENDIAN__
63471	65685	#define vst3q_lane_p64(__p0, __p1, __p2) __extension__ ({ \
63472	65686	poly64x2x3_t __s1 = __p1; \
63473		__builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 38); \
	65687	__builtin_neon_vst3q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 38); \
63474	65688	})
63475	65689	#else
63476	65690	#define vst3q_lane_p64(__p0, __p1, __p2) __extension__ ({ \
63477	65691	poly64x2x3_t __s1 = __p1; \
63478	65692	poly64x2x3_t __rev1; \
63479		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
63480		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
63481		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
63482		__builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 38); \
	65693	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	65694	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	65695	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	65696	__builtin_neon_vst3q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 38); \
63483	65697	})
63484	65698	#endif
63485	65699
63486	65700	#ifdef __LITTLE_ENDIAN__
63487	65701	#define vst3q_lane_u8(__p0, __p1, __p2) __extension__ ({ \
63488	65702	uint8x16x3_t __s1 = __p1; \
63489		__builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 48); \
	65703	__builtin_neon_vst3q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 48); \
63490	65704	})
63491	65705	#else
63492	65706	#define vst3q_lane_u8(__p0, __p1, __p2) __extension__ ({ \
63493	65707	uint8x16x3_t __s1 = __p1; \
63494	65708	uint8x16x3_t __rev1; \
63495		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
63496		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
63497		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
63498		__builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 48); \
	65709	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	65710	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	65711	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	65712	__builtin_neon_vst3q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 48); \
63499	65713	})
63500	65714	#endif
63501	65715
63502	65716	#ifdef __LITTLE_ENDIAN__
63503	65717	#define vst3q_lane_u64(__p0, __p1, __p2) __extension__ ({ \
63504	65718	uint64x2x3_t __s1 = __p1; \
63505		__builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 51); \
	65719	__builtin_neon_vst3q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 51); \
63506	65720	})
63507	65721	#else
63508	65722	#define vst3q_lane_u64(__p0, __p1, __p2) __extension__ ({ \
63509	65723	uint64x2x3_t __s1 = __p1; \
63510	65724	uint64x2x3_t __rev1; \
63511		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
63512		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
63513		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
63514		__builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 51); \
	65725	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	65726	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	65727	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	65728	__builtin_neon_vst3q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 51); \
63515	65729	})
63516	65730	#endif
63517	65731
63518	65732	#ifdef __LITTLE_ENDIAN__
63519	65733	#define vst3q_lane_s8(__p0, __p1, __p2) __extension__ ({ \
63520	65734	int8x16x3_t __s1 = __p1; \
63521		__builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 32); \
	65735	__builtin_neon_vst3q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 32); \
63522	65736	})
63523	65737	#else
63524	65738	#define vst3q_lane_s8(__p0, __p1, __p2) __extension__ ({ \
63525	65739	int8x16x3_t __s1 = __p1; \
63526	65740	int8x16x3_t __rev1; \
63527		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
63528		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
63529		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
63530		__builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 32); \
	65741	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	65742	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	65743	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	65744	__builtin_neon_vst3q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 32); \
63531	65745	})
63532	65746	#endif
63533	65747
63534	65748	#ifdef __LITTLE_ENDIAN__
63535	65749	#define vst3q_lane_f64(__p0, __p1, __p2) __extension__ ({ \
63536	65750	float64x2x3_t __s1 = __p1; \
63537		__builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 42); \
	65751	__builtin_neon_vst3q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 42); \
63538	65752	})
63539	65753	#else
63540	65754	#define vst3q_lane_f64(__p0, __p1, __p2) __extension__ ({ \
63541	65755	float64x2x3_t __s1 = __p1; \
63542	65756	float64x2x3_t __rev1; \
63543		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
63544		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
63545		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
63546		__builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 42); \
	65757	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	65758	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	65759	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	65760	__builtin_neon_vst3q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 42); \
63547	65761	})
63548	65762	#endif
63549	65763
63550	65764	#ifdef __LITTLE_ENDIAN__
63551	65765	#define vst3q_lane_s64(__p0, __p1, __p2) __extension__ ({ \
63552	65766	int64x2x3_t __s1 = __p1; \
63553		__builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 35); \
	65767	__builtin_neon_vst3q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 35); \
63554	65768	})
63555	65769	#else
63556	65770	#define vst3q_lane_s64(__p0, __p1, __p2) __extension__ ({ \
63557	65771	int64x2x3_t __s1 = __p1; \
63558	65772	int64x2x3_t __rev1; \
63559		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
63560		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
63561		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
63562		__builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 35); \
	65773	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	65774	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	65775	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	65776	__builtin_neon_vst3q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 35); \
	65777	})
	65778	#endif
	65779
	65780	#ifdef __LITTLE_ENDIAN__
	65781	#define vst3q_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	65782	mfloat8x16x3_t __s1 = __p1; \
	65783	__builtin_neon_vst3q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __p2, 44); \
	65784	})
	65785	#else
	65786	#define vst3q_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	65787	mfloat8x16x3_t __s1 = __p1; \
	65788	mfloat8x16x3_t __rev1; \
	65789	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	65790	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	65791	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	65792	__builtin_neon_vst3q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __p2, 44); \
63563	65793	})
63564	65794	#endif
63565	65795
63566	65796	#define vst3_lane_u64(__p0, __p1, __p2) __extension__ ({ \
63567	65797	uint64x1x3_t __s1 = __p1; \
63568		__builtin_neon_vst3_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 19); \
	65798	__builtin_neon_vst3_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 19); \
63569	65799	})
63570	65800	#define vst3_lane_f64(__p0, __p1, __p2) __extension__ ({ \
63571	65801	float64x1x3_t __s1 = __p1; \
63572		__builtin_neon_vst3_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 10); \
	65802	__builtin_neon_vst3_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 10); \
63573	65803	})
63574	65804	#define vst3_lane_s64(__p0, __p1, __p2) __extension__ ({ \
63575	65805	int64x1x3_t __s1 = __p1; \
63576		__builtin_neon_vst3_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 3); \
	65806	__builtin_neon_vst3_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 3); \
	65807	})
	65808	#ifdef __LITTLE_ENDIAN__
	65809	#define vst3_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	65810	mfloat8x8x3_t __s1 = __p1; \
	65811	__builtin_neon_vst3_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __p2, 12); \
63577	65812	})
	65813	#else
	65814	#define vst3_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	65815	mfloat8x8x3_t __s1 = __p1; \
	65816	mfloat8x8x3_t __rev1; \
	65817	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	65818	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	65819	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	65820	__builtin_neon_vst3_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __p2, 12); \
	65821	})
	65822	#endif
	65823
63578	65824	#define vst4_p64(__p0, __p1) __extension__ ({ \
63579	65825	poly64x1x4_t __s1 = __p1; \
63580		__builtin_neon_vst4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 6); \
	65826	__builtin_neon_vst4_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 6); \
63581	65827	})
63582	65828	#ifdef __LITTLE_ENDIAN__
63583	65829	#define vst4q_p64(__p0, __p1) __extension__ ({ \
63584	65830	poly64x2x4_t __s1 = __p1; \
63585		__builtin_neon_vst4q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 38); \
	65831	__builtin_neon_vst4q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 38); \
63586	65832	})
63587	65833	#else
63588	65834	#define vst4q_p64(__p0, __p1) __extension__ ({ \
63589	65835	poly64x2x4_t __s1 = __p1; \
63590	65836	poly64x2x4_t __rev1; \
63591		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
63592		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
63593		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
63594		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
63595		__builtin_neon_vst4q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 38); \
	65837	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	65838	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	65839	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	65840	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_64); \
	65841	__builtin_neon_vst4q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 38); \
63596	65842	})
63597	65843	#endif
63598	65844
63599	65845	#ifdef __LITTLE_ENDIAN__
63600	65846	#define vst4q_u64(__p0, __p1) __extension__ ({ \
63601	65847	uint64x2x4_t __s1 = __p1; \
63602		__builtin_neon_vst4q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 51); \
	65848	__builtin_neon_vst4q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 51); \
63603	65849	})
63604	65850	#else
63605	65851	#define vst4q_u64(__p0, __p1) __extension__ ({ \
63606	65852	uint64x2x4_t __s1 = __p1; \
63607	65853	uint64x2x4_t __rev1; \
63608		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
63609		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
63610		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
63611		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
63612		__builtin_neon_vst4q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 51); \
	65854	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	65855	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	65856	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	65857	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_64); \
	65858	__builtin_neon_vst4q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 51); \
63613	65859	})
63614	65860	#endif
63615	65861
63616	65862	#ifdef __LITTLE_ENDIAN__
63617	65863	#define vst4q_f64(__p0, __p1) __extension__ ({ \
63618	65864	float64x2x4_t __s1 = __p1; \
63619		__builtin_neon_vst4q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 42); \
	65865	__builtin_neon_vst4q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 42); \
63620	65866	})
63621	65867	#else
63622	65868	#define vst4q_f64(__p0, __p1) __extension__ ({ \
63623	65869	float64x2x4_t __s1 = __p1; \
63624	65870	float64x2x4_t __rev1; \
63625		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
63626		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
63627		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
63628		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
63629		__builtin_neon_vst4q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 42); \
	65871	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	65872	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	65873	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	65874	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_64); \
	65875	__builtin_neon_vst4q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 42); \
63630	65876	})
63631	65877	#endif
63632	65878
63633	65879	#ifdef __LITTLE_ENDIAN__
63634	65880	#define vst4q_s64(__p0, __p1) __extension__ ({ \
63635	65881	int64x2x4_t __s1 = __p1; \
63636		__builtin_neon_vst4q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 35); \
	65882	__builtin_neon_vst4q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 35); \
63637	65883	})
63638	65884	#else
63639	65885	#define vst4q_s64(__p0, __p1) __extension__ ({ \
63640	65886	int64x2x4_t __s1 = __p1; \
63641	65887	int64x2x4_t __rev1; \
63642		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
63643		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
63644		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
63645		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
63646		__builtin_neon_vst4q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 35); \
	65888	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	65889	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	65890	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	65891	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_64); \
	65892	__builtin_neon_vst4q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 35); \
	65893	})
	65894	#endif
	65895
	65896	#ifdef __LITTLE_ENDIAN__
	65897	#define vst4q_mf8(__p0, __p1) __extension__ ({ \
	65898	mfloat8x16x4_t __s1 = __p1; \
	65899	__builtin_neon_vst4q_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), 44); \
	65900	})
	65901	#else
	65902	#define vst4q_mf8(__p0, __p1) __extension__ ({ \
	65903	mfloat8x16x4_t __s1 = __p1; \
	65904	mfloat8x16x4_t __rev1; \
	65905	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	65906	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	65907	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	65908	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_8); \
	65909	__builtin_neon_vst4q_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), 44); \
63647	65910	})
63648	65911	#endif
63649	65912
63650	65913	#define vst4_f64(__p0, __p1) __extension__ ({ \
63651	65914	float64x1x4_t __s1 = __p1; \
63652		__builtin_neon_vst4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 10); \
	65915	__builtin_neon_vst4_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 10); \
	65916	})
	65917	#ifdef __LITTLE_ENDIAN__
	65918	#define vst4_mf8(__p0, __p1) __extension__ ({ \
	65919	mfloat8x8x4_t __s1 = __p1; \
	65920	__builtin_neon_vst4_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), 12); \
63653	65921	})
	65922	#else
	65923	#define vst4_mf8(__p0, __p1) __extension__ ({ \
	65924	mfloat8x8x4_t __s1 = __p1; \
	65925	mfloat8x8x4_t __rev1; \
	65926	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	65927	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	65928	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	65929	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_8); \
	65930	__builtin_neon_vst4_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), 12); \
	65931	})
	65932	#endif
	65933
63654	65934	#define vst4_lane_p64(__p0, __p1, __p2) __extension__ ({ \
63655	65935	poly64x1x4_t __s1 = __p1; \
63656		__builtin_neon_vst4_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 6); \
	65936	__builtin_neon_vst4_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 6); \
63657	65937	})
63658	65938	#ifdef __LITTLE_ENDIAN__
63659	65939	#define vst4q_lane_p8(__p0, __p1, __p2) __extension__ ({ \
63660	65940	poly8x16x4_t __s1 = __p1; \
63661		__builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 36); \
	65941	__builtin_neon_vst4q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 36); \
63662	65942	})
63663	65943	#else
63664	65944	#define vst4q_lane_p8(__p0, __p1, __p2) __extension__ ({ \
63665	65945	poly8x16x4_t __s1 = __p1; \
63666	65946	poly8x16x4_t __rev1; \
63667		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
63668		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
63669		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
63670		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
63671		__builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 36); \
	65947	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	65948	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	65949	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	65950	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_8); \
	65951	__builtin_neon_vst4q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 36); \
63672	65952	})
63673	65953	#endif
63674	65954
63675	65955	#ifdef __LITTLE_ENDIAN__
63676	65956	#define vst4q_lane_p64(__p0, __p1, __p2) __extension__ ({ \
63677	65957	poly64x2x4_t __s1 = __p1; \
63678		__builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 38); \
	65958	__builtin_neon_vst4q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 38); \
63679	65959	})
63680	65960	#else
63681	65961	#define vst4q_lane_p64(__p0, __p1, __p2) __extension__ ({ \
63682	65962	poly64x2x4_t __s1 = __p1; \
63683	65963	poly64x2x4_t __rev1; \
63684		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
63685		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
63686		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
63687		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
63688		__builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 38); \
	65964	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	65965	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	65966	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	65967	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_64); \
	65968	__builtin_neon_vst4q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 38); \
63689	65969	})
63690	65970	#endif
63691	65971
63692	65972	#ifdef __LITTLE_ENDIAN__
63693	65973	#define vst4q_lane_u8(__p0, __p1, __p2) __extension__ ({ \
63694	65974	uint8x16x4_t __s1 = __p1; \
63695		__builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 48); \
	65975	__builtin_neon_vst4q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 48); \
63696	65976	})
63697	65977	#else
63698	65978	#define vst4q_lane_u8(__p0, __p1, __p2) __extension__ ({ \
63699	65979	uint8x16x4_t __s1 = __p1; \
63700	65980	uint8x16x4_t __rev1; \
63701		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
63702		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
63703		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
63704		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
63705		__builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 48); \
	65981	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	65982	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	65983	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	65984	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_8); \
	65985	__builtin_neon_vst4q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 48); \
63706	65986	})
63707	65987	#endif
63708	65988
63709	65989	#ifdef __LITTLE_ENDIAN__
63710	65990	#define vst4q_lane_u64(__p0, __p1, __p2) __extension__ ({ \
63711	65991	uint64x2x4_t __s1 = __p1; \
63712		__builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 51); \
	65992	__builtin_neon_vst4q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 51); \
63713	65993	})
63714	65994	#else
63715	65995	#define vst4q_lane_u64(__p0, __p1, __p2) __extension__ ({ \
63716	65996	uint64x2x4_t __s1 = __p1; \
63717	65997	uint64x2x4_t __rev1; \
63718		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
63719		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
63720		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
63721		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
63722		__builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 51); \
	65998	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	65999	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	66000	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	66001	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_64); \
	66002	__builtin_neon_vst4q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 51); \
63723	66003	})
63724	66004	#endif
63725	66005
63726	66006	#ifdef __LITTLE_ENDIAN__
63727	66007	#define vst4q_lane_s8(__p0, __p1, __p2) __extension__ ({ \
63728	66008	int8x16x4_t __s1 = __p1; \
63729		__builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 32); \
	66009	__builtin_neon_vst4q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 32); \
63730	66010	})
63731	66011	#else
63732	66012	#define vst4q_lane_s8(__p0, __p1, __p2) __extension__ ({ \
63733	66013	int8x16x4_t __s1 = __p1; \
63734	66014	int8x16x4_t __rev1; \
63735		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
63736		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
63737		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
63738		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
63739		__builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 32); \
	66015	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	66016	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	66017	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	66018	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_8); \
	66019	__builtin_neon_vst4q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 32); \
63740	66020	})
63741	66021	#endif
63742	66022
63743	66023	#ifdef __LITTLE_ENDIAN__
63744	66024	#define vst4q_lane_f64(__p0, __p1, __p2) __extension__ ({ \
63745	66025	float64x2x4_t __s1 = __p1; \
63746		__builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 42); \
	66026	__builtin_neon_vst4q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 42); \
63747	66027	})
63748	66028	#else
63749	66029	#define vst4q_lane_f64(__p0, __p1, __p2) __extension__ ({ \
63750	66030	float64x2x4_t __s1 = __p1; \
63751	66031	float64x2x4_t __rev1; \
63752		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
63753		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
63754		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
63755		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
63756		__builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 42); \
	66032	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	66033	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	66034	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	66035	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_64); \
	66036	__builtin_neon_vst4q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 42); \
63757	66037	})
63758	66038	#endif
63759	66039
63760	66040	#ifdef __LITTLE_ENDIAN__
63761	66041	#define vst4q_lane_s64(__p0, __p1, __p2) __extension__ ({ \
63762	66042	int64x2x4_t __s1 = __p1; \
63763		__builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 35); \
	66043	__builtin_neon_vst4q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 35); \
63764	66044	})
63765	66045	#else
63766	66046	#define vst4q_lane_s64(__p0, __p1, __p2) __extension__ ({ \
63767	66047	int64x2x4_t __s1 = __p1; \
63768	66048	int64x2x4_t __rev1; \
63769		__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
63770		__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
63771		__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
63772		__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
63773		__builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 35); \
	66049	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_64); \
	66050	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_64); \
	66051	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_64); \
	66052	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_64); \
	66053	__builtin_neon_vst4q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 35); \
	66054	})
	66055	#endif
	66056
	66057	#ifdef __LITTLE_ENDIAN__
	66058	#define vst4q_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	66059	mfloat8x16x4_t __s1 = __p1; \
	66060	__builtin_neon_vst4q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __s1.val[0]), __builtin_bit_cast(int8x16_t, __s1.val[1]), __builtin_bit_cast(int8x16_t, __s1.val[2]), __builtin_bit_cast(int8x16_t, __s1.val[3]), __p2, 44); \
	66061	})
	66062	#else
	66063	#define vst4q_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	66064	mfloat8x16x4_t __s1 = __p1; \
	66065	mfloat8x16x4_t __rev1; \
	66066	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_128_8); \
	66067	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_128_8); \
	66068	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_128_8); \
	66069	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_128_8); \
	66070	__builtin_neon_vst4q_lane_v(__p0, __builtin_bit_cast(int8x16_t, __rev1.val[0]), __builtin_bit_cast(int8x16_t, __rev1.val[1]), __builtin_bit_cast(int8x16_t, __rev1.val[2]), __builtin_bit_cast(int8x16_t, __rev1.val[3]), __p2, 44); \
63774	66071	})
63775	66072	#endif
63776	66073
63777	66074	#define vst4_lane_u64(__p0, __p1, __p2) __extension__ ({ \
63778	66075	uint64x1x4_t __s1 = __p1; \
63779		__builtin_neon_vst4_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 19); \
	66076	__builtin_neon_vst4_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 19); \
63780	66077	})
63781	66078	#define vst4_lane_f64(__p0, __p1, __p2) __extension__ ({ \
63782	66079	float64x1x4_t __s1 = __p1; \
63783		__builtin_neon_vst4_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 10); \
	66080	__builtin_neon_vst4_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 10); \
63784	66081	})
63785	66082	#define vst4_lane_s64(__p0, __p1, __p2) __extension__ ({ \
63786	66083	int64x1x4_t __s1 = __p1; \
63787		__builtin_neon_vst4_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 3); \
	66084	__builtin_neon_vst4_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 3); \
	66085	})
	66086	#ifdef __LITTLE_ENDIAN__
	66087	#define vst4_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	66088	mfloat8x8x4_t __s1 = __p1; \
	66089	__builtin_neon_vst4_lane_v(__p0, __builtin_bit_cast(int8x8_t, __s1.val[0]), __builtin_bit_cast(int8x8_t, __s1.val[1]), __builtin_bit_cast(int8x8_t, __s1.val[2]), __builtin_bit_cast(int8x8_t, __s1.val[3]), __p2, 12); \
63788	66090	})
	66091	#else
	66092	#define vst4_lane_mf8(__p0, __p1, __p2) __extension__ ({ \
	66093	mfloat8x8x4_t __s1 = __p1; \
	66094	mfloat8x8x4_t __rev1; \
	66095	__rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], __lane_reverse_64_8); \
	66096	__rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], __lane_reverse_64_8); \
	66097	__rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], __lane_reverse_64_8); \
	66098	__rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], __lane_reverse_64_8); \
	66099	__builtin_neon_vst4_lane_v(__p0, __builtin_bit_cast(int8x8_t, __rev1.val[0]), __builtin_bit_cast(int8x8_t, __rev1.val[1]), __builtin_bit_cast(int8x8_t, __rev1.val[2]), __builtin_bit_cast(int8x8_t, __rev1.val[3]), __p2, 12); \
	66100	})
	66101	#endif
	66102
63789	66103	#define vstrq_p128(__p0, __p1) __extension__ ({ \
63790	66104	poly128_t __s1 = __p1; \
63791	66105	__builtin_neon_vstrq_p128(__p0, __s1); \
63792	66106	})
63793	66107	__ai __attribute__((target("neon"))) uint64_t vsubd_u64(uint64_t __p0, uint64_t __p1) {
63794	66108	uint64_t __ret;
63795		__ret = (uint64_t) __builtin_neon_vsubd_u64(__p0, __p1);
	66109	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vsubd_u64(__p0, __p1));
63796	66110	return __ret;
63797	66111	}
63798	66112	__ai __attribute__((target("neon"))) int64_t vsubd_s64(int64_t __p0, int64_t __p1) {
63799	66113	int64_t __ret;
63800		__ret = (int64_t) __builtin_neon_vsubd_s64(__p0, __p1);
	66114	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vsubd_s64(__p0, __p1));
63801	66115	return __ret;
63802	66116	}
63803	66117	#ifdef __LITTLE_ENDIAN__
...	...	@@ -63809,10 +66123,10 @@ __ai __attribute__((target("neon"))) float64x2_t vsubq_f64(float64x2_t __p0, flo
63809	66123	#else
63810	66124	__ai __attribute__((target("neon"))) float64x2_t vsubq_f64(float64x2_t __p0, float64x2_t __p1) {
63811	66125	float64x2_t __ret;
63812		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
63813		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	66126	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	66127	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
63814	66128	__ret = __rev0 - __rev1;
63815		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	66129	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
63816	66130	return __ret;
63817	66131	}
63818	66132	#endif
...	...	@@ -63831,11 +66145,11 @@ __ai __attribute__((target("neon"))) uint16x8_t vsubhn_high_u32(uint16x4_t __p0,
63831	66145	#else
63832	66146	__ai __attribute__((target("neon"))) uint16x8_t vsubhn_high_u32(uint16x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
63833	66147	uint16x8_t __ret;
63834		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
63835		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
63836		uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	66148	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	66149	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	66150	uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
63837	66151	__ret = __noswap_vcombine_u16(__rev0, __noswap_vsubhn_u32(__rev1, __rev2));
63838		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	66152	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
63839	66153	return __ret;
63840	66154	}
63841	66155	#endif
...	...	@@ -63849,11 +66163,11 @@ __ai __attribute__((target("neon"))) uint32x4_t vsubhn_high_u64(uint32x2_t __p0,
63849	66163	#else
63850	66164	__ai __attribute__((target("neon"))) uint32x4_t vsubhn_high_u64(uint32x2_t __p0, uint64x2_t __p1, uint64x2_t __p2) {
63851	66165	uint32x4_t __ret;
63852		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
63853		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
63854		uint64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
	66166	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	66167	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	66168	uint64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_64);
63855	66169	__ret = __noswap_vcombine_u32(__rev0, __noswap_vsubhn_u64(__rev1, __rev2));
63856		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	66170	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
63857	66171	return __ret;
63858	66172	}
63859	66173	#endif
...	...	@@ -63867,11 +66181,11 @@ __ai __attribute__((target("neon"))) uint8x16_t vsubhn_high_u16(uint8x8_t __p0,
63867	66181	#else
63868	66182	__ai __attribute__((target("neon"))) uint8x16_t vsubhn_high_u16(uint8x8_t __p0, uint16x8_t __p1, uint16x8_t __p2) {
63869	66183	uint8x16_t __ret;
63870		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
63871		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
63872		uint16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	66184	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	66185	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	66186	uint16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
63873	66187	__ret = __noswap_vcombine_u8(__rev0, __noswap_vsubhn_u16(__rev1, __rev2));
63874		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	66188	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
63875	66189	return __ret;
63876	66190	}
63877	66191	#endif
...	...	@@ -63885,11 +66199,11 @@ __ai __attribute__((target("neon"))) int16x8_t vsubhn_high_s32(int16x4_t __p0, i
63885	66199	#else
63886	66200	__ai __attribute__((target("neon"))) int16x8_t vsubhn_high_s32(int16x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
63887	66201	int16x8_t __ret;
63888		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
63889		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
63890		int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	66202	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	66203	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	66204	int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
63891	66205	__ret = __noswap_vcombine_s16(__rev0, __noswap_vsubhn_s32(__rev1, __rev2));
63892		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	66206	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
63893	66207	return __ret;
63894	66208	}
63895	66209	#endif
...	...	@@ -63903,11 +66217,11 @@ __ai __attribute__((target("neon"))) int32x4_t vsubhn_high_s64(int32x2_t __p0, i
63903	66217	#else
63904	66218	__ai __attribute__((target("neon"))) int32x4_t vsubhn_high_s64(int32x2_t __p0, int64x2_t __p1, int64x2_t __p2) {
63905	66219	int32x4_t __ret;
63906		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
63907		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
63908		int64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
	66220	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	66221	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	66222	int64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_64);
63909	66223	__ret = __noswap_vcombine_s32(__rev0, __noswap_vsubhn_s64(__rev1, __rev2));
63910		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	66224	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
63911	66225	return __ret;
63912	66226	}
63913	66227	#endif
...	...	@@ -63921,11 +66235,11 @@ __ai __attribute__((target("neon"))) int8x16_t vsubhn_high_s16(int8x8_t __p0, in
63921	66235	#else
63922	66236	__ai __attribute__((target("neon"))) int8x16_t vsubhn_high_s16(int8x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
63923	66237	int8x16_t __ret;
63924		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
63925		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
63926		int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	66238	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	66239	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	66240	int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
63927	66241	__ret = __noswap_vcombine_s8(__rev0, __noswap_vsubhn_s16(__rev1, __rev2));
63928		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	66242	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
63929	66243	return __ret;
63930	66244	}
63931	66245	#endif
...	...	@@ -63939,10 +66253,10 @@ __ai __attribute__((target("neon"))) uint16x8_t vsubl_high_u8(uint8x16_t __p0, u
63939	66253	#else
63940	66254	__ai __attribute__((target("neon"))) uint16x8_t vsubl_high_u8(uint8x16_t __p0, uint8x16_t __p1) {
63941	66255	uint16x8_t __ret;
63942		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
63943		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	66256	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	66257	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
63944	66258	__ret = __noswap_vmovl_high_u8(__rev0) - __noswap_vmovl_high_u8(__rev1);
63945		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	66259	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
63946	66260	return __ret;
63947	66261	}
63948	66262	#endif
...	...	@@ -63956,10 +66270,10 @@ __ai __attribute__((target("neon"))) uint64x2_t vsubl_high_u32(uint32x4_t __p0,
63956	66270	#else
63957	66271	__ai __attribute__((target("neon"))) uint64x2_t vsubl_high_u32(uint32x4_t __p0, uint32x4_t __p1) {
63958	66272	uint64x2_t __ret;
63959		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
63960		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	66273	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	66274	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
63961	66275	__ret = __noswap_vmovl_high_u32(__rev0) - __noswap_vmovl_high_u32(__rev1);
63962		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	66276	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
63963	66277	return __ret;
63964	66278	}
63965	66279	#endif
...	...	@@ -63973,10 +66287,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vsubl_high_u16(uint16x8_t __p0,
63973	66287	#else
63974	66288	__ai __attribute__((target("neon"))) uint32x4_t vsubl_high_u16(uint16x8_t __p0, uint16x8_t __p1) {
63975	66289	uint32x4_t __ret;
63976		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
63977		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	66290	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	66291	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
63978	66292	__ret = __noswap_vmovl_high_u16(__rev0) - __noswap_vmovl_high_u16(__rev1);
63979		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	66293	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
63980	66294	return __ret;
63981	66295	}
63982	66296	#endif
...	...	@@ -63990,10 +66304,10 @@ __ai __attribute__((target("neon"))) int16x8_t vsubl_high_s8(int8x16_t __p0, int
63990	66304	#else
63991	66305	__ai __attribute__((target("neon"))) int16x8_t vsubl_high_s8(int8x16_t __p0, int8x16_t __p1) {
63992	66306	int16x8_t __ret;
63993		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
63994		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	66307	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	66308	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
63995	66309	__ret = __noswap_vmovl_high_s8(__rev0) - __noswap_vmovl_high_s8(__rev1);
63996		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	66310	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
63997	66311	return __ret;
63998	66312	}
63999	66313	#endif
...	...	@@ -64007,10 +66321,10 @@ __ai __attribute__((target("neon"))) int64x2_t vsubl_high_s32(int32x4_t __p0, in
64007	66321	#else
64008	66322	__ai __attribute__((target("neon"))) int64x2_t vsubl_high_s32(int32x4_t __p0, int32x4_t __p1) {
64009	66323	int64x2_t __ret;
64010		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
64011		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	66324	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	66325	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
64012	66326	__ret = __noswap_vmovl_high_s32(__rev0) - __noswap_vmovl_high_s32(__rev1);
64013		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	66327	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
64014	66328	return __ret;
64015	66329	}
64016	66330	#endif
...	...	@@ -64024,10 +66338,10 @@ __ai __attribute__((target("neon"))) int32x4_t vsubl_high_s16(int16x8_t __p0, in
64024	66338	#else
64025	66339	__ai __attribute__((target("neon"))) int32x4_t vsubl_high_s16(int16x8_t __p0, int16x8_t __p1) {
64026	66340	int32x4_t __ret;
64027		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
64028		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	66341	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	66342	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
64029	66343	__ret = __noswap_vmovl_high_s16(__rev0) - __noswap_vmovl_high_s16(__rev1);
64030		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	66344	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
64031	66345	return __ret;
64032	66346	}
64033	66347	#endif
...	...	@@ -64041,10 +66355,10 @@ __ai __attribute__((target("neon"))) uint16x8_t vsubw_high_u8(uint16x8_t __p0, u
64041	66355	#else
64042	66356	__ai __attribute__((target("neon"))) uint16x8_t vsubw_high_u8(uint16x8_t __p0, uint8x16_t __p1) {
64043	66357	uint16x8_t __ret;
64044		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
64045		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	66358	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	66359	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
64046	66360	__ret = __rev0 - __noswap_vmovl_high_u8(__rev1);
64047		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	66361	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
64048	66362	return __ret;
64049	66363	}
64050	66364	#endif
...	...	@@ -64058,10 +66372,10 @@ __ai __attribute__((target("neon"))) uint64x2_t vsubw_high_u32(uint64x2_t __p0,
64058	66372	#else
64059	66373	__ai __attribute__((target("neon"))) uint64x2_t vsubw_high_u32(uint64x2_t __p0, uint32x4_t __p1) {
64060	66374	uint64x2_t __ret;
64061		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
64062		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	66375	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	66376	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
64063	66377	__ret = __rev0 - __noswap_vmovl_high_u32(__rev1);
64064		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	66378	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
64065	66379	return __ret;
64066	66380	}
64067	66381	#endif
...	...	@@ -64075,10 +66389,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vsubw_high_u16(uint32x4_t __p0,
64075	66389	#else
64076	66390	__ai __attribute__((target("neon"))) uint32x4_t vsubw_high_u16(uint32x4_t __p0, uint16x8_t __p1) {
64077	66391	uint32x4_t __ret;
64078		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
64079		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	66392	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	66393	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
64080	66394	__ret = __rev0 - __noswap_vmovl_high_u16(__rev1);
64081		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	66395	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
64082	66396	return __ret;
64083	66397	}
64084	66398	#endif
...	...	@@ -64092,10 +66406,10 @@ __ai __attribute__((target("neon"))) int16x8_t vsubw_high_s8(int16x8_t __p0, int
64092	66406	#else
64093	66407	__ai __attribute__((target("neon"))) int16x8_t vsubw_high_s8(int16x8_t __p0, int8x16_t __p1) {
64094	66408	int16x8_t __ret;
64095		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
64096		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	66409	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	66410	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
64097	66411	__ret = __rev0 - __noswap_vmovl_high_s8(__rev1);
64098		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	66412	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
64099	66413	return __ret;
64100	66414	}
64101	66415	#endif
...	...	@@ -64109,10 +66423,10 @@ __ai __attribute__((target("neon"))) int64x2_t vsubw_high_s32(int64x2_t __p0, in
64109	66423	#else
64110	66424	__ai __attribute__((target("neon"))) int64x2_t vsubw_high_s32(int64x2_t __p0, int32x4_t __p1) {
64111	66425	int64x2_t __ret;
64112		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
64113		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	66426	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	66427	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
64114	66428	__ret = __rev0 - __noswap_vmovl_high_s32(__rev1);
64115		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	66429	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
64116	66430	return __ret;
64117	66431	}
64118	66432	#endif
...	...	@@ -64126,10 +66440,10 @@ __ai __attribute__((target("neon"))) int32x4_t vsubw_high_s16(int32x4_t __p0, in
64126	66440	#else
64127	66441	__ai __attribute__((target("neon"))) int32x4_t vsubw_high_s16(int32x4_t __p0, int16x8_t __p1) {
64128	66442	int32x4_t __ret;
64129		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
64130		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	66443	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	66444	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
64131	66445	__ret = __rev0 - __noswap_vmovl_high_s16(__rev1);
64132		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	66446	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
64133	66447	return __ret;
64134	66448	}
64135	66449	#endif
...	...	@@ -64143,10 +66457,10 @@ __ai __attribute__((target("neon"))) poly8x8_t vtrn1_p8(poly8x8_t __p0, poly8x8_
64143	66457	#else
64144	66458	__ai __attribute__((target("neon"))) poly8x8_t vtrn1_p8(poly8x8_t __p0, poly8x8_t __p1) {
64145	66459	poly8x8_t __ret;
64146		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
64147		poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	66460	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	66461	poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
64148	66462	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 8, 2, 10, 4, 12, 6, 14);
64149		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	66463	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
64150	66464	return __ret;
64151	66465	}
64152	66466	#endif
...	...	@@ -64160,10 +66474,10 @@ __ai __attribute__((target("neon"))) poly16x4_t vtrn1_p16(poly16x4_t __p0, poly1
64160	66474	#else
64161	66475	__ai __attribute__((target("neon"))) poly16x4_t vtrn1_p16(poly16x4_t __p0, poly16x4_t __p1) {
64162	66476	poly16x4_t __ret;
64163		poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
64164		poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	66477	poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	66478	poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
64165	66479	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 4, 2, 6);
64166		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	66480	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
64167	66481	return __ret;
64168	66482	}
64169	66483	#endif
...	...	@@ -64177,10 +66491,10 @@ __ai __attribute__((target("neon"))) poly8x16_t vtrn1q_p8(poly8x16_t __p0, poly8
64177	66491	#else
64178	66492	__ai __attribute__((target("neon"))) poly8x16_t vtrn1q_p8(poly8x16_t __p0, poly8x16_t __p1) {
64179	66493	poly8x16_t __ret;
64180		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
64181		poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	66494	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	66495	poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
64182	66496	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 16, 2, 18, 4, 20, 6, 22, 8, 24, 10, 26, 12, 28, 14, 30);
64183		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	66497	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
64184	66498	return __ret;
64185	66499	}
64186	66500	#endif
...	...	@@ -64194,10 +66508,10 @@ __ai __attribute__((target("neon"))) poly64x2_t vtrn1q_p64(poly64x2_t __p0, poly
64194	66508	#else
64195	66509	__ai __attribute__((target("neon"))) poly64x2_t vtrn1q_p64(poly64x2_t __p0, poly64x2_t __p1) {
64196	66510	poly64x2_t __ret;
64197		poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
64198		poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	66511	poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	66512	poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
64199	66513	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2);
64200		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	66514	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
64201	66515	return __ret;
64202	66516	}
64203	66517	#endif
...	...	@@ -64211,10 +66525,10 @@ __ai __attribute__((target("neon"))) poly16x8_t vtrn1q_p16(poly16x8_t __p0, poly
64211	66525	#else
64212	66526	__ai __attribute__((target("neon"))) poly16x8_t vtrn1q_p16(poly16x8_t __p0, poly16x8_t __p1) {
64213	66527	poly16x8_t __ret;
64214		poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
64215		poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	66528	poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	66529	poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
64216	66530	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 8, 2, 10, 4, 12, 6, 14);
64217		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	66531	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
64218	66532	return __ret;
64219	66533	}
64220	66534	#endif
...	...	@@ -64228,10 +66542,10 @@ __ai __attribute__((target("neon"))) uint8x16_t vtrn1q_u8(uint8x16_t __p0, uint8
64228	66542	#else
64229	66543	__ai __attribute__((target("neon"))) uint8x16_t vtrn1q_u8(uint8x16_t __p0, uint8x16_t __p1) {
64230	66544	uint8x16_t __ret;
64231		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
64232		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	66545	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	66546	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
64233	66547	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 16, 2, 18, 4, 20, 6, 22, 8, 24, 10, 26, 12, 28, 14, 30);
64234		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	66548	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
64235	66549	return __ret;
64236	66550	}
64237	66551	#endif
...	...	@@ -64245,10 +66559,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vtrn1q_u32(uint32x4_t __p0, uint
64245	66559	#else
64246	66560	__ai __attribute__((target("neon"))) uint32x4_t vtrn1q_u32(uint32x4_t __p0, uint32x4_t __p1) {
64247	66561	uint32x4_t __ret;
64248		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
64249		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	66562	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	66563	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
64250	66564	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 4, 2, 6);
64251		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	66565	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
64252	66566	return __ret;
64253	66567	}
64254	66568	#endif
...	...	@@ -64262,10 +66576,10 @@ __ai __attribute__((target("neon"))) uint64x2_t vtrn1q_u64(uint64x2_t __p0, uint
64262	66576	#else
64263	66577	__ai __attribute__((target("neon"))) uint64x2_t vtrn1q_u64(uint64x2_t __p0, uint64x2_t __p1) {
64264	66578	uint64x2_t __ret;
64265		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
64266		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	66579	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	66580	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
64267	66581	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2);
64268		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	66582	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
64269	66583	return __ret;
64270	66584	}
64271	66585	#endif
...	...	@@ -64279,10 +66593,10 @@ __ai __attribute__((target("neon"))) uint16x8_t vtrn1q_u16(uint16x8_t __p0, uint
64279	66593	#else
64280	66594	__ai __attribute__((target("neon"))) uint16x8_t vtrn1q_u16(uint16x8_t __p0, uint16x8_t __p1) {
64281	66595	uint16x8_t __ret;
64282		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
64283		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	66596	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	66597	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
64284	66598	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 8, 2, 10, 4, 12, 6, 14);
64285		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	66599	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
64286	66600	return __ret;
64287	66601	}
64288	66602	#endif
...	...	@@ -64296,10 +66610,10 @@ __ai __attribute__((target("neon"))) int8x16_t vtrn1q_s8(int8x16_t __p0, int8x16
64296	66610	#else
64297	66611	__ai __attribute__((target("neon"))) int8x16_t vtrn1q_s8(int8x16_t __p0, int8x16_t __p1) {
64298	66612	int8x16_t __ret;
64299		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
64300		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	66613	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	66614	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
64301	66615	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 16, 2, 18, 4, 20, 6, 22, 8, 24, 10, 26, 12, 28, 14, 30);
64302		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	66616	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
64303	66617	return __ret;
64304	66618	}
64305	66619	#endif
...	...	@@ -64313,10 +66627,10 @@ __ai __attribute__((target("neon"))) float64x2_t vtrn1q_f64(float64x2_t __p0, fl
64313	66627	#else
64314	66628	__ai __attribute__((target("neon"))) float64x2_t vtrn1q_f64(float64x2_t __p0, float64x2_t __p1) {
64315	66629	float64x2_t __ret;
64316		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
64317		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	66630	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	66631	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
64318	66632	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2);
64319		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	66633	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
64320	66634	return __ret;
64321	66635	}
64322	66636	#endif
...	...	@@ -64330,10 +66644,10 @@ __ai __attribute__((target("neon"))) float32x4_t vtrn1q_f32(float32x4_t __p0, fl
64330	66644	#else
64331	66645	__ai __attribute__((target("neon"))) float32x4_t vtrn1q_f32(float32x4_t __p0, float32x4_t __p1) {
64332	66646	float32x4_t __ret;
64333		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
64334		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	66647	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	66648	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
64335	66649	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 4, 2, 6);
64336		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	66650	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
64337	66651	return __ret;
64338	66652	}
64339	66653	#endif
...	...	@@ -64347,10 +66661,10 @@ __ai __attribute__((target("neon"))) int32x4_t vtrn1q_s32(int32x4_t __p0, int32x
64347	66661	#else
64348	66662	__ai __attribute__((target("neon"))) int32x4_t vtrn1q_s32(int32x4_t __p0, int32x4_t __p1) {
64349	66663	int32x4_t __ret;
64350		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
64351		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	66664	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	66665	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
64352	66666	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 4, 2, 6);
64353		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	66667	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
64354	66668	return __ret;
64355	66669	}
64356	66670	#endif
...	...	@@ -64364,10 +66678,27 @@ __ai __attribute__((target("neon"))) int64x2_t vtrn1q_s64(int64x2_t __p0, int64x
64364	66678	#else
64365	66679	__ai __attribute__((target("neon"))) int64x2_t vtrn1q_s64(int64x2_t __p0, int64x2_t __p1) {
64366	66680	int64x2_t __ret;
64367		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
64368		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	66681	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	66682	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
64369	66683	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2);
64370		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	66684	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
	66685	return __ret;
	66686	}
	66687	#endif
	66688
	66689	#ifdef __LITTLE_ENDIAN__
	66690	__ai __attribute__((target("neon"))) mfloat8x16_t vtrn1q_mf8(mfloat8x16_t __p0, mfloat8x16_t __p1) {
	66691	mfloat8x16_t __ret;
	66692	__ret = __builtin_shufflevector(__p0, __p1, 0, 16, 2, 18, 4, 20, 6, 22, 8, 24, 10, 26, 12, 28, 14, 30);
	66693	return __ret;
	66694	}
	66695	#else
	66696	__ai __attribute__((target("neon"))) mfloat8x16_t vtrn1q_mf8(mfloat8x16_t __p0, mfloat8x16_t __p1) {
	66697	mfloat8x16_t __ret;
	66698	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	66699	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	66700	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 16, 2, 18, 4, 20, 6, 22, 8, 24, 10, 26, 12, 28, 14, 30);
	66701	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
64371	66702	return __ret;
64372	66703	}
64373	66704	#endif
...	...	@@ -64381,10 +66712,10 @@ __ai __attribute__((target("neon"))) int16x8_t vtrn1q_s16(int16x8_t __p0, int16x
64381	66712	#else
64382	66713	__ai __attribute__((target("neon"))) int16x8_t vtrn1q_s16(int16x8_t __p0, int16x8_t __p1) {
64383	66714	int16x8_t __ret;
64384		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
64385		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	66715	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	66716	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
64386	66717	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 8, 2, 10, 4, 12, 6, 14);
64387		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	66718	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
64388	66719	return __ret;
64389	66720	}
64390	66721	#endif
...	...	@@ -64398,10 +66729,10 @@ __ai __attribute__((target("neon"))) uint8x8_t vtrn1_u8(uint8x8_t __p0, uint8x8_
64398	66729	#else
64399	66730	__ai __attribute__((target("neon"))) uint8x8_t vtrn1_u8(uint8x8_t __p0, uint8x8_t __p1) {
64400	66731	uint8x8_t __ret;
64401		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
64402		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	66732	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	66733	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
64403	66734	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 8, 2, 10, 4, 12, 6, 14);
64404		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	66735	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
64405	66736	return __ret;
64406	66737	}
64407	66738	#endif
...	...	@@ -64415,10 +66746,10 @@ __ai __attribute__((target("neon"))) uint32x2_t vtrn1_u32(uint32x2_t __p0, uint3
64415	66746	#else
64416	66747	__ai __attribute__((target("neon"))) uint32x2_t vtrn1_u32(uint32x2_t __p0, uint32x2_t __p1) {
64417	66748	uint32x2_t __ret;
64418		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
64419		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	66749	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	66750	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
64420	66751	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2);
64421		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	66752	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
64422	66753	return __ret;
64423	66754	}
64424	66755	#endif
...	...	@@ -64432,10 +66763,10 @@ __ai __attribute__((target("neon"))) uint16x4_t vtrn1_u16(uint16x4_t __p0, uint1
64432	66763	#else
64433	66764	__ai __attribute__((target("neon"))) uint16x4_t vtrn1_u16(uint16x4_t __p0, uint16x4_t __p1) {
64434	66765	uint16x4_t __ret;
64435		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
64436		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	66766	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	66767	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
64437	66768	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 4, 2, 6);
64438		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	66769	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
64439	66770	return __ret;
64440	66771	}
64441	66772	#endif
...	...	@@ -64449,10 +66780,10 @@ __ai __attribute__((target("neon"))) int8x8_t vtrn1_s8(int8x8_t __p0, int8x8_t _
64449	66780	#else
64450	66781	__ai __attribute__((target("neon"))) int8x8_t vtrn1_s8(int8x8_t __p0, int8x8_t __p1) {
64451	66782	int8x8_t __ret;
64452		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
64453		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	66783	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	66784	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
64454	66785	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 8, 2, 10, 4, 12, 6, 14);
64455		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	66786	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
64456	66787	return __ret;
64457	66788	}
64458	66789	#endif
...	...	@@ -64466,10 +66797,10 @@ __ai __attribute__((target("neon"))) float32x2_t vtrn1_f32(float32x2_t __p0, flo
64466	66797	#else
64467	66798	__ai __attribute__((target("neon"))) float32x2_t vtrn1_f32(float32x2_t __p0, float32x2_t __p1) {
64468	66799	float32x2_t __ret;
64469		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
64470		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	66800	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	66801	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
64471	66802	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2);
64472		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	66803	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
64473	66804	return __ret;
64474	66805	}
64475	66806	#endif
...	...	@@ -64483,10 +66814,27 @@ __ai __attribute__((target("neon"))) int32x2_t vtrn1_s32(int32x2_t __p0, int32x2
64483	66814	#else
64484	66815	__ai __attribute__((target("neon"))) int32x2_t vtrn1_s32(int32x2_t __p0, int32x2_t __p1) {
64485	66816	int32x2_t __ret;
64486		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
64487		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	66817	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	66818	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
64488	66819	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2);
64489		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	66820	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
	66821	return __ret;
	66822	}
	66823	#endif
	66824
	66825	#ifdef __LITTLE_ENDIAN__
	66826	__ai __attribute__((target("neon"))) mfloat8x8_t vtrn1_mf8(mfloat8x8_t __p0, mfloat8x8_t __p1) {
	66827	mfloat8x8_t __ret;
	66828	__ret = __builtin_shufflevector(__p0, __p1, 0, 8, 2, 10, 4, 12, 6, 14);
	66829	return __ret;
	66830	}
	66831	#else
	66832	__ai __attribute__((target("neon"))) mfloat8x8_t vtrn1_mf8(mfloat8x8_t __p0, mfloat8x8_t __p1) {
	66833	mfloat8x8_t __ret;
	66834	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	66835	mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	66836	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 8, 2, 10, 4, 12, 6, 14);
	66837	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
64490	66838	return __ret;
64491	66839	}
64492	66840	#endif
...	...	@@ -64500,10 +66848,10 @@ __ai __attribute__((target("neon"))) int16x4_t vtrn1_s16(int16x4_t __p0, int16x4
64500	66848	#else
64501	66849	__ai __attribute__((target("neon"))) int16x4_t vtrn1_s16(int16x4_t __p0, int16x4_t __p1) {
64502	66850	int16x4_t __ret;
64503		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
64504		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	66851	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	66852	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
64505	66853	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 4, 2, 6);
64506		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	66854	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
64507	66855	return __ret;
64508	66856	}
64509	66857	#endif
...	...	@@ -64517,10 +66865,10 @@ __ai __attribute__((target("neon"))) float16x8_t vtrn1q_f16(float16x8_t __p0, fl
64517	66865	#else
64518	66866	__ai __attribute__((target("neon"))) float16x8_t vtrn1q_f16(float16x8_t __p0, float16x8_t __p1) {
64519	66867	float16x8_t __ret;
64520		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
64521		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	66868	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	66869	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
64522	66870	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 8, 2, 10, 4, 12, 6, 14);
64523		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	66871	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
64524	66872	return __ret;
64525	66873	}
64526	66874	#endif
...	...	@@ -64534,10 +66882,10 @@ __ai __attribute__((target("neon"))) float16x4_t vtrn1_f16(float16x4_t __p0, flo
64534	66882	#else
64535	66883	__ai __attribute__((target("neon"))) float16x4_t vtrn1_f16(float16x4_t __p0, float16x4_t __p1) {
64536	66884	float16x4_t __ret;
64537		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
64538		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	66885	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	66886	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
64539	66887	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 4, 2, 6);
64540		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	66888	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
64541	66889	return __ret;
64542	66890	}
64543	66891	#endif
...	...	@@ -64551,10 +66899,10 @@ __ai __attribute__((target("neon"))) poly8x8_t vtrn2_p8(poly8x8_t __p0, poly8x8_
64551	66899	#else
64552	66900	__ai __attribute__((target("neon"))) poly8x8_t vtrn2_p8(poly8x8_t __p0, poly8x8_t __p1) {
64553	66901	poly8x8_t __ret;
64554		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
64555		poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	66902	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	66903	poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
64556	66904	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 9, 3, 11, 5, 13, 7, 15);
64557		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	66905	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
64558	66906	return __ret;
64559	66907	}
64560	66908	#endif
...	...	@@ -64568,10 +66916,10 @@ __ai __attribute__((target("neon"))) poly16x4_t vtrn2_p16(poly16x4_t __p0, poly1
64568	66916	#else
64569	66917	__ai __attribute__((target("neon"))) poly16x4_t vtrn2_p16(poly16x4_t __p0, poly16x4_t __p1) {
64570	66918	poly16x4_t __ret;
64571		poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
64572		poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	66919	poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	66920	poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
64573	66921	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 5, 3, 7);
64574		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	66922	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
64575	66923	return __ret;
64576	66924	}
64577	66925	#endif
...	...	@@ -64585,10 +66933,10 @@ __ai __attribute__((target("neon"))) poly8x16_t vtrn2q_p8(poly8x16_t __p0, poly8
64585	66933	#else
64586	66934	__ai __attribute__((target("neon"))) poly8x16_t vtrn2q_p8(poly8x16_t __p0, poly8x16_t __p1) {
64587	66935	poly8x16_t __ret;
64588		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
64589		poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	66936	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	66937	poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
64590	66938	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 17, 3, 19, 5, 21, 7, 23, 9, 25, 11, 27, 13, 29, 15, 31);
64591		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	66939	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
64592	66940	return __ret;
64593	66941	}
64594	66942	#endif
...	...	@@ -64602,10 +66950,10 @@ __ai __attribute__((target("neon"))) poly64x2_t vtrn2q_p64(poly64x2_t __p0, poly
64602	66950	#else
64603	66951	__ai __attribute__((target("neon"))) poly64x2_t vtrn2q_p64(poly64x2_t __p0, poly64x2_t __p1) {
64604	66952	poly64x2_t __ret;
64605		poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
64606		poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	66953	poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	66954	poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
64607	66955	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3);
64608		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	66956	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
64609	66957	return __ret;
64610	66958	}
64611	66959	#endif
...	...	@@ -64619,10 +66967,10 @@ __ai __attribute__((target("neon"))) poly16x8_t vtrn2q_p16(poly16x8_t __p0, poly
64619	66967	#else
64620	66968	__ai __attribute__((target("neon"))) poly16x8_t vtrn2q_p16(poly16x8_t __p0, poly16x8_t __p1) {
64621	66969	poly16x8_t __ret;
64622		poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
64623		poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	66970	poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	66971	poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
64624	66972	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 9, 3, 11, 5, 13, 7, 15);
64625		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	66973	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
64626	66974	return __ret;
64627	66975	}
64628	66976	#endif
...	...	@@ -64636,10 +66984,10 @@ __ai __attribute__((target("neon"))) uint8x16_t vtrn2q_u8(uint8x16_t __p0, uint8
64636	66984	#else
64637	66985	__ai __attribute__((target("neon"))) uint8x16_t vtrn2q_u8(uint8x16_t __p0, uint8x16_t __p1) {
64638	66986	uint8x16_t __ret;
64639		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
64640		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	66987	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	66988	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
64641	66989	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 17, 3, 19, 5, 21, 7, 23, 9, 25, 11, 27, 13, 29, 15, 31);
64642		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	66990	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
64643	66991	return __ret;
64644	66992	}
64645	66993	#endif
...	...	@@ -64653,10 +67001,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vtrn2q_u32(uint32x4_t __p0, uint
64653	67001	#else
64654	67002	__ai __attribute__((target("neon"))) uint32x4_t vtrn2q_u32(uint32x4_t __p0, uint32x4_t __p1) {
64655	67003	uint32x4_t __ret;
64656		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
64657		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	67004	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	67005	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
64658	67006	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 5, 3, 7);
64659		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	67007	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
64660	67008	return __ret;
64661	67009	}
64662	67010	#endif
...	...	@@ -64670,10 +67018,10 @@ __ai __attribute__((target("neon"))) uint64x2_t vtrn2q_u64(uint64x2_t __p0, uint
64670	67018	#else
64671	67019	__ai __attribute__((target("neon"))) uint64x2_t vtrn2q_u64(uint64x2_t __p0, uint64x2_t __p1) {
64672	67020	uint64x2_t __ret;
64673		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
64674		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	67021	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	67022	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
64675	67023	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3);
64676		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	67024	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
64677	67025	return __ret;
64678	67026	}
64679	67027	#endif
...	...	@@ -64687,10 +67035,10 @@ __ai __attribute__((target("neon"))) uint16x8_t vtrn2q_u16(uint16x8_t __p0, uint
64687	67035	#else
64688	67036	__ai __attribute__((target("neon"))) uint16x8_t vtrn2q_u16(uint16x8_t __p0, uint16x8_t __p1) {
64689	67037	uint16x8_t __ret;
64690		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
64691		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	67038	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	67039	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
64692	67040	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 9, 3, 11, 5, 13, 7, 15);
64693		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	67041	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
64694	67042	return __ret;
64695	67043	}
64696	67044	#endif
...	...	@@ -64704,10 +67052,10 @@ __ai __attribute__((target("neon"))) int8x16_t vtrn2q_s8(int8x16_t __p0, int8x16
64704	67052	#else
64705	67053	__ai __attribute__((target("neon"))) int8x16_t vtrn2q_s8(int8x16_t __p0, int8x16_t __p1) {
64706	67054	int8x16_t __ret;
64707		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
64708		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	67055	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	67056	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
64709	67057	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 17, 3, 19, 5, 21, 7, 23, 9, 25, 11, 27, 13, 29, 15, 31);
64710		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	67058	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
64711	67059	return __ret;
64712	67060	}
64713	67061	#endif
...	...	@@ -64721,10 +67069,10 @@ __ai __attribute__((target("neon"))) float64x2_t vtrn2q_f64(float64x2_t __p0, fl
64721	67069	#else
64722	67070	__ai __attribute__((target("neon"))) float64x2_t vtrn2q_f64(float64x2_t __p0, float64x2_t __p1) {
64723	67071	float64x2_t __ret;
64724		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
64725		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	67072	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	67073	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
64726	67074	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3);
64727		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	67075	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
64728	67076	return __ret;
64729	67077	}
64730	67078	#endif
...	...	@@ -64738,10 +67086,10 @@ __ai __attribute__((target("neon"))) float32x4_t vtrn2q_f32(float32x4_t __p0, fl
64738	67086	#else
64739	67087	__ai __attribute__((target("neon"))) float32x4_t vtrn2q_f32(float32x4_t __p0, float32x4_t __p1) {
64740	67088	float32x4_t __ret;
64741		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
64742		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	67089	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	67090	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
64743	67091	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 5, 3, 7);
64744		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	67092	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
64745	67093	return __ret;
64746	67094	}
64747	67095	#endif
...	...	@@ -64755,10 +67103,10 @@ __ai __attribute__((target("neon"))) int32x4_t vtrn2q_s32(int32x4_t __p0, int32x
64755	67103	#else
64756	67104	__ai __attribute__((target("neon"))) int32x4_t vtrn2q_s32(int32x4_t __p0, int32x4_t __p1) {
64757	67105	int32x4_t __ret;
64758		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
64759		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	67106	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	67107	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
64760	67108	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 5, 3, 7);
64761		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	67109	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
64762	67110	return __ret;
64763	67111	}
64764	67112	#endif
...	...	@@ -64772,10 +67120,27 @@ __ai __attribute__((target("neon"))) int64x2_t vtrn2q_s64(int64x2_t __p0, int64x
64772	67120	#else
64773	67121	__ai __attribute__((target("neon"))) int64x2_t vtrn2q_s64(int64x2_t __p0, int64x2_t __p1) {
64774	67122	int64x2_t __ret;
64775		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
64776		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	67123	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	67124	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
64777	67125	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3);
64778		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	67126	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
	67127	return __ret;
	67128	}
	67129	#endif
	67130
	67131	#ifdef __LITTLE_ENDIAN__
	67132	__ai __attribute__((target("neon"))) mfloat8x16_t vtrn2q_mf8(mfloat8x16_t __p0, mfloat8x16_t __p1) {
	67133	mfloat8x16_t __ret;
	67134	__ret = __builtin_shufflevector(__p0, __p1, 1, 17, 3, 19, 5, 21, 7, 23, 9, 25, 11, 27, 13, 29, 15, 31);
	67135	return __ret;
	67136	}
	67137	#else
	67138	__ai __attribute__((target("neon"))) mfloat8x16_t vtrn2q_mf8(mfloat8x16_t __p0, mfloat8x16_t __p1) {
	67139	mfloat8x16_t __ret;
	67140	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	67141	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	67142	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 17, 3, 19, 5, 21, 7, 23, 9, 25, 11, 27, 13, 29, 15, 31);
	67143	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
64779	67144	return __ret;
64780	67145	}
64781	67146	#endif
...	...	@@ -64789,10 +67154,10 @@ __ai __attribute__((target("neon"))) int16x8_t vtrn2q_s16(int16x8_t __p0, int16x
64789	67154	#else
64790	67155	__ai __attribute__((target("neon"))) int16x8_t vtrn2q_s16(int16x8_t __p0, int16x8_t __p1) {
64791	67156	int16x8_t __ret;
64792		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
64793		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	67157	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	67158	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
64794	67159	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 9, 3, 11, 5, 13, 7, 15);
64795		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	67160	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
64796	67161	return __ret;
64797	67162	}
64798	67163	#endif
...	...	@@ -64806,10 +67171,10 @@ __ai __attribute__((target("neon"))) uint8x8_t vtrn2_u8(uint8x8_t __p0, uint8x8_
64806	67171	#else
64807	67172	__ai __attribute__((target("neon"))) uint8x8_t vtrn2_u8(uint8x8_t __p0, uint8x8_t __p1) {
64808	67173	uint8x8_t __ret;
64809		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
64810		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	67174	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	67175	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
64811	67176	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 9, 3, 11, 5, 13, 7, 15);
64812		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	67177	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
64813	67178	return __ret;
64814	67179	}
64815	67180	#endif
...	...	@@ -64823,10 +67188,10 @@ __ai __attribute__((target("neon"))) uint32x2_t vtrn2_u32(uint32x2_t __p0, uint3
64823	67188	#else
64824	67189	__ai __attribute__((target("neon"))) uint32x2_t vtrn2_u32(uint32x2_t __p0, uint32x2_t __p1) {
64825	67190	uint32x2_t __ret;
64826		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
64827		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	67191	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	67192	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
64828	67193	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3);
64829		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	67194	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
64830	67195	return __ret;
64831	67196	}
64832	67197	#endif
...	...	@@ -64840,10 +67205,10 @@ __ai __attribute__((target("neon"))) uint16x4_t vtrn2_u16(uint16x4_t __p0, uint1
64840	67205	#else
64841	67206	__ai __attribute__((target("neon"))) uint16x4_t vtrn2_u16(uint16x4_t __p0, uint16x4_t __p1) {
64842	67207	uint16x4_t __ret;
64843		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
64844		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	67208	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	67209	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
64845	67210	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 5, 3, 7);
64846		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	67211	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
64847	67212	return __ret;
64848	67213	}
64849	67214	#endif
...	...	@@ -64857,10 +67222,10 @@ __ai __attribute__((target("neon"))) int8x8_t vtrn2_s8(int8x8_t __p0, int8x8_t _
64857	67222	#else
64858	67223	__ai __attribute__((target("neon"))) int8x8_t vtrn2_s8(int8x8_t __p0, int8x8_t __p1) {
64859	67224	int8x8_t __ret;
64860		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
64861		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	67225	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	67226	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
64862	67227	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 9, 3, 11, 5, 13, 7, 15);
64863		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	67228	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
64864	67229	return __ret;
64865	67230	}
64866	67231	#endif
...	...	@@ -64874,10 +67239,10 @@ __ai __attribute__((target("neon"))) float32x2_t vtrn2_f32(float32x2_t __p0, flo
64874	67239	#else
64875	67240	__ai __attribute__((target("neon"))) float32x2_t vtrn2_f32(float32x2_t __p0, float32x2_t __p1) {
64876	67241	float32x2_t __ret;
64877		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
64878		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	67242	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	67243	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
64879	67244	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3);
64880		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	67245	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
64881	67246	return __ret;
64882	67247	}
64883	67248	#endif
...	...	@@ -64891,10 +67256,27 @@ __ai __attribute__((target("neon"))) int32x2_t vtrn2_s32(int32x2_t __p0, int32x2
64891	67256	#else
64892	67257	__ai __attribute__((target("neon"))) int32x2_t vtrn2_s32(int32x2_t __p0, int32x2_t __p1) {
64893	67258	int32x2_t __ret;
64894		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
64895		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	67259	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	67260	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
64896	67261	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3);
64897		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	67262	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
	67263	return __ret;
	67264	}
	67265	#endif
	67266
	67267	#ifdef __LITTLE_ENDIAN__
	67268	__ai __attribute__((target("neon"))) mfloat8x8_t vtrn2_mf8(mfloat8x8_t __p0, mfloat8x8_t __p1) {
	67269	mfloat8x8_t __ret;
	67270	__ret = __builtin_shufflevector(__p0, __p1, 1, 9, 3, 11, 5, 13, 7, 15);
	67271	return __ret;
	67272	}
	67273	#else
	67274	__ai __attribute__((target("neon"))) mfloat8x8_t vtrn2_mf8(mfloat8x8_t __p0, mfloat8x8_t __p1) {
	67275	mfloat8x8_t __ret;
	67276	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	67277	mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	67278	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 9, 3, 11, 5, 13, 7, 15);
	67279	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
64898	67280	return __ret;
64899	67281	}
64900	67282	#endif
...	...	@@ -64908,10 +67290,10 @@ __ai __attribute__((target("neon"))) int16x4_t vtrn2_s16(int16x4_t __p0, int16x4
64908	67290	#else
64909	67291	__ai __attribute__((target("neon"))) int16x4_t vtrn2_s16(int16x4_t __p0, int16x4_t __p1) {
64910	67292	int16x4_t __ret;
64911		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
64912		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	67293	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	67294	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
64913	67295	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 5, 3, 7);
64914		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	67296	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
64915	67297	return __ret;
64916	67298	}
64917	67299	#endif
...	...	@@ -64925,10 +67307,10 @@ __ai __attribute__((target("neon"))) float16x8_t vtrn2q_f16(float16x8_t __p0, fl
64925	67307	#else
64926	67308	__ai __attribute__((target("neon"))) float16x8_t vtrn2q_f16(float16x8_t __p0, float16x8_t __p1) {
64927	67309	float16x8_t __ret;
64928		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
64929		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	67310	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	67311	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
64930	67312	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 9, 3, 11, 5, 13, 7, 15);
64931		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	67313	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
64932	67314	return __ret;
64933	67315	}
64934	67316	#endif
...	...	@@ -64942,32 +67324,32 @@ __ai __attribute__((target("neon"))) float16x4_t vtrn2_f16(float16x4_t __p0, flo
64942	67324	#else
64943	67325	__ai __attribute__((target("neon"))) float16x4_t vtrn2_f16(float16x4_t __p0, float16x4_t __p1) {
64944	67326	float16x4_t __ret;
64945		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
64946		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	67327	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	67328	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
64947	67329	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 5, 3, 7);
64948		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	67330	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
64949	67331	return __ret;
64950	67332	}
64951	67333	#endif
64952	67334
64953	67335	__ai __attribute__((target("neon"))) uint64x1_t vtst_p64(poly64x1_t __p0, poly64x1_t __p1) {
64954	67336	uint64x1_t __ret;
64955		__ret = (uint64x1_t) __builtin_neon_vtst_v((int8x8_t)__p0, (int8x8_t)__p1, 19);
	67337	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vtst_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 19));
64956	67338	return __ret;
64957	67339	}
64958	67340	#ifdef __LITTLE_ENDIAN__
64959	67341	__ai __attribute__((target("neon"))) uint64x2_t vtstq_p64(poly64x2_t __p0, poly64x2_t __p1) {
64960	67342	uint64x2_t __ret;
64961		__ret = (uint64x2_t) __builtin_neon_vtstq_v((int8x16_t)__p0, (int8x16_t)__p1, 51);
	67343	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vtstq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 51));
64962	67344	return __ret;
64963	67345	}
64964	67346	#else
64965	67347	__ai __attribute__((target("neon"))) uint64x2_t vtstq_p64(poly64x2_t __p0, poly64x2_t __p1) {
64966	67348	uint64x2_t __ret;
64967		poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
64968		poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
64969		__ret = (uint64x2_t) __builtin_neon_vtstq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 51);
64970		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	67349	poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	67350	poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	67351	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vtstq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 51));
	67352	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
64971	67353	return __ret;
64972	67354	}
64973	67355	#endif
...	...	@@ -64975,16 +67357,16 @@ __ai __attribute__((target("neon"))) uint64x2_t vtstq_p64(poly64x2_t __p0, poly6
64975	67357	#ifdef __LITTLE_ENDIAN__
64976	67358	__ai __attribute__((target("neon"))) uint64x2_t vtstq_u64(uint64x2_t __p0, uint64x2_t __p1) {
64977	67359	uint64x2_t __ret;
64978		__ret = (uint64x2_t) __builtin_neon_vtstq_v((int8x16_t)__p0, (int8x16_t)__p1, 51);
	67360	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vtstq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 51));
64979	67361	return __ret;
64980	67362	}
64981	67363	#else
64982	67364	__ai __attribute__((target("neon"))) uint64x2_t vtstq_u64(uint64x2_t __p0, uint64x2_t __p1) {
64983	67365	uint64x2_t __ret;
64984		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
64985		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
64986		__ret = (uint64x2_t) __builtin_neon_vtstq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 51);
64987		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	67366	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	67367	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	67368	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vtstq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 51));
	67369	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
64988	67370	return __ret;
64989	67371	}
64990	67372	#endif
...	...	@@ -64992,73 +67374,73 @@ __ai __attribute__((target("neon"))) uint64x2_t vtstq_u64(uint64x2_t __p0, uint6
64992	67374	#ifdef __LITTLE_ENDIAN__
64993	67375	__ai __attribute__((target("neon"))) uint64x2_t vtstq_s64(int64x2_t __p0, int64x2_t __p1) {
64994	67376	uint64x2_t __ret;
64995		__ret = (uint64x2_t) __builtin_neon_vtstq_v((int8x16_t)__p0, (int8x16_t)__p1, 51);
	67377	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vtstq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 51));
64996	67378	return __ret;
64997	67379	}
64998	67380	#else
64999	67381	__ai __attribute__((target("neon"))) uint64x2_t vtstq_s64(int64x2_t __p0, int64x2_t __p1) {
65000	67382	uint64x2_t __ret;
65001		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
65002		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
65003		__ret = (uint64x2_t) __builtin_neon_vtstq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 51);
65004		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	67383	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	67384	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	67385	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vtstq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 51));
	67386	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
65005	67387	return __ret;
65006	67388	}
65007	67389	#endif
65008	67390
65009	67391	__ai __attribute__((target("neon"))) uint64x1_t vtst_u64(uint64x1_t __p0, uint64x1_t __p1) {
65010	67392	uint64x1_t __ret;
65011		__ret = (uint64x1_t) __builtin_neon_vtst_v((int8x8_t)__p0, (int8x8_t)__p1, 19);
	67393	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vtst_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 19));
65012	67394	return __ret;
65013	67395	}
65014	67396	__ai __attribute__((target("neon"))) uint64x1_t vtst_s64(int64x1_t __p0, int64x1_t __p1) {
65015	67397	uint64x1_t __ret;
65016		__ret = (uint64x1_t) __builtin_neon_vtst_v((int8x8_t)__p0, (int8x8_t)__p1, 19);
	67398	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vtst_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 19));
65017	67399	return __ret;
65018	67400	}
65019	67401	__ai __attribute__((target("neon"))) uint64_t vtstd_u64(uint64_t __p0, uint64_t __p1) {
65020	67402	uint64_t __ret;
65021		__ret = (uint64_t) __builtin_neon_vtstd_u64(__p0, __p1);
	67403	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vtstd_u64(__p0, __p1));
65022	67404	return __ret;
65023	67405	}
65024	67406	__ai __attribute__((target("neon"))) uint64_t vtstd_s64(int64_t __p0, int64_t __p1) {
65025	67407	uint64_t __ret;
65026		__ret = (uint64_t) __builtin_neon_vtstd_s64(__p0, __p1);
	67408	__ret = __builtin_bit_cast(uint64_t, __builtin_neon_vtstd_s64(__p0, __p1));
65027	67409	return __ret;
65028	67410	}
65029	67411	__ai __attribute__((target("neon"))) int8_t vuqaddb_s8(int8_t __p0, uint8_t __p1) {
65030	67412	int8_t __ret;
65031		__ret = (int8_t) __builtin_neon_vuqaddb_s8(__p0, __p1);
	67413	__ret = __builtin_bit_cast(int8_t, __builtin_neon_vuqaddb_s8(__p0, __p1));
65032	67414	return __ret;
65033	67415	}
65034	67416	__ai __attribute__((target("neon"))) int32_t vuqadds_s32(int32_t __p0, uint32_t __p1) {
65035	67417	int32_t __ret;
65036		__ret = (int32_t) __builtin_neon_vuqadds_s32(__p0, __p1);
	67418	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vuqadds_s32(__p0, __p1));
65037	67419	return __ret;
65038	67420	}
65039	67421	__ai __attribute__((target("neon"))) int64_t vuqaddd_s64(int64_t __p0, uint64_t __p1) {
65040	67422	int64_t __ret;
65041		__ret = (int64_t) __builtin_neon_vuqaddd_s64(__p0, __p1);
	67423	__ret = __builtin_bit_cast(int64_t, __builtin_neon_vuqaddd_s64(__p0, __p1));
65042	67424	return __ret;
65043	67425	}
65044	67426	__ai __attribute__((target("neon"))) int16_t vuqaddh_s16(int16_t __p0, uint16_t __p1) {
65045	67427	int16_t __ret;
65046		__ret = (int16_t) __builtin_neon_vuqaddh_s16(__p0, __p1);
	67428	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vuqaddh_s16(__p0, __p1));
65047	67429	return __ret;
65048	67430	}
65049	67431	#ifdef __LITTLE_ENDIAN__
65050	67432	__ai __attribute__((target("neon"))) int8x16_t vuqaddq_s8(int8x16_t __p0, uint8x16_t __p1) {
65051	67433	int8x16_t __ret;
65052		__ret = (int8x16_t) __builtin_neon_vuqaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
	67434	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vuqaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 32));
65053	67435	return __ret;
65054	67436	}
65055	67437	#else
65056	67438	__ai __attribute__((target("neon"))) int8x16_t vuqaddq_s8(int8x16_t __p0, uint8x16_t __p1) {
65057	67439	int8x16_t __ret;
65058		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
65059		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
65060		__ret = (int8x16_t) __builtin_neon_vuqaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 32);
65061		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	67440	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	67441	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	67442	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vuqaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 32));
	67443	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
65062	67444	return __ret;
65063	67445	}
65064	67446	#endif
...	...	@@ -65066,16 +67448,16 @@ __ai __attribute__((target("neon"))) int8x16_t vuqaddq_s8(int8x16_t __p0, uint8x
65066	67448	#ifdef __LITTLE_ENDIAN__
65067	67449	__ai __attribute__((target("neon"))) int32x4_t vuqaddq_s32(int32x4_t __p0, uint32x4_t __p1) {
65068	67450	int32x4_t __ret;
65069		__ret = (int32x4_t) __builtin_neon_vuqaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
	67451	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vuqaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 34));
65070	67452	return __ret;
65071	67453	}
65072	67454	#else
65073	67455	__ai __attribute__((target("neon"))) int32x4_t vuqaddq_s32(int32x4_t __p0, uint32x4_t __p1) {
65074	67456	int32x4_t __ret;
65075		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
65076		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
65077		__ret = (int32x4_t) __builtin_neon_vuqaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
65078		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	67457	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	67458	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	67459	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vuqaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 34));
	67460	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
65079	67461	return __ret;
65080	67462	}
65081	67463	#endif
...	...	@@ -65083,16 +67465,16 @@ __ai __attribute__((target("neon"))) int32x4_t vuqaddq_s32(int32x4_t __p0, uint3
65083	67465	#ifdef __LITTLE_ENDIAN__
65084	67466	__ai __attribute__((target("neon"))) int64x2_t vuqaddq_s64(int64x2_t __p0, uint64x2_t __p1) {
65085	67467	int64x2_t __ret;
65086		__ret = (int64x2_t) __builtin_neon_vuqaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 35);
	67468	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vuqaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 35));
65087	67469	return __ret;
65088	67470	}
65089	67471	#else
65090	67472	__ai __attribute__((target("neon"))) int64x2_t vuqaddq_s64(int64x2_t __p0, uint64x2_t __p1) {
65091	67473	int64x2_t __ret;
65092		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
65093		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
65094		__ret = (int64x2_t) __builtin_neon_vuqaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 35);
65095		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	67474	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	67475	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	67476	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vuqaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 35));
	67477	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
65096	67478	return __ret;
65097	67479	}
65098	67480	#endif
...	...	@@ -65100,16 +67482,16 @@ __ai __attribute__((target("neon"))) int64x2_t vuqaddq_s64(int64x2_t __p0, uint6
65100	67482	#ifdef __LITTLE_ENDIAN__
65101	67483	__ai __attribute__((target("neon"))) int16x8_t vuqaddq_s16(int16x8_t __p0, uint16x8_t __p1) {
65102	67484	int16x8_t __ret;
65103		__ret = (int16x8_t) __builtin_neon_vuqaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
	67485	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vuqaddq_v(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 33));
65104	67486	return __ret;
65105	67487	}
65106	67488	#else
65107	67489	__ai __attribute__((target("neon"))) int16x8_t vuqaddq_s16(int16x8_t __p0, uint16x8_t __p1) {
65108	67490	int16x8_t __ret;
65109		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
65110		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
65111		__ret = (int16x8_t) __builtin_neon_vuqaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
65112		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	67491	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	67492	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	67493	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vuqaddq_v(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 33));
	67494	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
65113	67495	return __ret;
65114	67496	}
65115	67497	#endif
...	...	@@ -65117,16 +67499,16 @@ __ai __attribute__((target("neon"))) int16x8_t vuqaddq_s16(int16x8_t __p0, uint1
65117	67499	#ifdef __LITTLE_ENDIAN__
65118	67500	__ai __attribute__((target("neon"))) int8x8_t vuqadd_s8(int8x8_t __p0, uint8x8_t __p1) {
65119	67501	int8x8_t __ret;
65120		__ret = (int8x8_t) __builtin_neon_vuqadd_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
	67502	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vuqadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 0));
65121	67503	return __ret;
65122	67504	}
65123	67505	#else
65124	67506	__ai __attribute__((target("neon"))) int8x8_t vuqadd_s8(int8x8_t __p0, uint8x8_t __p1) {
65125	67507	int8x8_t __ret;
65126		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
65127		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
65128		__ret = (int8x8_t) __builtin_neon_vuqadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
65129		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	67508	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	67509	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	67510	__ret = __builtin_bit_cast(int8x8_t, __builtin_neon_vuqadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 0));
	67511	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
65130	67512	return __ret;
65131	67513	}
65132	67514	#endif
...	...	@@ -65134,38 +67516,38 @@ __ai __attribute__((target("neon"))) int8x8_t vuqadd_s8(int8x8_t __p0, uint8x8_t
65134	67516	#ifdef __LITTLE_ENDIAN__
65135	67517	__ai __attribute__((target("neon"))) int32x2_t vuqadd_s32(int32x2_t __p0, uint32x2_t __p1) {
65136	67518	int32x2_t __ret;
65137		__ret = (int32x2_t) __builtin_neon_vuqadd_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
	67519	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vuqadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 2));
65138	67520	return __ret;
65139	67521	}
65140	67522	#else
65141	67523	__ai __attribute__((target("neon"))) int32x2_t vuqadd_s32(int32x2_t __p0, uint32x2_t __p1) {
65142	67524	int32x2_t __ret;
65143		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
65144		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
65145		__ret = (int32x2_t) __builtin_neon_vuqadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
65146		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	67525	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	67526	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	67527	__ret = __builtin_bit_cast(int32x2_t, __builtin_neon_vuqadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 2));
	67528	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
65147	67529	return __ret;
65148	67530	}
65149	67531	#endif
65150	67532
65151	67533	__ai __attribute__((target("neon"))) int64x1_t vuqadd_s64(int64x1_t __p0, uint64x1_t __p1) {
65152	67534	int64x1_t __ret;
65153		__ret = (int64x1_t) __builtin_neon_vuqadd_v((int8x8_t)__p0, (int8x8_t)__p1, 3);
	67535	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vuqadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 3));
65154	67536	return __ret;
65155	67537	}
65156	67538	#ifdef __LITTLE_ENDIAN__
65157	67539	__ai __attribute__((target("neon"))) int16x4_t vuqadd_s16(int16x4_t __p0, uint16x4_t __p1) {
65158	67540	int16x4_t __ret;
65159		__ret = (int16x4_t) __builtin_neon_vuqadd_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
	67541	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vuqadd_v(__builtin_bit_cast(int8x8_t, __p0), __builtin_bit_cast(int8x8_t, __p1), 1));
65160	67542	return __ret;
65161	67543	}
65162	67544	#else
65163	67545	__ai __attribute__((target("neon"))) int16x4_t vuqadd_s16(int16x4_t __p0, uint16x4_t __p1) {
65164	67546	int16x4_t __ret;
65165		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
65166		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
65167		__ret = (int16x4_t) __builtin_neon_vuqadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
65168		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	67547	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	67548	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	67549	__ret = __builtin_bit_cast(int16x4_t, __builtin_neon_vuqadd_v(__builtin_bit_cast(int8x8_t, __rev0), __builtin_bit_cast(int8x8_t, __rev1), 1));
	67550	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
65169	67551	return __ret;
65170	67552	}
65171	67553	#endif
...	...	@@ -65179,10 +67561,10 @@ __ai __attribute__((target("neon"))) poly8x8_t vuzp1_p8(poly8x8_t __p0, poly8x8_
65179	67561	#else
65180	67562	__ai __attribute__((target("neon"))) poly8x8_t vuzp1_p8(poly8x8_t __p0, poly8x8_t __p1) {
65181	67563	poly8x8_t __ret;
65182		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
65183		poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	67564	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	67565	poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
65184	67566	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2, 4, 6, 8, 10, 12, 14);
65185		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	67567	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
65186	67568	return __ret;
65187	67569	}
65188	67570	#endif
...	...	@@ -65196,10 +67578,10 @@ __ai __attribute__((target("neon"))) poly16x4_t vuzp1_p16(poly16x4_t __p0, poly1
65196	67578	#else
65197	67579	__ai __attribute__((target("neon"))) poly16x4_t vuzp1_p16(poly16x4_t __p0, poly16x4_t __p1) {
65198	67580	poly16x4_t __ret;
65199		poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
65200		poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	67581	poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	67582	poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
65201	67583	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2, 4, 6);
65202		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	67584	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
65203	67585	return __ret;
65204	67586	}
65205	67587	#endif
...	...	@@ -65213,10 +67595,10 @@ __ai __attribute__((target("neon"))) poly8x16_t vuzp1q_p8(poly8x16_t __p0, poly8
65213	67595	#else
65214	67596	__ai __attribute__((target("neon"))) poly8x16_t vuzp1q_p8(poly8x16_t __p0, poly8x16_t __p1) {
65215	67597	poly8x16_t __ret;
65216		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
65217		poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	67598	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	67599	poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
65218	67600	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30);
65219		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	67601	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
65220	67602	return __ret;
65221	67603	}
65222	67604	#endif
...	...	@@ -65230,10 +67612,10 @@ __ai __attribute__((target("neon"))) poly64x2_t vuzp1q_p64(poly64x2_t __p0, poly
65230	67612	#else
65231	67613	__ai __attribute__((target("neon"))) poly64x2_t vuzp1q_p64(poly64x2_t __p0, poly64x2_t __p1) {
65232	67614	poly64x2_t __ret;
65233		poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
65234		poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	67615	poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	67616	poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
65235	67617	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2);
65236		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	67618	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
65237	67619	return __ret;
65238	67620	}
65239	67621	#endif
...	...	@@ -65247,10 +67629,10 @@ __ai __attribute__((target("neon"))) poly16x8_t vuzp1q_p16(poly16x8_t __p0, poly
65247	67629	#else
65248	67630	__ai __attribute__((target("neon"))) poly16x8_t vuzp1q_p16(poly16x8_t __p0, poly16x8_t __p1) {
65249	67631	poly16x8_t __ret;
65250		poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
65251		poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	67632	poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	67633	poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
65252	67634	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2, 4, 6, 8, 10, 12, 14);
65253		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	67635	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
65254	67636	return __ret;
65255	67637	}
65256	67638	#endif
...	...	@@ -65264,10 +67646,10 @@ __ai __attribute__((target("neon"))) uint8x16_t vuzp1q_u8(uint8x16_t __p0, uint8
65264	67646	#else
65265	67647	__ai __attribute__((target("neon"))) uint8x16_t vuzp1q_u8(uint8x16_t __p0, uint8x16_t __p1) {
65266	67648	uint8x16_t __ret;
65267		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
65268		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	67649	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	67650	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
65269	67651	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30);
65270		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	67652	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
65271	67653	return __ret;
65272	67654	}
65273	67655	#endif
...	...	@@ -65281,10 +67663,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vuzp1q_u32(uint32x4_t __p0, uint
65281	67663	#else
65282	67664	__ai __attribute__((target("neon"))) uint32x4_t vuzp1q_u32(uint32x4_t __p0, uint32x4_t __p1) {
65283	67665	uint32x4_t __ret;
65284		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
65285		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	67666	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	67667	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
65286	67668	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2, 4, 6);
65287		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	67669	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
65288	67670	return __ret;
65289	67671	}
65290	67672	#endif
...	...	@@ -65298,10 +67680,10 @@ __ai __attribute__((target("neon"))) uint64x2_t vuzp1q_u64(uint64x2_t __p0, uint
65298	67680	#else
65299	67681	__ai __attribute__((target("neon"))) uint64x2_t vuzp1q_u64(uint64x2_t __p0, uint64x2_t __p1) {
65300	67682	uint64x2_t __ret;
65301		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
65302		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	67683	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	67684	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
65303	67685	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2);
65304		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	67686	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
65305	67687	return __ret;
65306	67688	}
65307	67689	#endif
...	...	@@ -65315,10 +67697,10 @@ __ai __attribute__((target("neon"))) uint16x8_t vuzp1q_u16(uint16x8_t __p0, uint
65315	67697	#else
65316	67698	__ai __attribute__((target("neon"))) uint16x8_t vuzp1q_u16(uint16x8_t __p0, uint16x8_t __p1) {
65317	67699	uint16x8_t __ret;
65318		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
65319		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	67700	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	67701	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
65320	67702	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2, 4, 6, 8, 10, 12, 14);
65321		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	67703	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
65322	67704	return __ret;
65323	67705	}
65324	67706	#endif
...	...	@@ -65332,10 +67714,10 @@ __ai __attribute__((target("neon"))) int8x16_t vuzp1q_s8(int8x16_t __p0, int8x16
65332	67714	#else
65333	67715	__ai __attribute__((target("neon"))) int8x16_t vuzp1q_s8(int8x16_t __p0, int8x16_t __p1) {
65334	67716	int8x16_t __ret;
65335		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
65336		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	67717	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	67718	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
65337	67719	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30);
65338		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	67720	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
65339	67721	return __ret;
65340	67722	}
65341	67723	#endif
...	...	@@ -65349,10 +67731,10 @@ __ai __attribute__((target("neon"))) float64x2_t vuzp1q_f64(float64x2_t __p0, fl
65349	67731	#else
65350	67732	__ai __attribute__((target("neon"))) float64x2_t vuzp1q_f64(float64x2_t __p0, float64x2_t __p1) {
65351	67733	float64x2_t __ret;
65352		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
65353		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	67734	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	67735	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
65354	67736	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2);
65355		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	67737	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
65356	67738	return __ret;
65357	67739	}
65358	67740	#endif
...	...	@@ -65366,10 +67748,10 @@ __ai __attribute__((target("neon"))) float32x4_t vuzp1q_f32(float32x4_t __p0, fl
65366	67748	#else
65367	67749	__ai __attribute__((target("neon"))) float32x4_t vuzp1q_f32(float32x4_t __p0, float32x4_t __p1) {
65368	67750	float32x4_t __ret;
65369		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
65370		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	67751	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	67752	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
65371	67753	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2, 4, 6);
65372		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	67754	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
65373	67755	return __ret;
65374	67756	}
65375	67757	#endif
...	...	@@ -65383,10 +67765,10 @@ __ai __attribute__((target("neon"))) int32x4_t vuzp1q_s32(int32x4_t __p0, int32x
65383	67765	#else
65384	67766	__ai __attribute__((target("neon"))) int32x4_t vuzp1q_s32(int32x4_t __p0, int32x4_t __p1) {
65385	67767	int32x4_t __ret;
65386		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
65387		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	67768	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	67769	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
65388	67770	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2, 4, 6);
65389		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	67771	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
65390	67772	return __ret;
65391	67773	}
65392	67774	#endif
...	...	@@ -65400,10 +67782,27 @@ __ai __attribute__((target("neon"))) int64x2_t vuzp1q_s64(int64x2_t __p0, int64x
65400	67782	#else
65401	67783	__ai __attribute__((target("neon"))) int64x2_t vuzp1q_s64(int64x2_t __p0, int64x2_t __p1) {
65402	67784	int64x2_t __ret;
65403		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
65404		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	67785	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	67786	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
65405	67787	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2);
65406		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	67788	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
	67789	return __ret;
	67790	}
	67791	#endif
	67792
	67793	#ifdef __LITTLE_ENDIAN__
	67794	__ai __attribute__((target("neon"))) mfloat8x16_t vuzp1q_mf8(mfloat8x16_t __p0, mfloat8x16_t __p1) {
	67795	mfloat8x16_t __ret;
	67796	__ret = __builtin_shufflevector(__p0, __p1, 0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30);
	67797	return __ret;
	67798	}
	67799	#else
	67800	__ai __attribute__((target("neon"))) mfloat8x16_t vuzp1q_mf8(mfloat8x16_t __p0, mfloat8x16_t __p1) {
	67801	mfloat8x16_t __ret;
	67802	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	67803	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	67804	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30);
	67805	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
65407	67806	return __ret;
65408	67807	}
65409	67808	#endif
...	...	@@ -65417,10 +67816,10 @@ __ai __attribute__((target("neon"))) int16x8_t vuzp1q_s16(int16x8_t __p0, int16x
65417	67816	#else
65418	67817	__ai __attribute__((target("neon"))) int16x8_t vuzp1q_s16(int16x8_t __p0, int16x8_t __p1) {
65419	67818	int16x8_t __ret;
65420		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
65421		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	67819	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	67820	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
65422	67821	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2, 4, 6, 8, 10, 12, 14);
65423		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	67822	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
65424	67823	return __ret;
65425	67824	}
65426	67825	#endif
...	...	@@ -65434,10 +67833,10 @@ __ai __attribute__((target("neon"))) uint8x8_t vuzp1_u8(uint8x8_t __p0, uint8x8_
65434	67833	#else
65435	67834	__ai __attribute__((target("neon"))) uint8x8_t vuzp1_u8(uint8x8_t __p0, uint8x8_t __p1) {
65436	67835	uint8x8_t __ret;
65437		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
65438		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	67836	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	67837	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
65439	67838	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2, 4, 6, 8, 10, 12, 14);
65440		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	67839	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
65441	67840	return __ret;
65442	67841	}
65443	67842	#endif
...	...	@@ -65451,10 +67850,10 @@ __ai __attribute__((target("neon"))) uint32x2_t vuzp1_u32(uint32x2_t __p0, uint3
65451	67850	#else
65452	67851	__ai __attribute__((target("neon"))) uint32x2_t vuzp1_u32(uint32x2_t __p0, uint32x2_t __p1) {
65453	67852	uint32x2_t __ret;
65454		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
65455		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	67853	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	67854	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
65456	67855	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2);
65457		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	67856	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
65458	67857	return __ret;
65459	67858	}
65460	67859	#endif
...	...	@@ -65468,10 +67867,10 @@ __ai __attribute__((target("neon"))) uint16x4_t vuzp1_u16(uint16x4_t __p0, uint1
65468	67867	#else
65469	67868	__ai __attribute__((target("neon"))) uint16x4_t vuzp1_u16(uint16x4_t __p0, uint16x4_t __p1) {
65470	67869	uint16x4_t __ret;
65471		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
65472		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	67870	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	67871	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
65473	67872	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2, 4, 6);
65474		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	67873	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
65475	67874	return __ret;
65476	67875	}
65477	67876	#endif
...	...	@@ -65485,10 +67884,10 @@ __ai __attribute__((target("neon"))) int8x8_t vuzp1_s8(int8x8_t __p0, int8x8_t _
65485	67884	#else
65486	67885	__ai __attribute__((target("neon"))) int8x8_t vuzp1_s8(int8x8_t __p0, int8x8_t __p1) {
65487	67886	int8x8_t __ret;
65488		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
65489		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	67887	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	67888	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
65490	67889	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2, 4, 6, 8, 10, 12, 14);
65491		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	67890	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
65492	67891	return __ret;
65493	67892	}
65494	67893	#endif
...	...	@@ -65502,10 +67901,10 @@ __ai __attribute__((target("neon"))) float32x2_t vuzp1_f32(float32x2_t __p0, flo
65502	67901	#else
65503	67902	__ai __attribute__((target("neon"))) float32x2_t vuzp1_f32(float32x2_t __p0, float32x2_t __p1) {
65504	67903	float32x2_t __ret;
65505		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
65506		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	67904	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	67905	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
65507	67906	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2);
65508		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	67907	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
65509	67908	return __ret;
65510	67909	}
65511	67910	#endif
...	...	@@ -65519,10 +67918,27 @@ __ai __attribute__((target("neon"))) int32x2_t vuzp1_s32(int32x2_t __p0, int32x2
65519	67918	#else
65520	67919	__ai __attribute__((target("neon"))) int32x2_t vuzp1_s32(int32x2_t __p0, int32x2_t __p1) {
65521	67920	int32x2_t __ret;
65522		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
65523		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	67921	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	67922	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
65524	67923	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2);
65525		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	67924	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
	67925	return __ret;
	67926	}
	67927	#endif
	67928
	67929	#ifdef __LITTLE_ENDIAN__
	67930	__ai __attribute__((target("neon"))) mfloat8x8_t vuzp1_mf8(mfloat8x8_t __p0, mfloat8x8_t __p1) {
	67931	mfloat8x8_t __ret;
	67932	__ret = __builtin_shufflevector(__p0, __p1, 0, 2, 4, 6, 8, 10, 12, 14);
	67933	return __ret;
	67934	}
	67935	#else
	67936	__ai __attribute__((target("neon"))) mfloat8x8_t vuzp1_mf8(mfloat8x8_t __p0, mfloat8x8_t __p1) {
	67937	mfloat8x8_t __ret;
	67938	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	67939	mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	67940	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2, 4, 6, 8, 10, 12, 14);
	67941	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
65526	67942	return __ret;
65527	67943	}
65528	67944	#endif
...	...	@@ -65536,10 +67952,10 @@ __ai __attribute__((target("neon"))) int16x4_t vuzp1_s16(int16x4_t __p0, int16x4
65536	67952	#else
65537	67953	__ai __attribute__((target("neon"))) int16x4_t vuzp1_s16(int16x4_t __p0, int16x4_t __p1) {
65538	67954	int16x4_t __ret;
65539		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
65540		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	67955	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	67956	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
65541	67957	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2, 4, 6);
65542		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	67958	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
65543	67959	return __ret;
65544	67960	}
65545	67961	#endif
...	...	@@ -65553,10 +67969,10 @@ __ai __attribute__((target("neon"))) float16x8_t vuzp1q_f16(float16x8_t __p0, fl
65553	67969	#else
65554	67970	__ai __attribute__((target("neon"))) float16x8_t vuzp1q_f16(float16x8_t __p0, float16x8_t __p1) {
65555	67971	float16x8_t __ret;
65556		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
65557		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	67972	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	67973	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
65558	67974	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2, 4, 6, 8, 10, 12, 14);
65559		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	67975	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
65560	67976	return __ret;
65561	67977	}
65562	67978	#endif
...	...	@@ -65570,10 +67986,10 @@ __ai __attribute__((target("neon"))) float16x4_t vuzp1_f16(float16x4_t __p0, flo
65570	67986	#else
65571	67987	__ai __attribute__((target("neon"))) float16x4_t vuzp1_f16(float16x4_t __p0, float16x4_t __p1) {
65572	67988	float16x4_t __ret;
65573		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
65574		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	67989	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	67990	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
65575	67991	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2, 4, 6);
65576		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	67992	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
65577	67993	return __ret;
65578	67994	}
65579	67995	#endif
...	...	@@ -65587,10 +68003,10 @@ __ai __attribute__((target("neon"))) poly8x8_t vuzp2_p8(poly8x8_t __p0, poly8x8_
65587	68003	#else
65588	68004	__ai __attribute__((target("neon"))) poly8x8_t vuzp2_p8(poly8x8_t __p0, poly8x8_t __p1) {
65589	68005	poly8x8_t __ret;
65590		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
65591		poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	68006	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	68007	poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
65592	68008	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3, 5, 7, 9, 11, 13, 15);
65593		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	68009	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
65594	68010	return __ret;
65595	68011	}
65596	68012	#endif
...	...	@@ -65604,10 +68020,10 @@ __ai __attribute__((target("neon"))) poly16x4_t vuzp2_p16(poly16x4_t __p0, poly1
65604	68020	#else
65605	68021	__ai __attribute__((target("neon"))) poly16x4_t vuzp2_p16(poly16x4_t __p0, poly16x4_t __p1) {
65606	68022	poly16x4_t __ret;
65607		poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
65608		poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	68023	poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	68024	poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
65609	68025	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3, 5, 7);
65610		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	68026	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
65611	68027	return __ret;
65612	68028	}
65613	68029	#endif
...	...	@@ -65621,10 +68037,10 @@ __ai __attribute__((target("neon"))) poly8x16_t vuzp2q_p8(poly8x16_t __p0, poly8
65621	68037	#else
65622	68038	__ai __attribute__((target("neon"))) poly8x16_t vuzp2q_p8(poly8x16_t __p0, poly8x16_t __p1) {
65623	68039	poly8x16_t __ret;
65624		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
65625		poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	68040	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	68041	poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
65626	68042	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3, 5, 7, 9, 11, 13, 15, 17, 19, 21, 23, 25, 27, 29, 31);
65627		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	68043	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
65628	68044	return __ret;
65629	68045	}
65630	68046	#endif
...	...	@@ -65638,10 +68054,10 @@ __ai __attribute__((target("neon"))) poly64x2_t vuzp2q_p64(poly64x2_t __p0, poly
65638	68054	#else
65639	68055	__ai __attribute__((target("neon"))) poly64x2_t vuzp2q_p64(poly64x2_t __p0, poly64x2_t __p1) {
65640	68056	poly64x2_t __ret;
65641		poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
65642		poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	68057	poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	68058	poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
65643	68059	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3);
65644		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	68060	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
65645	68061	return __ret;
65646	68062	}
65647	68063	#endif
...	...	@@ -65655,10 +68071,10 @@ __ai __attribute__((target("neon"))) poly16x8_t vuzp2q_p16(poly16x8_t __p0, poly
65655	68071	#else
65656	68072	__ai __attribute__((target("neon"))) poly16x8_t vuzp2q_p16(poly16x8_t __p0, poly16x8_t __p1) {
65657	68073	poly16x8_t __ret;
65658		poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
65659		poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	68074	poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	68075	poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
65660	68076	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3, 5, 7, 9, 11, 13, 15);
65661		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	68077	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
65662	68078	return __ret;
65663	68079	}
65664	68080	#endif
...	...	@@ -65672,10 +68088,10 @@ __ai __attribute__((target("neon"))) uint8x16_t vuzp2q_u8(uint8x16_t __p0, uint8
65672	68088	#else
65673	68089	__ai __attribute__((target("neon"))) uint8x16_t vuzp2q_u8(uint8x16_t __p0, uint8x16_t __p1) {
65674	68090	uint8x16_t __ret;
65675		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
65676		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	68091	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	68092	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
65677	68093	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3, 5, 7, 9, 11, 13, 15, 17, 19, 21, 23, 25, 27, 29, 31);
65678		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	68094	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
65679	68095	return __ret;
65680	68096	}
65681	68097	#endif
...	...	@@ -65689,10 +68105,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vuzp2q_u32(uint32x4_t __p0, uint
65689	68105	#else
65690	68106	__ai __attribute__((target("neon"))) uint32x4_t vuzp2q_u32(uint32x4_t __p0, uint32x4_t __p1) {
65691	68107	uint32x4_t __ret;
65692		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
65693		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	68108	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	68109	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
65694	68110	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3, 5, 7);
65695		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	68111	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
65696	68112	return __ret;
65697	68113	}
65698	68114	#endif
...	...	@@ -65706,10 +68122,10 @@ __ai __attribute__((target("neon"))) uint64x2_t vuzp2q_u64(uint64x2_t __p0, uint
65706	68122	#else
65707	68123	__ai __attribute__((target("neon"))) uint64x2_t vuzp2q_u64(uint64x2_t __p0, uint64x2_t __p1) {
65708	68124	uint64x2_t __ret;
65709		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
65710		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	68125	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	68126	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
65711	68127	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3);
65712		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	68128	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
65713	68129	return __ret;
65714	68130	}
65715	68131	#endif
...	...	@@ -65723,10 +68139,10 @@ __ai __attribute__((target("neon"))) uint16x8_t vuzp2q_u16(uint16x8_t __p0, uint
65723	68139	#else
65724	68140	__ai __attribute__((target("neon"))) uint16x8_t vuzp2q_u16(uint16x8_t __p0, uint16x8_t __p1) {
65725	68141	uint16x8_t __ret;
65726		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
65727		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	68142	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	68143	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
65728	68144	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3, 5, 7, 9, 11, 13, 15);
65729		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	68145	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
65730	68146	return __ret;
65731	68147	}
65732	68148	#endif
...	...	@@ -65740,10 +68156,10 @@ __ai __attribute__((target("neon"))) int8x16_t vuzp2q_s8(int8x16_t __p0, int8x16
65740	68156	#else
65741	68157	__ai __attribute__((target("neon"))) int8x16_t vuzp2q_s8(int8x16_t __p0, int8x16_t __p1) {
65742	68158	int8x16_t __ret;
65743		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
65744		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	68159	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	68160	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
65745	68161	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3, 5, 7, 9, 11, 13, 15, 17, 19, 21, 23, 25, 27, 29, 31);
65746		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	68162	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
65747	68163	return __ret;
65748	68164	}
65749	68165	#endif
...	...	@@ -65757,10 +68173,10 @@ __ai __attribute__((target("neon"))) float64x2_t vuzp2q_f64(float64x2_t __p0, fl
65757	68173	#else
65758	68174	__ai __attribute__((target("neon"))) float64x2_t vuzp2q_f64(float64x2_t __p0, float64x2_t __p1) {
65759	68175	float64x2_t __ret;
65760		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
65761		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	68176	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	68177	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
65762	68178	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3);
65763		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	68179	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
65764	68180	return __ret;
65765	68181	}
65766	68182	#endif
...	...	@@ -65774,10 +68190,10 @@ __ai __attribute__((target("neon"))) float32x4_t vuzp2q_f32(float32x4_t __p0, fl
65774	68190	#else
65775	68191	__ai __attribute__((target("neon"))) float32x4_t vuzp2q_f32(float32x4_t __p0, float32x4_t __p1) {
65776	68192	float32x4_t __ret;
65777		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
65778		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	68193	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	68194	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
65779	68195	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3, 5, 7);
65780		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	68196	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
65781	68197	return __ret;
65782	68198	}
65783	68199	#endif
...	...	@@ -65791,10 +68207,10 @@ __ai __attribute__((target("neon"))) int32x4_t vuzp2q_s32(int32x4_t __p0, int32x
65791	68207	#else
65792	68208	__ai __attribute__((target("neon"))) int32x4_t vuzp2q_s32(int32x4_t __p0, int32x4_t __p1) {
65793	68209	int32x4_t __ret;
65794		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
65795		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	68210	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	68211	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
65796	68212	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3, 5, 7);
65797		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	68213	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
65798	68214	return __ret;
65799	68215	}
65800	68216	#endif
...	...	@@ -65808,10 +68224,27 @@ __ai __attribute__((target("neon"))) int64x2_t vuzp2q_s64(int64x2_t __p0, int64x
65808	68224	#else
65809	68225	__ai __attribute__((target("neon"))) int64x2_t vuzp2q_s64(int64x2_t __p0, int64x2_t __p1) {
65810	68226	int64x2_t __ret;
65811		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
65812		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	68227	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	68228	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
65813	68229	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3);
65814		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	68230	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
	68231	return __ret;
	68232	}
	68233	#endif
	68234
	68235	#ifdef __LITTLE_ENDIAN__
	68236	__ai __attribute__((target("neon"))) mfloat8x16_t vuzp2q_mf8(mfloat8x16_t __p0, mfloat8x16_t __p1) {
	68237	mfloat8x16_t __ret;
	68238	__ret = __builtin_shufflevector(__p0, __p1, 1, 3, 5, 7, 9, 11, 13, 15, 17, 19, 21, 23, 25, 27, 29, 31);
	68239	return __ret;
	68240	}
	68241	#else
	68242	__ai __attribute__((target("neon"))) mfloat8x16_t vuzp2q_mf8(mfloat8x16_t __p0, mfloat8x16_t __p1) {
	68243	mfloat8x16_t __ret;
	68244	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	68245	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	68246	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3, 5, 7, 9, 11, 13, 15, 17, 19, 21, 23, 25, 27, 29, 31);
	68247	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
65815	68248	return __ret;
65816	68249	}
65817	68250	#endif
...	...	@@ -65825,10 +68258,10 @@ __ai __attribute__((target("neon"))) int16x8_t vuzp2q_s16(int16x8_t __p0, int16x
65825	68258	#else
65826	68259	__ai __attribute__((target("neon"))) int16x8_t vuzp2q_s16(int16x8_t __p0, int16x8_t __p1) {
65827	68260	int16x8_t __ret;
65828		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
65829		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	68261	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	68262	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
65830	68263	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3, 5, 7, 9, 11, 13, 15);
65831		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	68264	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
65832	68265	return __ret;
65833	68266	}
65834	68267	#endif
...	...	@@ -65842,10 +68275,10 @@ __ai __attribute__((target("neon"))) uint8x8_t vuzp2_u8(uint8x8_t __p0, uint8x8_
65842	68275	#else
65843	68276	__ai __attribute__((target("neon"))) uint8x8_t vuzp2_u8(uint8x8_t __p0, uint8x8_t __p1) {
65844	68277	uint8x8_t __ret;
65845		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
65846		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	68278	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	68279	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
65847	68280	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3, 5, 7, 9, 11, 13, 15);
65848		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	68281	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
65849	68282	return __ret;
65850	68283	}
65851	68284	#endif
...	...	@@ -65859,10 +68292,10 @@ __ai __attribute__((target("neon"))) uint32x2_t vuzp2_u32(uint32x2_t __p0, uint3
65859	68292	#else
65860	68293	__ai __attribute__((target("neon"))) uint32x2_t vuzp2_u32(uint32x2_t __p0, uint32x2_t __p1) {
65861	68294	uint32x2_t __ret;
65862		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
65863		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	68295	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	68296	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
65864	68297	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3);
65865		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	68298	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
65866	68299	return __ret;
65867	68300	}
65868	68301	#endif
...	...	@@ -65876,10 +68309,10 @@ __ai __attribute__((target("neon"))) uint16x4_t vuzp2_u16(uint16x4_t __p0, uint1
65876	68309	#else
65877	68310	__ai __attribute__((target("neon"))) uint16x4_t vuzp2_u16(uint16x4_t __p0, uint16x4_t __p1) {
65878	68311	uint16x4_t __ret;
65879		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
65880		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	68312	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	68313	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
65881	68314	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3, 5, 7);
65882		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	68315	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
65883	68316	return __ret;
65884	68317	}
65885	68318	#endif
...	...	@@ -65893,10 +68326,10 @@ __ai __attribute__((target("neon"))) int8x8_t vuzp2_s8(int8x8_t __p0, int8x8_t _
65893	68326	#else
65894	68327	__ai __attribute__((target("neon"))) int8x8_t vuzp2_s8(int8x8_t __p0, int8x8_t __p1) {
65895	68328	int8x8_t __ret;
65896		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
65897		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	68329	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	68330	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
65898	68331	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3, 5, 7, 9, 11, 13, 15);
65899		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	68332	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
65900	68333	return __ret;
65901	68334	}
65902	68335	#endif
...	...	@@ -65910,10 +68343,10 @@ __ai __attribute__((target("neon"))) float32x2_t vuzp2_f32(float32x2_t __p0, flo
65910	68343	#else
65911	68344	__ai __attribute__((target("neon"))) float32x2_t vuzp2_f32(float32x2_t __p0, float32x2_t __p1) {
65912	68345	float32x2_t __ret;
65913		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
65914		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	68346	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	68347	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
65915	68348	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3);
65916		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	68349	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
65917	68350	return __ret;
65918	68351	}
65919	68352	#endif
...	...	@@ -65927,10 +68360,27 @@ __ai __attribute__((target("neon"))) int32x2_t vuzp2_s32(int32x2_t __p0, int32x2
65927	68360	#else
65928	68361	__ai __attribute__((target("neon"))) int32x2_t vuzp2_s32(int32x2_t __p0, int32x2_t __p1) {
65929	68362	int32x2_t __ret;
65930		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
65931		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	68363	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	68364	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
65932	68365	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3);
65933		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	68366	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
	68367	return __ret;
	68368	}
	68369	#endif
	68370
	68371	#ifdef __LITTLE_ENDIAN__
	68372	__ai __attribute__((target("neon"))) mfloat8x8_t vuzp2_mf8(mfloat8x8_t __p0, mfloat8x8_t __p1) {
	68373	mfloat8x8_t __ret;
	68374	__ret = __builtin_shufflevector(__p0, __p1, 1, 3, 5, 7, 9, 11, 13, 15);
	68375	return __ret;
	68376	}
	68377	#else
	68378	__ai __attribute__((target("neon"))) mfloat8x8_t vuzp2_mf8(mfloat8x8_t __p0, mfloat8x8_t __p1) {
	68379	mfloat8x8_t __ret;
	68380	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	68381	mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	68382	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3, 5, 7, 9, 11, 13, 15);
	68383	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
65934	68384	return __ret;
65935	68385	}
65936	68386	#endif
...	...	@@ -65944,10 +68394,10 @@ __ai __attribute__((target("neon"))) int16x4_t vuzp2_s16(int16x4_t __p0, int16x4
65944	68394	#else
65945	68395	__ai __attribute__((target("neon"))) int16x4_t vuzp2_s16(int16x4_t __p0, int16x4_t __p1) {
65946	68396	int16x4_t __ret;
65947		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
65948		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	68397	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	68398	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
65949	68399	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3, 5, 7);
65950		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	68400	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
65951	68401	return __ret;
65952	68402	}
65953	68403	#endif
...	...	@@ -65961,10 +68411,10 @@ __ai __attribute__((target("neon"))) float16x8_t vuzp2q_f16(float16x8_t __p0, fl
65961	68411	#else
65962	68412	__ai __attribute__((target("neon"))) float16x8_t vuzp2q_f16(float16x8_t __p0, float16x8_t __p1) {
65963	68413	float16x8_t __ret;
65964		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
65965		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	68414	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	68415	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
65966	68416	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3, 5, 7, 9, 11, 13, 15);
65967		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	68417	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
65968	68418	return __ret;
65969	68419	}
65970	68420	#endif
...	...	@@ -65978,10 +68428,10 @@ __ai __attribute__((target("neon"))) float16x4_t vuzp2_f16(float16x4_t __p0, flo
65978	68428	#else
65979	68429	__ai __attribute__((target("neon"))) float16x4_t vuzp2_f16(float16x4_t __p0, float16x4_t __p1) {
65980	68430	float16x4_t __ret;
65981		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
65982		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	68431	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	68432	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
65983	68433	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3, 5, 7);
65984		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	68434	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
65985	68435	return __ret;
65986	68436	}
65987	68437	#endif
...	...	@@ -65995,10 +68445,10 @@ __ai __attribute__((target("neon"))) poly8x8_t vzip1_p8(poly8x8_t __p0, poly8x8_
65995	68445	#else
65996	68446	__ai __attribute__((target("neon"))) poly8x8_t vzip1_p8(poly8x8_t __p0, poly8x8_t __p1) {
65997	68447	poly8x8_t __ret;
65998		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
65999		poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	68448	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	68449	poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
66000	68450	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 8, 1, 9, 2, 10, 3, 11);
66001		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	68451	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
66002	68452	return __ret;
66003	68453	}
66004	68454	#endif
...	...	@@ -66012,10 +68462,10 @@ __ai __attribute__((target("neon"))) poly16x4_t vzip1_p16(poly16x4_t __p0, poly1
66012	68462	#else
66013	68463	__ai __attribute__((target("neon"))) poly16x4_t vzip1_p16(poly16x4_t __p0, poly16x4_t __p1) {
66014	68464	poly16x4_t __ret;
66015		poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
66016		poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	68465	poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	68466	poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
66017	68467	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 4, 1, 5);
66018		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	68468	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
66019	68469	return __ret;
66020	68470	}
66021	68471	#endif
...	...	@@ -66029,10 +68479,10 @@ __ai __attribute__((target("neon"))) poly8x16_t vzip1q_p8(poly8x16_t __p0, poly8
66029	68479	#else
66030	68480	__ai __attribute__((target("neon"))) poly8x16_t vzip1q_p8(poly8x16_t __p0, poly8x16_t __p1) {
66031	68481	poly8x16_t __ret;
66032		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
66033		poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	68482	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	68483	poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
66034	68484	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 16, 1, 17, 2, 18, 3, 19, 4, 20, 5, 21, 6, 22, 7, 23);
66035		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	68485	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
66036	68486	return __ret;
66037	68487	}
66038	68488	#endif
...	...	@@ -66046,10 +68496,10 @@ __ai __attribute__((target("neon"))) poly64x2_t vzip1q_p64(poly64x2_t __p0, poly
66046	68496	#else
66047	68497	__ai __attribute__((target("neon"))) poly64x2_t vzip1q_p64(poly64x2_t __p0, poly64x2_t __p1) {
66048	68498	poly64x2_t __ret;
66049		poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
66050		poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	68499	poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	68500	poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
66051	68501	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2);
66052		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	68502	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
66053	68503	return __ret;
66054	68504	}
66055	68505	#endif
...	...	@@ -66063,10 +68513,10 @@ __ai __attribute__((target("neon"))) poly16x8_t vzip1q_p16(poly16x8_t __p0, poly
66063	68513	#else
66064	68514	__ai __attribute__((target("neon"))) poly16x8_t vzip1q_p16(poly16x8_t __p0, poly16x8_t __p1) {
66065	68515	poly16x8_t __ret;
66066		poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
66067		poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	68516	poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	68517	poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
66068	68518	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 8, 1, 9, 2, 10, 3, 11);
66069		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	68519	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
66070	68520	return __ret;
66071	68521	}
66072	68522	#endif
...	...	@@ -66080,10 +68530,10 @@ __ai __attribute__((target("neon"))) uint8x16_t vzip1q_u8(uint8x16_t __p0, uint8
66080	68530	#else
66081	68531	__ai __attribute__((target("neon"))) uint8x16_t vzip1q_u8(uint8x16_t __p0, uint8x16_t __p1) {
66082	68532	uint8x16_t __ret;
66083		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
66084		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	68533	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	68534	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
66085	68535	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 16, 1, 17, 2, 18, 3, 19, 4, 20, 5, 21, 6, 22, 7, 23);
66086		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	68536	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
66087	68537	return __ret;
66088	68538	}
66089	68539	#endif
...	...	@@ -66097,10 +68547,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vzip1q_u32(uint32x4_t __p0, uint
66097	68547	#else
66098	68548	__ai __attribute__((target("neon"))) uint32x4_t vzip1q_u32(uint32x4_t __p0, uint32x4_t __p1) {
66099	68549	uint32x4_t __ret;
66100		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
66101		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	68550	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	68551	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
66102	68552	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 4, 1, 5);
66103		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	68553	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
66104	68554	return __ret;
66105	68555	}
66106	68556	#endif
...	...	@@ -66114,10 +68564,10 @@ __ai __attribute__((target("neon"))) uint64x2_t vzip1q_u64(uint64x2_t __p0, uint
66114	68564	#else
66115	68565	__ai __attribute__((target("neon"))) uint64x2_t vzip1q_u64(uint64x2_t __p0, uint64x2_t __p1) {
66116	68566	uint64x2_t __ret;
66117		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
66118		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	68567	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	68568	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
66119	68569	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2);
66120		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	68570	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
66121	68571	return __ret;
66122	68572	}
66123	68573	#endif
...	...	@@ -66131,10 +68581,10 @@ __ai __attribute__((target("neon"))) uint16x8_t vzip1q_u16(uint16x8_t __p0, uint
66131	68581	#else
66132	68582	__ai __attribute__((target("neon"))) uint16x8_t vzip1q_u16(uint16x8_t __p0, uint16x8_t __p1) {
66133	68583	uint16x8_t __ret;
66134		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
66135		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	68584	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	68585	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
66136	68586	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 8, 1, 9, 2, 10, 3, 11);
66137		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	68587	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
66138	68588	return __ret;
66139	68589	}
66140	68590	#endif
...	...	@@ -66148,10 +68598,10 @@ __ai __attribute__((target("neon"))) int8x16_t vzip1q_s8(int8x16_t __p0, int8x16
66148	68598	#else
66149	68599	__ai __attribute__((target("neon"))) int8x16_t vzip1q_s8(int8x16_t __p0, int8x16_t __p1) {
66150	68600	int8x16_t __ret;
66151		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
66152		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	68601	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	68602	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
66153	68603	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 16, 1, 17, 2, 18, 3, 19, 4, 20, 5, 21, 6, 22, 7, 23);
66154		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	68604	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
66155	68605	return __ret;
66156	68606	}
66157	68607	#endif
...	...	@@ -66165,10 +68615,10 @@ __ai __attribute__((target("neon"))) float64x2_t vzip1q_f64(float64x2_t __p0, fl
66165	68615	#else
66166	68616	__ai __attribute__((target("neon"))) float64x2_t vzip1q_f64(float64x2_t __p0, float64x2_t __p1) {
66167	68617	float64x2_t __ret;
66168		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
66169		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	68618	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	68619	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
66170	68620	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2);
66171		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	68621	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
66172	68622	return __ret;
66173	68623	}
66174	68624	#endif
...	...	@@ -66182,10 +68632,10 @@ __ai __attribute__((target("neon"))) float32x4_t vzip1q_f32(float32x4_t __p0, fl
66182	68632	#else
66183	68633	__ai __attribute__((target("neon"))) float32x4_t vzip1q_f32(float32x4_t __p0, float32x4_t __p1) {
66184	68634	float32x4_t __ret;
66185		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
66186		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	68635	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	68636	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
66187	68637	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 4, 1, 5);
66188		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	68638	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
66189	68639	return __ret;
66190	68640	}
66191	68641	#endif
...	...	@@ -66199,10 +68649,10 @@ __ai __attribute__((target("neon"))) int32x4_t vzip1q_s32(int32x4_t __p0, int32x
66199	68649	#else
66200	68650	__ai __attribute__((target("neon"))) int32x4_t vzip1q_s32(int32x4_t __p0, int32x4_t __p1) {
66201	68651	int32x4_t __ret;
66202		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
66203		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	68652	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	68653	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
66204	68654	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 4, 1, 5);
66205		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	68655	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
66206	68656	return __ret;
66207	68657	}
66208	68658	#endif
...	...	@@ -66216,10 +68666,27 @@ __ai __attribute__((target("neon"))) int64x2_t vzip1q_s64(int64x2_t __p0, int64x
66216	68666	#else
66217	68667	__ai __attribute__((target("neon"))) int64x2_t vzip1q_s64(int64x2_t __p0, int64x2_t __p1) {
66218	68668	int64x2_t __ret;
66219		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
66220		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	68669	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	68670	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
66221	68671	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2);
66222		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	68672	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
	68673	return __ret;
	68674	}
	68675	#endif
	68676
	68677	#ifdef __LITTLE_ENDIAN__
	68678	__ai __attribute__((target("neon"))) mfloat8x16_t vzip1q_mf8(mfloat8x16_t __p0, mfloat8x16_t __p1) {
	68679	mfloat8x16_t __ret;
	68680	__ret = __builtin_shufflevector(__p0, __p1, 0, 16, 1, 17, 2, 18, 3, 19, 4, 20, 5, 21, 6, 22, 7, 23);
	68681	return __ret;
	68682	}
	68683	#else
	68684	__ai __attribute__((target("neon"))) mfloat8x16_t vzip1q_mf8(mfloat8x16_t __p0, mfloat8x16_t __p1) {
	68685	mfloat8x16_t __ret;
	68686	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	68687	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	68688	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 16, 1, 17, 2, 18, 3, 19, 4, 20, 5, 21, 6, 22, 7, 23);
	68689	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
66223	68690	return __ret;
66224	68691	}
66225	68692	#endif
...	...	@@ -66233,10 +68700,10 @@ __ai __attribute__((target("neon"))) int16x8_t vzip1q_s16(int16x8_t __p0, int16x
66233	68700	#else
66234	68701	__ai __attribute__((target("neon"))) int16x8_t vzip1q_s16(int16x8_t __p0, int16x8_t __p1) {
66235	68702	int16x8_t __ret;
66236		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
66237		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	68703	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	68704	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
66238	68705	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 8, 1, 9, 2, 10, 3, 11);
66239		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	68706	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
66240	68707	return __ret;
66241	68708	}
66242	68709	#endif
...	...	@@ -66250,10 +68717,10 @@ __ai __attribute__((target("neon"))) uint8x8_t vzip1_u8(uint8x8_t __p0, uint8x8_
66250	68717	#else
66251	68718	__ai __attribute__((target("neon"))) uint8x8_t vzip1_u8(uint8x8_t __p0, uint8x8_t __p1) {
66252	68719	uint8x8_t __ret;
66253		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
66254		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	68720	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	68721	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
66255	68722	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 8, 1, 9, 2, 10, 3, 11);
66256		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	68723	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
66257	68724	return __ret;
66258	68725	}
66259	68726	#endif
...	...	@@ -66267,10 +68734,10 @@ __ai __attribute__((target("neon"))) uint32x2_t vzip1_u32(uint32x2_t __p0, uint3
66267	68734	#else
66268	68735	__ai __attribute__((target("neon"))) uint32x2_t vzip1_u32(uint32x2_t __p0, uint32x2_t __p1) {
66269	68736	uint32x2_t __ret;
66270		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
66271		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	68737	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	68738	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
66272	68739	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2);
66273		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	68740	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
66274	68741	return __ret;
66275	68742	}
66276	68743	#endif
...	...	@@ -66284,10 +68751,10 @@ __ai __attribute__((target("neon"))) uint16x4_t vzip1_u16(uint16x4_t __p0, uint1
66284	68751	#else
66285	68752	__ai __attribute__((target("neon"))) uint16x4_t vzip1_u16(uint16x4_t __p0, uint16x4_t __p1) {
66286	68753	uint16x4_t __ret;
66287		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
66288		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	68754	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	68755	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
66289	68756	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 4, 1, 5);
66290		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	68757	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
66291	68758	return __ret;
66292	68759	}
66293	68760	#endif
...	...	@@ -66301,10 +68768,10 @@ __ai __attribute__((target("neon"))) int8x8_t vzip1_s8(int8x8_t __p0, int8x8_t _
66301	68768	#else
66302	68769	__ai __attribute__((target("neon"))) int8x8_t vzip1_s8(int8x8_t __p0, int8x8_t __p1) {
66303	68770	int8x8_t __ret;
66304		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
66305		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	68771	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	68772	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
66306	68773	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 8, 1, 9, 2, 10, 3, 11);
66307		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	68774	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
66308	68775	return __ret;
66309	68776	}
66310	68777	#endif
...	...	@@ -66318,10 +68785,10 @@ __ai __attribute__((target("neon"))) float32x2_t vzip1_f32(float32x2_t __p0, flo
66318	68785	#else
66319	68786	__ai __attribute__((target("neon"))) float32x2_t vzip1_f32(float32x2_t __p0, float32x2_t __p1) {
66320	68787	float32x2_t __ret;
66321		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
66322		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	68788	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	68789	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
66323	68790	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2);
66324		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	68791	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
66325	68792	return __ret;
66326	68793	}
66327	68794	#endif
...	...	@@ -66335,10 +68802,27 @@ __ai __attribute__((target("neon"))) int32x2_t vzip1_s32(int32x2_t __p0, int32x2
66335	68802	#else
66336	68803	__ai __attribute__((target("neon"))) int32x2_t vzip1_s32(int32x2_t __p0, int32x2_t __p1) {
66337	68804	int32x2_t __ret;
66338		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
66339		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	68805	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	68806	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
66340	68807	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 2);
66341		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	68808	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
	68809	return __ret;
	68810	}
	68811	#endif
	68812
	68813	#ifdef __LITTLE_ENDIAN__
	68814	__ai __attribute__((target("neon"))) mfloat8x8_t vzip1_mf8(mfloat8x8_t __p0, mfloat8x8_t __p1) {
	68815	mfloat8x8_t __ret;
	68816	__ret = __builtin_shufflevector(__p0, __p1, 0, 8, 1, 9, 2, 10, 3, 11);
	68817	return __ret;
	68818	}
	68819	#else
	68820	__ai __attribute__((target("neon"))) mfloat8x8_t vzip1_mf8(mfloat8x8_t __p0, mfloat8x8_t __p1) {
	68821	mfloat8x8_t __ret;
	68822	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	68823	mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	68824	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 8, 1, 9, 2, 10, 3, 11);
	68825	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
66342	68826	return __ret;
66343	68827	}
66344	68828	#endif
...	...	@@ -66352,10 +68836,10 @@ __ai __attribute__((target("neon"))) int16x4_t vzip1_s16(int16x4_t __p0, int16x4
66352	68836	#else
66353	68837	__ai __attribute__((target("neon"))) int16x4_t vzip1_s16(int16x4_t __p0, int16x4_t __p1) {
66354	68838	int16x4_t __ret;
66355		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
66356		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	68839	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	68840	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
66357	68841	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 4, 1, 5);
66358		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	68842	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
66359	68843	return __ret;
66360	68844	}
66361	68845	#endif
...	...	@@ -66369,10 +68853,10 @@ __ai __attribute__((target("neon"))) float16x8_t vzip1q_f16(float16x8_t __p0, fl
66369	68853	#else
66370	68854	__ai __attribute__((target("neon"))) float16x8_t vzip1q_f16(float16x8_t __p0, float16x8_t __p1) {
66371	68855	float16x8_t __ret;
66372		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
66373		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	68856	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	68857	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
66374	68858	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 8, 1, 9, 2, 10, 3, 11);
66375		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	68859	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
66376	68860	return __ret;
66377	68861	}
66378	68862	#endif
...	...	@@ -66386,10 +68870,10 @@ __ai __attribute__((target("neon"))) float16x4_t vzip1_f16(float16x4_t __p0, flo
66386	68870	#else
66387	68871	__ai __attribute__((target("neon"))) float16x4_t vzip1_f16(float16x4_t __p0, float16x4_t __p1) {
66388	68872	float16x4_t __ret;
66389		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
66390		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	68873	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	68874	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
66391	68875	__ret = __builtin_shufflevector(__rev0, __rev1, 0, 4, 1, 5);
66392		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	68876	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
66393	68877	return __ret;
66394	68878	}
66395	68879	#endif
...	...	@@ -66403,10 +68887,10 @@ __ai __attribute__((target("neon"))) poly8x8_t vzip2_p8(poly8x8_t __p0, poly8x8_
66403	68887	#else
66404	68888	__ai __attribute__((target("neon"))) poly8x8_t vzip2_p8(poly8x8_t __p0, poly8x8_t __p1) {
66405	68889	poly8x8_t __ret;
66406		poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
66407		poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	68890	poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	68891	poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
66408	68892	__ret = __builtin_shufflevector(__rev0, __rev1, 4, 12, 5, 13, 6, 14, 7, 15);
66409		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	68893	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
66410	68894	return __ret;
66411	68895	}
66412	68896	#endif
...	...	@@ -66420,10 +68904,10 @@ __ai __attribute__((target("neon"))) poly16x4_t vzip2_p16(poly16x4_t __p0, poly1
66420	68904	#else
66421	68905	__ai __attribute__((target("neon"))) poly16x4_t vzip2_p16(poly16x4_t __p0, poly16x4_t __p1) {
66422	68906	poly16x4_t __ret;
66423		poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
66424		poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	68907	poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	68908	poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
66425	68909	__ret = __builtin_shufflevector(__rev0, __rev1, 2, 6, 3, 7);
66426		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	68910	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
66427	68911	return __ret;
66428	68912	}
66429	68913	#endif
...	...	@@ -66437,10 +68921,10 @@ __ai __attribute__((target("neon"))) poly8x16_t vzip2q_p8(poly8x16_t __p0, poly8
66437	68921	#else
66438	68922	__ai __attribute__((target("neon"))) poly8x16_t vzip2q_p8(poly8x16_t __p0, poly8x16_t __p1) {
66439	68923	poly8x16_t __ret;
66440		poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
66441		poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	68924	poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	68925	poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
66442	68926	__ret = __builtin_shufflevector(__rev0, __rev1, 8, 24, 9, 25, 10, 26, 11, 27, 12, 28, 13, 29, 14, 30, 15, 31);
66443		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	68927	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
66444	68928	return __ret;
66445	68929	}
66446	68930	#endif
...	...	@@ -66454,10 +68938,10 @@ __ai __attribute__((target("neon"))) poly64x2_t vzip2q_p64(poly64x2_t __p0, poly
66454	68938	#else
66455	68939	__ai __attribute__((target("neon"))) poly64x2_t vzip2q_p64(poly64x2_t __p0, poly64x2_t __p1) {
66456	68940	poly64x2_t __ret;
66457		poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
66458		poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	68941	poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	68942	poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
66459	68943	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3);
66460		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	68944	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
66461	68945	return __ret;
66462	68946	}
66463	68947	#endif
...	...	@@ -66471,10 +68955,10 @@ __ai __attribute__((target("neon"))) poly16x8_t vzip2q_p16(poly16x8_t __p0, poly
66471	68955	#else
66472	68956	__ai __attribute__((target("neon"))) poly16x8_t vzip2q_p16(poly16x8_t __p0, poly16x8_t __p1) {
66473	68957	poly16x8_t __ret;
66474		poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
66475		poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	68958	poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	68959	poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
66476	68960	__ret = __builtin_shufflevector(__rev0, __rev1, 4, 12, 5, 13, 6, 14, 7, 15);
66477		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	68961	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
66478	68962	return __ret;
66479	68963	}
66480	68964	#endif
...	...	@@ -66488,10 +68972,10 @@ __ai __attribute__((target("neon"))) uint8x16_t vzip2q_u8(uint8x16_t __p0, uint8
66488	68972	#else
66489	68973	__ai __attribute__((target("neon"))) uint8x16_t vzip2q_u8(uint8x16_t __p0, uint8x16_t __p1) {
66490	68974	uint8x16_t __ret;
66491		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
66492		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	68975	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	68976	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
66493	68977	__ret = __builtin_shufflevector(__rev0, __rev1, 8, 24, 9, 25, 10, 26, 11, 27, 12, 28, 13, 29, 14, 30, 15, 31);
66494		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	68978	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
66495	68979	return __ret;
66496	68980	}
66497	68981	#endif
...	...	@@ -66505,10 +68989,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vzip2q_u32(uint32x4_t __p0, uint
66505	68989	#else
66506	68990	__ai __attribute__((target("neon"))) uint32x4_t vzip2q_u32(uint32x4_t __p0, uint32x4_t __p1) {
66507	68991	uint32x4_t __ret;
66508		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
66509		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	68992	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	68993	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
66510	68994	__ret = __builtin_shufflevector(__rev0, __rev1, 2, 6, 3, 7);
66511		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	68995	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
66512	68996	return __ret;
66513	68997	}
66514	68998	#endif
...	...	@@ -66522,10 +69006,10 @@ __ai __attribute__((target("neon"))) uint64x2_t vzip2q_u64(uint64x2_t __p0, uint
66522	69006	#else
66523	69007	__ai __attribute__((target("neon"))) uint64x2_t vzip2q_u64(uint64x2_t __p0, uint64x2_t __p1) {
66524	69008	uint64x2_t __ret;
66525		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
66526		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	69009	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	69010	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
66527	69011	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3);
66528		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	69012	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
66529	69013	return __ret;
66530	69014	}
66531	69015	#endif
...	...	@@ -66539,10 +69023,10 @@ __ai __attribute__((target("neon"))) uint16x8_t vzip2q_u16(uint16x8_t __p0, uint
66539	69023	#else
66540	69024	__ai __attribute__((target("neon"))) uint16x8_t vzip2q_u16(uint16x8_t __p0, uint16x8_t __p1) {
66541	69025	uint16x8_t __ret;
66542		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
66543		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	69026	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	69027	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
66544	69028	__ret = __builtin_shufflevector(__rev0, __rev1, 4, 12, 5, 13, 6, 14, 7, 15);
66545		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	69029	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
66546	69030	return __ret;
66547	69031	}
66548	69032	#endif
...	...	@@ -66556,10 +69040,10 @@ __ai __attribute__((target("neon"))) int8x16_t vzip2q_s8(int8x16_t __p0, int8x16
66556	69040	#else
66557	69041	__ai __attribute__((target("neon"))) int8x16_t vzip2q_s8(int8x16_t __p0, int8x16_t __p1) {
66558	69042	int8x16_t __ret;
66559		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
66560		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	69043	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	69044	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
66561	69045	__ret = __builtin_shufflevector(__rev0, __rev1, 8, 24, 9, 25, 10, 26, 11, 27, 12, 28, 13, 29, 14, 30, 15, 31);
66562		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	69046	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
66563	69047	return __ret;
66564	69048	}
66565	69049	#endif
...	...	@@ -66573,10 +69057,10 @@ __ai __attribute__((target("neon"))) float64x2_t vzip2q_f64(float64x2_t __p0, fl
66573	69057	#else
66574	69058	__ai __attribute__((target("neon"))) float64x2_t vzip2q_f64(float64x2_t __p0, float64x2_t __p1) {
66575	69059	float64x2_t __ret;
66576		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
66577		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	69060	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	69061	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
66578	69062	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3);
66579		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	69063	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
66580	69064	return __ret;
66581	69065	}
66582	69066	#endif
...	...	@@ -66590,10 +69074,10 @@ __ai __attribute__((target("neon"))) float32x4_t vzip2q_f32(float32x4_t __p0, fl
66590	69074	#else
66591	69075	__ai __attribute__((target("neon"))) float32x4_t vzip2q_f32(float32x4_t __p0, float32x4_t __p1) {
66592	69076	float32x4_t __ret;
66593		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
66594		float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	69077	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	69078	float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
66595	69079	__ret = __builtin_shufflevector(__rev0, __rev1, 2, 6, 3, 7);
66596		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	69080	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
66597	69081	return __ret;
66598	69082	}
66599	69083	#endif
...	...	@@ -66607,10 +69091,10 @@ __ai __attribute__((target("neon"))) int32x4_t vzip2q_s32(int32x4_t __p0, int32x
66607	69091	#else
66608	69092	__ai __attribute__((target("neon"))) int32x4_t vzip2q_s32(int32x4_t __p0, int32x4_t __p1) {
66609	69093	int32x4_t __ret;
66610		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
66611		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	69094	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	69095	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
66612	69096	__ret = __builtin_shufflevector(__rev0, __rev1, 2, 6, 3, 7);
66613		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	69097	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
66614	69098	return __ret;
66615	69099	}
66616	69100	#endif
...	...	@@ -66624,10 +69108,27 @@ __ai __attribute__((target("neon"))) int64x2_t vzip2q_s64(int64x2_t __p0, int64x
66624	69108	#else
66625	69109	__ai __attribute__((target("neon"))) int64x2_t vzip2q_s64(int64x2_t __p0, int64x2_t __p1) {
66626	69110	int64x2_t __ret;
66627		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
66628		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	69111	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	69112	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
66629	69113	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3);
66630		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	69114	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
	69115	return __ret;
	69116	}
	69117	#endif
	69118
	69119	#ifdef __LITTLE_ENDIAN__
	69120	__ai __attribute__((target("neon"))) mfloat8x16_t vzip2q_mf8(mfloat8x16_t __p0, mfloat8x16_t __p1) {
	69121	mfloat8x16_t __ret;
	69122	__ret = __builtin_shufflevector(__p0, __p1, 8, 24, 9, 25, 10, 26, 11, 27, 12, 28, 13, 29, 14, 30, 15, 31);
	69123	return __ret;
	69124	}
	69125	#else
	69126	__ai __attribute__((target("neon"))) mfloat8x16_t vzip2q_mf8(mfloat8x16_t __p0, mfloat8x16_t __p1) {
	69127	mfloat8x16_t __ret;
	69128	mfloat8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	69129	mfloat8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	69130	__ret = __builtin_shufflevector(__rev0, __rev1, 8, 24, 9, 25, 10, 26, 11, 27, 12, 28, 13, 29, 14, 30, 15, 31);
	69131	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
66631	69132	return __ret;
66632	69133	}
66633	69134	#endif
...	...	@@ -66641,10 +69142,10 @@ __ai __attribute__((target("neon"))) int16x8_t vzip2q_s16(int16x8_t __p0, int16x
66641	69142	#else
66642	69143	__ai __attribute__((target("neon"))) int16x8_t vzip2q_s16(int16x8_t __p0, int16x8_t __p1) {
66643	69144	int16x8_t __ret;
66644		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
66645		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	69145	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	69146	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
66646	69147	__ret = __builtin_shufflevector(__rev0, __rev1, 4, 12, 5, 13, 6, 14, 7, 15);
66647		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	69148	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
66648	69149	return __ret;
66649	69150	}
66650	69151	#endif
...	...	@@ -66658,10 +69159,10 @@ __ai __attribute__((target("neon"))) uint8x8_t vzip2_u8(uint8x8_t __p0, uint8x8_
66658	69159	#else
66659	69160	__ai __attribute__((target("neon"))) uint8x8_t vzip2_u8(uint8x8_t __p0, uint8x8_t __p1) {
66660	69161	uint8x8_t __ret;
66661		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
66662		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	69162	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	69163	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
66663	69164	__ret = __builtin_shufflevector(__rev0, __rev1, 4, 12, 5, 13, 6, 14, 7, 15);
66664		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	69165	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
66665	69166	return __ret;
66666	69167	}
66667	69168	#endif
...	...	@@ -66675,10 +69176,10 @@ __ai __attribute__((target("neon"))) uint32x2_t vzip2_u32(uint32x2_t __p0, uint3
66675	69176	#else
66676	69177	__ai __attribute__((target("neon"))) uint32x2_t vzip2_u32(uint32x2_t __p0, uint32x2_t __p1) {
66677	69178	uint32x2_t __ret;
66678		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
66679		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	69179	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	69180	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
66680	69181	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3);
66681		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	69182	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
66682	69183	return __ret;
66683	69184	}
66684	69185	#endif
...	...	@@ -66692,10 +69193,10 @@ __ai __attribute__((target("neon"))) uint16x4_t vzip2_u16(uint16x4_t __p0, uint1
66692	69193	#else
66693	69194	__ai __attribute__((target("neon"))) uint16x4_t vzip2_u16(uint16x4_t __p0, uint16x4_t __p1) {
66694	69195	uint16x4_t __ret;
66695		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
66696		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	69196	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	69197	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
66697	69198	__ret = __builtin_shufflevector(__rev0, __rev1, 2, 6, 3, 7);
66698		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	69199	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
66699	69200	return __ret;
66700	69201	}
66701	69202	#endif
...	...	@@ -66709,10 +69210,10 @@ __ai __attribute__((target("neon"))) int8x8_t vzip2_s8(int8x8_t __p0, int8x8_t _
66709	69210	#else
66710	69211	__ai __attribute__((target("neon"))) int8x8_t vzip2_s8(int8x8_t __p0, int8x8_t __p1) {
66711	69212	int8x8_t __ret;
66712		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
66713		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	69213	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	69214	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
66714	69215	__ret = __builtin_shufflevector(__rev0, __rev1, 4, 12, 5, 13, 6, 14, 7, 15);
66715		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	69216	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
66716	69217	return __ret;
66717	69218	}
66718	69219	#endif
...	...	@@ -66726,10 +69227,10 @@ __ai __attribute__((target("neon"))) float32x2_t vzip2_f32(float32x2_t __p0, flo
66726	69227	#else
66727	69228	__ai __attribute__((target("neon"))) float32x2_t vzip2_f32(float32x2_t __p0, float32x2_t __p1) {
66728	69229	float32x2_t __ret;
66729		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
66730		float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	69230	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	69231	float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
66731	69232	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3);
66732		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	69233	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
66733	69234	return __ret;
66734	69235	}
66735	69236	#endif
...	...	@@ -66743,10 +69244,27 @@ __ai __attribute__((target("neon"))) int32x2_t vzip2_s32(int32x2_t __p0, int32x2
66743	69244	#else
66744	69245	__ai __attribute__((target("neon"))) int32x2_t vzip2_s32(int32x2_t __p0, int32x2_t __p1) {
66745	69246	int32x2_t __ret;
66746		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
66747		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	69247	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	69248	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
66748	69249	__ret = __builtin_shufflevector(__rev0, __rev1, 1, 3);
66749		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	69250	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
	69251	return __ret;
	69252	}
	69253	#endif
	69254
	69255	#ifdef __LITTLE_ENDIAN__
	69256	__ai __attribute__((target("neon"))) mfloat8x8_t vzip2_mf8(mfloat8x8_t __p0, mfloat8x8_t __p1) {
	69257	mfloat8x8_t __ret;
	69258	__ret = __builtin_shufflevector(__p0, __p1, 4, 12, 5, 13, 6, 14, 7, 15);
	69259	return __ret;
	69260	}
	69261	#else
	69262	__ai __attribute__((target("neon"))) mfloat8x8_t vzip2_mf8(mfloat8x8_t __p0, mfloat8x8_t __p1) {
	69263	mfloat8x8_t __ret;
	69264	mfloat8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	69265	mfloat8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	69266	__ret = __builtin_shufflevector(__rev0, __rev1, 4, 12, 5, 13, 6, 14, 7, 15);
	69267	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
66750	69268	return __ret;
66751	69269	}
66752	69270	#endif
...	...	@@ -66760,10 +69278,10 @@ __ai __attribute__((target("neon"))) int16x4_t vzip2_s16(int16x4_t __p0, int16x4
66760	69278	#else
66761	69279	__ai __attribute__((target("neon"))) int16x4_t vzip2_s16(int16x4_t __p0, int16x4_t __p1) {
66762	69280	int16x4_t __ret;
66763		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
66764		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	69281	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	69282	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
66765	69283	__ret = __builtin_shufflevector(__rev0, __rev1, 2, 6, 3, 7);
66766		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	69284	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
66767	69285	return __ret;
66768	69286	}
66769	69287	#endif
...	...	@@ -66777,10 +69295,10 @@ __ai __attribute__((target("neon"))) float16x8_t vzip2q_f16(float16x8_t __p0, fl
66777	69295	#else
66778	69296	__ai __attribute__((target("neon"))) float16x8_t vzip2q_f16(float16x8_t __p0, float16x8_t __p1) {
66779	69297	float16x8_t __ret;
66780		float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
66781		float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	69298	float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	69299	float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
66782	69300	__ret = __builtin_shufflevector(__rev0, __rev1, 4, 12, 5, 13, 6, 14, 7, 15);
66783		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	69301	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
66784	69302	return __ret;
66785	69303	}
66786	69304	#endif
...	...	@@ -66794,10 +69312,10 @@ __ai __attribute__((target("neon"))) float16x4_t vzip2_f16(float16x4_t __p0, flo
66794	69312	#else
66795	69313	__ai __attribute__((target("neon"))) float16x4_t vzip2_f16(float16x4_t __p0, float16x4_t __p1) {
66796	69314	float16x4_t __ret;
66797		float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
66798		float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	69315	float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	69316	float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
66799	69317	__ret = __builtin_shufflevector(__rev0, __rev1, 2, 6, 3, 7);
66800		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	69318	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
66801	69319	return __ret;
66802	69320	}
66803	69321	#endif
...	...	@@ -66805,23 +69323,23 @@ __ai __attribute__((target("neon"))) float16x4_t vzip2_f16(float16x4_t __p0, flo
66805	69323	#define vldap1_lane_p64(__p0, __p1, __p2) __extension__ ({ \
66806	69324	poly64x1_t __ret; \
66807	69325	poly64x1_t __s1 = __p1; \
66808		__ret = (poly64x1_t) __builtin_neon_vldap1_lane_p64(__p0, (int8x8_t)__s1, __p2, 6); \
	69326	__ret = __builtin_bit_cast(poly64x1_t, __builtin_neon_vldap1_lane_p64(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 6)); \
66809	69327	__ret; \
66810	69328	})
66811	69329	#ifdef __LITTLE_ENDIAN__
66812	69330	#define vldap1q_lane_p64(__p0, __p1, __p2) __extension__ ({ \
66813	69331	poly64x2_t __ret; \
66814	69332	poly64x2_t __s1 = __p1; \
66815		__ret = (poly64x2_t) __builtin_neon_vldap1q_lane_p64(__p0, (int8x16_t)__s1, __p2, 38); \
	69333	__ret = __builtin_bit_cast(poly64x2_t, __builtin_neon_vldap1q_lane_p64(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 38)); \
66816	69334	__ret; \
66817	69335	})
66818	69336	#else
66819	69337	#define vldap1q_lane_p64(__p0, __p1, __p2) __extension__ ({ \
66820	69338	poly64x2_t __ret; \
66821	69339	poly64x2_t __s1 = __p1; \
66822		poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
66823		__ret = (poly64x2_t) __builtin_neon_vldap1q_lane_p64(__p0, (int8x16_t)__rev1, __p2, 38); \
66824		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	69340	poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	69341	__ret = __builtin_bit_cast(poly64x2_t, __builtin_neon_vldap1q_lane_p64(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 38)); \
	69342	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
66825	69343	__ret; \
66826	69344	})
66827	69345	#endif
...	...	@@ -66830,16 +69348,16 @@ __ai __attribute__((target("neon"))) float16x4_t vzip2_f16(float16x4_t __p0, flo
66830	69348	#define vldap1q_lane_u64(__p0, __p1, __p2) __extension__ ({ \
66831	69349	uint64x2_t __ret; \
66832	69350	uint64x2_t __s1 = __p1; \
66833		__ret = (uint64x2_t) __builtin_neon_vldap1q_lane_u64(__p0, (int8x16_t)__s1, __p2, 51); \
	69351	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vldap1q_lane_u64(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 51)); \
66834	69352	__ret; \
66835	69353	})
66836	69354	#else
66837	69355	#define vldap1q_lane_u64(__p0, __p1, __p2) __extension__ ({ \
66838	69356	uint64x2_t __ret; \
66839	69357	uint64x2_t __s1 = __p1; \
66840		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
66841		__ret = (uint64x2_t) __builtin_neon_vldap1q_lane_u64(__p0, (int8x16_t)__rev1, __p2, 51); \
66842		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	69358	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	69359	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vldap1q_lane_u64(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 51)); \
	69360	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
66843	69361	__ret; \
66844	69362	})
66845	69363	#endif
...	...	@@ -66848,16 +69366,16 @@ __ai __attribute__((target("neon"))) float16x4_t vzip2_f16(float16x4_t __p0, flo
66848	69366	#define vldap1q_lane_f64(__p0, __p1, __p2) __extension__ ({ \
66849	69367	float64x2_t __ret; \
66850	69368	float64x2_t __s1 = __p1; \
66851		__ret = (float64x2_t) __builtin_neon_vldap1q_lane_f64(__p0, (int8x16_t)__s1, __p2, 42); \
	69369	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vldap1q_lane_f64(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 42)); \
66852	69370	__ret; \
66853	69371	})
66854	69372	#else
66855	69373	#define vldap1q_lane_f64(__p0, __p1, __p2) __extension__ ({ \
66856	69374	float64x2_t __ret; \
66857	69375	float64x2_t __s1 = __p1; \
66858		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
66859		__ret = (float64x2_t) __builtin_neon_vldap1q_lane_f64(__p0, (int8x16_t)__rev1, __p2, 42); \
66860		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	69376	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	69377	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vldap1q_lane_f64(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 42)); \
	69378	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
66861	69379	__ret; \
66862	69380	})
66863	69381	#endif
...	...	@@ -66866,16 +69384,16 @@ __ai __attribute__((target("neon"))) float16x4_t vzip2_f16(float16x4_t __p0, flo
66866	69384	#define vldap1q_lane_s64(__p0, __p1, __p2) __extension__ ({ \
66867	69385	int64x2_t __ret; \
66868	69386	int64x2_t __s1 = __p1; \
66869		__ret = (int64x2_t) __builtin_neon_vldap1q_lane_s64(__p0, (int8x16_t)__s1, __p2, 35); \
	69387	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vldap1q_lane_s64(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 35)); \
66870	69388	__ret; \
66871	69389	})
66872	69390	#else
66873	69391	#define vldap1q_lane_s64(__p0, __p1, __p2) __extension__ ({ \
66874	69392	int64x2_t __ret; \
66875	69393	int64x2_t __s1 = __p1; \
66876		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
66877		__ret = (int64x2_t) __builtin_neon_vldap1q_lane_s64(__p0, (int8x16_t)__rev1, __p2, 35); \
66878		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	69394	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	69395	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vldap1q_lane_s64(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 35)); \
	69396	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
66879	69397	__ret; \
66880	69398	})
66881	69399	#endif
...	...	@@ -66883,103 +69401,103 @@ __ai __attribute__((target("neon"))) float16x4_t vzip2_f16(float16x4_t __p0, flo
66883	69401	#define vldap1_lane_u64(__p0, __p1, __p2) __extension__ ({ \
66884	69402	uint64x1_t __ret; \
66885	69403	uint64x1_t __s1 = __p1; \
66886		__ret = (uint64x1_t) __builtin_neon_vldap1_lane_u64(__p0, (int8x8_t)__s1, __p2, 19); \
	69404	__ret = __builtin_bit_cast(uint64x1_t, __builtin_neon_vldap1_lane_u64(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 19)); \
66887	69405	__ret; \
66888	69406	})
66889	69407	#define vldap1_lane_f64(__p0, __p1, __p2) __extension__ ({ \
66890	69408	float64x1_t __ret; \
66891	69409	float64x1_t __s1 = __p1; \
66892		__ret = (float64x1_t) __builtin_neon_vldap1_lane_f64(__p0, (int8x8_t)__s1, __p2, 10); \
	69410	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vldap1_lane_f64(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 10)); \
66893	69411	__ret; \
66894	69412	})
66895	69413	#define vldap1_lane_s64(__p0, __p1, __p2) __extension__ ({ \
66896	69414	int64x1_t __ret; \
66897	69415	int64x1_t __s1 = __p1; \
66898		__ret = (int64x1_t) __builtin_neon_vldap1_lane_s64(__p0, (int8x8_t)__s1, __p2, 3); \
	69416	__ret = __builtin_bit_cast(int64x1_t, __builtin_neon_vldap1_lane_s64(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 3)); \
66899	69417	__ret; \
66900	69418	})
66901	69419	#define vstl1_lane_p64(__p0, __p1, __p2) __extension__ ({ \
66902	69420	poly64x1_t __s1 = __p1; \
66903		__builtin_neon_vstl1_lane_p64(__p0, (int8x8_t)__s1, __p2, 6); \
	69421	__builtin_neon_vstl1_lane_p64(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 6); \
66904	69422	})
66905	69423	#ifdef __LITTLE_ENDIAN__
66906	69424	#define vstl1q_lane_p64(__p0, __p1, __p2) __extension__ ({ \
66907	69425	poly64x2_t __s1 = __p1; \
66908		__builtin_neon_vstl1q_lane_p64(__p0, (int8x16_t)__s1, __p2, 38); \
	69426	__builtin_neon_vstl1q_lane_p64(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 38); \
66909	69427	})
66910	69428	#else
66911	69429	#define vstl1q_lane_p64(__p0, __p1, __p2) __extension__ ({ \
66912	69430	poly64x2_t __s1 = __p1; \
66913		poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
66914		__builtin_neon_vstl1q_lane_p64(__p0, (int8x16_t)__rev1, __p2, 38); \
	69431	poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	69432	__builtin_neon_vstl1q_lane_p64(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 38); \
66915	69433	})
66916	69434	#endif
66917	69435
66918	69436	#ifdef __LITTLE_ENDIAN__
66919	69437	#define vstl1q_lane_u64(__p0, __p1, __p2) __extension__ ({ \
66920	69438	uint64x2_t __s1 = __p1; \
66921		__builtin_neon_vstl1q_lane_u64(__p0, (int8x16_t)__s1, __p2, 51); \
	69439	__builtin_neon_vstl1q_lane_u64(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 51); \
66922	69440	})
66923	69441	#else
66924	69442	#define vstl1q_lane_u64(__p0, __p1, __p2) __extension__ ({ \
66925	69443	uint64x2_t __s1 = __p1; \
66926		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
66927		__builtin_neon_vstl1q_lane_u64(__p0, (int8x16_t)__rev1, __p2, 51); \
	69444	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	69445	__builtin_neon_vstl1q_lane_u64(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 51); \
66928	69446	})
66929	69447	#endif
66930	69448
66931	69449	#ifdef __LITTLE_ENDIAN__
66932	69450	#define vstl1q_lane_f64(__p0, __p1, __p2) __extension__ ({ \
66933	69451	float64x2_t __s1 = __p1; \
66934		__builtin_neon_vstl1q_lane_f64(__p0, (int8x16_t)__s1, __p2, 42); \
	69452	__builtin_neon_vstl1q_lane_f64(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 42); \
66935	69453	})
66936	69454	#else
66937	69455	#define vstl1q_lane_f64(__p0, __p1, __p2) __extension__ ({ \
66938	69456	float64x2_t __s1 = __p1; \
66939		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
66940		__builtin_neon_vstl1q_lane_f64(__p0, (int8x16_t)__rev1, __p2, 42); \
	69457	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	69458	__builtin_neon_vstl1q_lane_f64(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 42); \
66941	69459	})
66942	69460	#endif
66943	69461
66944	69462	#ifdef __LITTLE_ENDIAN__
66945	69463	#define vstl1q_lane_s64(__p0, __p1, __p2) __extension__ ({ \
66946	69464	int64x2_t __s1 = __p1; \
66947		__builtin_neon_vstl1q_lane_s64(__p0, (int8x16_t)__s1, __p2, 35); \
	69465	__builtin_neon_vstl1q_lane_s64(__p0, __builtin_bit_cast(int8x16_t, __s1), __p2, 35); \
66948	69466	})
66949	69467	#else
66950	69468	#define vstl1q_lane_s64(__p0, __p1, __p2) __extension__ ({ \
66951	69469	int64x2_t __s1 = __p1; \
66952		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
66953		__builtin_neon_vstl1q_lane_s64(__p0, (int8x16_t)__rev1, __p2, 35); \
	69470	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	69471	__builtin_neon_vstl1q_lane_s64(__p0, __builtin_bit_cast(int8x16_t, __rev1), __p2, 35); \
66954	69472	})
66955	69473	#endif
66956	69474
66957	69475	#define vstl1_lane_u64(__p0, __p1, __p2) __extension__ ({ \
66958	69476	uint64x1_t __s1 = __p1; \
66959		__builtin_neon_vstl1_lane_u64(__p0, (int8x8_t)__s1, __p2, 19); \
	69477	__builtin_neon_vstl1_lane_u64(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 19); \
66960	69478	})
66961	69479	#define vstl1_lane_f64(__p0, __p1, __p2) __extension__ ({ \
66962	69480	float64x1_t __s1 = __p1; \
66963		__builtin_neon_vstl1_lane_f64(__p0, (int8x8_t)__s1, __p2, 10); \
	69481	__builtin_neon_vstl1_lane_f64(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 10); \
66964	69482	})
66965	69483	#define vstl1_lane_s64(__p0, __p1, __p2) __extension__ ({ \
66966	69484	int64x1_t __s1 = __p1; \
66967		__builtin_neon_vstl1_lane_s64(__p0, (int8x8_t)__s1, __p2, 3); \
	69485	__builtin_neon_vstl1_lane_s64(__p0, __builtin_bit_cast(int8x8_t, __s1), __p2, 3); \
66968	69486	})
66969	69487	#ifdef __LITTLE_ENDIAN__
66970	69488	__ai __attribute__((target("sha3,neon"))) uint8x16_t vbcaxq_u8(uint8x16_t __p0, uint8x16_t __p1, uint8x16_t __p2) {
66971	69489	uint8x16_t __ret;
66972		__ret = (uint8x16_t) __builtin_neon_vbcaxq_u8((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 48);
	69490	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vbcaxq_u8(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 48));
66973	69491	return __ret;
66974	69492	}
66975	69493	#else
66976	69494	__ai __attribute__((target("sha3,neon"))) uint8x16_t vbcaxq_u8(uint8x16_t __p0, uint8x16_t __p1, uint8x16_t __p2) {
66977	69495	uint8x16_t __ret;
66978		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
66979		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
66980		uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
66981		__ret = (uint8x16_t) __builtin_neon_vbcaxq_u8((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 48);
66982		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	69496	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	69497	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	69498	uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	69499	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_vbcaxq_u8(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 48));
	69500	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
66983	69501	return __ret;
66984	69502	}
66985	69503	#endif
...	...	@@ -66987,17 +69505,17 @@ __ai __attribute__((target("sha3,neon"))) uint8x16_t vbcaxq_u8(uint8x16_t __p0,
66987	69505	#ifdef __LITTLE_ENDIAN__
66988	69506	__ai __attribute__((target("sha3,neon"))) uint32x4_t vbcaxq_u32(uint32x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
66989	69507	uint32x4_t __ret;
66990		__ret = (uint32x4_t) __builtin_neon_vbcaxq_u32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 50);
	69508	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vbcaxq_u32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 50));
66991	69509	return __ret;
66992	69510	}
66993	69511	#else
66994	69512	__ai __attribute__((target("sha3,neon"))) uint32x4_t vbcaxq_u32(uint32x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
66995	69513	uint32x4_t __ret;
66996		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
66997		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
66998		uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
66999		__ret = (uint32x4_t) __builtin_neon_vbcaxq_u32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 50);
67000		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	69514	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	69515	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	69516	uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
	69517	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vbcaxq_u32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 50));
	69518	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
67001	69519	return __ret;
67002	69520	}
67003	69521	#endif
...	...	@@ -67005,17 +69523,17 @@ __ai __attribute__((target("sha3,neon"))) uint32x4_t vbcaxq_u32(uint32x4_t __p0,
67005	69523	#ifdef __LITTLE_ENDIAN__
67006	69524	__ai __attribute__((target("sha3,neon"))) uint64x2_t vbcaxq_u64(uint64x2_t __p0, uint64x2_t __p1, uint64x2_t __p2) {
67007	69525	uint64x2_t __ret;
67008		__ret = (uint64x2_t) __builtin_neon_vbcaxq_u64((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 51);
	69526	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vbcaxq_u64(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 51));
67009	69527	return __ret;
67010	69528	}
67011	69529	#else
67012	69530	__ai __attribute__((target("sha3,neon"))) uint64x2_t vbcaxq_u64(uint64x2_t __p0, uint64x2_t __p1, uint64x2_t __p2) {
67013	69531	uint64x2_t __ret;
67014		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
67015		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
67016		uint64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
67017		__ret = (uint64x2_t) __builtin_neon_vbcaxq_u64((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 51);
67018		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	69532	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	69533	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	69534	uint64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_64);
	69535	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vbcaxq_u64(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 51));
	69536	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
67019	69537	return __ret;
67020	69538	}
67021	69539	#endif
...	...	@@ -67023,17 +69541,17 @@ __ai __attribute__((target("sha3,neon"))) uint64x2_t vbcaxq_u64(uint64x2_t __p0,
67023	69541	#ifdef __LITTLE_ENDIAN__
67024	69542	__ai __attribute__((target("sha3,neon"))) uint16x8_t vbcaxq_u16(uint16x8_t __p0, uint16x8_t __p1, uint16x8_t __p2) {
67025	69543	uint16x8_t __ret;
67026		__ret = (uint16x8_t) __builtin_neon_vbcaxq_u16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 49);
	69544	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vbcaxq_u16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 49));
67027	69545	return __ret;
67028	69546	}
67029	69547	#else
67030	69548	__ai __attribute__((target("sha3,neon"))) uint16x8_t vbcaxq_u16(uint16x8_t __p0, uint16x8_t __p1, uint16x8_t __p2) {
67031	69549	uint16x8_t __ret;
67032		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
67033		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
67034		uint16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
67035		__ret = (uint16x8_t) __builtin_neon_vbcaxq_u16((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 49);
67036		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	69550	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	69551	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	69552	uint16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
	69553	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_vbcaxq_u16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 49));
	69554	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
67037	69555	return __ret;
67038	69556	}
67039	69557	#endif
...	...	@@ -67041,17 +69559,17 @@ __ai __attribute__((target("sha3,neon"))) uint16x8_t vbcaxq_u16(uint16x8_t __p0,
67041	69559	#ifdef __LITTLE_ENDIAN__
67042	69560	__ai __attribute__((target("sha3,neon"))) int8x16_t vbcaxq_s8(int8x16_t __p0, int8x16_t __p1, int8x16_t __p2) {
67043	69561	int8x16_t __ret;
67044		__ret = (int8x16_t) __builtin_neon_vbcaxq_s8((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 32);
	69562	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vbcaxq_s8(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 32));
67045	69563	return __ret;
67046	69564	}
67047	69565	#else
67048	69566	__ai __attribute__((target("sha3,neon"))) int8x16_t vbcaxq_s8(int8x16_t __p0, int8x16_t __p1, int8x16_t __p2) {
67049	69567	int8x16_t __ret;
67050		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
67051		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
67052		int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
67053		__ret = (int8x16_t) __builtin_neon_vbcaxq_s8((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 32);
67054		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	69568	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	69569	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	69570	int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	69571	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_vbcaxq_s8(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 32));
	69572	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
67055	69573	return __ret;
67056	69574	}
67057	69575	#endif
...	...	@@ -67059,17 +69577,17 @@ __ai __attribute__((target("sha3,neon"))) int8x16_t vbcaxq_s8(int8x16_t __p0, in
67059	69577	#ifdef __LITTLE_ENDIAN__
67060	69578	__ai __attribute__((target("sha3,neon"))) int32x4_t vbcaxq_s32(int32x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
67061	69579	int32x4_t __ret;
67062		__ret = (int32x4_t) __builtin_neon_vbcaxq_s32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 34);
	69580	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vbcaxq_s32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 34));
67063	69581	return __ret;
67064	69582	}
67065	69583	#else
67066	69584	__ai __attribute__((target("sha3,neon"))) int32x4_t vbcaxq_s32(int32x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
67067	69585	int32x4_t __ret;
67068		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
67069		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
67070		int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
67071		__ret = (int32x4_t) __builtin_neon_vbcaxq_s32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 34);
67072		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	69586	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	69587	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	69588	int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
	69589	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_vbcaxq_s32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 34));
	69590	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
67073	69591	return __ret;
67074	69592	}
67075	69593	#endif
...	...	@@ -67077,17 +69595,17 @@ __ai __attribute__((target("sha3,neon"))) int32x4_t vbcaxq_s32(int32x4_t __p0, i
67077	69595	#ifdef __LITTLE_ENDIAN__
67078	69596	__ai __attribute__((target("sha3,neon"))) int64x2_t vbcaxq_s64(int64x2_t __p0, int64x2_t __p1, int64x2_t __p2) {
67079	69597	int64x2_t __ret;
67080		__ret = (int64x2_t) __builtin_neon_vbcaxq_s64((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 35);
	69598	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vbcaxq_s64(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 35));
67081	69599	return __ret;
67082	69600	}
67083	69601	#else
67084	69602	__ai __attribute__((target("sha3,neon"))) int64x2_t vbcaxq_s64(int64x2_t __p0, int64x2_t __p1, int64x2_t __p2) {
67085	69603	int64x2_t __ret;
67086		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
67087		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
67088		int64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
67089		__ret = (int64x2_t) __builtin_neon_vbcaxq_s64((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 35);
67090		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	69604	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	69605	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	69606	int64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_64);
	69607	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_vbcaxq_s64(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 35));
	69608	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
67091	69609	return __ret;
67092	69610	}
67093	69611	#endif
...	...	@@ -67095,17 +69613,17 @@ __ai __attribute__((target("sha3,neon"))) int64x2_t vbcaxq_s64(int64x2_t __p0, i
67095	69613	#ifdef __LITTLE_ENDIAN__
67096	69614	__ai __attribute__((target("sha3,neon"))) int16x8_t vbcaxq_s16(int16x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
67097	69615	int16x8_t __ret;
67098		__ret = (int16x8_t) __builtin_neon_vbcaxq_s16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 33);
	69616	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vbcaxq_s16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 33));
67099	69617	return __ret;
67100	69618	}
67101	69619	#else
67102	69620	__ai __attribute__((target("sha3,neon"))) int16x8_t vbcaxq_s16(int16x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
67103	69621	int16x8_t __ret;
67104		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
67105		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
67106		int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
67107		__ret = (int16x8_t) __builtin_neon_vbcaxq_s16((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 33);
67108		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	69622	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	69623	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	69624	int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
	69625	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_vbcaxq_s16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 33));
	69626	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
67109	69627	return __ret;
67110	69628	}
67111	69629	#endif
...	...	@@ -67113,17 +69631,17 @@ __ai __attribute__((target("sha3,neon"))) int16x8_t vbcaxq_s16(int16x8_t __p0, i
67113	69631	#ifdef __LITTLE_ENDIAN__
67114	69632	__ai __attribute__((target("sha3,neon"))) uint8x16_t veor3q_u8(uint8x16_t __p0, uint8x16_t __p1, uint8x16_t __p2) {
67115	69633	uint8x16_t __ret;
67116		__ret = (uint8x16_t) __builtin_neon_veor3q_u8((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 48);
	69634	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_veor3q_u8(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 48));
67117	69635	return __ret;
67118	69636	}
67119	69637	#else
67120	69638	__ai __attribute__((target("sha3,neon"))) uint8x16_t veor3q_u8(uint8x16_t __p0, uint8x16_t __p1, uint8x16_t __p2) {
67121	69639	uint8x16_t __ret;
67122		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
67123		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
67124		uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
67125		__ret = (uint8x16_t) __builtin_neon_veor3q_u8((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 48);
67126		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	69640	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	69641	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	69642	uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	69643	__ret = __builtin_bit_cast(uint8x16_t, __builtin_neon_veor3q_u8(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 48));
	69644	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
67127	69645	return __ret;
67128	69646	}
67129	69647	#endif
...	...	@@ -67131,17 +69649,17 @@ __ai __attribute__((target("sha3,neon"))) uint8x16_t veor3q_u8(uint8x16_t __p0,
67131	69649	#ifdef __LITTLE_ENDIAN__
67132	69650	__ai __attribute__((target("sha3,neon"))) uint32x4_t veor3q_u32(uint32x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
67133	69651	uint32x4_t __ret;
67134		__ret = (uint32x4_t) __builtin_neon_veor3q_u32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 50);
	69652	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_veor3q_u32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 50));
67135	69653	return __ret;
67136	69654	}
67137	69655	#else
67138	69656	__ai __attribute__((target("sha3,neon"))) uint32x4_t veor3q_u32(uint32x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
67139	69657	uint32x4_t __ret;
67140		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
67141		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
67142		uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
67143		__ret = (uint32x4_t) __builtin_neon_veor3q_u32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 50);
67144		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	69658	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	69659	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	69660	uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
	69661	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_veor3q_u32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 50));
	69662	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
67145	69663	return __ret;
67146	69664	}
67147	69665	#endif
...	...	@@ -67149,17 +69667,17 @@ __ai __attribute__((target("sha3,neon"))) uint32x4_t veor3q_u32(uint32x4_t __p0,
67149	69667	#ifdef __LITTLE_ENDIAN__
67150	69668	__ai __attribute__((target("sha3,neon"))) uint64x2_t veor3q_u64(uint64x2_t __p0, uint64x2_t __p1, uint64x2_t __p2) {
67151	69669	uint64x2_t __ret;
67152		__ret = (uint64x2_t) __builtin_neon_veor3q_u64((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 51);
	69670	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_veor3q_u64(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 51));
67153	69671	return __ret;
67154	69672	}
67155	69673	#else
67156	69674	__ai __attribute__((target("sha3,neon"))) uint64x2_t veor3q_u64(uint64x2_t __p0, uint64x2_t __p1, uint64x2_t __p2) {
67157	69675	uint64x2_t __ret;
67158		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
67159		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
67160		uint64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
67161		__ret = (uint64x2_t) __builtin_neon_veor3q_u64((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 51);
67162		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	69676	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	69677	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	69678	uint64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_64);
	69679	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_veor3q_u64(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 51));
	69680	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
67163	69681	return __ret;
67164	69682	}
67165	69683	#endif
...	...	@@ -67167,17 +69685,17 @@ __ai __attribute__((target("sha3,neon"))) uint64x2_t veor3q_u64(uint64x2_t __p0,
67167	69685	#ifdef __LITTLE_ENDIAN__
67168	69686	__ai __attribute__((target("sha3,neon"))) uint16x8_t veor3q_u16(uint16x8_t __p0, uint16x8_t __p1, uint16x8_t __p2) {
67169	69687	uint16x8_t __ret;
67170		__ret = (uint16x8_t) __builtin_neon_veor3q_u16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 49);
	69688	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_veor3q_u16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 49));
67171	69689	return __ret;
67172	69690	}
67173	69691	#else
67174	69692	__ai __attribute__((target("sha3,neon"))) uint16x8_t veor3q_u16(uint16x8_t __p0, uint16x8_t __p1, uint16x8_t __p2) {
67175	69693	uint16x8_t __ret;
67176		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
67177		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
67178		uint16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
67179		__ret = (uint16x8_t) __builtin_neon_veor3q_u16((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 49);
67180		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	69694	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	69695	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	69696	uint16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
	69697	__ret = __builtin_bit_cast(uint16x8_t, __builtin_neon_veor3q_u16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 49));
	69698	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
67181	69699	return __ret;
67182	69700	}
67183	69701	#endif
...	...	@@ -67185,17 +69703,17 @@ __ai __attribute__((target("sha3,neon"))) uint16x8_t veor3q_u16(uint16x8_t __p0,
67185	69703	#ifdef __LITTLE_ENDIAN__
67186	69704	__ai __attribute__((target("sha3,neon"))) int8x16_t veor3q_s8(int8x16_t __p0, int8x16_t __p1, int8x16_t __p2) {
67187	69705	int8x16_t __ret;
67188		__ret = (int8x16_t) __builtin_neon_veor3q_s8((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 32);
	69706	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_veor3q_s8(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 32));
67189	69707	return __ret;
67190	69708	}
67191	69709	#else
67192	69710	__ai __attribute__((target("sha3,neon"))) int8x16_t veor3q_s8(int8x16_t __p0, int8x16_t __p1, int8x16_t __p2) {
67193	69711	int8x16_t __ret;
67194		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
67195		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
67196		int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
67197		__ret = (int8x16_t) __builtin_neon_veor3q_s8((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 32);
67198		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	69712	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	69713	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	69714	int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
	69715	__ret = __builtin_bit_cast(int8x16_t, __builtin_neon_veor3q_s8(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 32));
	69716	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
67199	69717	return __ret;
67200	69718	}
67201	69719	#endif
...	...	@@ -67203,17 +69721,17 @@ __ai __attribute__((target("sha3,neon"))) int8x16_t veor3q_s8(int8x16_t __p0, in
67203	69721	#ifdef __LITTLE_ENDIAN__
67204	69722	__ai __attribute__((target("sha3,neon"))) int32x4_t veor3q_s32(int32x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
67205	69723	int32x4_t __ret;
67206		__ret = (int32x4_t) __builtin_neon_veor3q_s32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 34);
	69724	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_veor3q_s32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 34));
67207	69725	return __ret;
67208	69726	}
67209	69727	#else
67210	69728	__ai __attribute__((target("sha3,neon"))) int32x4_t veor3q_s32(int32x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
67211	69729	int32x4_t __ret;
67212		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
67213		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
67214		int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
67215		__ret = (int32x4_t) __builtin_neon_veor3q_s32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 34);
67216		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	69730	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	69731	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	69732	int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
	69733	__ret = __builtin_bit_cast(int32x4_t, __builtin_neon_veor3q_s32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 34));
	69734	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
67217	69735	return __ret;
67218	69736	}
67219	69737	#endif
...	...	@@ -67221,17 +69739,17 @@ __ai __attribute__((target("sha3,neon"))) int32x4_t veor3q_s32(int32x4_t __p0, i
67221	69739	#ifdef __LITTLE_ENDIAN__
67222	69740	__ai __attribute__((target("sha3,neon"))) int64x2_t veor3q_s64(int64x2_t __p0, int64x2_t __p1, int64x2_t __p2) {
67223	69741	int64x2_t __ret;
67224		__ret = (int64x2_t) __builtin_neon_veor3q_s64((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 35);
	69742	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_veor3q_s64(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 35));
67225	69743	return __ret;
67226	69744	}
67227	69745	#else
67228	69746	__ai __attribute__((target("sha3,neon"))) int64x2_t veor3q_s64(int64x2_t __p0, int64x2_t __p1, int64x2_t __p2) {
67229	69747	int64x2_t __ret;
67230		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
67231		int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
67232		int64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
67233		__ret = (int64x2_t) __builtin_neon_veor3q_s64((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 35);
67234		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	69748	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	69749	int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	69750	int64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_64);
	69751	__ret = __builtin_bit_cast(int64x2_t, __builtin_neon_veor3q_s64(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 35));
	69752	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
67235	69753	return __ret;
67236	69754	}
67237	69755	#endif
...	...	@@ -67239,17 +69757,17 @@ __ai __attribute__((target("sha3,neon"))) int64x2_t veor3q_s64(int64x2_t __p0, i
67239	69757	#ifdef __LITTLE_ENDIAN__
67240	69758	__ai __attribute__((target("sha3,neon"))) int16x8_t veor3q_s16(int16x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
67241	69759	int16x8_t __ret;
67242		__ret = (int16x8_t) __builtin_neon_veor3q_s16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 33);
	69760	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_veor3q_s16(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 33));
67243	69761	return __ret;
67244	69762	}
67245	69763	#else
67246	69764	__ai __attribute__((target("sha3,neon"))) int16x8_t veor3q_s16(int16x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
67247	69765	int16x8_t __ret;
67248		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
67249		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
67250		int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
67251		__ret = (int16x8_t) __builtin_neon_veor3q_s16((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 33);
67252		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	69766	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	69767	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	69768	int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
	69769	__ret = __builtin_bit_cast(int16x8_t, __builtin_neon_veor3q_s16(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 33));
	69770	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
67253	69771	return __ret;
67254	69772	}
67255	69773	#endif
...	...	@@ -67257,16 +69775,16 @@ __ai __attribute__((target("sha3,neon"))) int16x8_t veor3q_s16(int16x8_t __p0, i
67257	69775	#ifdef __LITTLE_ENDIAN__
67258	69776	__ai __attribute__((target("sha3,neon"))) uint64x2_t vrax1q_u64(uint64x2_t __p0, uint64x2_t __p1) {
67259	69777	uint64x2_t __ret;
67260		__ret = (uint64x2_t) __builtin_neon_vrax1q_u64((int8x16_t)__p0, (int8x16_t)__p1, 51);
	69778	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vrax1q_u64(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 51));
67261	69779	return __ret;
67262	69780	}
67263	69781	#else
67264	69782	__ai __attribute__((target("sha3,neon"))) uint64x2_t vrax1q_u64(uint64x2_t __p0, uint64x2_t __p1) {
67265	69783	uint64x2_t __ret;
67266		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
67267		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
67268		__ret = (uint64x2_t) __builtin_neon_vrax1q_u64((int8x16_t)__rev0, (int8x16_t)__rev1, 51);
67269		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	69784	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	69785	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	69786	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vrax1q_u64(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 51));
	69787	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
67270	69788	return __ret;
67271	69789	}
67272	69790	#endif
...	...	@@ -67274,17 +69792,17 @@ __ai __attribute__((target("sha3,neon"))) uint64x2_t vrax1q_u64(uint64x2_t __p0,
67274	69792	#ifdef __LITTLE_ENDIAN__
67275	69793	__ai __attribute__((target("sha3,neon"))) uint64x2_t vsha512hq_u64(uint64x2_t __p0, uint64x2_t __p1, uint64x2_t __p2) {
67276	69794	uint64x2_t __ret;
67277		__ret = (uint64x2_t) __builtin_neon_vsha512hq_u64((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 51);
	69795	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vsha512hq_u64(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 51));
67278	69796	return __ret;
67279	69797	}
67280	69798	#else
67281	69799	__ai __attribute__((target("sha3,neon"))) uint64x2_t vsha512hq_u64(uint64x2_t __p0, uint64x2_t __p1, uint64x2_t __p2) {
67282	69800	uint64x2_t __ret;
67283		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
67284		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
67285		uint64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
67286		__ret = (uint64x2_t) __builtin_neon_vsha512hq_u64((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 51);
67287		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	69801	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	69802	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	69803	uint64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_64);
	69804	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vsha512hq_u64(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 51));
	69805	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
67288	69806	return __ret;
67289	69807	}
67290	69808	#endif
...	...	@@ -67292,17 +69810,17 @@ __ai __attribute__((target("sha3,neon"))) uint64x2_t vsha512hq_u64(uint64x2_t __
67292	69810	#ifdef __LITTLE_ENDIAN__
67293	69811	__ai __attribute__((target("sha3,neon"))) uint64x2_t vsha512h2q_u64(uint64x2_t __p0, uint64x2_t __p1, uint64x2_t __p2) {
67294	69812	uint64x2_t __ret;
67295		__ret = (uint64x2_t) __builtin_neon_vsha512h2q_u64((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 51);
	69813	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vsha512h2q_u64(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 51));
67296	69814	return __ret;
67297	69815	}
67298	69816	#else
67299	69817	__ai __attribute__((target("sha3,neon"))) uint64x2_t vsha512h2q_u64(uint64x2_t __p0, uint64x2_t __p1, uint64x2_t __p2) {
67300	69818	uint64x2_t __ret;
67301		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
67302		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
67303		uint64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
67304		__ret = (uint64x2_t) __builtin_neon_vsha512h2q_u64((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 51);
67305		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	69819	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	69820	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	69821	uint64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_64);
	69822	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vsha512h2q_u64(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 51));
	69823	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
67306	69824	return __ret;
67307	69825	}
67308	69826	#endif
...	...	@@ -67310,16 +69828,16 @@ __ai __attribute__((target("sha3,neon"))) uint64x2_t vsha512h2q_u64(uint64x2_t _
67310	69828	#ifdef __LITTLE_ENDIAN__
67311	69829	__ai __attribute__((target("sha3,neon"))) uint64x2_t vsha512su0q_u64(uint64x2_t __p0, uint64x2_t __p1) {
67312	69830	uint64x2_t __ret;
67313		__ret = (uint64x2_t) __builtin_neon_vsha512su0q_u64((int8x16_t)__p0, (int8x16_t)__p1, 51);
	69831	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vsha512su0q_u64(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 51));
67314	69832	return __ret;
67315	69833	}
67316	69834	#else
67317	69835	__ai __attribute__((target("sha3,neon"))) uint64x2_t vsha512su0q_u64(uint64x2_t __p0, uint64x2_t __p1) {
67318	69836	uint64x2_t __ret;
67319		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
67320		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
67321		__ret = (uint64x2_t) __builtin_neon_vsha512su0q_u64((int8x16_t)__rev0, (int8x16_t)__rev1, 51);
67322		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	69837	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	69838	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	69839	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vsha512su0q_u64(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 51));
	69840	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
67323	69841	return __ret;
67324	69842	}
67325	69843	#endif
...	...	@@ -67327,17 +69845,17 @@ __ai __attribute__((target("sha3,neon"))) uint64x2_t vsha512su0q_u64(uint64x2_t
67327	69845	#ifdef __LITTLE_ENDIAN__
67328	69846	__ai __attribute__((target("sha3,neon"))) uint64x2_t vsha512su1q_u64(uint64x2_t __p0, uint64x2_t __p1, uint64x2_t __p2) {
67329	69847	uint64x2_t __ret;
67330		__ret = (uint64x2_t) __builtin_neon_vsha512su1q_u64((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 51);
	69848	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vsha512su1q_u64(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 51));
67331	69849	return __ret;
67332	69850	}
67333	69851	#else
67334	69852	__ai __attribute__((target("sha3,neon"))) uint64x2_t vsha512su1q_u64(uint64x2_t __p0, uint64x2_t __p1, uint64x2_t __p2) {
67335	69853	uint64x2_t __ret;
67336		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
67337		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
67338		uint64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
67339		__ret = (uint64x2_t) __builtin_neon_vsha512su1q_u64((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 51);
67340		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	69854	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	69855	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	69856	uint64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_64);
	69857	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vsha512su1q_u64(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 51));
	69858	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
67341	69859	return __ret;
67342	69860	}
67343	69861	#endif
...	...	@@ -67347,7 +69865,7 @@ __ai __attribute__((target("sha3,neon"))) uint64x2_t vsha512su1q_u64(uint64x2_t
67347	69865	uint64x2_t __ret; \
67348	69866	uint64x2_t __s0 = __p0; \
67349	69867	uint64x2_t __s1 = __p1; \
67350		__ret = (uint64x2_t) __builtin_neon_vxarq_u64((int8x16_t)__s0, (int8x16_t)__s1, __p2, 51); \
	69868	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vxarq_u64(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __p2, 51)); \
67351	69869	__ret; \
67352	69870	})
67353	69871	#else
...	...	@@ -67355,10 +69873,10 @@ __ai __attribute__((target("sha3,neon"))) uint64x2_t vsha512su1q_u64(uint64x2_t
67355	69873	uint64x2_t __ret; \
67356	69874	uint64x2_t __s0 = __p0; \
67357	69875	uint64x2_t __s1 = __p1; \
67358		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
67359		uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
67360		__ret = (uint64x2_t) __builtin_neon_vxarq_u64((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 51); \
67361		__ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
	69876	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_64); \
	69877	uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_64); \
	69878	__ret = __builtin_bit_cast(uint64x2_t, __builtin_neon_vxarq_u64(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __p2, 51)); \
	69879	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64); \
67362	69880	__ret; \
67363	69881	})
67364	69882	#endif
...	...	@@ -67366,17 +69884,17 @@ __ai __attribute__((target("sha3,neon"))) uint64x2_t vsha512su1q_u64(uint64x2_t
67366	69884	#ifdef __LITTLE_ENDIAN__
67367	69885	__ai __attribute__((target("sm4,neon"))) uint32x4_t vsm3partw1q_u32(uint32x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
67368	69886	uint32x4_t __ret;
67369		__ret = (uint32x4_t) __builtin_neon_vsm3partw1q_u32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 50);
	69887	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsm3partw1q_u32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 50));
67370	69888	return __ret;
67371	69889	}
67372	69890	#else
67373	69891	__ai __attribute__((target("sm4,neon"))) uint32x4_t vsm3partw1q_u32(uint32x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
67374	69892	uint32x4_t __ret;
67375		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
67376		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
67377		uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
67378		__ret = (uint32x4_t) __builtin_neon_vsm3partw1q_u32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 50);
67379		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	69893	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	69894	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	69895	uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
	69896	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsm3partw1q_u32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 50));
	69897	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
67380	69898	return __ret;
67381	69899	}
67382	69900	#endif
...	...	@@ -67384,17 +69902,17 @@ __ai __attribute__((target("sm4,neon"))) uint32x4_t vsm3partw1q_u32(uint32x4_t _
67384	69902	#ifdef __LITTLE_ENDIAN__
67385	69903	__ai __attribute__((target("sm4,neon"))) uint32x4_t vsm3partw2q_u32(uint32x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
67386	69904	uint32x4_t __ret;
67387		__ret = (uint32x4_t) __builtin_neon_vsm3partw2q_u32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 50);
	69905	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsm3partw2q_u32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 50));
67388	69906	return __ret;
67389	69907	}
67390	69908	#else
67391	69909	__ai __attribute__((target("sm4,neon"))) uint32x4_t vsm3partw2q_u32(uint32x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
67392	69910	uint32x4_t __ret;
67393		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
67394		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
67395		uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
67396		__ret = (uint32x4_t) __builtin_neon_vsm3partw2q_u32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 50);
67397		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	69911	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	69912	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	69913	uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
	69914	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsm3partw2q_u32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 50));
	69915	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
67398	69916	return __ret;
67399	69917	}
67400	69918	#endif
...	...	@@ -67402,17 +69920,17 @@ __ai __attribute__((target("sm4,neon"))) uint32x4_t vsm3partw2q_u32(uint32x4_t _
67402	69920	#ifdef __LITTLE_ENDIAN__
67403	69921	__ai __attribute__((target("sm4,neon"))) uint32x4_t vsm3ss1q_u32(uint32x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
67404	69922	uint32x4_t __ret;
67405		__ret = (uint32x4_t) __builtin_neon_vsm3ss1q_u32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 50);
	69923	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsm3ss1q_u32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 50));
67406	69924	return __ret;
67407	69925	}
67408	69926	#else
67409	69927	__ai __attribute__((target("sm4,neon"))) uint32x4_t vsm3ss1q_u32(uint32x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
67410	69928	uint32x4_t __ret;
67411		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
67412		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
67413		uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
67414		__ret = (uint32x4_t) __builtin_neon_vsm3ss1q_u32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 50);
67415		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	69929	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	69930	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	69931	uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
	69932	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsm3ss1q_u32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 50));
	69933	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
67416	69934	return __ret;
67417	69935	}
67418	69936	#endif
...	...	@@ -67423,7 +69941,7 @@ __ai __attribute__((target("sm4,neon"))) uint32x4_t vsm3ss1q_u32(uint32x4_t __p0
67423	69941	uint32x4_t __s0 = __p0; \
67424	69942	uint32x4_t __s1 = __p1; \
67425	69943	uint32x4_t __s2 = __p2; \
67426		__ret = (uint32x4_t) __builtin_neon_vsm3tt1aq_u32((int8x16_t)__s0, (int8x16_t)__s1, (int8x16_t)__s2, __p3, 50); \
	69944	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsm3tt1aq_u32(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __builtin_bit_cast(int8x16_t, __s2), __p3, 50)); \
67427	69945	__ret; \
67428	69946	})
67429	69947	#else
...	...	@@ -67432,11 +69950,11 @@ __ai __attribute__((target("sm4,neon"))) uint32x4_t vsm3ss1q_u32(uint32x4_t __p0
67432	69950	uint32x4_t __s0 = __p0; \
67433	69951	uint32x4_t __s1 = __p1; \
67434	69952	uint32x4_t __s2 = __p2; \
67435		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
67436		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
67437		uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 3, 2, 1, 0); \
67438		__ret = (uint32x4_t) __builtin_neon_vsm3tt1aq_u32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, __p3, 50); \
67439		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	69953	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	69954	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	69955	uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_128_32); \
	69956	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsm3tt1aq_u32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), __p3, 50)); \
	69957	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
67440	69958	__ret; \
67441	69959	})
67442	69960	#endif
...	...	@@ -67447,7 +69965,7 @@ __ai __attribute__((target("sm4,neon"))) uint32x4_t vsm3ss1q_u32(uint32x4_t __p0
67447	69965	uint32x4_t __s0 = __p0; \
67448	69966	uint32x4_t __s1 = __p1; \
67449	69967	uint32x4_t __s2 = __p2; \
67450		__ret = (uint32x4_t) __builtin_neon_vsm3tt1bq_u32((int8x16_t)__s0, (int8x16_t)__s1, (int8x16_t)__s2, __p3, 50); \
	69968	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsm3tt1bq_u32(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __builtin_bit_cast(int8x16_t, __s2), __p3, 50)); \
67451	69969	__ret; \
67452	69970	})
67453	69971	#else
...	...	@@ -67456,11 +69974,11 @@ __ai __attribute__((target("sm4,neon"))) uint32x4_t vsm3ss1q_u32(uint32x4_t __p0
67456	69974	uint32x4_t __s0 = __p0; \
67457	69975	uint32x4_t __s1 = __p1; \
67458	69976	uint32x4_t __s2 = __p2; \
67459		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
67460		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
67461		uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 3, 2, 1, 0); \
67462		__ret = (uint32x4_t) __builtin_neon_vsm3tt1bq_u32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, __p3, 50); \
67463		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	69977	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	69978	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	69979	uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_128_32); \
	69980	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsm3tt1bq_u32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), __p3, 50)); \
	69981	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
67464	69982	__ret; \
67465	69983	})
67466	69984	#endif
...	...	@@ -67471,7 +69989,7 @@ __ai __attribute__((target("sm4,neon"))) uint32x4_t vsm3ss1q_u32(uint32x4_t __p0
67471	69989	uint32x4_t __s0 = __p0; \
67472	69990	uint32x4_t __s1 = __p1; \
67473	69991	uint32x4_t __s2 = __p2; \
67474		__ret = (uint32x4_t) __builtin_neon_vsm3tt2aq_u32((int8x16_t)__s0, (int8x16_t)__s1, (int8x16_t)__s2, __p3, 50); \
	69992	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsm3tt2aq_u32(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __builtin_bit_cast(int8x16_t, __s2), __p3, 50)); \
67475	69993	__ret; \
67476	69994	})
67477	69995	#else
...	...	@@ -67480,11 +69998,11 @@ __ai __attribute__((target("sm4,neon"))) uint32x4_t vsm3ss1q_u32(uint32x4_t __p0
67480	69998	uint32x4_t __s0 = __p0; \
67481	69999	uint32x4_t __s1 = __p1; \
67482	70000	uint32x4_t __s2 = __p2; \
67483		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
67484		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
67485		uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 3, 2, 1, 0); \
67486		__ret = (uint32x4_t) __builtin_neon_vsm3tt2aq_u32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, __p3, 50); \
67487		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	70001	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	70002	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	70003	uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_128_32); \
	70004	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsm3tt2aq_u32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), __p3, 50)); \
	70005	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
67488	70006	__ret; \
67489	70007	})
67490	70008	#endif
...	...	@@ -67495,7 +70013,7 @@ __ai __attribute__((target("sm4,neon"))) uint32x4_t vsm3ss1q_u32(uint32x4_t __p0
67495	70013	uint32x4_t __s0 = __p0; \
67496	70014	uint32x4_t __s1 = __p1; \
67497	70015	uint32x4_t __s2 = __p2; \
67498		__ret = (uint32x4_t) __builtin_neon_vsm3tt2bq_u32((int8x16_t)__s0, (int8x16_t)__s1, (int8x16_t)__s2, __p3, 50); \
	70016	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsm3tt2bq_u32(__builtin_bit_cast(int8x16_t, __s0), __builtin_bit_cast(int8x16_t, __s1), __builtin_bit_cast(int8x16_t, __s2), __p3, 50)); \
67499	70017	__ret; \
67500	70018	})
67501	70019	#else
...	...	@@ -67504,11 +70022,11 @@ __ai __attribute__((target("sm4,neon"))) uint32x4_t vsm3ss1q_u32(uint32x4_t __p0
67504	70022	uint32x4_t __s0 = __p0; \
67505	70023	uint32x4_t __s1 = __p1; \
67506	70024	uint32x4_t __s2 = __p2; \
67507		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
67508		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
67509		uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, 3, 2, 1, 0); \
67510		__ret = (uint32x4_t) __builtin_neon_vsm3tt2bq_u32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, __p3, 50); \
67511		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
	70025	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, __lane_reverse_128_32); \
	70026	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, __lane_reverse_128_32); \
	70027	uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__s2, __s2, __lane_reverse_128_32); \
	70028	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsm3tt2bq_u32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), __p3, 50)); \
	70029	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32); \
67512	70030	__ret; \
67513	70031	})
67514	70032	#endif
...	...	@@ -67516,16 +70034,16 @@ __ai __attribute__((target("sm4,neon"))) uint32x4_t vsm3ss1q_u32(uint32x4_t __p0
67516	70034	#ifdef __LITTLE_ENDIAN__
67517	70035	__ai __attribute__((target("sm4,neon"))) uint32x4_t vsm4eq_u32(uint32x4_t __p0, uint32x4_t __p1) {
67518	70036	uint32x4_t __ret;
67519		__ret = (uint32x4_t) __builtin_neon_vsm4eq_u32((int8x16_t)__p0, (int8x16_t)__p1, 50);
	70037	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsm4eq_u32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50));
67520	70038	return __ret;
67521	70039	}
67522	70040	#else
67523	70041	__ai __attribute__((target("sm4,neon"))) uint32x4_t vsm4eq_u32(uint32x4_t __p0, uint32x4_t __p1) {
67524	70042	uint32x4_t __ret;
67525		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
67526		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
67527		__ret = (uint32x4_t) __builtin_neon_vsm4eq_u32((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
67528		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	70043	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	70044	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	70045	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsm4eq_u32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 50));
	70046	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
67529	70047	return __ret;
67530	70048	}
67531	70049	#endif
...	...	@@ -67533,413 +70051,413 @@ __ai __attribute__((target("sm4,neon"))) uint32x4_t vsm4eq_u32(uint32x4_t __p0,
67533	70051	#ifdef __LITTLE_ENDIAN__
67534	70052	__ai __attribute__((target("sm4,neon"))) uint32x4_t vsm4ekeyq_u32(uint32x4_t __p0, uint32x4_t __p1) {
67535	70053	uint32x4_t __ret;
67536		__ret = (uint32x4_t) __builtin_neon_vsm4ekeyq_u32((int8x16_t)__p0, (int8x16_t)__p1, 50);
	70054	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsm4ekeyq_u32(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 50));
67537	70055	return __ret;
67538	70056	}
67539	70057	#else
67540	70058	__ai __attribute__((target("sm4,neon"))) uint32x4_t vsm4ekeyq_u32(uint32x4_t __p0, uint32x4_t __p1) {
67541	70059	uint32x4_t __ret;
67542		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
67543		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
67544		__ret = (uint32x4_t) __builtin_neon_vsm4ekeyq_u32((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
67545		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	70060	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	70061	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	70062	__ret = __builtin_bit_cast(uint32x4_t, __builtin_neon_vsm4ekeyq_u32(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 50));
	70063	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
67546	70064	return __ret;
67547	70065	}
67548	70066	#endif
67549	70067
67550	70068	__ai __attribute__((target("v8.1a,neon"))) int32_t vqrdmlahs_s32(int32_t __p0, int32_t __p1, int32_t __p2) {
67551	70069	int32_t __ret;
67552		__ret = (int32_t) __builtin_neon_vqrdmlahs_s32(__p0, __p1, __p2);
	70070	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vqrdmlahs_s32(__p0, __p1, __p2));
67553	70071	return __ret;
67554	70072	}
67555	70073	__ai __attribute__((target("v8.1a,neon"))) int16_t vqrdmlahh_s16(int16_t __p0, int16_t __p1, int16_t __p2) {
67556	70074	int16_t __ret;
67557		__ret = (int16_t) __builtin_neon_vqrdmlahh_s16(__p0, __p1, __p2);
67558		return __ret;
67559		}
67560		#ifdef __LITTLE_ENDIAN__
67561		#define vqrdmlahs_lane_s32(__p0_760, __p1_760, __p2_760, __p3_760) __extension__ ({ \
67562		int32_t __ret_760; \
67563		int32_t __s0_760 = __p0_760; \
67564		int32_t __s1_760 = __p1_760; \
67565		int32x2_t __s2_760 = __p2_760; \
67566		__ret_760 = vqrdmlahs_s32(__s0_760, __s1_760, vget_lane_s32(__s2_760, __p3_760)); \
67567		__ret_760; \
67568		})
67569		#else
67570		#define vqrdmlahs_lane_s32(__p0_761, __p1_761, __p2_761, __p3_761) __extension__ ({ \
67571		int32_t __ret_761; \
67572		int32_t __s0_761 = __p0_761; \
67573		int32_t __s1_761 = __p1_761; \
67574		int32x2_t __s2_761 = __p2_761; \
67575		int32x2_t __rev2_761; __rev2_761 = __builtin_shufflevector(__s2_761, __s2_761, 1, 0); \
67576		__ret_761 = vqrdmlahs_s32(__s0_761, __s1_761, __noswap_vget_lane_s32(__rev2_761, __p3_761)); \
67577		__ret_761; \
67578		})
67579		#endif
67580
67581		#ifdef __LITTLE_ENDIAN__
67582		#define vqrdmlahh_lane_s16(__p0_762, __p1_762, __p2_762, __p3_762) __extension__ ({ \
67583		int16_t __ret_762; \
67584		int16_t __s0_762 = __p0_762; \
67585		int16_t __s1_762 = __p1_762; \
67586		int16x4_t __s2_762 = __p2_762; \
67587		__ret_762 = vqrdmlahh_s16(__s0_762, __s1_762, vget_lane_s16(__s2_762, __p3_762)); \
67588		__ret_762; \
67589		})
67590		#else
67591		#define vqrdmlahh_lane_s16(__p0_763, __p1_763, __p2_763, __p3_763) __extension__ ({ \
67592		int16_t __ret_763; \
67593		int16_t __s0_763 = __p0_763; \
67594		int16_t __s1_763 = __p1_763; \
67595		int16x4_t __s2_763 = __p2_763; \
67596		int16x4_t __rev2_763; __rev2_763 = __builtin_shufflevector(__s2_763, __s2_763, 3, 2, 1, 0); \
67597		__ret_763 = vqrdmlahh_s16(__s0_763, __s1_763, __noswap_vget_lane_s16(__rev2_763, __p3_763)); \
67598		__ret_763; \
67599		})
67600		#endif
67601
67602		#ifdef __LITTLE_ENDIAN__
67603		#define vqrdmlahs_laneq_s32(__p0_764, __p1_764, __p2_764, __p3_764) __extension__ ({ \
67604		int32_t __ret_764; \
67605		int32_t __s0_764 = __p0_764; \
67606		int32_t __s1_764 = __p1_764; \
67607		int32x4_t __s2_764 = __p2_764; \
67608		__ret_764 = vqrdmlahs_s32(__s0_764, __s1_764, vgetq_lane_s32(__s2_764, __p3_764)); \
67609		__ret_764; \
67610		})
67611		#else
67612		#define vqrdmlahs_laneq_s32(__p0_765, __p1_765, __p2_765, __p3_765) __extension__ ({ \
67613		int32_t __ret_765; \
67614		int32_t __s0_765 = __p0_765; \
67615		int32_t __s1_765 = __p1_765; \
67616		int32x4_t __s2_765 = __p2_765; \
67617		int32x4_t __rev2_765; __rev2_765 = __builtin_shufflevector(__s2_765, __s2_765, 3, 2, 1, 0); \
67618		__ret_765 = vqrdmlahs_s32(__s0_765, __s1_765, __noswap_vgetq_lane_s32(__rev2_765, __p3_765)); \
67619		__ret_765; \
67620		})
67621		#endif
67622
67623		#ifdef __LITTLE_ENDIAN__
67624		#define vqrdmlahh_laneq_s16(__p0_766, __p1_766, __p2_766, __p3_766) __extension__ ({ \
67625		int16_t __ret_766; \
67626		int16_t __s0_766 = __p0_766; \
67627		int16_t __s1_766 = __p1_766; \
67628		int16x8_t __s2_766 = __p2_766; \
67629		__ret_766 = vqrdmlahh_s16(__s0_766, __s1_766, vgetq_lane_s16(__s2_766, __p3_766)); \
67630		__ret_766; \
67631		})
67632		#else
67633		#define vqrdmlahh_laneq_s16(__p0_767, __p1_767, __p2_767, __p3_767) __extension__ ({ \
67634		int16_t __ret_767; \
67635		int16_t __s0_767 = __p0_767; \
67636		int16_t __s1_767 = __p1_767; \
67637		int16x8_t __s2_767 = __p2_767; \
67638		int16x8_t __rev2_767; __rev2_767 = __builtin_shufflevector(__s2_767, __s2_767, 7, 6, 5, 4, 3, 2, 1, 0); \
67639		__ret_767 = vqrdmlahh_s16(__s0_767, __s1_767, __noswap_vgetq_lane_s16(__rev2_767, __p3_767)); \
67640		__ret_767; \
67641		})
67642		#endif
67643
67644		#ifdef __LITTLE_ENDIAN__
67645		#define vqrdmlahq_laneq_s32(__p0_768, __p1_768, __p2_768, __p3_768) __extension__ ({ \
67646		int32x4_t __ret_768; \
67647		int32x4_t __s0_768 = __p0_768; \
67648		int32x4_t __s1_768 = __p1_768; \
67649		int32x4_t __s2_768 = __p2_768; \
67650		__ret_768 = vqrdmlahq_s32(__s0_768, __s1_768, splatq_laneq_s32(__s2_768, __p3_768)); \
67651		__ret_768; \
67652		})
67653		#else
67654		#define vqrdmlahq_laneq_s32(__p0_769, __p1_769, __p2_769, __p3_769) __extension__ ({ \
67655		int32x4_t __ret_769; \
67656		int32x4_t __s0_769 = __p0_769; \
67657		int32x4_t __s1_769 = __p1_769; \
67658		int32x4_t __s2_769 = __p2_769; \
67659		int32x4_t __rev0_769; __rev0_769 = __builtin_shufflevector(__s0_769, __s0_769, 3, 2, 1, 0); \
67660		int32x4_t __rev1_769; __rev1_769 = __builtin_shufflevector(__s1_769, __s1_769, 3, 2, 1, 0); \
67661		int32x4_t __rev2_769; __rev2_769 = __builtin_shufflevector(__s2_769, __s2_769, 3, 2, 1, 0); \
67662		__ret_769 = __noswap_vqrdmlahq_s32(__rev0_769, __rev1_769, __noswap_splatq_laneq_s32(__rev2_769, __p3_769)); \
67663		__ret_769 = __builtin_shufflevector(__ret_769, __ret_769, 3, 2, 1, 0); \
67664		__ret_769; \
67665		})
67666		#endif
67667
67668		#ifdef __LITTLE_ENDIAN__
67669		#define vqrdmlahq_laneq_s16(__p0_770, __p1_770, __p2_770, __p3_770) __extension__ ({ \
67670		int16x8_t __ret_770; \
67671		int16x8_t __s0_770 = __p0_770; \
67672		int16x8_t __s1_770 = __p1_770; \
67673		int16x8_t __s2_770 = __p2_770; \
67674		__ret_770 = vqrdmlahq_s16(__s0_770, __s1_770, splatq_laneq_s16(__s2_770, __p3_770)); \
67675		__ret_770; \
67676		})
67677		#else
67678		#define vqrdmlahq_laneq_s16(__p0_771, __p1_771, __p2_771, __p3_771) __extension__ ({ \
67679		int16x8_t __ret_771; \
67680		int16x8_t __s0_771 = __p0_771; \
67681		int16x8_t __s1_771 = __p1_771; \
67682		int16x8_t __s2_771 = __p2_771; \
67683		int16x8_t __rev0_771; __rev0_771 = __builtin_shufflevector(__s0_771, __s0_771, 7, 6, 5, 4, 3, 2, 1, 0); \
67684		int16x8_t __rev1_771; __rev1_771 = __builtin_shufflevector(__s1_771, __s1_771, 7, 6, 5, 4, 3, 2, 1, 0); \
67685		int16x8_t __rev2_771; __rev2_771 = __builtin_shufflevector(__s2_771, __s2_771, 7, 6, 5, 4, 3, 2, 1, 0); \
67686		__ret_771 = __noswap_vqrdmlahq_s16(__rev0_771, __rev1_771, __noswap_splatq_laneq_s16(__rev2_771, __p3_771)); \
67687		__ret_771 = __builtin_shufflevector(__ret_771, __ret_771, 7, 6, 5, 4, 3, 2, 1, 0); \
67688		__ret_771; \
67689		})
67690		#endif
67691
67692		#ifdef __LITTLE_ENDIAN__
67693		#define vqrdmlah_laneq_s32(__p0_772, __p1_772, __p2_772, __p3_772) __extension__ ({ \
67694		int32x2_t __ret_772; \
67695		int32x2_t __s0_772 = __p0_772; \
67696		int32x2_t __s1_772 = __p1_772; \
67697		int32x4_t __s2_772 = __p2_772; \
67698		__ret_772 = vqrdmlah_s32(__s0_772, __s1_772, splat_laneq_s32(__s2_772, __p3_772)); \
67699		__ret_772; \
67700		})
67701		#else
67702		#define vqrdmlah_laneq_s32(__p0_773, __p1_773, __p2_773, __p3_773) __extension__ ({ \
67703		int32x2_t __ret_773; \
67704		int32x2_t __s0_773 = __p0_773; \
67705		int32x2_t __s1_773 = __p1_773; \
67706		int32x4_t __s2_773 = __p2_773; \
67707		int32x2_t __rev0_773; __rev0_773 = __builtin_shufflevector(__s0_773, __s0_773, 1, 0); \
67708		int32x2_t __rev1_773; __rev1_773 = __builtin_shufflevector(__s1_773, __s1_773, 1, 0); \
67709		int32x4_t __rev2_773; __rev2_773 = __builtin_shufflevector(__s2_773, __s2_773, 3, 2, 1, 0); \
67710		__ret_773 = __noswap_vqrdmlah_s32(__rev0_773, __rev1_773, __noswap_splat_laneq_s32(__rev2_773, __p3_773)); \
67711		__ret_773 = __builtin_shufflevector(__ret_773, __ret_773, 1, 0); \
67712		__ret_773; \
67713		})
67714		#endif
67715
67716		#ifdef __LITTLE_ENDIAN__
67717		#define vqrdmlah_laneq_s16(__p0_774, __p1_774, __p2_774, __p3_774) __extension__ ({ \
67718		int16x4_t __ret_774; \
67719		int16x4_t __s0_774 = __p0_774; \
67720		int16x4_t __s1_774 = __p1_774; \
67721		int16x8_t __s2_774 = __p2_774; \
67722		__ret_774 = vqrdmlah_s16(__s0_774, __s1_774, splat_laneq_s16(__s2_774, __p3_774)); \
67723		__ret_774; \
67724		})
67725		#else
67726		#define vqrdmlah_laneq_s16(__p0_775, __p1_775, __p2_775, __p3_775) __extension__ ({ \
67727		int16x4_t __ret_775; \
67728		int16x4_t __s0_775 = __p0_775; \
67729		int16x4_t __s1_775 = __p1_775; \
67730		int16x8_t __s2_775 = __p2_775; \
67731		int16x4_t __rev0_775; __rev0_775 = __builtin_shufflevector(__s0_775, __s0_775, 3, 2, 1, 0); \
67732		int16x4_t __rev1_775; __rev1_775 = __builtin_shufflevector(__s1_775, __s1_775, 3, 2, 1, 0); \
67733		int16x8_t __rev2_775; __rev2_775 = __builtin_shufflevector(__s2_775, __s2_775, 7, 6, 5, 4, 3, 2, 1, 0); \
67734		__ret_775 = __noswap_vqrdmlah_s16(__rev0_775, __rev1_775, __noswap_splat_laneq_s16(__rev2_775, __p3_775)); \
67735		__ret_775 = __builtin_shufflevector(__ret_775, __ret_775, 3, 2, 1, 0); \
67736		__ret_775; \
67737		})
67738		#endif
67739
67740		__ai __attribute__((target("v8.1a,neon"))) int32_t vqrdmlshs_s32(int32_t __p0, int32_t __p1, int32_t __p2) {
67741		int32_t __ret;
67742		__ret = (int32_t) __builtin_neon_vqrdmlshs_s32(__p0, __p1, __p2);
67743		return __ret;
67744		}
67745		__ai __attribute__((target("v8.1a,neon"))) int16_t vqrdmlshh_s16(int16_t __p0, int16_t __p1, int16_t __p2) {
67746		int16_t __ret;
67747		__ret = (int16_t) __builtin_neon_vqrdmlshh_s16(__p0, __p1, __p2);
	70075	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vqrdmlahh_s16(__p0, __p1, __p2));
67748	70076	return __ret;
67749	70077	}
67750	70078	#ifdef __LITTLE_ENDIAN__
67751		#define vqrdmlshs_lane_s32(__p0_776, __p1_776, __p2_776, __p3_776) __extension__ ({ \
	70079	#define vqrdmlahs_lane_s32(__p0_776, __p1_776, __p2_776, __p3_776) __extension__ ({ \
67752	70080	int32_t __ret_776; \
67753	70081	int32_t __s0_776 = __p0_776; \
67754	70082	int32_t __s1_776 = __p1_776; \
67755	70083	int32x2_t __s2_776 = __p2_776; \
67756		__ret_776 = vqrdmlshs_s32(__s0_776, __s1_776, vget_lane_s32(__s2_776, __p3_776)); \
	70084	__ret_776 = vqrdmlahs_s32(__s0_776, __s1_776, vget_lane_s32(__s2_776, __p3_776)); \
67757	70085	__ret_776; \
67758	70086	})
67759	70087	#else
67760		#define vqrdmlshs_lane_s32(__p0_777, __p1_777, __p2_777, __p3_777) __extension__ ({ \
	70088	#define vqrdmlahs_lane_s32(__p0_777, __p1_777, __p2_777, __p3_777) __extension__ ({ \
67761	70089	int32_t __ret_777; \
67762	70090	int32_t __s0_777 = __p0_777; \
67763	70091	int32_t __s1_777 = __p1_777; \
67764	70092	int32x2_t __s2_777 = __p2_777; \
67765		int32x2_t __rev2_777; __rev2_777 = __builtin_shufflevector(__s2_777, __s2_777, 1, 0); \
67766		__ret_777 = vqrdmlshs_s32(__s0_777, __s1_777, __noswap_vget_lane_s32(__rev2_777, __p3_777)); \
	70093	int32x2_t __rev2_777; __rev2_777 = __builtin_shufflevector(__s2_777, __s2_777, __lane_reverse_64_32); \
	70094	__ret_777 = vqrdmlahs_s32(__s0_777, __s1_777, __noswap_vget_lane_s32(__rev2_777, __p3_777)); \
67767	70095	__ret_777; \
67768	70096	})
67769	70097	#endif
67770	70098
67771	70099	#ifdef __LITTLE_ENDIAN__
67772		#define vqrdmlshh_lane_s16(__p0_778, __p1_778, __p2_778, __p3_778) __extension__ ({ \
	70100	#define vqrdmlahh_lane_s16(__p0_778, __p1_778, __p2_778, __p3_778) __extension__ ({ \
67773	70101	int16_t __ret_778; \
67774	70102	int16_t __s0_778 = __p0_778; \
67775	70103	int16_t __s1_778 = __p1_778; \
67776	70104	int16x4_t __s2_778 = __p2_778; \
67777		__ret_778 = vqrdmlshh_s16(__s0_778, __s1_778, vget_lane_s16(__s2_778, __p3_778)); \
	70105	__ret_778 = vqrdmlahh_s16(__s0_778, __s1_778, vget_lane_s16(__s2_778, __p3_778)); \
67778	70106	__ret_778; \
67779	70107	})
67780	70108	#else
67781		#define vqrdmlshh_lane_s16(__p0_779, __p1_779, __p2_779, __p3_779) __extension__ ({ \
	70109	#define vqrdmlahh_lane_s16(__p0_779, __p1_779, __p2_779, __p3_779) __extension__ ({ \
67782	70110	int16_t __ret_779; \
67783	70111	int16_t __s0_779 = __p0_779; \
67784	70112	int16_t __s1_779 = __p1_779; \
67785	70113	int16x4_t __s2_779 = __p2_779; \
67786		int16x4_t __rev2_779; __rev2_779 = __builtin_shufflevector(__s2_779, __s2_779, 3, 2, 1, 0); \
67787		__ret_779 = vqrdmlshh_s16(__s0_779, __s1_779, __noswap_vget_lane_s16(__rev2_779, __p3_779)); \
	70114	int16x4_t __rev2_779; __rev2_779 = __builtin_shufflevector(__s2_779, __s2_779, __lane_reverse_64_16); \
	70115	__ret_779 = vqrdmlahh_s16(__s0_779, __s1_779, __noswap_vget_lane_s16(__rev2_779, __p3_779)); \
67788	70116	__ret_779; \
67789	70117	})
67790	70118	#endif
67791	70119
67792	70120	#ifdef __LITTLE_ENDIAN__
67793		#define vqrdmlshs_laneq_s32(__p0_780, __p1_780, __p2_780, __p3_780) __extension__ ({ \
	70121	#define vqrdmlahs_laneq_s32(__p0_780, __p1_780, __p2_780, __p3_780) __extension__ ({ \
67794	70122	int32_t __ret_780; \
67795	70123	int32_t __s0_780 = __p0_780; \
67796	70124	int32_t __s1_780 = __p1_780; \
67797	70125	int32x4_t __s2_780 = __p2_780; \
67798		__ret_780 = vqrdmlshs_s32(__s0_780, __s1_780, vgetq_lane_s32(__s2_780, __p3_780)); \
	70126	__ret_780 = vqrdmlahs_s32(__s0_780, __s1_780, vgetq_lane_s32(__s2_780, __p3_780)); \
67799	70127	__ret_780; \
67800	70128	})
67801	70129	#else
67802		#define vqrdmlshs_laneq_s32(__p0_781, __p1_781, __p2_781, __p3_781) __extension__ ({ \
	70130	#define vqrdmlahs_laneq_s32(__p0_781, __p1_781, __p2_781, __p3_781) __extension__ ({ \
67803	70131	int32_t __ret_781; \
67804	70132	int32_t __s0_781 = __p0_781; \
67805	70133	int32_t __s1_781 = __p1_781; \
67806	70134	int32x4_t __s2_781 = __p2_781; \
67807		int32x4_t __rev2_781; __rev2_781 = __builtin_shufflevector(__s2_781, __s2_781, 3, 2, 1, 0); \
67808		__ret_781 = vqrdmlshs_s32(__s0_781, __s1_781, __noswap_vgetq_lane_s32(__rev2_781, __p3_781)); \
	70135	int32x4_t __rev2_781; __rev2_781 = __builtin_shufflevector(__s2_781, __s2_781, __lane_reverse_128_32); \
	70136	__ret_781 = vqrdmlahs_s32(__s0_781, __s1_781, __noswap_vgetq_lane_s32(__rev2_781, __p3_781)); \
67809	70137	__ret_781; \
67810	70138	})
67811	70139	#endif
67812	70140
67813	70141	#ifdef __LITTLE_ENDIAN__
67814		#define vqrdmlshh_laneq_s16(__p0_782, __p1_782, __p2_782, __p3_782) __extension__ ({ \
	70142	#define vqrdmlahh_laneq_s16(__p0_782, __p1_782, __p2_782, __p3_782) __extension__ ({ \
67815	70143	int16_t __ret_782; \
67816	70144	int16_t __s0_782 = __p0_782; \
67817	70145	int16_t __s1_782 = __p1_782; \
67818	70146	int16x8_t __s2_782 = __p2_782; \
67819		__ret_782 = vqrdmlshh_s16(__s0_782, __s1_782, vgetq_lane_s16(__s2_782, __p3_782)); \
	70147	__ret_782 = vqrdmlahh_s16(__s0_782, __s1_782, vgetq_lane_s16(__s2_782, __p3_782)); \
67820	70148	__ret_782; \
67821	70149	})
67822	70150	#else
67823		#define vqrdmlshh_laneq_s16(__p0_783, __p1_783, __p2_783, __p3_783) __extension__ ({ \
	70151	#define vqrdmlahh_laneq_s16(__p0_783, __p1_783, __p2_783, __p3_783) __extension__ ({ \
67824	70152	int16_t __ret_783; \
67825	70153	int16_t __s0_783 = __p0_783; \
67826	70154	int16_t __s1_783 = __p1_783; \
67827	70155	int16x8_t __s2_783 = __p2_783; \
67828		int16x8_t __rev2_783; __rev2_783 = __builtin_shufflevector(__s2_783, __s2_783, 7, 6, 5, 4, 3, 2, 1, 0); \
67829		__ret_783 = vqrdmlshh_s16(__s0_783, __s1_783, __noswap_vgetq_lane_s16(__rev2_783, __p3_783)); \
	70156	int16x8_t __rev2_783; __rev2_783 = __builtin_shufflevector(__s2_783, __s2_783, __lane_reverse_128_16); \
	70157	__ret_783 = vqrdmlahh_s16(__s0_783, __s1_783, __noswap_vgetq_lane_s16(__rev2_783, __p3_783)); \
67830	70158	__ret_783; \
67831	70159	})
67832	70160	#endif
67833	70161
67834	70162	#ifdef __LITTLE_ENDIAN__
67835		#define vqrdmlshq_laneq_s32(__p0_784, __p1_784, __p2_784, __p3_784) __extension__ ({ \
	70163	#define vqrdmlahq_laneq_s32(__p0_784, __p1_784, __p2_784, __p3_784) __extension__ ({ \
67836	70164	int32x4_t __ret_784; \
67837	70165	int32x4_t __s0_784 = __p0_784; \
67838	70166	int32x4_t __s1_784 = __p1_784; \
67839	70167	int32x4_t __s2_784 = __p2_784; \
67840		__ret_784 = vqrdmlshq_s32(__s0_784, __s1_784, splatq_laneq_s32(__s2_784, __p3_784)); \
	70168	__ret_784 = vqrdmlahq_s32(__s0_784, __s1_784, splatq_laneq_s32(__s2_784, __p3_784)); \
67841	70169	__ret_784; \
67842	70170	})
67843	70171	#else
67844		#define vqrdmlshq_laneq_s32(__p0_785, __p1_785, __p2_785, __p3_785) __extension__ ({ \
	70172	#define vqrdmlahq_laneq_s32(__p0_785, __p1_785, __p2_785, __p3_785) __extension__ ({ \
67845	70173	int32x4_t __ret_785; \
67846	70174	int32x4_t __s0_785 = __p0_785; \
67847	70175	int32x4_t __s1_785 = __p1_785; \
67848	70176	int32x4_t __s2_785 = __p2_785; \
67849		int32x4_t __rev0_785; __rev0_785 = __builtin_shufflevector(__s0_785, __s0_785, 3, 2, 1, 0); \
67850		int32x4_t __rev1_785; __rev1_785 = __builtin_shufflevector(__s1_785, __s1_785, 3, 2, 1, 0); \
67851		int32x4_t __rev2_785; __rev2_785 = __builtin_shufflevector(__s2_785, __s2_785, 3, 2, 1, 0); \
67852		__ret_785 = __noswap_vqrdmlshq_s32(__rev0_785, __rev1_785, __noswap_splatq_laneq_s32(__rev2_785, __p3_785)); \
67853		__ret_785 = __builtin_shufflevector(__ret_785, __ret_785, 3, 2, 1, 0); \
	70177	int32x4_t __rev0_785; __rev0_785 = __builtin_shufflevector(__s0_785, __s0_785, __lane_reverse_128_32); \
	70178	int32x4_t __rev1_785; __rev1_785 = __builtin_shufflevector(__s1_785, __s1_785, __lane_reverse_128_32); \
	70179	int32x4_t __rev2_785; __rev2_785 = __builtin_shufflevector(__s2_785, __s2_785, __lane_reverse_128_32); \
	70180	__ret_785 = __noswap_vqrdmlahq_s32(__rev0_785, __rev1_785, __noswap_splatq_laneq_s32(__rev2_785, __p3_785)); \
	70181	__ret_785 = __builtin_shufflevector(__ret_785, __ret_785, __lane_reverse_128_32); \
67854	70182	__ret_785; \
67855	70183	})
67856	70184	#endif
67857	70185
67858	70186	#ifdef __LITTLE_ENDIAN__
67859		#define vqrdmlshq_laneq_s16(__p0_786, __p1_786, __p2_786, __p3_786) __extension__ ({ \
	70187	#define vqrdmlahq_laneq_s16(__p0_786, __p1_786, __p2_786, __p3_786) __extension__ ({ \
67860	70188	int16x8_t __ret_786; \
67861	70189	int16x8_t __s0_786 = __p0_786; \
67862	70190	int16x8_t __s1_786 = __p1_786; \
67863	70191	int16x8_t __s2_786 = __p2_786; \
67864		__ret_786 = vqrdmlshq_s16(__s0_786, __s1_786, splatq_laneq_s16(__s2_786, __p3_786)); \
	70192	__ret_786 = vqrdmlahq_s16(__s0_786, __s1_786, splatq_laneq_s16(__s2_786, __p3_786)); \
67865	70193	__ret_786; \
67866	70194	})
67867	70195	#else
67868		#define vqrdmlshq_laneq_s16(__p0_787, __p1_787, __p2_787, __p3_787) __extension__ ({ \
	70196	#define vqrdmlahq_laneq_s16(__p0_787, __p1_787, __p2_787, __p3_787) __extension__ ({ \
67869	70197	int16x8_t __ret_787; \
67870	70198	int16x8_t __s0_787 = __p0_787; \
67871	70199	int16x8_t __s1_787 = __p1_787; \
67872	70200	int16x8_t __s2_787 = __p2_787; \
67873		int16x8_t __rev0_787; __rev0_787 = __builtin_shufflevector(__s0_787, __s0_787, 7, 6, 5, 4, 3, 2, 1, 0); \
67874		int16x8_t __rev1_787; __rev1_787 = __builtin_shufflevector(__s1_787, __s1_787, 7, 6, 5, 4, 3, 2, 1, 0); \
67875		int16x8_t __rev2_787; __rev2_787 = __builtin_shufflevector(__s2_787, __s2_787, 7, 6, 5, 4, 3, 2, 1, 0); \
67876		__ret_787 = __noswap_vqrdmlshq_s16(__rev0_787, __rev1_787, __noswap_splatq_laneq_s16(__rev2_787, __p3_787)); \
67877		__ret_787 = __builtin_shufflevector(__ret_787, __ret_787, 7, 6, 5, 4, 3, 2, 1, 0); \
	70201	int16x8_t __rev0_787; __rev0_787 = __builtin_shufflevector(__s0_787, __s0_787, __lane_reverse_128_16); \
	70202	int16x8_t __rev1_787; __rev1_787 = __builtin_shufflevector(__s1_787, __s1_787, __lane_reverse_128_16); \
	70203	int16x8_t __rev2_787; __rev2_787 = __builtin_shufflevector(__s2_787, __s2_787, __lane_reverse_128_16); \
	70204	__ret_787 = __noswap_vqrdmlahq_s16(__rev0_787, __rev1_787, __noswap_splatq_laneq_s16(__rev2_787, __p3_787)); \
	70205	__ret_787 = __builtin_shufflevector(__ret_787, __ret_787, __lane_reverse_128_16); \
67878	70206	__ret_787; \
67879	70207	})
67880	70208	#endif
67881	70209
67882	70210	#ifdef __LITTLE_ENDIAN__
67883		#define vqrdmlsh_laneq_s32(__p0_788, __p1_788, __p2_788, __p3_788) __extension__ ({ \
	70211	#define vqrdmlah_laneq_s32(__p0_788, __p1_788, __p2_788, __p3_788) __extension__ ({ \
67884	70212	int32x2_t __ret_788; \
67885	70213	int32x2_t __s0_788 = __p0_788; \
67886	70214	int32x2_t __s1_788 = __p1_788; \
67887	70215	int32x4_t __s2_788 = __p2_788; \
67888		__ret_788 = vqrdmlsh_s32(__s0_788, __s1_788, splat_laneq_s32(__s2_788, __p3_788)); \
	70216	__ret_788 = vqrdmlah_s32(__s0_788, __s1_788, splat_laneq_s32(__s2_788, __p3_788)); \
67889	70217	__ret_788; \
67890	70218	})
67891	70219	#else
67892		#define vqrdmlsh_laneq_s32(__p0_789, __p1_789, __p2_789, __p3_789) __extension__ ({ \
	70220	#define vqrdmlah_laneq_s32(__p0_789, __p1_789, __p2_789, __p3_789) __extension__ ({ \
67893	70221	int32x2_t __ret_789; \
67894	70222	int32x2_t __s0_789 = __p0_789; \
67895	70223	int32x2_t __s1_789 = __p1_789; \
67896	70224	int32x4_t __s2_789 = __p2_789; \
67897		int32x2_t __rev0_789; __rev0_789 = __builtin_shufflevector(__s0_789, __s0_789, 1, 0); \
67898		int32x2_t __rev1_789; __rev1_789 = __builtin_shufflevector(__s1_789, __s1_789, 1, 0); \
67899		int32x4_t __rev2_789; __rev2_789 = __builtin_shufflevector(__s2_789, __s2_789, 3, 2, 1, 0); \
67900		__ret_789 = __noswap_vqrdmlsh_s32(__rev0_789, __rev1_789, __noswap_splat_laneq_s32(__rev2_789, __p3_789)); \
67901		__ret_789 = __builtin_shufflevector(__ret_789, __ret_789, 1, 0); \
	70225	int32x2_t __rev0_789; __rev0_789 = __builtin_shufflevector(__s0_789, __s0_789, __lane_reverse_64_32); \
	70226	int32x2_t __rev1_789; __rev1_789 = __builtin_shufflevector(__s1_789, __s1_789, __lane_reverse_64_32); \
	70227	int32x4_t __rev2_789; __rev2_789 = __builtin_shufflevector(__s2_789, __s2_789, __lane_reverse_128_32); \
	70228	__ret_789 = __noswap_vqrdmlah_s32(__rev0_789, __rev1_789, __noswap_splat_laneq_s32(__rev2_789, __p3_789)); \
	70229	__ret_789 = __builtin_shufflevector(__ret_789, __ret_789, __lane_reverse_64_32); \
67902	70230	__ret_789; \
67903	70231	})
67904	70232	#endif
67905	70233
67906	70234	#ifdef __LITTLE_ENDIAN__
67907		#define vqrdmlsh_laneq_s16(__p0_790, __p1_790, __p2_790, __p3_790) __extension__ ({ \
	70235	#define vqrdmlah_laneq_s16(__p0_790, __p1_790, __p2_790, __p3_790) __extension__ ({ \
67908	70236	int16x4_t __ret_790; \
67909	70237	int16x4_t __s0_790 = __p0_790; \
67910	70238	int16x4_t __s1_790 = __p1_790; \
67911	70239	int16x8_t __s2_790 = __p2_790; \
67912		__ret_790 = vqrdmlsh_s16(__s0_790, __s1_790, splat_laneq_s16(__s2_790, __p3_790)); \
	70240	__ret_790 = vqrdmlah_s16(__s0_790, __s1_790, splat_laneq_s16(__s2_790, __p3_790)); \
67913	70241	__ret_790; \
67914	70242	})
67915	70243	#else
67916		#define vqrdmlsh_laneq_s16(__p0_791, __p1_791, __p2_791, __p3_791) __extension__ ({ \
	70244	#define vqrdmlah_laneq_s16(__p0_791, __p1_791, __p2_791, __p3_791) __extension__ ({ \
67917	70245	int16x4_t __ret_791; \
67918	70246	int16x4_t __s0_791 = __p0_791; \
67919	70247	int16x4_t __s1_791 = __p1_791; \
67920	70248	int16x8_t __s2_791 = __p2_791; \
67921		int16x4_t __rev0_791; __rev0_791 = __builtin_shufflevector(__s0_791, __s0_791, 3, 2, 1, 0); \
67922		int16x4_t __rev1_791; __rev1_791 = __builtin_shufflevector(__s1_791, __s1_791, 3, 2, 1, 0); \
67923		int16x8_t __rev2_791; __rev2_791 = __builtin_shufflevector(__s2_791, __s2_791, 7, 6, 5, 4, 3, 2, 1, 0); \
67924		__ret_791 = __noswap_vqrdmlsh_s16(__rev0_791, __rev1_791, __noswap_splat_laneq_s16(__rev2_791, __p3_791)); \
67925		__ret_791 = __builtin_shufflevector(__ret_791, __ret_791, 3, 2, 1, 0); \
	70249	int16x4_t __rev0_791; __rev0_791 = __builtin_shufflevector(__s0_791, __s0_791, __lane_reverse_64_16); \
	70250	int16x4_t __rev1_791; __rev1_791 = __builtin_shufflevector(__s1_791, __s1_791, __lane_reverse_64_16); \
	70251	int16x8_t __rev2_791; __rev2_791 = __builtin_shufflevector(__s2_791, __s2_791, __lane_reverse_128_16); \
	70252	__ret_791 = __noswap_vqrdmlah_s16(__rev0_791, __rev1_791, __noswap_splat_laneq_s16(__rev2_791, __p3_791)); \
	70253	__ret_791 = __builtin_shufflevector(__ret_791, __ret_791, __lane_reverse_64_16); \
67926	70254	__ret_791; \
67927	70255	})
67928	70256	#endif
67929	70257
	70258	__ai __attribute__((target("v8.1a,neon"))) int32_t vqrdmlshs_s32(int32_t __p0, int32_t __p1, int32_t __p2) {
	70259	int32_t __ret;
	70260	__ret = __builtin_bit_cast(int32_t, __builtin_neon_vqrdmlshs_s32(__p0, __p1, __p2));
	70261	return __ret;
	70262	}
	70263	__ai __attribute__((target("v8.1a,neon"))) int16_t vqrdmlshh_s16(int16_t __p0, int16_t __p1, int16_t __p2) {
	70264	int16_t __ret;
	70265	__ret = __builtin_bit_cast(int16_t, __builtin_neon_vqrdmlshh_s16(__p0, __p1, __p2));
	70266	return __ret;
	70267	}
	70268	#ifdef __LITTLE_ENDIAN__
	70269	#define vqrdmlshs_lane_s32(__p0_792, __p1_792, __p2_792, __p3_792) __extension__ ({ \
	70270	int32_t __ret_792; \
	70271	int32_t __s0_792 = __p0_792; \
	70272	int32_t __s1_792 = __p1_792; \
	70273	int32x2_t __s2_792 = __p2_792; \
	70274	__ret_792 = vqrdmlshs_s32(__s0_792, __s1_792, vget_lane_s32(__s2_792, __p3_792)); \
	70275	__ret_792; \
	70276	})
	70277	#else
	70278	#define vqrdmlshs_lane_s32(__p0_793, __p1_793, __p2_793, __p3_793) __extension__ ({ \
	70279	int32_t __ret_793; \
	70280	int32_t __s0_793 = __p0_793; \
	70281	int32_t __s1_793 = __p1_793; \
	70282	int32x2_t __s2_793 = __p2_793; \
	70283	int32x2_t __rev2_793; __rev2_793 = __builtin_shufflevector(__s2_793, __s2_793, __lane_reverse_64_32); \
	70284	__ret_793 = vqrdmlshs_s32(__s0_793, __s1_793, __noswap_vget_lane_s32(__rev2_793, __p3_793)); \
	70285	__ret_793; \
	70286	})
	70287	#endif
	70288
	70289	#ifdef __LITTLE_ENDIAN__
	70290	#define vqrdmlshh_lane_s16(__p0_794, __p1_794, __p2_794, __p3_794) __extension__ ({ \
	70291	int16_t __ret_794; \
	70292	int16_t __s0_794 = __p0_794; \
	70293	int16_t __s1_794 = __p1_794; \
	70294	int16x4_t __s2_794 = __p2_794; \
	70295	__ret_794 = vqrdmlshh_s16(__s0_794, __s1_794, vget_lane_s16(__s2_794, __p3_794)); \
	70296	__ret_794; \
	70297	})
	70298	#else
	70299	#define vqrdmlshh_lane_s16(__p0_795, __p1_795, __p2_795, __p3_795) __extension__ ({ \
	70300	int16_t __ret_795; \
	70301	int16_t __s0_795 = __p0_795; \
	70302	int16_t __s1_795 = __p1_795; \
	70303	int16x4_t __s2_795 = __p2_795; \
	70304	int16x4_t __rev2_795; __rev2_795 = __builtin_shufflevector(__s2_795, __s2_795, __lane_reverse_64_16); \
	70305	__ret_795 = vqrdmlshh_s16(__s0_795, __s1_795, __noswap_vget_lane_s16(__rev2_795, __p3_795)); \
	70306	__ret_795; \
	70307	})
	70308	#endif
	70309
	70310	#ifdef __LITTLE_ENDIAN__
	70311	#define vqrdmlshs_laneq_s32(__p0_796, __p1_796, __p2_796, __p3_796) __extension__ ({ \
	70312	int32_t __ret_796; \
	70313	int32_t __s0_796 = __p0_796; \
	70314	int32_t __s1_796 = __p1_796; \
	70315	int32x4_t __s2_796 = __p2_796; \
	70316	__ret_796 = vqrdmlshs_s32(__s0_796, __s1_796, vgetq_lane_s32(__s2_796, __p3_796)); \
	70317	__ret_796; \
	70318	})
	70319	#else
	70320	#define vqrdmlshs_laneq_s32(__p0_797, __p1_797, __p2_797, __p3_797) __extension__ ({ \
	70321	int32_t __ret_797; \
	70322	int32_t __s0_797 = __p0_797; \
	70323	int32_t __s1_797 = __p1_797; \
	70324	int32x4_t __s2_797 = __p2_797; \
	70325	int32x4_t __rev2_797; __rev2_797 = __builtin_shufflevector(__s2_797, __s2_797, __lane_reverse_128_32); \
	70326	__ret_797 = vqrdmlshs_s32(__s0_797, __s1_797, __noswap_vgetq_lane_s32(__rev2_797, __p3_797)); \
	70327	__ret_797; \
	70328	})
	70329	#endif
	70330
	70331	#ifdef __LITTLE_ENDIAN__
	70332	#define vqrdmlshh_laneq_s16(__p0_798, __p1_798, __p2_798, __p3_798) __extension__ ({ \
	70333	int16_t __ret_798; \
	70334	int16_t __s0_798 = __p0_798; \
	70335	int16_t __s1_798 = __p1_798; \
	70336	int16x8_t __s2_798 = __p2_798; \
	70337	__ret_798 = vqrdmlshh_s16(__s0_798, __s1_798, vgetq_lane_s16(__s2_798, __p3_798)); \
	70338	__ret_798; \
	70339	})
	70340	#else
	70341	#define vqrdmlshh_laneq_s16(__p0_799, __p1_799, __p2_799, __p3_799) __extension__ ({ \
	70342	int16_t __ret_799; \
	70343	int16_t __s0_799 = __p0_799; \
	70344	int16_t __s1_799 = __p1_799; \
	70345	int16x8_t __s2_799 = __p2_799; \
	70346	int16x8_t __rev2_799; __rev2_799 = __builtin_shufflevector(__s2_799, __s2_799, __lane_reverse_128_16); \
	70347	__ret_799 = vqrdmlshh_s16(__s0_799, __s1_799, __noswap_vgetq_lane_s16(__rev2_799, __p3_799)); \
	70348	__ret_799; \
	70349	})
	70350	#endif
	70351
	70352	#ifdef __LITTLE_ENDIAN__
	70353	#define vqrdmlshq_laneq_s32(__p0_800, __p1_800, __p2_800, __p3_800) __extension__ ({ \
	70354	int32x4_t __ret_800; \
	70355	int32x4_t __s0_800 = __p0_800; \
	70356	int32x4_t __s1_800 = __p1_800; \
	70357	int32x4_t __s2_800 = __p2_800; \
	70358	__ret_800 = vqrdmlshq_s32(__s0_800, __s1_800, splatq_laneq_s32(__s2_800, __p3_800)); \
	70359	__ret_800; \
	70360	})
	70361	#else
	70362	#define vqrdmlshq_laneq_s32(__p0_801, __p1_801, __p2_801, __p3_801) __extension__ ({ \
	70363	int32x4_t __ret_801; \
	70364	int32x4_t __s0_801 = __p0_801; \
	70365	int32x4_t __s1_801 = __p1_801; \
	70366	int32x4_t __s2_801 = __p2_801; \
	70367	int32x4_t __rev0_801; __rev0_801 = __builtin_shufflevector(__s0_801, __s0_801, __lane_reverse_128_32); \
	70368	int32x4_t __rev1_801; __rev1_801 = __builtin_shufflevector(__s1_801, __s1_801, __lane_reverse_128_32); \
	70369	int32x4_t __rev2_801; __rev2_801 = __builtin_shufflevector(__s2_801, __s2_801, __lane_reverse_128_32); \
	70370	__ret_801 = __noswap_vqrdmlshq_s32(__rev0_801, __rev1_801, __noswap_splatq_laneq_s32(__rev2_801, __p3_801)); \
	70371	__ret_801 = __builtin_shufflevector(__ret_801, __ret_801, __lane_reverse_128_32); \
	70372	__ret_801; \
	70373	})
	70374	#endif
	70375
	70376	#ifdef __LITTLE_ENDIAN__
	70377	#define vqrdmlshq_laneq_s16(__p0_802, __p1_802, __p2_802, __p3_802) __extension__ ({ \
	70378	int16x8_t __ret_802; \
	70379	int16x8_t __s0_802 = __p0_802; \
	70380	int16x8_t __s1_802 = __p1_802; \
	70381	int16x8_t __s2_802 = __p2_802; \
	70382	__ret_802 = vqrdmlshq_s16(__s0_802, __s1_802, splatq_laneq_s16(__s2_802, __p3_802)); \
	70383	__ret_802; \
	70384	})
	70385	#else
	70386	#define vqrdmlshq_laneq_s16(__p0_803, __p1_803, __p2_803, __p3_803) __extension__ ({ \
	70387	int16x8_t __ret_803; \
	70388	int16x8_t __s0_803 = __p0_803; \
	70389	int16x8_t __s1_803 = __p1_803; \
	70390	int16x8_t __s2_803 = __p2_803; \
	70391	int16x8_t __rev0_803; __rev0_803 = __builtin_shufflevector(__s0_803, __s0_803, __lane_reverse_128_16); \
	70392	int16x8_t __rev1_803; __rev1_803 = __builtin_shufflevector(__s1_803, __s1_803, __lane_reverse_128_16); \
	70393	int16x8_t __rev2_803; __rev2_803 = __builtin_shufflevector(__s2_803, __s2_803, __lane_reverse_128_16); \
	70394	__ret_803 = __noswap_vqrdmlshq_s16(__rev0_803, __rev1_803, __noswap_splatq_laneq_s16(__rev2_803, __p3_803)); \
	70395	__ret_803 = __builtin_shufflevector(__ret_803, __ret_803, __lane_reverse_128_16); \
	70396	__ret_803; \
	70397	})
	70398	#endif
	70399
	70400	#ifdef __LITTLE_ENDIAN__
	70401	#define vqrdmlsh_laneq_s32(__p0_804, __p1_804, __p2_804, __p3_804) __extension__ ({ \
	70402	int32x2_t __ret_804; \
	70403	int32x2_t __s0_804 = __p0_804; \
	70404	int32x2_t __s1_804 = __p1_804; \
	70405	int32x4_t __s2_804 = __p2_804; \
	70406	__ret_804 = vqrdmlsh_s32(__s0_804, __s1_804, splat_laneq_s32(__s2_804, __p3_804)); \
	70407	__ret_804; \
	70408	})
	70409	#else
	70410	#define vqrdmlsh_laneq_s32(__p0_805, __p1_805, __p2_805, __p3_805) __extension__ ({ \
	70411	int32x2_t __ret_805; \
	70412	int32x2_t __s0_805 = __p0_805; \
	70413	int32x2_t __s1_805 = __p1_805; \
	70414	int32x4_t __s2_805 = __p2_805; \
	70415	int32x2_t __rev0_805; __rev0_805 = __builtin_shufflevector(__s0_805, __s0_805, __lane_reverse_64_32); \
	70416	int32x2_t __rev1_805; __rev1_805 = __builtin_shufflevector(__s1_805, __s1_805, __lane_reverse_64_32); \
	70417	int32x4_t __rev2_805; __rev2_805 = __builtin_shufflevector(__s2_805, __s2_805, __lane_reverse_128_32); \
	70418	__ret_805 = __noswap_vqrdmlsh_s32(__rev0_805, __rev1_805, __noswap_splat_laneq_s32(__rev2_805, __p3_805)); \
	70419	__ret_805 = __builtin_shufflevector(__ret_805, __ret_805, __lane_reverse_64_32); \
	70420	__ret_805; \
	70421	})
	70422	#endif
	70423
	70424	#ifdef __LITTLE_ENDIAN__
	70425	#define vqrdmlsh_laneq_s16(__p0_806, __p1_806, __p2_806, __p3_806) __extension__ ({ \
	70426	int16x4_t __ret_806; \
	70427	int16x4_t __s0_806 = __p0_806; \
	70428	int16x4_t __s1_806 = __p1_806; \
	70429	int16x8_t __s2_806 = __p2_806; \
	70430	__ret_806 = vqrdmlsh_s16(__s0_806, __s1_806, splat_laneq_s16(__s2_806, __p3_806)); \
	70431	__ret_806; \
	70432	})
	70433	#else
	70434	#define vqrdmlsh_laneq_s16(__p0_807, __p1_807, __p2_807, __p3_807) __extension__ ({ \
	70435	int16x4_t __ret_807; \
	70436	int16x4_t __s0_807 = __p0_807; \
	70437	int16x4_t __s1_807 = __p1_807; \
	70438	int16x8_t __s2_807 = __p2_807; \
	70439	int16x4_t __rev0_807; __rev0_807 = __builtin_shufflevector(__s0_807, __s0_807, __lane_reverse_64_16); \
	70440	int16x4_t __rev1_807; __rev1_807 = __builtin_shufflevector(__s1_807, __s1_807, __lane_reverse_64_16); \
	70441	int16x8_t __rev2_807; __rev2_807 = __builtin_shufflevector(__s2_807, __s2_807, __lane_reverse_128_16); \
	70442	__ret_807 = __noswap_vqrdmlsh_s16(__rev0_807, __rev1_807, __noswap_splat_laneq_s16(__rev2_807, __p3_807)); \
	70443	__ret_807 = __builtin_shufflevector(__ret_807, __ret_807, __lane_reverse_64_16); \
	70444	__ret_807; \
	70445	})
	70446	#endif
	70447
67930	70448	#ifdef __LITTLE_ENDIAN__
67931	70449	__ai __attribute__((target("v8.3a,neon"))) float64x2_t vcaddq_rot270_f64(float64x2_t __p0, float64x2_t __p1) {
67932	70450	float64x2_t __ret;
67933		__ret = (float64x2_t) __builtin_neon_vcaddq_rot270_f64((int8x16_t)__p0, (int8x16_t)__p1, 42);
	70451	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vcaddq_rot270_f64(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 42));
67934	70452	return __ret;
67935	70453	}
67936	70454	#else
67937	70455	__ai __attribute__((target("v8.3a,neon"))) float64x2_t vcaddq_rot270_f64(float64x2_t __p0, float64x2_t __p1) {
67938	70456	float64x2_t __ret;
67939		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
67940		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
67941		__ret = (float64x2_t) __builtin_neon_vcaddq_rot270_f64((int8x16_t)__rev0, (int8x16_t)__rev1, 42);
67942		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	70457	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	70458	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	70459	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vcaddq_rot270_f64(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 42));
	70460	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
67943	70461	return __ret;
67944	70462	}
67945	70463	#endif
...	...	@@ -67947,16 +70465,16 @@ __ai __attribute__((target("v8.3a,neon"))) float64x2_t vcaddq_rot270_f64(float64
67947	70465	#ifdef __LITTLE_ENDIAN__
67948	70466	__ai __attribute__((target("v8.3a,neon"))) float64x2_t vcaddq_rot90_f64(float64x2_t __p0, float64x2_t __p1) {
67949	70467	float64x2_t __ret;
67950		__ret = (float64x2_t) __builtin_neon_vcaddq_rot90_f64((int8x16_t)__p0, (int8x16_t)__p1, 42);
	70468	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vcaddq_rot90_f64(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), 42));
67951	70469	return __ret;
67952	70470	}
67953	70471	#else
67954	70472	__ai __attribute__((target("v8.3a,neon"))) float64x2_t vcaddq_rot90_f64(float64x2_t __p0, float64x2_t __p1) {
67955	70473	float64x2_t __ret;
67956		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
67957		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
67958		__ret = (float64x2_t) __builtin_neon_vcaddq_rot90_f64((int8x16_t)__rev0, (int8x16_t)__rev1, 42);
67959		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	70474	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	70475	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	70476	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vcaddq_rot90_f64(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), 42));
	70477	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
67960	70478	return __ret;
67961	70479	}
67962	70480	#endif
...	...	@@ -67964,17 +70482,17 @@ __ai __attribute__((target("v8.3a,neon"))) float64x2_t vcaddq_rot90_f64(float64x
67964	70482	#ifdef __LITTLE_ENDIAN__
67965	70483	__ai __attribute__((target("v8.3a,neon"))) float64x2_t vcmlaq_f64(float64x2_t __p0, float64x2_t __p1, float64x2_t __p2) {
67966	70484	float64x2_t __ret;
67967		__ret = (float64x2_t) __builtin_neon_vcmlaq_f64((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 42);
	70485	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vcmlaq_f64(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 42));
67968	70486	return __ret;
67969	70487	}
67970	70488	#else
67971	70489	__ai __attribute__((target("v8.3a,neon"))) float64x2_t vcmlaq_f64(float64x2_t __p0, float64x2_t __p1, float64x2_t __p2) {
67972	70490	float64x2_t __ret;
67973		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
67974		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
67975		float64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
67976		__ret = (float64x2_t) __builtin_neon_vcmlaq_f64((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 42);
67977		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	70491	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	70492	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	70493	float64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_64);
	70494	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vcmlaq_f64(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 42));
	70495	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
67978	70496	return __ret;
67979	70497	}
67980	70498	#endif
...	...	@@ -67982,17 +70500,17 @@ __ai __attribute__((target("v8.3a,neon"))) float64x2_t vcmlaq_f64(float64x2_t __
67982	70500	#ifdef __LITTLE_ENDIAN__
67983	70501	__ai __attribute__((target("v8.3a,neon"))) float64x2_t vcmlaq_rot180_f64(float64x2_t __p0, float64x2_t __p1, float64x2_t __p2) {
67984	70502	float64x2_t __ret;
67985		__ret = (float64x2_t) __builtin_neon_vcmlaq_rot180_f64((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 42);
	70503	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vcmlaq_rot180_f64(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 42));
67986	70504	return __ret;
67987	70505	}
67988	70506	#else
67989	70507	__ai __attribute__((target("v8.3a,neon"))) float64x2_t vcmlaq_rot180_f64(float64x2_t __p0, float64x2_t __p1, float64x2_t __p2) {
67990	70508	float64x2_t __ret;
67991		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
67992		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
67993		float64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
67994		__ret = (float64x2_t) __builtin_neon_vcmlaq_rot180_f64((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 42);
67995		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	70509	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	70510	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	70511	float64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_64);
	70512	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vcmlaq_rot180_f64(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 42));
	70513	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
67996	70514	return __ret;
67997	70515	}
67998	70516	#endif
...	...	@@ -68000,17 +70518,17 @@ __ai __attribute__((target("v8.3a,neon"))) float64x2_t vcmlaq_rot180_f64(float64
68000	70518	#ifdef __LITTLE_ENDIAN__
68001	70519	__ai __attribute__((target("v8.3a,neon"))) float64x2_t vcmlaq_rot270_f64(float64x2_t __p0, float64x2_t __p1, float64x2_t __p2) {
68002	70520	float64x2_t __ret;
68003		__ret = (float64x2_t) __builtin_neon_vcmlaq_rot270_f64((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 42);
	70521	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vcmlaq_rot270_f64(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 42));
68004	70522	return __ret;
68005	70523	}
68006	70524	#else
68007	70525	__ai __attribute__((target("v8.3a,neon"))) float64x2_t vcmlaq_rot270_f64(float64x2_t __p0, float64x2_t __p1, float64x2_t __p2) {
68008	70526	float64x2_t __ret;
68009		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
68010		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
68011		float64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
68012		__ret = (float64x2_t) __builtin_neon_vcmlaq_rot270_f64((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 42);
68013		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	70527	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	70528	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	70529	float64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_64);
	70530	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vcmlaq_rot270_f64(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 42));
	70531	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
68014	70532	return __ret;
68015	70533	}
68016	70534	#endif
...	...	@@ -68018,17 +70536,17 @@ __ai __attribute__((target("v8.3a,neon"))) float64x2_t vcmlaq_rot270_f64(float64
68018	70536	#ifdef __LITTLE_ENDIAN__
68019	70537	__ai __attribute__((target("v8.3a,neon"))) float64x2_t vcmlaq_rot90_f64(float64x2_t __p0, float64x2_t __p1, float64x2_t __p2) {
68020	70538	float64x2_t __ret;
68021		__ret = (float64x2_t) __builtin_neon_vcmlaq_rot90_f64((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 42);
	70539	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vcmlaq_rot90_f64(__builtin_bit_cast(int8x16_t, __p0), __builtin_bit_cast(int8x16_t, __p1), __builtin_bit_cast(int8x16_t, __p2), 42));
68022	70540	return __ret;
68023	70541	}
68024	70542	#else
68025	70543	__ai __attribute__((target("v8.3a,neon"))) float64x2_t vcmlaq_rot90_f64(float64x2_t __p0, float64x2_t __p1, float64x2_t __p2) {
68026	70544	float64x2_t __ret;
68027		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
68028		float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
68029		float64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
68030		__ret = (float64x2_t) __builtin_neon_vcmlaq_rot90_f64((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 42);
68031		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	70545	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	70546	float64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	70547	float64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_64);
	70548	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vcmlaq_rot90_f64(__builtin_bit_cast(int8x16_t, __rev0), __builtin_bit_cast(int8x16_t, __rev1), __builtin_bit_cast(int8x16_t, __rev2), 42));
	70549	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
68032	70550	return __ret;
68033	70551	}
68034	70552	#endif
...	...	@@ -68036,15 +70554,15 @@ __ai __attribute__((target("v8.3a,neon"))) float64x2_t vcmlaq_rot90_f64(float64x
68036	70554	#ifdef __LITTLE_ENDIAN__
68037	70555	__ai __attribute__((target("v8.5a,neon"))) float32x4_t vrnd32xq_f32(float32x4_t __p0) {
68038	70556	float32x4_t __ret;
68039		__ret = (float32x4_t) __builtin_neon_vrnd32xq_f32((int8x16_t)__p0, 41);
	70557	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vrnd32xq_f32(__builtin_bit_cast(int8x16_t, __p0), 41));
68040	70558	return __ret;
68041	70559	}
68042	70560	#else
68043	70561	__ai __attribute__((target("v8.5a,neon"))) float32x4_t vrnd32xq_f32(float32x4_t __p0) {
68044	70562	float32x4_t __ret;
68045		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
68046		__ret = (float32x4_t) __builtin_neon_vrnd32xq_f32((int8x16_t)__rev0, 41);
68047		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	70563	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	70564	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vrnd32xq_f32(__builtin_bit_cast(int8x16_t, __rev0), 41));
	70565	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
68048	70566	return __ret;
68049	70567	}
68050	70568	#endif
...	...	@@ -68052,15 +70570,15 @@ __ai __attribute__((target("v8.5a,neon"))) float32x4_t vrnd32xq_f32(float32x4_t
68052	70570	#ifdef __LITTLE_ENDIAN__
68053	70571	__ai __attribute__((target("v8.5a,neon"))) float32x2_t vrnd32x_f32(float32x2_t __p0) {
68054	70572	float32x2_t __ret;
68055		__ret = (float32x2_t) __builtin_neon_vrnd32x_f32((int8x8_t)__p0, 9);
	70573	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vrnd32x_f32(__builtin_bit_cast(int8x8_t, __p0), 9));
68056	70574	return __ret;
68057	70575	}
68058	70576	#else
68059	70577	__ai __attribute__((target("v8.5a,neon"))) float32x2_t vrnd32x_f32(float32x2_t __p0) {
68060	70578	float32x2_t __ret;
68061		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
68062		__ret = (float32x2_t) __builtin_neon_vrnd32x_f32((int8x8_t)__rev0, 9);
68063		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	70579	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	70580	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vrnd32x_f32(__builtin_bit_cast(int8x8_t, __rev0), 9));
	70581	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
68064	70582	return __ret;
68065	70583	}
68066	70584	#endif
...	...	@@ -68068,36 +70586,36 @@ __ai __attribute__((target("v8.5a,neon"))) float32x2_t vrnd32x_f32(float32x2_t _
68068	70586	#ifdef __LITTLE_ENDIAN__
68069	70587	__ai __attribute__((target("v8.5a,neon"))) float64x2_t vrnd32xq_f64(float64x2_t __p0) {
68070	70588	float64x2_t __ret;
68071		__ret = (float64x2_t) __builtin_neon_vrnd32xq_f64((int8x16_t)__p0, 42);
	70589	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vrnd32xq_f64(__builtin_bit_cast(int8x16_t, __p0), 42));
68072	70590	return __ret;
68073	70591	}
68074	70592	#else
68075	70593	__ai __attribute__((target("v8.5a,neon"))) float64x2_t vrnd32xq_f64(float64x2_t __p0) {
68076	70594	float64x2_t __ret;
68077		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
68078		__ret = (float64x2_t) __builtin_neon_vrnd32xq_f64((int8x16_t)__rev0, 42);
68079		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	70595	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	70596	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vrnd32xq_f64(__builtin_bit_cast(int8x16_t, __rev0), 42));
	70597	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
68080	70598	return __ret;
68081	70599	}
68082	70600	#endif
68083	70601
68084	70602	__ai __attribute__((target("v8.5a,neon"))) float64x1_t vrnd32x_f64(float64x1_t __p0) {
68085	70603	float64x1_t __ret;
68086		__ret = (float64x1_t) __builtin_neon_vrnd32x_f64((int8x8_t)__p0, 10);
	70604	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vrnd32x_f64(__builtin_bit_cast(int8x8_t, __p0), 10));
68087	70605	return __ret;
68088	70606	}
68089	70607	#ifdef __LITTLE_ENDIAN__
68090	70608	__ai __attribute__((target("v8.5a,neon"))) float32x4_t vrnd32zq_f32(float32x4_t __p0) {
68091	70609	float32x4_t __ret;
68092		__ret = (float32x4_t) __builtin_neon_vrnd32zq_f32((int8x16_t)__p0, 41);
	70610	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vrnd32zq_f32(__builtin_bit_cast(int8x16_t, __p0), 41));
68093	70611	return __ret;
68094	70612	}
68095	70613	#else
68096	70614	__ai __attribute__((target("v8.5a,neon"))) float32x4_t vrnd32zq_f32(float32x4_t __p0) {
68097	70615	float32x4_t __ret;
68098		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
68099		__ret = (float32x4_t) __builtin_neon_vrnd32zq_f32((int8x16_t)__rev0, 41);
68100		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	70616	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	70617	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vrnd32zq_f32(__builtin_bit_cast(int8x16_t, __rev0), 41));
	70618	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
68101	70619	return __ret;
68102	70620	}
68103	70621	#endif
...	...	@@ -68105,15 +70623,15 @@ __ai __attribute__((target("v8.5a,neon"))) float32x4_t vrnd32zq_f32(float32x4_t
68105	70623	#ifdef __LITTLE_ENDIAN__
68106	70624	__ai __attribute__((target("v8.5a,neon"))) float32x2_t vrnd32z_f32(float32x2_t __p0) {
68107	70625	float32x2_t __ret;
68108		__ret = (float32x2_t) __builtin_neon_vrnd32z_f32((int8x8_t)__p0, 9);
	70626	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vrnd32z_f32(__builtin_bit_cast(int8x8_t, __p0), 9));
68109	70627	return __ret;
68110	70628	}
68111	70629	#else
68112	70630	__ai __attribute__((target("v8.5a,neon"))) float32x2_t vrnd32z_f32(float32x2_t __p0) {
68113	70631	float32x2_t __ret;
68114		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
68115		__ret = (float32x2_t) __builtin_neon_vrnd32z_f32((int8x8_t)__rev0, 9);
68116		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	70632	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	70633	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vrnd32z_f32(__builtin_bit_cast(int8x8_t, __rev0), 9));
	70634	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
68117	70635	return __ret;
68118	70636	}
68119	70637	#endif
...	...	@@ -68121,36 +70639,36 @@ __ai __attribute__((target("v8.5a,neon"))) float32x2_t vrnd32z_f32(float32x2_t _
68121	70639	#ifdef __LITTLE_ENDIAN__
68122	70640	__ai __attribute__((target("v8.5a,neon"))) float64x2_t vrnd32zq_f64(float64x2_t __p0) {
68123	70641	float64x2_t __ret;
68124		__ret = (float64x2_t) __builtin_neon_vrnd32zq_f64((int8x16_t)__p0, 42);
	70642	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vrnd32zq_f64(__builtin_bit_cast(int8x16_t, __p0), 42));
68125	70643	return __ret;
68126	70644	}
68127	70645	#else
68128	70646	__ai __attribute__((target("v8.5a,neon"))) float64x2_t vrnd32zq_f64(float64x2_t __p0) {
68129	70647	float64x2_t __ret;
68130		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
68131		__ret = (float64x2_t) __builtin_neon_vrnd32zq_f64((int8x16_t)__rev0, 42);
68132		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	70648	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	70649	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vrnd32zq_f64(__builtin_bit_cast(int8x16_t, __rev0), 42));
	70650	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
68133	70651	return __ret;
68134	70652	}
68135	70653	#endif
68136	70654
68137	70655	__ai __attribute__((target("v8.5a,neon"))) float64x1_t vrnd32z_f64(float64x1_t __p0) {
68138	70656	float64x1_t __ret;
68139		__ret = (float64x1_t) __builtin_neon_vrnd32z_f64((int8x8_t)__p0, 10);
	70657	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vrnd32z_f64(__builtin_bit_cast(int8x8_t, __p0), 10));
68140	70658	return __ret;
68141	70659	}
68142	70660	#ifdef __LITTLE_ENDIAN__
68143	70661	__ai __attribute__((target("v8.5a,neon"))) float32x4_t vrnd64xq_f32(float32x4_t __p0) {
68144	70662	float32x4_t __ret;
68145		__ret = (float32x4_t) __builtin_neon_vrnd64xq_f32((int8x16_t)__p0, 41);
	70663	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vrnd64xq_f32(__builtin_bit_cast(int8x16_t, __p0), 41));
68146	70664	return __ret;
68147	70665	}
68148	70666	#else
68149	70667	__ai __attribute__((target("v8.5a,neon"))) float32x4_t vrnd64xq_f32(float32x4_t __p0) {
68150	70668	float32x4_t __ret;
68151		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
68152		__ret = (float32x4_t) __builtin_neon_vrnd64xq_f32((int8x16_t)__rev0, 41);
68153		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	70669	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	70670	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vrnd64xq_f32(__builtin_bit_cast(int8x16_t, __rev0), 41));
	70671	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
68154	70672	return __ret;
68155	70673	}
68156	70674	#endif
...	...	@@ -68158,15 +70676,15 @@ __ai __attribute__((target("v8.5a,neon"))) float32x4_t vrnd64xq_f32(float32x4_t
68158	70676	#ifdef __LITTLE_ENDIAN__
68159	70677	__ai __attribute__((target("v8.5a,neon"))) float32x2_t vrnd64x_f32(float32x2_t __p0) {
68160	70678	float32x2_t __ret;
68161		__ret = (float32x2_t) __builtin_neon_vrnd64x_f32((int8x8_t)__p0, 9);
	70679	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vrnd64x_f32(__builtin_bit_cast(int8x8_t, __p0), 9));
68162	70680	return __ret;
68163	70681	}
68164	70682	#else
68165	70683	__ai __attribute__((target("v8.5a,neon"))) float32x2_t vrnd64x_f32(float32x2_t __p0) {
68166	70684	float32x2_t __ret;
68167		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
68168		__ret = (float32x2_t) __builtin_neon_vrnd64x_f32((int8x8_t)__rev0, 9);
68169		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	70685	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	70686	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vrnd64x_f32(__builtin_bit_cast(int8x8_t, __rev0), 9));
	70687	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
68170	70688	return __ret;
68171	70689	}
68172	70690	#endif
...	...	@@ -68174,36 +70692,36 @@ __ai __attribute__((target("v8.5a,neon"))) float32x2_t vrnd64x_f32(float32x2_t _
68174	70692	#ifdef __LITTLE_ENDIAN__
68175	70693	__ai __attribute__((target("v8.5a,neon"))) float64x2_t vrnd64xq_f64(float64x2_t __p0) {
68176	70694	float64x2_t __ret;
68177		__ret = (float64x2_t) __builtin_neon_vrnd64xq_f64((int8x16_t)__p0, 42);
	70695	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vrnd64xq_f64(__builtin_bit_cast(int8x16_t, __p0), 42));
68178	70696	return __ret;
68179	70697	}
68180	70698	#else
68181	70699	__ai __attribute__((target("v8.5a,neon"))) float64x2_t vrnd64xq_f64(float64x2_t __p0) {
68182	70700	float64x2_t __ret;
68183		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
68184		__ret = (float64x2_t) __builtin_neon_vrnd64xq_f64((int8x16_t)__rev0, 42);
68185		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	70701	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	70702	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vrnd64xq_f64(__builtin_bit_cast(int8x16_t, __rev0), 42));
	70703	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
68186	70704	return __ret;
68187	70705	}
68188	70706	#endif
68189	70707
68190	70708	__ai __attribute__((target("v8.5a,neon"))) float64x1_t vrnd64x_f64(float64x1_t __p0) {
68191	70709	float64x1_t __ret;
68192		__ret = (float64x1_t) __builtin_neon_vrnd64x_f64((int8x8_t)__p0, 10);
	70710	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vrnd64x_f64(__builtin_bit_cast(int8x8_t, __p0), 10));
68193	70711	return __ret;
68194	70712	}
68195	70713	#ifdef __LITTLE_ENDIAN__
68196	70714	__ai __attribute__((target("v8.5a,neon"))) float32x4_t vrnd64zq_f32(float32x4_t __p0) {
68197	70715	float32x4_t __ret;
68198		__ret = (float32x4_t) __builtin_neon_vrnd64zq_f32((int8x16_t)__p0, 41);
	70716	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vrnd64zq_f32(__builtin_bit_cast(int8x16_t, __p0), 41));
68199	70717	return __ret;
68200	70718	}
68201	70719	#else
68202	70720	__ai __attribute__((target("v8.5a,neon"))) float32x4_t vrnd64zq_f32(float32x4_t __p0) {
68203	70721	float32x4_t __ret;
68204		float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
68205		__ret = (float32x4_t) __builtin_neon_vrnd64zq_f32((int8x16_t)__rev0, 41);
68206		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	70722	float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	70723	__ret = __builtin_bit_cast(float32x4_t, __builtin_neon_vrnd64zq_f32(__builtin_bit_cast(int8x16_t, __rev0), 41));
	70724	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
68207	70725	return __ret;
68208	70726	}
68209	70727	#endif
...	...	@@ -68211,15 +70729,15 @@ __ai __attribute__((target("v8.5a,neon"))) float32x4_t vrnd64zq_f32(float32x4_t
68211	70729	#ifdef __LITTLE_ENDIAN__
68212	70730	__ai __attribute__((target("v8.5a,neon"))) float32x2_t vrnd64z_f32(float32x2_t __p0) {
68213	70731	float32x2_t __ret;
68214		__ret = (float32x2_t) __builtin_neon_vrnd64z_f32((int8x8_t)__p0, 9);
	70732	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vrnd64z_f32(__builtin_bit_cast(int8x8_t, __p0), 9));
68215	70733	return __ret;
68216	70734	}
68217	70735	#else
68218	70736	__ai __attribute__((target("v8.5a,neon"))) float32x2_t vrnd64z_f32(float32x2_t __p0) {
68219	70737	float32x2_t __ret;
68220		float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
68221		__ret = (float32x2_t) __builtin_neon_vrnd64z_f32((int8x8_t)__rev0, 9);
68222		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	70738	float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	70739	__ret = __builtin_bit_cast(float32x2_t, __builtin_neon_vrnd64z_f32(__builtin_bit_cast(int8x8_t, __rev0), 9));
	70740	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
68223	70741	return __ret;
68224	70742	}
68225	70743	#endif
...	...	@@ -68227,235 +70745,235 @@ __ai __attribute__((target("v8.5a,neon"))) float32x2_t vrnd64z_f32(float32x2_t _
68227	70745	#ifdef __LITTLE_ENDIAN__
68228	70746	__ai __attribute__((target("v8.5a,neon"))) float64x2_t vrnd64zq_f64(float64x2_t __p0) {
68229	70747	float64x2_t __ret;
68230		__ret = (float64x2_t) __builtin_neon_vrnd64zq_f64((int8x16_t)__p0, 42);
	70748	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vrnd64zq_f64(__builtin_bit_cast(int8x16_t, __p0), 42));
68231	70749	return __ret;
68232	70750	}
68233	70751	#else
68234	70752	__ai __attribute__((target("v8.5a,neon"))) float64x2_t vrnd64zq_f64(float64x2_t __p0) {
68235	70753	float64x2_t __ret;
68236		float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
68237		__ret = (float64x2_t) __builtin_neon_vrnd64zq_f64((int8x16_t)__rev0, 42);
68238		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	70754	float64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	70755	__ret = __builtin_bit_cast(float64x2_t, __builtin_neon_vrnd64zq_f64(__builtin_bit_cast(int8x16_t, __rev0), 42));
	70756	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
68239	70757	return __ret;
68240	70758	}
68241	70759	#endif
68242	70760
68243	70761	__ai __attribute__((target("v8.5a,neon"))) float64x1_t vrnd64z_f64(float64x1_t __p0) {
68244	70762	float64x1_t __ret;
68245		__ret = (float64x1_t) __builtin_neon_vrnd64z_f64((int8x8_t)__p0, 10);
	70763	__ret = __builtin_bit_cast(float64x1_t, __builtin_neon_vrnd64z_f64(__builtin_bit_cast(int8x8_t, __p0), 10));
68246	70764	return __ret;
68247	70765	}
68248	70766	#endif
68249	70767	#ifdef __LITTLE_ENDIAN__
68250		#define vbfdotq_lane_f32(__p0_792, __p1_792, __p2_792, __p3_792) __extension__ ({ \
68251		float32x4_t __ret_792; \
68252		float32x4_t __s0_792 = __p0_792; \
68253		bfloat16x8_t __s1_792 = __p1_792; \
68254		bfloat16x4_t __s2_792 = __p2_792; \
68255		__ret_792 = vbfdotq_f32(__s0_792, __s1_792, __builtin_bit_cast(bfloat16x8_t, splatq_lane_f32(__builtin_bit_cast(float32x2_t, __s2_792), __p3_792))); \
68256		__ret_792; \
	70768	#define vbfdotq_lane_f32(__p0_808, __p1_808, __p2_808, __p3_808) __extension__ ({ \
	70769	float32x4_t __ret_808; \
	70770	float32x4_t __s0_808 = __p0_808; \
	70771	bfloat16x8_t __s1_808 = __p1_808; \
	70772	bfloat16x4_t __s2_808 = __p2_808; \
	70773	__ret_808 = vbfdotq_f32(__s0_808, __s1_808, __builtin_bit_cast(bfloat16x8_t, splatq_lane_f32(__builtin_bit_cast(float32x2_t, __s2_808), __p3_808))); \
	70774	__ret_808; \
68257	70775	})
68258	70776	#else
68259		#define vbfdotq_lane_f32(__p0_793, __p1_793, __p2_793, __p3_793) __extension__ ({ \
68260		float32x4_t __ret_793; \
68261		float32x4_t __s0_793 = __p0_793; \
68262		bfloat16x8_t __s1_793 = __p1_793; \
68263		bfloat16x4_t __s2_793 = __p2_793; \
68264		float32x4_t __rev0_793; __rev0_793 = __builtin_shufflevector(__s0_793, __s0_793, 3, 2, 1, 0); \
68265		bfloat16x8_t __rev1_793; __rev1_793 = __builtin_shufflevector(__s1_793, __s1_793, 7, 6, 5, 4, 3, 2, 1, 0); \
68266		bfloat16x4_t __rev2_793; __rev2_793 = __builtin_shufflevector(__s2_793, __s2_793, 3, 2, 1, 0); \
68267		__ret_793 = __noswap_vbfdotq_f32(__rev0_793, __rev1_793, __builtin_bit_cast(bfloat16x8_t, __noswap_splatq_lane_f32(__builtin_bit_cast(float32x2_t, __rev2_793), __p3_793))); \
68268		__ret_793 = __builtin_shufflevector(__ret_793, __ret_793, 3, 2, 1, 0); \
68269		__ret_793; \
	70777	#define vbfdotq_lane_f32(__p0_809, __p1_809, __p2_809, __p3_809) __extension__ ({ \
	70778	float32x4_t __ret_809; \
	70779	float32x4_t __s0_809 = __p0_809; \
	70780	bfloat16x8_t __s1_809 = __p1_809; \
	70781	bfloat16x4_t __s2_809 = __p2_809; \
	70782	float32x4_t __rev0_809; __rev0_809 = __builtin_shufflevector(__s0_809, __s0_809, __lane_reverse_128_32); \
	70783	bfloat16x8_t __rev1_809; __rev1_809 = __builtin_shufflevector(__s1_809, __s1_809, __lane_reverse_128_16); \
	70784	bfloat16x4_t __rev2_809; __rev2_809 = __builtin_shufflevector(__s2_809, __s2_809, __lane_reverse_64_16); \
	70785	__ret_809 = __noswap_vbfdotq_f32(__rev0_809, __rev1_809, __builtin_bit_cast(bfloat16x8_t, __noswap_splatq_lane_f32(__builtin_bit_cast(float32x2_t, __rev2_809), __p3_809))); \
	70786	__ret_809 = __builtin_shufflevector(__ret_809, __ret_809, __lane_reverse_128_32); \
	70787	__ret_809; \
68270	70788	})
68271	70789	#endif
68272	70790
68273	70791	#ifdef __LITTLE_ENDIAN__
68274		#define vbfdot_lane_f32(__p0_794, __p1_794, __p2_794, __p3_794) __extension__ ({ \
68275		float32x2_t __ret_794; \
68276		float32x2_t __s0_794 = __p0_794; \
68277		bfloat16x4_t __s1_794 = __p1_794; \
68278		bfloat16x4_t __s2_794 = __p2_794; \
68279		__ret_794 = vbfdot_f32(__s0_794, __s1_794, __builtin_bit_cast(bfloat16x4_t, splat_lane_f32(__builtin_bit_cast(float32x2_t, __s2_794), __p3_794))); \
68280		__ret_794; \
	70792	#define vbfdot_lane_f32(__p0_810, __p1_810, __p2_810, __p3_810) __extension__ ({ \
	70793	float32x2_t __ret_810; \
	70794	float32x2_t __s0_810 = __p0_810; \
	70795	bfloat16x4_t __s1_810 = __p1_810; \
	70796	bfloat16x4_t __s2_810 = __p2_810; \
	70797	__ret_810 = vbfdot_f32(__s0_810, __s1_810, __builtin_bit_cast(bfloat16x4_t, splat_lane_f32(__builtin_bit_cast(float32x2_t, __s2_810), __p3_810))); \
	70798	__ret_810; \
68281	70799	})
68282	70800	#else
68283		#define vbfdot_lane_f32(__p0_795, __p1_795, __p2_795, __p3_795) __extension__ ({ \
68284		float32x2_t __ret_795; \
68285		float32x2_t __s0_795 = __p0_795; \
68286		bfloat16x4_t __s1_795 = __p1_795; \
68287		bfloat16x4_t __s2_795 = __p2_795; \
68288		float32x2_t __rev0_795; __rev0_795 = __builtin_shufflevector(__s0_795, __s0_795, 1, 0); \
68289		bfloat16x4_t __rev1_795; __rev1_795 = __builtin_shufflevector(__s1_795, __s1_795, 3, 2, 1, 0); \
68290		bfloat16x4_t __rev2_795; __rev2_795 = __builtin_shufflevector(__s2_795, __s2_795, 3, 2, 1, 0); \
68291		__ret_795 = __noswap_vbfdot_f32(__rev0_795, __rev1_795, __builtin_bit_cast(bfloat16x4_t, __noswap_splat_lane_f32(__builtin_bit_cast(float32x2_t, __rev2_795), __p3_795))); \
68292		__ret_795 = __builtin_shufflevector(__ret_795, __ret_795, 1, 0); \
68293		__ret_795; \
	70801	#define vbfdot_lane_f32(__p0_811, __p1_811, __p2_811, __p3_811) __extension__ ({ \
	70802	float32x2_t __ret_811; \
	70803	float32x2_t __s0_811 = __p0_811; \
	70804	bfloat16x4_t __s1_811 = __p1_811; \
	70805	bfloat16x4_t __s2_811 = __p2_811; \
	70806	float32x2_t __rev0_811; __rev0_811 = __builtin_shufflevector(__s0_811, __s0_811, __lane_reverse_64_32); \
	70807	bfloat16x4_t __rev1_811; __rev1_811 = __builtin_shufflevector(__s1_811, __s1_811, __lane_reverse_64_16); \
	70808	bfloat16x4_t __rev2_811; __rev2_811 = __builtin_shufflevector(__s2_811, __s2_811, __lane_reverse_64_16); \
	70809	__ret_811 = __noswap_vbfdot_f32(__rev0_811, __rev1_811, __builtin_bit_cast(bfloat16x4_t, __noswap_splat_lane_f32(__builtin_bit_cast(float32x2_t, __rev2_811), __p3_811))); \
	70810	__ret_811 = __builtin_shufflevector(__ret_811, __ret_811, __lane_reverse_64_32); \
	70811	__ret_811; \
68294	70812	})
68295	70813	#endif
68296	70814
68297	70815	#ifdef __LITTLE_ENDIAN__
68298		#define vbfdotq_laneq_f32(__p0_796, __p1_796, __p2_796, __p3_796) __extension__ ({ \
68299		float32x4_t __ret_796; \
68300		float32x4_t __s0_796 = __p0_796; \
68301		bfloat16x8_t __s1_796 = __p1_796; \
68302		bfloat16x8_t __s2_796 = __p2_796; \
68303		__ret_796 = vbfdotq_f32(__s0_796, __s1_796, __builtin_bit_cast(bfloat16x8_t, splatq_laneq_f32(__builtin_bit_cast(float32x4_t, __s2_796), __p3_796))); \
68304		__ret_796; \
	70816	#define vbfdotq_laneq_f32(__p0_812, __p1_812, __p2_812, __p3_812) __extension__ ({ \
	70817	float32x4_t __ret_812; \
	70818	float32x4_t __s0_812 = __p0_812; \
	70819	bfloat16x8_t __s1_812 = __p1_812; \
	70820	bfloat16x8_t __s2_812 = __p2_812; \
	70821	__ret_812 = vbfdotq_f32(__s0_812, __s1_812, __builtin_bit_cast(bfloat16x8_t, splatq_laneq_f32(__builtin_bit_cast(float32x4_t, __s2_812), __p3_812))); \
	70822	__ret_812; \
68305	70823	})
68306	70824	#else
68307		#define vbfdotq_laneq_f32(__p0_797, __p1_797, __p2_797, __p3_797) __extension__ ({ \
68308		float32x4_t __ret_797; \
68309		float32x4_t __s0_797 = __p0_797; \
68310		bfloat16x8_t __s1_797 = __p1_797; \
68311		bfloat16x8_t __s2_797 = __p2_797; \
68312		float32x4_t __rev0_797; __rev0_797 = __builtin_shufflevector(__s0_797, __s0_797, 3, 2, 1, 0); \
68313		bfloat16x8_t __rev1_797; __rev1_797 = __builtin_shufflevector(__s1_797, __s1_797, 7, 6, 5, 4, 3, 2, 1, 0); \
68314		bfloat16x8_t __rev2_797; __rev2_797 = __builtin_shufflevector(__s2_797, __s2_797, 7, 6, 5, 4, 3, 2, 1, 0); \
68315		__ret_797 = __noswap_vbfdotq_f32(__rev0_797, __rev1_797, __builtin_bit_cast(bfloat16x8_t, __noswap_splatq_laneq_f32(__builtin_bit_cast(float32x4_t, __rev2_797), __p3_797))); \
68316		__ret_797 = __builtin_shufflevector(__ret_797, __ret_797, 3, 2, 1, 0); \
68317		__ret_797; \
	70825	#define vbfdotq_laneq_f32(__p0_813, __p1_813, __p2_813, __p3_813) __extension__ ({ \
	70826	float32x4_t __ret_813; \
	70827	float32x4_t __s0_813 = __p0_813; \
	70828	bfloat16x8_t __s1_813 = __p1_813; \
	70829	bfloat16x8_t __s2_813 = __p2_813; \
	70830	float32x4_t __rev0_813; __rev0_813 = __builtin_shufflevector(__s0_813, __s0_813, __lane_reverse_128_32); \
	70831	bfloat16x8_t __rev1_813; __rev1_813 = __builtin_shufflevector(__s1_813, __s1_813, __lane_reverse_128_16); \
	70832	bfloat16x8_t __rev2_813; __rev2_813 = __builtin_shufflevector(__s2_813, __s2_813, __lane_reverse_128_16); \
	70833	__ret_813 = __noswap_vbfdotq_f32(__rev0_813, __rev1_813, __builtin_bit_cast(bfloat16x8_t, __noswap_splatq_laneq_f32(__builtin_bit_cast(float32x4_t, __rev2_813), __p3_813))); \
	70834	__ret_813 = __builtin_shufflevector(__ret_813, __ret_813, __lane_reverse_128_32); \
	70835	__ret_813; \
68318	70836	})
68319	70837	#endif
68320	70838
68321	70839	#ifdef __LITTLE_ENDIAN__
68322		#define vbfdot_laneq_f32(__p0_798, __p1_798, __p2_798, __p3_798) __extension__ ({ \
68323		float32x2_t __ret_798; \
68324		float32x2_t __s0_798 = __p0_798; \
68325		bfloat16x4_t __s1_798 = __p1_798; \
68326		bfloat16x8_t __s2_798 = __p2_798; \
68327		__ret_798 = vbfdot_f32(__s0_798, __s1_798, __builtin_bit_cast(bfloat16x4_t, splat_laneq_f32(__builtin_bit_cast(float32x4_t, __s2_798), __p3_798))); \
68328		__ret_798; \
	70840	#define vbfdot_laneq_f32(__p0_814, __p1_814, __p2_814, __p3_814) __extension__ ({ \
	70841	float32x2_t __ret_814; \
	70842	float32x2_t __s0_814 = __p0_814; \
	70843	bfloat16x4_t __s1_814 = __p1_814; \
	70844	bfloat16x8_t __s2_814 = __p2_814; \
	70845	__ret_814 = vbfdot_f32(__s0_814, __s1_814, __builtin_bit_cast(bfloat16x4_t, splat_laneq_f32(__builtin_bit_cast(float32x4_t, __s2_814), __p3_814))); \
	70846	__ret_814; \
68329	70847	})
68330	70848	#else
68331		#define vbfdot_laneq_f32(__p0_799, __p1_799, __p2_799, __p3_799) __extension__ ({ \
68332		float32x2_t __ret_799; \
68333		float32x2_t __s0_799 = __p0_799; \
68334		bfloat16x4_t __s1_799 = __p1_799; \
68335		bfloat16x8_t __s2_799 = __p2_799; \
68336		float32x2_t __rev0_799; __rev0_799 = __builtin_shufflevector(__s0_799, __s0_799, 1, 0); \
68337		bfloat16x4_t __rev1_799; __rev1_799 = __builtin_shufflevector(__s1_799, __s1_799, 3, 2, 1, 0); \
68338		bfloat16x8_t __rev2_799; __rev2_799 = __builtin_shufflevector(__s2_799, __s2_799, 7, 6, 5, 4, 3, 2, 1, 0); \
68339		__ret_799 = __noswap_vbfdot_f32(__rev0_799, __rev1_799, __builtin_bit_cast(bfloat16x4_t, __noswap_splat_laneq_f32(__builtin_bit_cast(float32x4_t, __rev2_799), __p3_799))); \
68340		__ret_799 = __builtin_shufflevector(__ret_799, __ret_799, 1, 0); \
68341		__ret_799; \
	70849	#define vbfdot_laneq_f32(__p0_815, __p1_815, __p2_815, __p3_815) __extension__ ({ \
	70850	float32x2_t __ret_815; \
	70851	float32x2_t __s0_815 = __p0_815; \
	70852	bfloat16x4_t __s1_815 = __p1_815; \
	70853	bfloat16x8_t __s2_815 = __p2_815; \
	70854	float32x2_t __rev0_815; __rev0_815 = __builtin_shufflevector(__s0_815, __s0_815, __lane_reverse_64_32); \
	70855	bfloat16x4_t __rev1_815; __rev1_815 = __builtin_shufflevector(__s1_815, __s1_815, __lane_reverse_64_16); \
	70856	bfloat16x8_t __rev2_815; __rev2_815 = __builtin_shufflevector(__s2_815, __s2_815, __lane_reverse_128_16); \
	70857	__ret_815 = __noswap_vbfdot_f32(__rev0_815, __rev1_815, __builtin_bit_cast(bfloat16x4_t, __noswap_splat_laneq_f32(__builtin_bit_cast(float32x4_t, __rev2_815), __p3_815))); \
	70858	__ret_815 = __builtin_shufflevector(__ret_815, __ret_815, __lane_reverse_64_32); \
	70859	__ret_815; \
68342	70860	})
68343	70861	#endif
68344	70862
68345	70863	#ifdef __LITTLE_ENDIAN__
68346		#define vbfmlalbq_lane_f32(__p0_800, __p1_800, __p2_800, __p3_800) __extension__ ({ \
68347		float32x4_t __ret_800; \
68348		float32x4_t __s0_800 = __p0_800; \
68349		bfloat16x8_t __s1_800 = __p1_800; \
68350		bfloat16x4_t __s2_800 = __p2_800; \
68351		__ret_800 = vbfmlalbq_f32(__s0_800, __s1_800, (bfloat16x8_t) {vget_lane_bf16(__s2_800, __p3_800), vget_lane_bf16(__s2_800, __p3_800), vget_lane_bf16(__s2_800, __p3_800), vget_lane_bf16(__s2_800, __p3_800), vget_lane_bf16(__s2_800, __p3_800), vget_lane_bf16(__s2_800, __p3_800), vget_lane_bf16(__s2_800, __p3_800), vget_lane_bf16(__s2_800, __p3_800)}); \
68352		__ret_800; \
	70864	#define vbfmlalbq_lane_f32(__p0_816, __p1_816, __p2_816, __p3_816) __extension__ ({ \
	70865	float32x4_t __ret_816; \
	70866	float32x4_t __s0_816 = __p0_816; \
	70867	bfloat16x8_t __s1_816 = __p1_816; \
	70868	bfloat16x4_t __s2_816 = __p2_816; \
	70869	__ret_816 = vbfmlalbq_f32(__s0_816, __s1_816, (bfloat16x8_t) {vget_lane_bf16(__s2_816, __p3_816), vget_lane_bf16(__s2_816, __p3_816), vget_lane_bf16(__s2_816, __p3_816), vget_lane_bf16(__s2_816, __p3_816), vget_lane_bf16(__s2_816, __p3_816), vget_lane_bf16(__s2_816, __p3_816), vget_lane_bf16(__s2_816, __p3_816), vget_lane_bf16(__s2_816, __p3_816)}); \
	70870	__ret_816; \
68353	70871	})
68354	70872	#else
68355		#define vbfmlalbq_lane_f32(__p0_801, __p1_801, __p2_801, __p3_801) __extension__ ({ \
68356		float32x4_t __ret_801; \
68357		float32x4_t __s0_801 = __p0_801; \
68358		bfloat16x8_t __s1_801 = __p1_801; \
68359		bfloat16x4_t __s2_801 = __p2_801; \
68360		float32x4_t __rev0_801; __rev0_801 = __builtin_shufflevector(__s0_801, __s0_801, 3, 2, 1, 0); \
68361		bfloat16x8_t __rev1_801; __rev1_801 = __builtin_shufflevector(__s1_801, __s1_801, 7, 6, 5, 4, 3, 2, 1, 0); \
68362		bfloat16x4_t __rev2_801; __rev2_801 = __builtin_shufflevector(__s2_801, __s2_801, 3, 2, 1, 0); \
68363		__ret_801 = __noswap_vbfmlalbq_f32(__rev0_801, __rev1_801, (bfloat16x8_t) {__noswap_vget_lane_bf16(__rev2_801, __p3_801), __noswap_vget_lane_bf16(__rev2_801, __p3_801), __noswap_vget_lane_bf16(__rev2_801, __p3_801), __noswap_vget_lane_bf16(__rev2_801, __p3_801), __noswap_vget_lane_bf16(__rev2_801, __p3_801), __noswap_vget_lane_bf16(__rev2_801, __p3_801), __noswap_vget_lane_bf16(__rev2_801, __p3_801), __noswap_vget_lane_bf16(__rev2_801, __p3_801)}); \
68364		__ret_801 = __builtin_shufflevector(__ret_801, __ret_801, 3, 2, 1, 0); \
68365		__ret_801; \
	70873	#define vbfmlalbq_lane_f32(__p0_817, __p1_817, __p2_817, __p3_817) __extension__ ({ \
	70874	float32x4_t __ret_817; \
	70875	float32x4_t __s0_817 = __p0_817; \
	70876	bfloat16x8_t __s1_817 = __p1_817; \
	70877	bfloat16x4_t __s2_817 = __p2_817; \
	70878	float32x4_t __rev0_817; __rev0_817 = __builtin_shufflevector(__s0_817, __s0_817, __lane_reverse_128_32); \
	70879	bfloat16x8_t __rev1_817; __rev1_817 = __builtin_shufflevector(__s1_817, __s1_817, __lane_reverse_128_16); \
	70880	bfloat16x4_t __rev2_817; __rev2_817 = __builtin_shufflevector(__s2_817, __s2_817, __lane_reverse_64_16); \
	70881	__ret_817 = __noswap_vbfmlalbq_f32(__rev0_817, __rev1_817, (bfloat16x8_t) {__noswap_vget_lane_bf16(__rev2_817, __p3_817), __noswap_vget_lane_bf16(__rev2_817, __p3_817), __noswap_vget_lane_bf16(__rev2_817, __p3_817), __noswap_vget_lane_bf16(__rev2_817, __p3_817), __noswap_vget_lane_bf16(__rev2_817, __p3_817), __noswap_vget_lane_bf16(__rev2_817, __p3_817), __noswap_vget_lane_bf16(__rev2_817, __p3_817), __noswap_vget_lane_bf16(__rev2_817, __p3_817)}); \
	70882	__ret_817 = __builtin_shufflevector(__ret_817, __ret_817, __lane_reverse_128_32); \
	70883	__ret_817; \
68366	70884	})
68367	70885	#endif
68368	70886
68369	70887	#ifdef __LITTLE_ENDIAN__
68370		#define vbfmlalbq_laneq_f32(__p0_802, __p1_802, __p2_802, __p3_802) __extension__ ({ \
68371		float32x4_t __ret_802; \
68372		float32x4_t __s0_802 = __p0_802; \
68373		bfloat16x8_t __s1_802 = __p1_802; \
68374		bfloat16x8_t __s2_802 = __p2_802; \
68375		__ret_802 = vbfmlalbq_f32(__s0_802, __s1_802, (bfloat16x8_t) {vgetq_lane_bf16(__s2_802, __p3_802), vgetq_lane_bf16(__s2_802, __p3_802), vgetq_lane_bf16(__s2_802, __p3_802), vgetq_lane_bf16(__s2_802, __p3_802), vgetq_lane_bf16(__s2_802, __p3_802), vgetq_lane_bf16(__s2_802, __p3_802), vgetq_lane_bf16(__s2_802, __p3_802), vgetq_lane_bf16(__s2_802, __p3_802)}); \
68376		__ret_802; \
	70888	#define vbfmlalbq_laneq_f32(__p0_818, __p1_818, __p2_818, __p3_818) __extension__ ({ \
	70889	float32x4_t __ret_818; \
	70890	float32x4_t __s0_818 = __p0_818; \
	70891	bfloat16x8_t __s1_818 = __p1_818; \
	70892	bfloat16x8_t __s2_818 = __p2_818; \
	70893	__ret_818 = vbfmlalbq_f32(__s0_818, __s1_818, (bfloat16x8_t) {vgetq_lane_bf16(__s2_818, __p3_818), vgetq_lane_bf16(__s2_818, __p3_818), vgetq_lane_bf16(__s2_818, __p3_818), vgetq_lane_bf16(__s2_818, __p3_818), vgetq_lane_bf16(__s2_818, __p3_818), vgetq_lane_bf16(__s2_818, __p3_818), vgetq_lane_bf16(__s2_818, __p3_818), vgetq_lane_bf16(__s2_818, __p3_818)}); \
	70894	__ret_818; \
68377	70895	})
68378	70896	#else
68379		#define vbfmlalbq_laneq_f32(__p0_803, __p1_803, __p2_803, __p3_803) __extension__ ({ \
68380		float32x4_t __ret_803; \
68381		float32x4_t __s0_803 = __p0_803; \
68382		bfloat16x8_t __s1_803 = __p1_803; \
68383		bfloat16x8_t __s2_803 = __p2_803; \
68384		float32x4_t __rev0_803; __rev0_803 = __builtin_shufflevector(__s0_803, __s0_803, 3, 2, 1, 0); \
68385		bfloat16x8_t __rev1_803; __rev1_803 = __builtin_shufflevector(__s1_803, __s1_803, 7, 6, 5, 4, 3, 2, 1, 0); \
68386		bfloat16x8_t __rev2_803; __rev2_803 = __builtin_shufflevector(__s2_803, __s2_803, 7, 6, 5, 4, 3, 2, 1, 0); \
68387		__ret_803 = __noswap_vbfmlalbq_f32(__rev0_803, __rev1_803, (bfloat16x8_t) {__noswap_vgetq_lane_bf16(__rev2_803, __p3_803), __noswap_vgetq_lane_bf16(__rev2_803, __p3_803), __noswap_vgetq_lane_bf16(__rev2_803, __p3_803), __noswap_vgetq_lane_bf16(__rev2_803, __p3_803), __noswap_vgetq_lane_bf16(__rev2_803, __p3_803), __noswap_vgetq_lane_bf16(__rev2_803, __p3_803), __noswap_vgetq_lane_bf16(__rev2_803, __p3_803), __noswap_vgetq_lane_bf16(__rev2_803, __p3_803)}); \
68388		__ret_803 = __builtin_shufflevector(__ret_803, __ret_803, 3, 2, 1, 0); \
68389		__ret_803; \
	70897	#define vbfmlalbq_laneq_f32(__p0_819, __p1_819, __p2_819, __p3_819) __extension__ ({ \
	70898	float32x4_t __ret_819; \
	70899	float32x4_t __s0_819 = __p0_819; \
	70900	bfloat16x8_t __s1_819 = __p1_819; \
	70901	bfloat16x8_t __s2_819 = __p2_819; \
	70902	float32x4_t __rev0_819; __rev0_819 = __builtin_shufflevector(__s0_819, __s0_819, __lane_reverse_128_32); \
	70903	bfloat16x8_t __rev1_819; __rev1_819 = __builtin_shufflevector(__s1_819, __s1_819, __lane_reverse_128_16); \
	70904	bfloat16x8_t __rev2_819; __rev2_819 = __builtin_shufflevector(__s2_819, __s2_819, __lane_reverse_128_16); \
	70905	__ret_819 = __noswap_vbfmlalbq_f32(__rev0_819, __rev1_819, (bfloat16x8_t) {__noswap_vgetq_lane_bf16(__rev2_819, __p3_819), __noswap_vgetq_lane_bf16(__rev2_819, __p3_819), __noswap_vgetq_lane_bf16(__rev2_819, __p3_819), __noswap_vgetq_lane_bf16(__rev2_819, __p3_819), __noswap_vgetq_lane_bf16(__rev2_819, __p3_819), __noswap_vgetq_lane_bf16(__rev2_819, __p3_819), __noswap_vgetq_lane_bf16(__rev2_819, __p3_819), __noswap_vgetq_lane_bf16(__rev2_819, __p3_819)}); \
	70906	__ret_819 = __builtin_shufflevector(__ret_819, __ret_819, __lane_reverse_128_32); \
	70907	__ret_819; \
68390	70908	})
68391	70909	#endif
68392	70910
68393	70911	#ifdef __LITTLE_ENDIAN__
68394		#define vbfmlaltq_lane_f32(__p0_804, __p1_804, __p2_804, __p3_804) __extension__ ({ \
68395		float32x4_t __ret_804; \
68396		float32x4_t __s0_804 = __p0_804; \
68397		bfloat16x8_t __s1_804 = __p1_804; \
68398		bfloat16x4_t __s2_804 = __p2_804; \
68399		__ret_804 = vbfmlaltq_f32(__s0_804, __s1_804, (bfloat16x8_t) {vget_lane_bf16(__s2_804, __p3_804), vget_lane_bf16(__s2_804, __p3_804), vget_lane_bf16(__s2_804, __p3_804), vget_lane_bf16(__s2_804, __p3_804), vget_lane_bf16(__s2_804, __p3_804), vget_lane_bf16(__s2_804, __p3_804), vget_lane_bf16(__s2_804, __p3_804), vget_lane_bf16(__s2_804, __p3_804)}); \
68400		__ret_804; \
	70912	#define vbfmlaltq_lane_f32(__p0_820, __p1_820, __p2_820, __p3_820) __extension__ ({ \
	70913	float32x4_t __ret_820; \
	70914	float32x4_t __s0_820 = __p0_820; \
	70915	bfloat16x8_t __s1_820 = __p1_820; \
	70916	bfloat16x4_t __s2_820 = __p2_820; \
	70917	__ret_820 = vbfmlaltq_f32(__s0_820, __s1_820, (bfloat16x8_t) {vget_lane_bf16(__s2_820, __p3_820), vget_lane_bf16(__s2_820, __p3_820), vget_lane_bf16(__s2_820, __p3_820), vget_lane_bf16(__s2_820, __p3_820), vget_lane_bf16(__s2_820, __p3_820), vget_lane_bf16(__s2_820, __p3_820), vget_lane_bf16(__s2_820, __p3_820), vget_lane_bf16(__s2_820, __p3_820)}); \
	70918	__ret_820; \
68401	70919	})
68402	70920	#else
68403		#define vbfmlaltq_lane_f32(__p0_805, __p1_805, __p2_805, __p3_805) __extension__ ({ \
68404		float32x4_t __ret_805; \
68405		float32x4_t __s0_805 = __p0_805; \
68406		bfloat16x8_t __s1_805 = __p1_805; \
68407		bfloat16x4_t __s2_805 = __p2_805; \
68408		float32x4_t __rev0_805; __rev0_805 = __builtin_shufflevector(__s0_805, __s0_805, 3, 2, 1, 0); \
68409		bfloat16x8_t __rev1_805; __rev1_805 = __builtin_shufflevector(__s1_805, __s1_805, 7, 6, 5, 4, 3, 2, 1, 0); \
68410		bfloat16x4_t __rev2_805; __rev2_805 = __builtin_shufflevector(__s2_805, __s2_805, 3, 2, 1, 0); \
68411		__ret_805 = __noswap_vbfmlaltq_f32(__rev0_805, __rev1_805, (bfloat16x8_t) {__noswap_vget_lane_bf16(__rev2_805, __p3_805), __noswap_vget_lane_bf16(__rev2_805, __p3_805), __noswap_vget_lane_bf16(__rev2_805, __p3_805), __noswap_vget_lane_bf16(__rev2_805, __p3_805), __noswap_vget_lane_bf16(__rev2_805, __p3_805), __noswap_vget_lane_bf16(__rev2_805, __p3_805), __noswap_vget_lane_bf16(__rev2_805, __p3_805), __noswap_vget_lane_bf16(__rev2_805, __p3_805)}); \
68412		__ret_805 = __builtin_shufflevector(__ret_805, __ret_805, 3, 2, 1, 0); \
68413		__ret_805; \
	70921	#define vbfmlaltq_lane_f32(__p0_821, __p1_821, __p2_821, __p3_821) __extension__ ({ \
	70922	float32x4_t __ret_821; \
	70923	float32x4_t __s0_821 = __p0_821; \
	70924	bfloat16x8_t __s1_821 = __p1_821; \
	70925	bfloat16x4_t __s2_821 = __p2_821; \
	70926	float32x4_t __rev0_821; __rev0_821 = __builtin_shufflevector(__s0_821, __s0_821, __lane_reverse_128_32); \
	70927	bfloat16x8_t __rev1_821; __rev1_821 = __builtin_shufflevector(__s1_821, __s1_821, __lane_reverse_128_16); \
	70928	bfloat16x4_t __rev2_821; __rev2_821 = __builtin_shufflevector(__s2_821, __s2_821, __lane_reverse_64_16); \
	70929	__ret_821 = __noswap_vbfmlaltq_f32(__rev0_821, __rev1_821, (bfloat16x8_t) {__noswap_vget_lane_bf16(__rev2_821, __p3_821), __noswap_vget_lane_bf16(__rev2_821, __p3_821), __noswap_vget_lane_bf16(__rev2_821, __p3_821), __noswap_vget_lane_bf16(__rev2_821, __p3_821), __noswap_vget_lane_bf16(__rev2_821, __p3_821), __noswap_vget_lane_bf16(__rev2_821, __p3_821), __noswap_vget_lane_bf16(__rev2_821, __p3_821), __noswap_vget_lane_bf16(__rev2_821, __p3_821)}); \
	70930	__ret_821 = __builtin_shufflevector(__ret_821, __ret_821, __lane_reverse_128_32); \
	70931	__ret_821; \
68414	70932	})
68415	70933	#endif
68416	70934
68417	70935	#ifdef __LITTLE_ENDIAN__
68418		#define vbfmlaltq_laneq_f32(__p0_806, __p1_806, __p2_806, __p3_806) __extension__ ({ \
68419		float32x4_t __ret_806; \
68420		float32x4_t __s0_806 = __p0_806; \
68421		bfloat16x8_t __s1_806 = __p1_806; \
68422		bfloat16x8_t __s2_806 = __p2_806; \
68423		__ret_806 = vbfmlaltq_f32(__s0_806, __s1_806, (bfloat16x8_t) {vgetq_lane_bf16(__s2_806, __p3_806), vgetq_lane_bf16(__s2_806, __p3_806), vgetq_lane_bf16(__s2_806, __p3_806), vgetq_lane_bf16(__s2_806, __p3_806), vgetq_lane_bf16(__s2_806, __p3_806), vgetq_lane_bf16(__s2_806, __p3_806), vgetq_lane_bf16(__s2_806, __p3_806), vgetq_lane_bf16(__s2_806, __p3_806)}); \
68424		__ret_806; \
	70936	#define vbfmlaltq_laneq_f32(__p0_822, __p1_822, __p2_822, __p3_822) __extension__ ({ \
	70937	float32x4_t __ret_822; \
	70938	float32x4_t __s0_822 = __p0_822; \
	70939	bfloat16x8_t __s1_822 = __p1_822; \
	70940	bfloat16x8_t __s2_822 = __p2_822; \
	70941	__ret_822 = vbfmlaltq_f32(__s0_822, __s1_822, (bfloat16x8_t) {vgetq_lane_bf16(__s2_822, __p3_822), vgetq_lane_bf16(__s2_822, __p3_822), vgetq_lane_bf16(__s2_822, __p3_822), vgetq_lane_bf16(__s2_822, __p3_822), vgetq_lane_bf16(__s2_822, __p3_822), vgetq_lane_bf16(__s2_822, __p3_822), vgetq_lane_bf16(__s2_822, __p3_822), vgetq_lane_bf16(__s2_822, __p3_822)}); \
	70942	__ret_822; \
68425	70943	})
68426	70944	#else
68427		#define vbfmlaltq_laneq_f32(__p0_807, __p1_807, __p2_807, __p3_807) __extension__ ({ \
68428		float32x4_t __ret_807; \
68429		float32x4_t __s0_807 = __p0_807; \
68430		bfloat16x8_t __s1_807 = __p1_807; \
68431		bfloat16x8_t __s2_807 = __p2_807; \
68432		float32x4_t __rev0_807; __rev0_807 = __builtin_shufflevector(__s0_807, __s0_807, 3, 2, 1, 0); \
68433		bfloat16x8_t __rev1_807; __rev1_807 = __builtin_shufflevector(__s1_807, __s1_807, 7, 6, 5, 4, 3, 2, 1, 0); \
68434		bfloat16x8_t __rev2_807; __rev2_807 = __builtin_shufflevector(__s2_807, __s2_807, 7, 6, 5, 4, 3, 2, 1, 0); \
68435		__ret_807 = __noswap_vbfmlaltq_f32(__rev0_807, __rev1_807, (bfloat16x8_t) {__noswap_vgetq_lane_bf16(__rev2_807, __p3_807), __noswap_vgetq_lane_bf16(__rev2_807, __p3_807), __noswap_vgetq_lane_bf16(__rev2_807, __p3_807), __noswap_vgetq_lane_bf16(__rev2_807, __p3_807), __noswap_vgetq_lane_bf16(__rev2_807, __p3_807), __noswap_vgetq_lane_bf16(__rev2_807, __p3_807), __noswap_vgetq_lane_bf16(__rev2_807, __p3_807), __noswap_vgetq_lane_bf16(__rev2_807, __p3_807)}); \
68436		__ret_807 = __builtin_shufflevector(__ret_807, __ret_807, 3, 2, 1, 0); \
68437		__ret_807; \
	70945	#define vbfmlaltq_laneq_f32(__p0_823, __p1_823, __p2_823, __p3_823) __extension__ ({ \
	70946	float32x4_t __ret_823; \
	70947	float32x4_t __s0_823 = __p0_823; \
	70948	bfloat16x8_t __s1_823 = __p1_823; \
	70949	bfloat16x8_t __s2_823 = __p2_823; \
	70950	float32x4_t __rev0_823; __rev0_823 = __builtin_shufflevector(__s0_823, __s0_823, __lane_reverse_128_32); \
	70951	bfloat16x8_t __rev1_823; __rev1_823 = __builtin_shufflevector(__s1_823, __s1_823, __lane_reverse_128_16); \
	70952	bfloat16x8_t __rev2_823; __rev2_823 = __builtin_shufflevector(__s2_823, __s2_823, __lane_reverse_128_16); \
	70953	__ret_823 = __noswap_vbfmlaltq_f32(__rev0_823, __rev1_823, (bfloat16x8_t) {__noswap_vgetq_lane_bf16(__rev2_823, __p3_823), __noswap_vgetq_lane_bf16(__rev2_823, __p3_823), __noswap_vgetq_lane_bf16(__rev2_823, __p3_823), __noswap_vgetq_lane_bf16(__rev2_823, __p3_823), __noswap_vgetq_lane_bf16(__rev2_823, __p3_823), __noswap_vgetq_lane_bf16(__rev2_823, __p3_823), __noswap_vgetq_lane_bf16(__rev2_823, __p3_823), __noswap_vgetq_lane_bf16(__rev2_823, __p3_823)}); \
	70954	__ret_823 = __builtin_shufflevector(__ret_823, __ret_823, __lane_reverse_128_32); \
	70955	__ret_823; \
68438	70956	})
68439	70957	#endif
68440	70958
68441	70959	#ifdef __LITTLE_ENDIAN__
68442		__ai __attribute__((target("bf16,neon"))) float32x4_t vcvt_f32_bf16(bfloat16x4_t __p0_808) {
68443		float32x4_t __ret_808;
68444		__ret_808 = __builtin_bit_cast(float32x4_t, vshll_n_u16(__builtin_bit_cast(uint16x4_t, __p0_808), 16));
68445		return __ret_808;
	70960	__ai __attribute__((target("bf16,neon"))) float32x4_t vcvt_f32_bf16(bfloat16x4_t __p0_824) {
	70961	float32x4_t __ret_824;
	70962	__ret_824 = __builtin_bit_cast(float32x4_t, vshll_n_u16(__builtin_bit_cast(uint16x4_t, __p0_824), 16));
	70963	return __ret_824;
68446	70964	}
68447	70965	#else
68448		__ai __attribute__((target("bf16,neon"))) float32x4_t vcvt_f32_bf16(bfloat16x4_t __p0_809) {
68449		float32x4_t __ret_809;
68450		bfloat16x4_t __rev0_809; __rev0_809 = __builtin_shufflevector(__p0_809, __p0_809, 3, 2, 1, 0);
68451		__ret_809 = __builtin_bit_cast(float32x4_t, __noswap_vshll_n_u16(__builtin_bit_cast(uint16x4_t, __rev0_809), 16));
68452		__ret_809 = __builtin_shufflevector(__ret_809, __ret_809, 3, 2, 1, 0);
68453		return __ret_809;
	70966	__ai __attribute__((target("bf16,neon"))) float32x4_t vcvt_f32_bf16(bfloat16x4_t __p0_825) {
	70967	float32x4_t __ret_825;
	70968	bfloat16x4_t __rev0_825; __rev0_825 = __builtin_shufflevector(__p0_825, __p0_825, __lane_reverse_64_16);
	70969	__ret_825 = __builtin_bit_cast(float32x4_t, __noswap_vshll_n_u16(__builtin_bit_cast(uint16x4_t, __rev0_825), 16));
	70970	__ret_825 = __builtin_shufflevector(__ret_825, __ret_825, __lane_reverse_128_32);
	70971	return __ret_825;
68454	70972	}
68455		__ai __attribute__((target("bf16,neon"))) float32x4_t __noswap_vcvt_f32_bf16(bfloat16x4_t __p0_810) {
68456		float32x4_t __ret_810;
68457		__ret_810 = __builtin_bit_cast(float32x4_t, __noswap_vshll_n_u16(__builtin_bit_cast(uint16x4_t, __p0_810), 16));
68458		return __ret_810;
	70973	__ai __attribute__((target("bf16,neon"))) float32x4_t __noswap_vcvt_f32_bf16(bfloat16x4_t __p0_826) {
	70974	float32x4_t __ret_826;
	70975	__ret_826 = __builtin_bit_cast(float32x4_t, __noswap_vshll_n_u16(__builtin_bit_cast(uint16x4_t, __p0_826), 16));
	70976	return __ret_826;
68459	70977	}
68460	70978	#endif
68461	70979
...	...	@@ -68468,9 +70986,9 @@ __ai __attribute__((target("bf16,neon"))) float32x4_t vcvtq_high_f32_bf16(bfloat
68468	70986	#else
68469	70987	__ai __attribute__((target("bf16,neon"))) float32x4_t vcvtq_high_f32_bf16(bfloat16x8_t __p0) {
68470	70988	float32x4_t __ret;
68471		bfloat16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	70989	bfloat16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
68472	70990	__ret = __noswap_vcvt_f32_bf16(__noswap_vget_high_bf16(__rev0));
68473		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	70991	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
68474	70992	return __ret;
68475	70993	}
68476	70994	#endif
...	...	@@ -68484,244 +71002,244 @@ __ai __attribute__((target("bf16,neon"))) float32x4_t vcvtq_low_f32_bf16(bfloat1
68484	71002	#else
68485	71003	__ai __attribute__((target("bf16,neon"))) float32x4_t vcvtq_low_f32_bf16(bfloat16x8_t __p0) {
68486	71004	float32x4_t __ret;
68487		bfloat16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
	71005	bfloat16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
68488	71006	__ret = __noswap_vcvt_f32_bf16(__noswap_vget_low_bf16(__rev0));
68489		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	71007	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
68490	71008	return __ret;
68491	71009	}
68492	71010	#endif
68493	71011
68494	71012	#ifdef __LITTLE_ENDIAN__
68495		#define vdotq_lane_u32(__p0_811, __p1_811, __p2_811, __p3_811) __extension__ ({ \
68496		uint32x4_t __ret_811; \
68497		uint32x4_t __s0_811 = __p0_811; \
68498		uint8x16_t __s1_811 = __p1_811; \
68499		uint8x8_t __s2_811 = __p2_811; \
68500		__ret_811 = vdotq_u32(__s0_811, __s1_811, __builtin_bit_cast(uint8x16_t, splatq_lane_u32(__builtin_bit_cast(uint32x2_t, __s2_811), __p3_811))); \
68501		__ret_811; \
	71013	#define vdotq_lane_u32(__p0_827, __p1_827, __p2_827, __p3_827) __extension__ ({ \
	71014	uint32x4_t __ret_827; \
	71015	uint32x4_t __s0_827 = __p0_827; \
	71016	uint8x16_t __s1_827 = __p1_827; \
	71017	uint8x8_t __s2_827 = __p2_827; \
	71018	__ret_827 = vdotq_u32(__s0_827, __s1_827, __builtin_bit_cast(uint8x16_t, splatq_lane_u32(__builtin_bit_cast(uint32x2_t, __s2_827), __p3_827))); \
	71019	__ret_827; \
68502	71020	})
68503	71021	#else
68504		#define vdotq_lane_u32(__p0_812, __p1_812, __p2_812, __p3_812) __extension__ ({ \
68505		uint32x4_t __ret_812; \
68506		uint32x4_t __s0_812 = __p0_812; \
68507		uint8x16_t __s1_812 = __p1_812; \
68508		uint8x8_t __s2_812 = __p2_812; \
68509		uint32x4_t __rev0_812; __rev0_812 = __builtin_shufflevector(__s0_812, __s0_812, 3, 2, 1, 0); \
68510		uint8x16_t __rev1_812; __rev1_812 = __builtin_shufflevector(__s1_812, __s1_812, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
68511		uint8x8_t __rev2_812; __rev2_812 = __builtin_shufflevector(__s2_812, __s2_812, 7, 6, 5, 4, 3, 2, 1, 0); \
68512		__ret_812 = __noswap_vdotq_u32(__rev0_812, __rev1_812, __builtin_bit_cast(uint8x16_t, __noswap_splatq_lane_u32(__builtin_bit_cast(uint32x2_t, __rev2_812), __p3_812))); \
68513		__ret_812 = __builtin_shufflevector(__ret_812, __ret_812, 3, 2, 1, 0); \
68514		__ret_812; \
	71022	#define vdotq_lane_u32(__p0_828, __p1_828, __p2_828, __p3_828) __extension__ ({ \
	71023	uint32x4_t __ret_828; \
	71024	uint32x4_t __s0_828 = __p0_828; \
	71025	uint8x16_t __s1_828 = __p1_828; \
	71026	uint8x8_t __s2_828 = __p2_828; \
	71027	uint32x4_t __rev0_828; __rev0_828 = __builtin_shufflevector(__s0_828, __s0_828, __lane_reverse_128_32); \
	71028	uint8x16_t __rev1_828; __rev1_828 = __builtin_shufflevector(__s1_828, __s1_828, __lane_reverse_128_8); \
	71029	uint8x8_t __rev2_828; __rev2_828 = __builtin_shufflevector(__s2_828, __s2_828, __lane_reverse_64_8); \
	71030	__ret_828 = __noswap_vdotq_u32(__rev0_828, __rev1_828, __builtin_bit_cast(uint8x16_t, __noswap_splatq_lane_u32(__builtin_bit_cast(uint32x2_t, __rev2_828), __p3_828))); \
	71031	__ret_828 = __builtin_shufflevector(__ret_828, __ret_828, __lane_reverse_128_32); \
	71032	__ret_828; \
68515	71033	})
68516	71034	#endif
68517	71035
68518	71036	#ifdef __LITTLE_ENDIAN__
68519		#define vdotq_lane_s32(__p0_813, __p1_813, __p2_813, __p3_813) __extension__ ({ \
68520		int32x4_t __ret_813; \
68521		int32x4_t __s0_813 = __p0_813; \
68522		int8x16_t __s1_813 = __p1_813; \
68523		int8x8_t __s2_813 = __p2_813; \
68524		__ret_813 = vdotq_s32(__s0_813, __s1_813, __builtin_bit_cast(int8x16_t, splatq_lane_s32(__builtin_bit_cast(int32x2_t, __s2_813), __p3_813))); \
68525		__ret_813; \
	71037	#define vdotq_lane_s32(__p0_829, __p1_829, __p2_829, __p3_829) __extension__ ({ \
	71038	int32x4_t __ret_829; \
	71039	int32x4_t __s0_829 = __p0_829; \
	71040	int8x16_t __s1_829 = __p1_829; \
	71041	int8x8_t __s2_829 = __p2_829; \
	71042	__ret_829 = vdotq_s32(__s0_829, __s1_829, __builtin_bit_cast(int8x16_t, splatq_lane_s32(__builtin_bit_cast(int32x2_t, __s2_829), __p3_829))); \
	71043	__ret_829; \
68526	71044	})
68527	71045	#else
68528		#define vdotq_lane_s32(__p0_814, __p1_814, __p2_814, __p3_814) __extension__ ({ \
68529		int32x4_t __ret_814; \
68530		int32x4_t __s0_814 = __p0_814; \
68531		int8x16_t __s1_814 = __p1_814; \
68532		int8x8_t __s2_814 = __p2_814; \
68533		int32x4_t __rev0_814; __rev0_814 = __builtin_shufflevector(__s0_814, __s0_814, 3, 2, 1, 0); \
68534		int8x16_t __rev1_814; __rev1_814 = __builtin_shufflevector(__s1_814, __s1_814, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
68535		int8x8_t __rev2_814; __rev2_814 = __builtin_shufflevector(__s2_814, __s2_814, 7, 6, 5, 4, 3, 2, 1, 0); \
68536		__ret_814 = __noswap_vdotq_s32(__rev0_814, __rev1_814, __builtin_bit_cast(int8x16_t, __noswap_splatq_lane_s32(__builtin_bit_cast(int32x2_t, __rev2_814), __p3_814))); \
68537		__ret_814 = __builtin_shufflevector(__ret_814, __ret_814, 3, 2, 1, 0); \
68538		__ret_814; \
	71046	#define vdotq_lane_s32(__p0_830, __p1_830, __p2_830, __p3_830) __extension__ ({ \
	71047	int32x4_t __ret_830; \
	71048	int32x4_t __s0_830 = __p0_830; \
	71049	int8x16_t __s1_830 = __p1_830; \
	71050	int8x8_t __s2_830 = __p2_830; \
	71051	int32x4_t __rev0_830; __rev0_830 = __builtin_shufflevector(__s0_830, __s0_830, __lane_reverse_128_32); \
	71052	int8x16_t __rev1_830; __rev1_830 = __builtin_shufflevector(__s1_830, __s1_830, __lane_reverse_128_8); \
	71053	int8x8_t __rev2_830; __rev2_830 = __builtin_shufflevector(__s2_830, __s2_830, __lane_reverse_64_8); \
	71054	__ret_830 = __noswap_vdotq_s32(__rev0_830, __rev1_830, __builtin_bit_cast(int8x16_t, __noswap_splatq_lane_s32(__builtin_bit_cast(int32x2_t, __rev2_830), __p3_830))); \
	71055	__ret_830 = __builtin_shufflevector(__ret_830, __ret_830, __lane_reverse_128_32); \
	71056	__ret_830; \
68539	71057	})
68540	71058	#endif
68541	71059
68542	71060	#ifdef __LITTLE_ENDIAN__
68543		#define vdot_lane_u32(__p0_815, __p1_815, __p2_815, __p3_815) __extension__ ({ \
68544		uint32x2_t __ret_815; \
68545		uint32x2_t __s0_815 = __p0_815; \
68546		uint8x8_t __s1_815 = __p1_815; \
68547		uint8x8_t __s2_815 = __p2_815; \
68548		__ret_815 = vdot_u32(__s0_815, __s1_815, __builtin_bit_cast(uint8x8_t, splat_lane_u32(__builtin_bit_cast(uint32x2_t, __s2_815), __p3_815))); \
68549		__ret_815; \
	71061	#define vdot_lane_u32(__p0_831, __p1_831, __p2_831, __p3_831) __extension__ ({ \
	71062	uint32x2_t __ret_831; \
	71063	uint32x2_t __s0_831 = __p0_831; \
	71064	uint8x8_t __s1_831 = __p1_831; \
	71065	uint8x8_t __s2_831 = __p2_831; \
	71066	__ret_831 = vdot_u32(__s0_831, __s1_831, __builtin_bit_cast(uint8x8_t, splat_lane_u32(__builtin_bit_cast(uint32x2_t, __s2_831), __p3_831))); \
	71067	__ret_831; \
68550	71068	})
68551	71069	#else
68552		#define vdot_lane_u32(__p0_816, __p1_816, __p2_816, __p3_816) __extension__ ({ \
68553		uint32x2_t __ret_816; \
68554		uint32x2_t __s0_816 = __p0_816; \
68555		uint8x8_t __s1_816 = __p1_816; \
68556		uint8x8_t __s2_816 = __p2_816; \
68557		uint32x2_t __rev0_816; __rev0_816 = __builtin_shufflevector(__s0_816, __s0_816, 1, 0); \
68558		uint8x8_t __rev1_816; __rev1_816 = __builtin_shufflevector(__s1_816, __s1_816, 7, 6, 5, 4, 3, 2, 1, 0); \
68559		uint8x8_t __rev2_816; __rev2_816 = __builtin_shufflevector(__s2_816, __s2_816, 7, 6, 5, 4, 3, 2, 1, 0); \
68560		__ret_816 = __noswap_vdot_u32(__rev0_816, __rev1_816, __builtin_bit_cast(uint8x8_t, __noswap_splat_lane_u32(__builtin_bit_cast(uint32x2_t, __rev2_816), __p3_816))); \
68561		__ret_816 = __builtin_shufflevector(__ret_816, __ret_816, 1, 0); \
68562		__ret_816; \
	71070	#define vdot_lane_u32(__p0_832, __p1_832, __p2_832, __p3_832) __extension__ ({ \
	71071	uint32x2_t __ret_832; \
	71072	uint32x2_t __s0_832 = __p0_832; \
	71073	uint8x8_t __s1_832 = __p1_832; \
	71074	uint8x8_t __s2_832 = __p2_832; \
	71075	uint32x2_t __rev0_832; __rev0_832 = __builtin_shufflevector(__s0_832, __s0_832, __lane_reverse_64_32); \
	71076	uint8x8_t __rev1_832; __rev1_832 = __builtin_shufflevector(__s1_832, __s1_832, __lane_reverse_64_8); \
	71077	uint8x8_t __rev2_832; __rev2_832 = __builtin_shufflevector(__s2_832, __s2_832, __lane_reverse_64_8); \
	71078	__ret_832 = __noswap_vdot_u32(__rev0_832, __rev1_832, __builtin_bit_cast(uint8x8_t, __noswap_splat_lane_u32(__builtin_bit_cast(uint32x2_t, __rev2_832), __p3_832))); \
	71079	__ret_832 = __builtin_shufflevector(__ret_832, __ret_832, __lane_reverse_64_32); \
	71080	__ret_832; \
68563	71081	})
68564	71082	#endif
68565	71083
68566	71084	#ifdef __LITTLE_ENDIAN__
68567		#define vdot_lane_s32(__p0_817, __p1_817, __p2_817, __p3_817) __extension__ ({ \
68568		int32x2_t __ret_817; \
68569		int32x2_t __s0_817 = __p0_817; \
68570		int8x8_t __s1_817 = __p1_817; \
68571		int8x8_t __s2_817 = __p2_817; \
68572		__ret_817 = vdot_s32(__s0_817, __s1_817, __builtin_bit_cast(int8x8_t, splat_lane_s32(__builtin_bit_cast(int32x2_t, __s2_817), __p3_817))); \
68573		__ret_817; \
	71085	#define vdot_lane_s32(__p0_833, __p1_833, __p2_833, __p3_833) __extension__ ({ \
	71086	int32x2_t __ret_833; \
	71087	int32x2_t __s0_833 = __p0_833; \
	71088	int8x8_t __s1_833 = __p1_833; \
	71089	int8x8_t __s2_833 = __p2_833; \
	71090	__ret_833 = vdot_s32(__s0_833, __s1_833, __builtin_bit_cast(int8x8_t, splat_lane_s32(__builtin_bit_cast(int32x2_t, __s2_833), __p3_833))); \
	71091	__ret_833; \
68574	71092	})
68575	71093	#else
68576		#define vdot_lane_s32(__p0_818, __p1_818, __p2_818, __p3_818) __extension__ ({ \
68577		int32x2_t __ret_818; \
68578		int32x2_t __s0_818 = __p0_818; \
68579		int8x8_t __s1_818 = __p1_818; \
68580		int8x8_t __s2_818 = __p2_818; \
68581		int32x2_t __rev0_818; __rev0_818 = __builtin_shufflevector(__s0_818, __s0_818, 1, 0); \
68582		int8x8_t __rev1_818; __rev1_818 = __builtin_shufflevector(__s1_818, __s1_818, 7, 6, 5, 4, 3, 2, 1, 0); \
68583		int8x8_t __rev2_818; __rev2_818 = __builtin_shufflevector(__s2_818, __s2_818, 7, 6, 5, 4, 3, 2, 1, 0); \
68584		__ret_818 = __noswap_vdot_s32(__rev0_818, __rev1_818, __builtin_bit_cast(int8x8_t, __noswap_splat_lane_s32(__builtin_bit_cast(int32x2_t, __rev2_818), __p3_818))); \
68585		__ret_818 = __builtin_shufflevector(__ret_818, __ret_818, 1, 0); \
68586		__ret_818; \
	71094	#define vdot_lane_s32(__p0_834, __p1_834, __p2_834, __p3_834) __extension__ ({ \
	71095	int32x2_t __ret_834; \
	71096	int32x2_t __s0_834 = __p0_834; \
	71097	int8x8_t __s1_834 = __p1_834; \
	71098	int8x8_t __s2_834 = __p2_834; \
	71099	int32x2_t __rev0_834; __rev0_834 = __builtin_shufflevector(__s0_834, __s0_834, __lane_reverse_64_32); \
	71100	int8x8_t __rev1_834; __rev1_834 = __builtin_shufflevector(__s1_834, __s1_834, __lane_reverse_64_8); \
	71101	int8x8_t __rev2_834; __rev2_834 = __builtin_shufflevector(__s2_834, __s2_834, __lane_reverse_64_8); \
	71102	__ret_834 = __noswap_vdot_s32(__rev0_834, __rev1_834, __builtin_bit_cast(int8x8_t, __noswap_splat_lane_s32(__builtin_bit_cast(int32x2_t, __rev2_834), __p3_834))); \
	71103	__ret_834 = __builtin_shufflevector(__ret_834, __ret_834, __lane_reverse_64_32); \
	71104	__ret_834; \
68587	71105	})
68588	71106	#endif
68589	71107
68590	71108	#ifdef __LITTLE_ENDIAN__
68591		#define vmulq_lane_f16(__p0_819, __p1_819, __p2_819) __extension__ ({ \
68592		float16x8_t __ret_819; \
68593		float16x8_t __s0_819 = __p0_819; \
68594		float16x4_t __s1_819 = __p1_819; \
68595		__ret_819 = __s0_819 * splatq_lane_f16(__s1_819, __p2_819); \
68596		__ret_819; \
	71109	#define vmulq_lane_f16(__p0_835, __p1_835, __p2_835) __extension__ ({ \
	71110	float16x8_t __ret_835; \
	71111	float16x8_t __s0_835 = __p0_835; \
	71112	float16x4_t __s1_835 = __p1_835; \
	71113	__ret_835 = __s0_835 * splatq_lane_f16(__s1_835, __p2_835); \
	71114	__ret_835; \
68597	71115	})
68598	71116	#else
68599		#define vmulq_lane_f16(__p0_820, __p1_820, __p2_820) __extension__ ({ \
68600		float16x8_t __ret_820; \
68601		float16x8_t __s0_820 = __p0_820; \
68602		float16x4_t __s1_820 = __p1_820; \
68603		float16x8_t __rev0_820; __rev0_820 = __builtin_shufflevector(__s0_820, __s0_820, 7, 6, 5, 4, 3, 2, 1, 0); \
68604		float16x4_t __rev1_820; __rev1_820 = __builtin_shufflevector(__s1_820, __s1_820, 3, 2, 1, 0); \
68605		__ret_820 = __rev0_820 * __noswap_splatq_lane_f16(__rev1_820, __p2_820); \
68606		__ret_820 = __builtin_shufflevector(__ret_820, __ret_820, 7, 6, 5, 4, 3, 2, 1, 0); \
68607		__ret_820; \
	71117	#define vmulq_lane_f16(__p0_836, __p1_836, __p2_836) __extension__ ({ \
	71118	float16x8_t __ret_836; \
	71119	float16x8_t __s0_836 = __p0_836; \
	71120	float16x4_t __s1_836 = __p1_836; \
	71121	float16x8_t __rev0_836; __rev0_836 = __builtin_shufflevector(__s0_836, __s0_836, __lane_reverse_128_16); \
	71122	float16x4_t __rev1_836; __rev1_836 = __builtin_shufflevector(__s1_836, __s1_836, __lane_reverse_64_16); \
	71123	__ret_836 = __rev0_836 * __noswap_splatq_lane_f16(__rev1_836, __p2_836); \
	71124	__ret_836 = __builtin_shufflevector(__ret_836, __ret_836, __lane_reverse_128_16); \
	71125	__ret_836; \
68608	71126	})
68609	71127	#endif
68610	71128
68611	71129	#ifdef __LITTLE_ENDIAN__
68612		#define vmul_lane_f16(__p0_821, __p1_821, __p2_821) __extension__ ({ \
68613		float16x4_t __ret_821; \
68614		float16x4_t __s0_821 = __p0_821; \
68615		float16x4_t __s1_821 = __p1_821; \
68616		__ret_821 = __s0_821 * splat_lane_f16(__s1_821, __p2_821); \
68617		__ret_821; \
	71130	#define vmul_lane_f16(__p0_837, __p1_837, __p2_837) __extension__ ({ \
	71131	float16x4_t __ret_837; \
	71132	float16x4_t __s0_837 = __p0_837; \
	71133	float16x4_t __s1_837 = __p1_837; \
	71134	__ret_837 = __s0_837 * splat_lane_f16(__s1_837, __p2_837); \
	71135	__ret_837; \
68618	71136	})
68619	71137	#else
68620		#define vmul_lane_f16(__p0_822, __p1_822, __p2_822) __extension__ ({ \
68621		float16x4_t __ret_822; \
68622		float16x4_t __s0_822 = __p0_822; \
68623		float16x4_t __s1_822 = __p1_822; \
68624		float16x4_t __rev0_822; __rev0_822 = __builtin_shufflevector(__s0_822, __s0_822, 3, 2, 1, 0); \
68625		float16x4_t __rev1_822; __rev1_822 = __builtin_shufflevector(__s1_822, __s1_822, 3, 2, 1, 0); \
68626		__ret_822 = __rev0_822 * __noswap_splat_lane_f16(__rev1_822, __p2_822); \
68627		__ret_822 = __builtin_shufflevector(__ret_822, __ret_822, 3, 2, 1, 0); \
68628		__ret_822; \
	71138	#define vmul_lane_f16(__p0_838, __p1_838, __p2_838) __extension__ ({ \
	71139	float16x4_t __ret_838; \
	71140	float16x4_t __s0_838 = __p0_838; \
	71141	float16x4_t __s1_838 = __p1_838; \
	71142	float16x4_t __rev0_838; __rev0_838 = __builtin_shufflevector(__s0_838, __s0_838, __lane_reverse_64_16); \
	71143	float16x4_t __rev1_838; __rev1_838 = __builtin_shufflevector(__s1_838, __s1_838, __lane_reverse_64_16); \
	71144	__ret_838 = __rev0_838 * __noswap_splat_lane_f16(__rev1_838, __p2_838); \
	71145	__ret_838 = __builtin_shufflevector(__ret_838, __ret_838, __lane_reverse_64_16); \
	71146	__ret_838; \
68629	71147	})
68630	71148	#endif
68631	71149
68632	71150	#ifdef __LITTLE_ENDIAN__
68633		#define vsudotq_lane_s32(__p0_823, __p1_823, __p2_823, __p3_823) __extension__ ({ \
68634		int32x4_t __ret_823; \
68635		int32x4_t __s0_823 = __p0_823; \
68636		int8x16_t __s1_823 = __p1_823; \
68637		uint8x8_t __s2_823 = __p2_823; \
68638		__ret_823 = vusdotq_s32(__s0_823, (uint8x16_t)(splatq_lane_s32(__builtin_bit_cast(int32x2_t, __s2_823), __p3_823)), __s1_823); \
68639		__ret_823; \
	71151	#define vsudotq_lane_s32(__p0_839, __p1_839, __p2_839, __p3_839) __extension__ ({ \
	71152	int32x4_t __ret_839; \
	71153	int32x4_t __s0_839 = __p0_839; \
	71154	int8x16_t __s1_839 = __p1_839; \
	71155	uint8x8_t __s2_839 = __p2_839; \
	71156	__ret_839 = vusdotq_s32(__s0_839, __builtin_bit_cast(uint8x16_t, splatq_lane_s32(__builtin_bit_cast(int32x2_t, __s2_839), __p3_839)), __s1_839); \
	71157	__ret_839; \
68640	71158	})
68641	71159	#else
68642		#define vsudotq_lane_s32(__p0_824, __p1_824, __p2_824, __p3_824) __extension__ ({ \
68643		int32x4_t __ret_824; \
68644		int32x4_t __s0_824 = __p0_824; \
68645		int8x16_t __s1_824 = __p1_824; \
68646		uint8x8_t __s2_824 = __p2_824; \
68647		int32x4_t __rev0_824; __rev0_824 = __builtin_shufflevector(__s0_824, __s0_824, 3, 2, 1, 0); \
68648		int8x16_t __rev1_824; __rev1_824 = __builtin_shufflevector(__s1_824, __s1_824, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
68649		uint8x8_t __rev2_824; __rev2_824 = __builtin_shufflevector(__s2_824, __s2_824, 7, 6, 5, 4, 3, 2, 1, 0); \
68650		__ret_824 = __noswap_vusdotq_s32(__rev0_824, (uint8x16_t)(__noswap_splatq_lane_s32(__builtin_bit_cast(int32x2_t, __rev2_824), __p3_824)), __rev1_824); \
68651		__ret_824 = __builtin_shufflevector(__ret_824, __ret_824, 3, 2, 1, 0); \
68652		__ret_824; \
	71160	#define vsudotq_lane_s32(__p0_840, __p1_840, __p2_840, __p3_840) __extension__ ({ \
	71161	int32x4_t __ret_840; \
	71162	int32x4_t __s0_840 = __p0_840; \
	71163	int8x16_t __s1_840 = __p1_840; \
	71164	uint8x8_t __s2_840 = __p2_840; \
	71165	int32x4_t __rev0_840; __rev0_840 = __builtin_shufflevector(__s0_840, __s0_840, __lane_reverse_128_32); \
	71166	int8x16_t __rev1_840; __rev1_840 = __builtin_shufflevector(__s1_840, __s1_840, __lane_reverse_128_8); \
	71167	uint8x8_t __rev2_840; __rev2_840 = __builtin_shufflevector(__s2_840, __s2_840, __lane_reverse_64_8); \
	71168	__ret_840 = __noswap_vusdotq_s32(__rev0_840, __builtin_bit_cast(uint8x16_t, __noswap_splatq_lane_s32(__builtin_bit_cast(int32x2_t, __rev2_840), __p3_840)), __rev1_840); \
	71169	__ret_840 = __builtin_shufflevector(__ret_840, __ret_840, __lane_reverse_128_32); \
	71170	__ret_840; \
68653	71171	})
68654	71172	#endif
68655	71173
68656	71174	#ifdef __LITTLE_ENDIAN__
68657		#define vsudot_lane_s32(__p0_825, __p1_825, __p2_825, __p3_825) __extension__ ({ \
68658		int32x2_t __ret_825; \
68659		int32x2_t __s0_825 = __p0_825; \
68660		int8x8_t __s1_825 = __p1_825; \
68661		uint8x8_t __s2_825 = __p2_825; \
68662		__ret_825 = vusdot_s32(__s0_825, (uint8x8_t)(splat_lane_s32(__builtin_bit_cast(int32x2_t, __s2_825), __p3_825)), __s1_825); \
68663		__ret_825; \
	71175	#define vsudot_lane_s32(__p0_841, __p1_841, __p2_841, __p3_841) __extension__ ({ \
	71176	int32x2_t __ret_841; \
	71177	int32x2_t __s0_841 = __p0_841; \
	71178	int8x8_t __s1_841 = __p1_841; \
	71179	uint8x8_t __s2_841 = __p2_841; \
	71180	__ret_841 = vusdot_s32(__s0_841, __builtin_bit_cast(uint8x8_t, splat_lane_s32(__builtin_bit_cast(int32x2_t, __s2_841), __p3_841)), __s1_841); \
	71181	__ret_841; \
68664	71182	})
68665	71183	#else
68666		#define vsudot_lane_s32(__p0_826, __p1_826, __p2_826, __p3_826) __extension__ ({ \
68667		int32x2_t __ret_826; \
68668		int32x2_t __s0_826 = __p0_826; \
68669		int8x8_t __s1_826 = __p1_826; \
68670		uint8x8_t __s2_826 = __p2_826; \
68671		int32x2_t __rev0_826; __rev0_826 = __builtin_shufflevector(__s0_826, __s0_826, 1, 0); \
68672		int8x8_t __rev1_826; __rev1_826 = __builtin_shufflevector(__s1_826, __s1_826, 7, 6, 5, 4, 3, 2, 1, 0); \
68673		uint8x8_t __rev2_826; __rev2_826 = __builtin_shufflevector(__s2_826, __s2_826, 7, 6, 5, 4, 3, 2, 1, 0); \
68674		__ret_826 = __noswap_vusdot_s32(__rev0_826, (uint8x8_t)(__noswap_splat_lane_s32(__builtin_bit_cast(int32x2_t, __rev2_826), __p3_826)), __rev1_826); \
68675		__ret_826 = __builtin_shufflevector(__ret_826, __ret_826, 1, 0); \
68676		__ret_826; \
	71184	#define vsudot_lane_s32(__p0_842, __p1_842, __p2_842, __p3_842) __extension__ ({ \
	71185	int32x2_t __ret_842; \
	71186	int32x2_t __s0_842 = __p0_842; \
	71187	int8x8_t __s1_842 = __p1_842; \
	71188	uint8x8_t __s2_842 = __p2_842; \
	71189	int32x2_t __rev0_842; __rev0_842 = __builtin_shufflevector(__s0_842, __s0_842, __lane_reverse_64_32); \
	71190	int8x8_t __rev1_842; __rev1_842 = __builtin_shufflevector(__s1_842, __s1_842, __lane_reverse_64_8); \
	71191	uint8x8_t __rev2_842; __rev2_842 = __builtin_shufflevector(__s2_842, __s2_842, __lane_reverse_64_8); \
	71192	__ret_842 = __noswap_vusdot_s32(__rev0_842, __builtin_bit_cast(uint8x8_t, __noswap_splat_lane_s32(__builtin_bit_cast(int32x2_t, __rev2_842), __p3_842)), __rev1_842); \
	71193	__ret_842 = __builtin_shufflevector(__ret_842, __ret_842, __lane_reverse_64_32); \
	71194	__ret_842; \
68677	71195	})
68678	71196	#endif
68679	71197
68680	71198	#ifdef __LITTLE_ENDIAN__
68681		#define vusdotq_lane_s32(__p0_827, __p1_827, __p2_827, __p3_827) __extension__ ({ \
68682		int32x4_t __ret_827; \
68683		int32x4_t __s0_827 = __p0_827; \
68684		uint8x16_t __s1_827 = __p1_827; \
68685		int8x8_t __s2_827 = __p2_827; \
68686		__ret_827 = vusdotq_s32(__s0_827, __s1_827, (int8x16_t)(splatq_lane_s32(__builtin_bit_cast(int32x2_t, __s2_827), __p3_827))); \
68687		__ret_827; \
	71199	#define vusdotq_lane_s32(__p0_843, __p1_843, __p2_843, __p3_843) __extension__ ({ \
	71200	int32x4_t __ret_843; \
	71201	int32x4_t __s0_843 = __p0_843; \
	71202	uint8x16_t __s1_843 = __p1_843; \
	71203	int8x8_t __s2_843 = __p2_843; \
	71204	__ret_843 = vusdotq_s32(__s0_843, __s1_843, __builtin_bit_cast(int8x16_t, splatq_lane_s32(__builtin_bit_cast(int32x2_t, __s2_843), __p3_843))); \
	71205	__ret_843; \
68688	71206	})
68689	71207	#else
68690		#define vusdotq_lane_s32(__p0_828, __p1_828, __p2_828, __p3_828) __extension__ ({ \
68691		int32x4_t __ret_828; \
68692		int32x4_t __s0_828 = __p0_828; \
68693		uint8x16_t __s1_828 = __p1_828; \
68694		int8x8_t __s2_828 = __p2_828; \
68695		int32x4_t __rev0_828; __rev0_828 = __builtin_shufflevector(__s0_828, __s0_828, 3, 2, 1, 0); \
68696		uint8x16_t __rev1_828; __rev1_828 = __builtin_shufflevector(__s1_828, __s1_828, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
68697		int8x8_t __rev2_828; __rev2_828 = __builtin_shufflevector(__s2_828, __s2_828, 7, 6, 5, 4, 3, 2, 1, 0); \
68698		__ret_828 = __noswap_vusdotq_s32(__rev0_828, __rev1_828, (int8x16_t)(__noswap_splatq_lane_s32(__builtin_bit_cast(int32x2_t, __rev2_828), __p3_828))); \
68699		__ret_828 = __builtin_shufflevector(__ret_828, __ret_828, 3, 2, 1, 0); \
68700		__ret_828; \
	71208	#define vusdotq_lane_s32(__p0_844, __p1_844, __p2_844, __p3_844) __extension__ ({ \
	71209	int32x4_t __ret_844; \
	71210	int32x4_t __s0_844 = __p0_844; \
	71211	uint8x16_t __s1_844 = __p1_844; \
	71212	int8x8_t __s2_844 = __p2_844; \
	71213	int32x4_t __rev0_844; __rev0_844 = __builtin_shufflevector(__s0_844, __s0_844, __lane_reverse_128_32); \
	71214	uint8x16_t __rev1_844; __rev1_844 = __builtin_shufflevector(__s1_844, __s1_844, __lane_reverse_128_8); \
	71215	int8x8_t __rev2_844; __rev2_844 = __builtin_shufflevector(__s2_844, __s2_844, __lane_reverse_64_8); \
	71216	__ret_844 = __noswap_vusdotq_s32(__rev0_844, __rev1_844, __builtin_bit_cast(int8x16_t, __noswap_splatq_lane_s32(__builtin_bit_cast(int32x2_t, __rev2_844), __p3_844))); \
	71217	__ret_844 = __builtin_shufflevector(__ret_844, __ret_844, __lane_reverse_128_32); \
	71218	__ret_844; \
68701	71219	})
68702	71220	#endif
68703	71221
68704	71222	#ifdef __LITTLE_ENDIAN__
68705		#define vusdot_lane_s32(__p0_829, __p1_829, __p2_829, __p3_829) __extension__ ({ \
68706		int32x2_t __ret_829; \
68707		int32x2_t __s0_829 = __p0_829; \
68708		uint8x8_t __s1_829 = __p1_829; \
68709		int8x8_t __s2_829 = __p2_829; \
68710		__ret_829 = vusdot_s32(__s0_829, __s1_829, (int8x8_t)(splat_lane_s32(__builtin_bit_cast(int32x2_t, __s2_829), __p3_829))); \
68711		__ret_829; \
	71223	#define vusdot_lane_s32(__p0_845, __p1_845, __p2_845, __p3_845) __extension__ ({ \
	71224	int32x2_t __ret_845; \
	71225	int32x2_t __s0_845 = __p0_845; \
	71226	uint8x8_t __s1_845 = __p1_845; \
	71227	int8x8_t __s2_845 = __p2_845; \
	71228	__ret_845 = vusdot_s32(__s0_845, __s1_845, __builtin_bit_cast(int8x8_t, splat_lane_s32(__builtin_bit_cast(int32x2_t, __s2_845), __p3_845))); \
	71229	__ret_845; \
68712	71230	})
68713	71231	#else
68714		#define vusdot_lane_s32(__p0_830, __p1_830, __p2_830, __p3_830) __extension__ ({ \
68715		int32x2_t __ret_830; \
68716		int32x2_t __s0_830 = __p0_830; \
68717		uint8x8_t __s1_830 = __p1_830; \
68718		int8x8_t __s2_830 = __p2_830; \
68719		int32x2_t __rev0_830; __rev0_830 = __builtin_shufflevector(__s0_830, __s0_830, 1, 0); \
68720		uint8x8_t __rev1_830; __rev1_830 = __builtin_shufflevector(__s1_830, __s1_830, 7, 6, 5, 4, 3, 2, 1, 0); \
68721		int8x8_t __rev2_830; __rev2_830 = __builtin_shufflevector(__s2_830, __s2_830, 7, 6, 5, 4, 3, 2, 1, 0); \
68722		__ret_830 = __noswap_vusdot_s32(__rev0_830, __rev1_830, (int8x8_t)(__noswap_splat_lane_s32(__builtin_bit_cast(int32x2_t, __rev2_830), __p3_830))); \
68723		__ret_830 = __builtin_shufflevector(__ret_830, __ret_830, 1, 0); \
68724		__ret_830; \
	71232	#define vusdot_lane_s32(__p0_846, __p1_846, __p2_846, __p3_846) __extension__ ({ \
	71233	int32x2_t __ret_846; \
	71234	int32x2_t __s0_846 = __p0_846; \
	71235	uint8x8_t __s1_846 = __p1_846; \
	71236	int8x8_t __s2_846 = __p2_846; \
	71237	int32x2_t __rev0_846; __rev0_846 = __builtin_shufflevector(__s0_846, __s0_846, __lane_reverse_64_32); \
	71238	uint8x8_t __rev1_846; __rev1_846 = __builtin_shufflevector(__s1_846, __s1_846, __lane_reverse_64_8); \
	71239	int8x8_t __rev2_846; __rev2_846 = __builtin_shufflevector(__s2_846, __s2_846, __lane_reverse_64_8); \
	71240	__ret_846 = __noswap_vusdot_s32(__rev0_846, __rev1_846, __builtin_bit_cast(int8x8_t, __noswap_splat_lane_s32(__builtin_bit_cast(int32x2_t, __rev2_846), __p3_846))); \
	71241	__ret_846 = __builtin_shufflevector(__ret_846, __ret_846, __lane_reverse_64_32); \
	71242	__ret_846; \
68725	71243	})
68726	71244	#endif
68727	71245
...	...	@@ -68734,11 +71252,11 @@ __ai __attribute__((target("neon"))) uint8x16_t vabaq_u8(uint8x16_t __p0, uint8x
68734	71252	#else
68735	71253	__ai __attribute__((target("neon"))) uint8x16_t vabaq_u8(uint8x16_t __p0, uint8x16_t __p1, uint8x16_t __p2) {
68736	71254	uint8x16_t __ret;
68737		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
68738		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
68739		uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	71255	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	71256	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	71257	uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
68740	71258	__ret = __rev0 + __noswap_vabdq_u8(__rev1, __rev2);
68741		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	71259	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
68742	71260	return __ret;
68743	71261	}
68744	71262	#endif
...	...	@@ -68752,11 +71270,11 @@ __ai __attribute__((target("neon"))) uint32x4_t vabaq_u32(uint32x4_t __p0, uint3
68752	71270	#else
68753	71271	__ai __attribute__((target("neon"))) uint32x4_t vabaq_u32(uint32x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
68754	71272	uint32x4_t __ret;
68755		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
68756		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
68757		uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	71273	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	71274	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	71275	uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
68758	71276	__ret = __rev0 + __noswap_vabdq_u32(__rev1, __rev2);
68759		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	71277	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
68760	71278	return __ret;
68761	71279	}
68762	71280	#endif
...	...	@@ -68770,11 +71288,11 @@ __ai __attribute__((target("neon"))) uint16x8_t vabaq_u16(uint16x8_t __p0, uint1
68770	71288	#else
68771	71289	__ai __attribute__((target("neon"))) uint16x8_t vabaq_u16(uint16x8_t __p0, uint16x8_t __p1, uint16x8_t __p2) {
68772	71290	uint16x8_t __ret;
68773		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
68774		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
68775		uint16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	71291	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	71292	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	71293	uint16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
68776	71294	__ret = __rev0 + __noswap_vabdq_u16(__rev1, __rev2);
68777		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	71295	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
68778	71296	return __ret;
68779	71297	}
68780	71298	#endif
...	...	@@ -68788,11 +71306,11 @@ __ai __attribute__((target("neon"))) int8x16_t vabaq_s8(int8x16_t __p0, int8x16_
68788	71306	#else
68789	71307	__ai __attribute__((target("neon"))) int8x16_t vabaq_s8(int8x16_t __p0, int8x16_t __p1, int8x16_t __p2) {
68790	71308	int8x16_t __ret;
68791		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
68792		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
68793		int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	71309	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	71310	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	71311	int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
68794	71312	__ret = __rev0 + __noswap_vabdq_s8(__rev1, __rev2);
68795		__ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	71313	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_8);
68796	71314	return __ret;
68797	71315	}
68798	71316	#endif
...	...	@@ -68806,11 +71324,11 @@ __ai __attribute__((target("neon"))) int32x4_t vabaq_s32(int32x4_t __p0, int32x4
68806	71324	#else
68807	71325	__ai __attribute__((target("neon"))) int32x4_t vabaq_s32(int32x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
68808	71326	int32x4_t __ret;
68809		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
68810		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
68811		int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	71327	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	71328	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	71329	int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
68812	71330	__ret = __rev0 + __noswap_vabdq_s32(__rev1, __rev2);
68813		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	71331	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
68814	71332	return __ret;
68815	71333	}
68816	71334	#endif
...	...	@@ -68824,11 +71342,11 @@ __ai __attribute__((target("neon"))) int16x8_t vabaq_s16(int16x8_t __p0, int16x8
68824	71342	#else
68825	71343	__ai __attribute__((target("neon"))) int16x8_t vabaq_s16(int16x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
68826	71344	int16x8_t __ret;
68827		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
68828		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
68829		int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	71345	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	71346	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	71347	int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
68830	71348	__ret = __rev0 + __noswap_vabdq_s16(__rev1, __rev2);
68831		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	71349	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
68832	71350	return __ret;
68833	71351	}
68834	71352	#endif
...	...	@@ -68842,11 +71360,11 @@ __ai __attribute__((target("neon"))) uint8x8_t vaba_u8(uint8x8_t __p0, uint8x8_t
68842	71360	#else
68843	71361	__ai __attribute__((target("neon"))) uint8x8_t vaba_u8(uint8x8_t __p0, uint8x8_t __p1, uint8x8_t __p2) {
68844	71362	uint8x8_t __ret;
68845		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
68846		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
68847		uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	71363	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	71364	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	71365	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
68848	71366	__ret = __rev0 + __noswap_vabd_u8(__rev1, __rev2);
68849		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	71367	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
68850	71368	return __ret;
68851	71369	}
68852	71370	#endif
...	...	@@ -68860,11 +71378,11 @@ __ai __attribute__((target("neon"))) uint32x2_t vaba_u32(uint32x2_t __p0, uint32
68860	71378	#else
68861	71379	__ai __attribute__((target("neon"))) uint32x2_t vaba_u32(uint32x2_t __p0, uint32x2_t __p1, uint32x2_t __p2) {
68862	71380	uint32x2_t __ret;
68863		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
68864		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
68865		uint32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
	71381	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	71382	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	71383	uint32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_32);
68866	71384	__ret = __rev0 + __noswap_vabd_u32(__rev1, __rev2);
68867		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	71385	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
68868	71386	return __ret;
68869	71387	}
68870	71388	#endif
...	...	@@ -68878,11 +71396,11 @@ __ai __attribute__((target("neon"))) uint16x4_t vaba_u16(uint16x4_t __p0, uint16
68878	71396	#else
68879	71397	__ai __attribute__((target("neon"))) uint16x4_t vaba_u16(uint16x4_t __p0, uint16x4_t __p1, uint16x4_t __p2) {
68880	71398	uint16x4_t __ret;
68881		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
68882		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
68883		uint16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	71399	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	71400	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	71401	uint16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
68884	71402	__ret = __rev0 + __noswap_vabd_u16(__rev1, __rev2);
68885		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	71403	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
68886	71404	return __ret;
68887	71405	}
68888	71406	#endif
...	...	@@ -68896,11 +71414,11 @@ __ai __attribute__((target("neon"))) int8x8_t vaba_s8(int8x8_t __p0, int8x8_t __
68896	71414	#else
68897	71415	__ai __attribute__((target("neon"))) int8x8_t vaba_s8(int8x8_t __p0, int8x8_t __p1, int8x8_t __p2) {
68898	71416	int8x8_t __ret;
68899		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
68900		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
68901		int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	71417	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	71418	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	71419	int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
68902	71420	__ret = __rev0 + __noswap_vabd_s8(__rev1, __rev2);
68903		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	71421	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_8);
68904	71422	return __ret;
68905	71423	}
68906	71424	#endif
...	...	@@ -68914,11 +71432,11 @@ __ai __attribute__((target("neon"))) int32x2_t vaba_s32(int32x2_t __p0, int32x2_
68914	71432	#else
68915	71433	__ai __attribute__((target("neon"))) int32x2_t vaba_s32(int32x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
68916	71434	int32x2_t __ret;
68917		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
68918		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
68919		int32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
	71435	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	71436	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	71437	int32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_32);
68920	71438	__ret = __rev0 + __noswap_vabd_s32(__rev1, __rev2);
68921		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	71439	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_32);
68922	71440	return __ret;
68923	71441	}
68924	71442	#endif
...	...	@@ -68932,11 +71450,11 @@ __ai __attribute__((target("neon"))) int16x4_t vaba_s16(int16x4_t __p0, int16x4_
68932	71450	#else
68933	71451	__ai __attribute__((target("neon"))) int16x4_t vaba_s16(int16x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
68934	71452	int16x4_t __ret;
68935		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
68936		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
68937		int16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	71453	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	71454	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	71455	int16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
68938	71456	__ret = __rev0 + __noswap_vabd_s16(__rev1, __rev2);
68939		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	71457	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_64_16);
68940	71458	return __ret;
68941	71459	}
68942	71460	#endif
...	...	@@ -68944,21 +71462,21 @@ __ai __attribute__((target("neon"))) int16x4_t vaba_s16(int16x4_t __p0, int16x4_
68944	71462	#ifdef __LITTLE_ENDIAN__
68945	71463	__ai __attribute__((target("neon"))) uint16x8_t vabdl_u8(uint8x8_t __p0, uint8x8_t __p1) {
68946	71464	uint16x8_t __ret;
68947		__ret = (uint16x8_t)(vmovl_u8((uint8x8_t)(vabd_u8(__p0, __p1))));
	71465	__ret = __builtin_bit_cast(uint16x8_t, vmovl_u8(__builtin_bit_cast(uint8x8_t, vabd_u8(__p0, __p1))));
68948	71466	return __ret;
68949	71467	}
68950	71468	#else
68951	71469	__ai __attribute__((target("neon"))) uint16x8_t vabdl_u8(uint8x8_t __p0, uint8x8_t __p1) {
68952	71470	uint16x8_t __ret;
68953		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
68954		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
68955		__ret = (uint16x8_t)(__noswap_vmovl_u8((uint8x8_t)(__noswap_vabd_u8(__rev0, __rev1))));
68956		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	71471	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	71472	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	71473	__ret = __builtin_bit_cast(uint16x8_t, __noswap_vmovl_u8(__builtin_bit_cast(uint8x8_t, __noswap_vabd_u8(__rev0, __rev1))));
	71474	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
68957	71475	return __ret;
68958	71476	}
68959	71477	__ai __attribute__((target("neon"))) uint16x8_t __noswap_vabdl_u8(uint8x8_t __p0, uint8x8_t __p1) {
68960	71478	uint16x8_t __ret;
68961		__ret = (uint16x8_t)(__noswap_vmovl_u8((uint8x8_t)(__noswap_vabd_u8(__p0, __p1))));
	71479	__ret = __builtin_bit_cast(uint16x8_t, __noswap_vmovl_u8(__builtin_bit_cast(uint8x8_t, __noswap_vabd_u8(__p0, __p1))));
68962	71480	return __ret;
68963	71481	}
68964	71482	#endif
...	...	@@ -68966,21 +71484,21 @@ __ai __attribute__((target("neon"))) uint16x8_t __noswap_vabdl_u8(uint8x8_t __p0
68966	71484	#ifdef __LITTLE_ENDIAN__
68967	71485	__ai __attribute__((target("neon"))) uint64x2_t vabdl_u32(uint32x2_t __p0, uint32x2_t __p1) {
68968	71486	uint64x2_t __ret;
68969		__ret = (uint64x2_t)(vmovl_u32((uint32x2_t)(vabd_u32(__p0, __p1))));
	71487	__ret = __builtin_bit_cast(uint64x2_t, vmovl_u32(__builtin_bit_cast(uint32x2_t, vabd_u32(__p0, __p1))));
68970	71488	return __ret;
68971	71489	}
68972	71490	#else
68973	71491	__ai __attribute__((target("neon"))) uint64x2_t vabdl_u32(uint32x2_t __p0, uint32x2_t __p1) {
68974	71492	uint64x2_t __ret;
68975		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
68976		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
68977		__ret = (uint64x2_t)(__noswap_vmovl_u32((uint32x2_t)(__noswap_vabd_u32(__rev0, __rev1))));
68978		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	71493	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	71494	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	71495	__ret = __builtin_bit_cast(uint64x2_t, __noswap_vmovl_u32(__builtin_bit_cast(uint32x2_t, __noswap_vabd_u32(__rev0, __rev1))));
	71496	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
68979	71497	return __ret;
68980	71498	}
68981	71499	__ai __attribute__((target("neon"))) uint64x2_t __noswap_vabdl_u32(uint32x2_t __p0, uint32x2_t __p1) {
68982	71500	uint64x2_t __ret;
68983		__ret = (uint64x2_t)(__noswap_vmovl_u32((uint32x2_t)(__noswap_vabd_u32(__p0, __p1))));
	71501	__ret = __builtin_bit_cast(uint64x2_t, __noswap_vmovl_u32(__builtin_bit_cast(uint32x2_t, __noswap_vabd_u32(__p0, __p1))));
68984	71502	return __ret;
68985	71503	}
68986	71504	#endif
...	...	@@ -68988,21 +71506,21 @@ __ai __attribute__((target("neon"))) uint64x2_t __noswap_vabdl_u32(uint32x2_t __
68988	71506	#ifdef __LITTLE_ENDIAN__
68989	71507	__ai __attribute__((target("neon"))) uint32x4_t vabdl_u16(uint16x4_t __p0, uint16x4_t __p1) {
68990	71508	uint32x4_t __ret;
68991		__ret = (uint32x4_t)(vmovl_u16((uint16x4_t)(vabd_u16(__p0, __p1))));
	71509	__ret = __builtin_bit_cast(uint32x4_t, vmovl_u16(__builtin_bit_cast(uint16x4_t, vabd_u16(__p0, __p1))));
68992	71510	return __ret;
68993	71511	}
68994	71512	#else
68995	71513	__ai __attribute__((target("neon"))) uint32x4_t vabdl_u16(uint16x4_t __p0, uint16x4_t __p1) {
68996	71514	uint32x4_t __ret;
68997		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
68998		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
68999		__ret = (uint32x4_t)(__noswap_vmovl_u16((uint16x4_t)(__noswap_vabd_u16(__rev0, __rev1))));
69000		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	71515	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	71516	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	71517	__ret = __builtin_bit_cast(uint32x4_t, __noswap_vmovl_u16(__builtin_bit_cast(uint16x4_t, __noswap_vabd_u16(__rev0, __rev1))));
	71518	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
69001	71519	return __ret;
69002	71520	}
69003	71521	__ai __attribute__((target("neon"))) uint32x4_t __noswap_vabdl_u16(uint16x4_t __p0, uint16x4_t __p1) {
69004	71522	uint32x4_t __ret;
69005		__ret = (uint32x4_t)(__noswap_vmovl_u16((uint16x4_t)(__noswap_vabd_u16(__p0, __p1))));
	71523	__ret = __builtin_bit_cast(uint32x4_t, __noswap_vmovl_u16(__builtin_bit_cast(uint16x4_t, __noswap_vabd_u16(__p0, __p1))));
69006	71524	return __ret;
69007	71525	}
69008	71526	#endif
...	...	@@ -69010,21 +71528,21 @@ __ai __attribute__((target("neon"))) uint32x4_t __noswap_vabdl_u16(uint16x4_t __
69010	71528	#ifdef __LITTLE_ENDIAN__
69011	71529	__ai __attribute__((target("neon"))) int16x8_t vabdl_s8(int8x8_t __p0, int8x8_t __p1) {
69012	71530	int16x8_t __ret;
69013		__ret = (int16x8_t)(vmovl_u8((uint8x8_t)(vabd_s8(__p0, __p1))));
	71531	__ret = __builtin_bit_cast(int16x8_t, vmovl_u8(__builtin_bit_cast(uint8x8_t, vabd_s8(__p0, __p1))));
69014	71532	return __ret;
69015	71533	}
69016	71534	#else
69017	71535	__ai __attribute__((target("neon"))) int16x8_t vabdl_s8(int8x8_t __p0, int8x8_t __p1) {
69018	71536	int16x8_t __ret;
69019		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
69020		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
69021		__ret = (int16x8_t)(__noswap_vmovl_u8((uint8x8_t)(__noswap_vabd_s8(__rev0, __rev1))));
69022		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	71537	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	71538	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	71539	__ret = __builtin_bit_cast(int16x8_t, __noswap_vmovl_u8(__builtin_bit_cast(uint8x8_t, __noswap_vabd_s8(__rev0, __rev1))));
	71540	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
69023	71541	return __ret;
69024	71542	}
69025	71543	__ai __attribute__((target("neon"))) int16x8_t __noswap_vabdl_s8(int8x8_t __p0, int8x8_t __p1) {
69026	71544	int16x8_t __ret;
69027		__ret = (int16x8_t)(__noswap_vmovl_u8((uint8x8_t)(__noswap_vabd_s8(__p0, __p1))));
	71545	__ret = __builtin_bit_cast(int16x8_t, __noswap_vmovl_u8(__builtin_bit_cast(uint8x8_t, __noswap_vabd_s8(__p0, __p1))));
69028	71546	return __ret;
69029	71547	}
69030	71548	#endif
...	...	@@ -69032,21 +71550,21 @@ __ai __attribute__((target("neon"))) int16x8_t __noswap_vabdl_s8(int8x8_t __p0,
69032	71550	#ifdef __LITTLE_ENDIAN__
69033	71551	__ai __attribute__((target("neon"))) int64x2_t vabdl_s32(int32x2_t __p0, int32x2_t __p1) {
69034	71552	int64x2_t __ret;
69035		__ret = (int64x2_t)(vmovl_u32((uint32x2_t)(vabd_s32(__p0, __p1))));
	71553	__ret = __builtin_bit_cast(int64x2_t, vmovl_u32(__builtin_bit_cast(uint32x2_t, vabd_s32(__p0, __p1))));
69036	71554	return __ret;
69037	71555	}
69038	71556	#else
69039	71557	__ai __attribute__((target("neon"))) int64x2_t vabdl_s32(int32x2_t __p0, int32x2_t __p1) {
69040	71558	int64x2_t __ret;
69041		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
69042		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
69043		__ret = (int64x2_t)(__noswap_vmovl_u32((uint32x2_t)(__noswap_vabd_s32(__rev0, __rev1))));
69044		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	71559	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	71560	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	71561	__ret = __builtin_bit_cast(int64x2_t, __noswap_vmovl_u32(__builtin_bit_cast(uint32x2_t, __noswap_vabd_s32(__rev0, __rev1))));
	71562	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
69045	71563	return __ret;
69046	71564	}
69047	71565	__ai __attribute__((target("neon"))) int64x2_t __noswap_vabdl_s32(int32x2_t __p0, int32x2_t __p1) {
69048	71566	int64x2_t __ret;
69049		__ret = (int64x2_t)(__noswap_vmovl_u32((uint32x2_t)(__noswap_vabd_s32(__p0, __p1))));
	71567	__ret = __builtin_bit_cast(int64x2_t, __noswap_vmovl_u32(__builtin_bit_cast(uint32x2_t, __noswap_vabd_s32(__p0, __p1))));
69050	71568	return __ret;
69051	71569	}
69052	71570	#endif
...	...	@@ -69054,21 +71572,21 @@ __ai __attribute__((target("neon"))) int64x2_t __noswap_vabdl_s32(int32x2_t __p0
69054	71572	#ifdef __LITTLE_ENDIAN__
69055	71573	__ai __attribute__((target("neon"))) int32x4_t vabdl_s16(int16x4_t __p0, int16x4_t __p1) {
69056	71574	int32x4_t __ret;
69057		__ret = (int32x4_t)(vmovl_u16((uint16x4_t)(vabd_s16(__p0, __p1))));
	71575	__ret = __builtin_bit_cast(int32x4_t, vmovl_u16(__builtin_bit_cast(uint16x4_t, vabd_s16(__p0, __p1))));
69058	71576	return __ret;
69059	71577	}
69060	71578	#else
69061	71579	__ai __attribute__((target("neon"))) int32x4_t vabdl_s16(int16x4_t __p0, int16x4_t __p1) {
69062	71580	int32x4_t __ret;
69063		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
69064		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
69065		__ret = (int32x4_t)(__noswap_vmovl_u16((uint16x4_t)(__noswap_vabd_s16(__rev0, __rev1))));
69066		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	71581	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	71582	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	71583	__ret = __builtin_bit_cast(int32x4_t, __noswap_vmovl_u16(__builtin_bit_cast(uint16x4_t, __noswap_vabd_s16(__rev0, __rev1))));
	71584	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
69067	71585	return __ret;
69068	71586	}
69069	71587	__ai __attribute__((target("neon"))) int32x4_t __noswap_vabdl_s16(int16x4_t __p0, int16x4_t __p1) {
69070	71588	int32x4_t __ret;
69071		__ret = (int32x4_t)(__noswap_vmovl_u16((uint16x4_t)(__noswap_vabd_s16(__p0, __p1))));
	71589	__ret = __builtin_bit_cast(int32x4_t, __noswap_vmovl_u16(__builtin_bit_cast(uint16x4_t, __noswap_vabd_s16(__p0, __p1))));
69072	71590	return __ret;
69073	71591	}
69074	71592	#endif
...	...	@@ -69082,10 +71600,10 @@ __ai __attribute__((target("neon"))) uint16x8_t vaddl_u8(uint8x8_t __p0, uint8x8
69082	71600	#else
69083	71601	__ai __attribute__((target("neon"))) uint16x8_t vaddl_u8(uint8x8_t __p0, uint8x8_t __p1) {
69084	71602	uint16x8_t __ret;
69085		uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
69086		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	71603	uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	71604	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
69087	71605	__ret = __noswap_vmovl_u8(__rev0) + __noswap_vmovl_u8(__rev1);
69088		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	71606	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
69089	71607	return __ret;
69090	71608	}
69091	71609	#endif
...	...	@@ -69099,10 +71617,10 @@ __ai __attribute__((target("neon"))) uint64x2_t vaddl_u32(uint32x2_t __p0, uint3
69099	71617	#else
69100	71618	__ai __attribute__((target("neon"))) uint64x2_t vaddl_u32(uint32x2_t __p0, uint32x2_t __p1) {
69101	71619	uint64x2_t __ret;
69102		uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
69103		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	71620	uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	71621	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
69104	71622	__ret = __noswap_vmovl_u32(__rev0) + __noswap_vmovl_u32(__rev1);
69105		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	71623	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
69106	71624	return __ret;
69107	71625	}
69108	71626	#endif
...	...	@@ -69116,10 +71634,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vaddl_u16(uint16x4_t __p0, uint1
69116	71634	#else
69117	71635	__ai __attribute__((target("neon"))) uint32x4_t vaddl_u16(uint16x4_t __p0, uint16x4_t __p1) {
69118	71636	uint32x4_t __ret;
69119		uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
69120		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	71637	uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	71638	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
69121	71639	__ret = __noswap_vmovl_u16(__rev0) + __noswap_vmovl_u16(__rev1);
69122		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	71640	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
69123	71641	return __ret;
69124	71642	}
69125	71643	#endif
...	...	@@ -69133,10 +71651,10 @@ __ai __attribute__((target("neon"))) int16x8_t vaddl_s8(int8x8_t __p0, int8x8_t
69133	71651	#else
69134	71652	__ai __attribute__((target("neon"))) int16x8_t vaddl_s8(int8x8_t __p0, int8x8_t __p1) {
69135	71653	int16x8_t __ret;
69136		int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
69137		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	71654	int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_8);
	71655	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
69138	71656	__ret = __noswap_vmovl_s8(__rev0) + __noswap_vmovl_s8(__rev1);
69139		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	71657	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
69140	71658	return __ret;
69141	71659	}
69142	71660	#endif
...	...	@@ -69150,10 +71668,10 @@ __ai __attribute__((target("neon"))) int64x2_t vaddl_s32(int32x2_t __p0, int32x2
69150	71668	#else
69151	71669	__ai __attribute__((target("neon"))) int64x2_t vaddl_s32(int32x2_t __p0, int32x2_t __p1) {
69152	71670	int64x2_t __ret;
69153		int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
69154		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	71671	int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_32);
	71672	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
69155	71673	__ret = __noswap_vmovl_s32(__rev0) + __noswap_vmovl_s32(__rev1);
69156		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	71674	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
69157	71675	return __ret;
69158	71676	}
69159	71677	#endif
...	...	@@ -69167,10 +71685,10 @@ __ai __attribute__((target("neon"))) int32x4_t vaddl_s16(int16x4_t __p0, int16x4
69167	71685	#else
69168	71686	__ai __attribute__((target("neon"))) int32x4_t vaddl_s16(int16x4_t __p0, int16x4_t __p1) {
69169	71687	int32x4_t __ret;
69170		int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
69171		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	71688	int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_64_16);
	71689	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
69172	71690	__ret = __noswap_vmovl_s16(__rev0) + __noswap_vmovl_s16(__rev1);
69173		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	71691	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
69174	71692	return __ret;
69175	71693	}
69176	71694	#endif
...	...	@@ -69184,10 +71702,10 @@ __ai __attribute__((target("neon"))) uint16x8_t vaddw_u8(uint16x8_t __p0, uint8x
69184	71702	#else
69185	71703	__ai __attribute__((target("neon"))) uint16x8_t vaddw_u8(uint16x8_t __p0, uint8x8_t __p1) {
69186	71704	uint16x8_t __ret;
69187		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
69188		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	71705	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	71706	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
69189	71707	__ret = __rev0 + __noswap_vmovl_u8(__rev1);
69190		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	71708	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
69191	71709	return __ret;
69192	71710	}
69193	71711	#endif
...	...	@@ -69201,10 +71719,10 @@ __ai __attribute__((target("neon"))) uint64x2_t vaddw_u32(uint64x2_t __p0, uint3
69201	71719	#else
69202	71720	__ai __attribute__((target("neon"))) uint64x2_t vaddw_u32(uint64x2_t __p0, uint32x2_t __p1) {
69203	71721	uint64x2_t __ret;
69204		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
69205		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	71722	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	71723	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
69206	71724	__ret = __rev0 + __noswap_vmovl_u32(__rev1);
69207		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	71725	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
69208	71726	return __ret;
69209	71727	}
69210	71728	#endif
...	...	@@ -69218,10 +71736,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vaddw_u16(uint32x4_t __p0, uint1
69218	71736	#else
69219	71737	__ai __attribute__((target("neon"))) uint32x4_t vaddw_u16(uint32x4_t __p0, uint16x4_t __p1) {
69220	71738	uint32x4_t __ret;
69221		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
69222		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	71739	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	71740	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
69223	71741	__ret = __rev0 + __noswap_vmovl_u16(__rev1);
69224		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	71742	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
69225	71743	return __ret;
69226	71744	}
69227	71745	#endif
...	...	@@ -69235,10 +71753,10 @@ __ai __attribute__((target("neon"))) int16x8_t vaddw_s8(int16x8_t __p0, int8x8_t
69235	71753	#else
69236	71754	__ai __attribute__((target("neon"))) int16x8_t vaddw_s8(int16x8_t __p0, int8x8_t __p1) {
69237	71755	int16x8_t __ret;
69238		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
69239		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	71756	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	71757	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
69240	71758	__ret = __rev0 + __noswap_vmovl_s8(__rev1);
69241		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	71759	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
69242	71760	return __ret;
69243	71761	}
69244	71762	#endif
...	...	@@ -69252,10 +71770,10 @@ __ai __attribute__((target("neon"))) int64x2_t vaddw_s32(int64x2_t __p0, int32x2
69252	71770	#else
69253	71771	__ai __attribute__((target("neon"))) int64x2_t vaddw_s32(int64x2_t __p0, int32x2_t __p1) {
69254	71772	int64x2_t __ret;
69255		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
69256		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	71773	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	71774	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
69257	71775	__ret = __rev0 + __noswap_vmovl_s32(__rev1);
69258		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	71776	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
69259	71777	return __ret;
69260	71778	}
69261	71779	#endif
...	...	@@ -69269,57 +71787,57 @@ __ai __attribute__((target("neon"))) int32x4_t vaddw_s16(int32x4_t __p0, int16x4
69269	71787	#else
69270	71788	__ai __attribute__((target("neon"))) int32x4_t vaddw_s16(int32x4_t __p0, int16x4_t __p1) {
69271	71789	int32x4_t __ret;
69272		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
69273		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	71790	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	71791	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
69274	71792	__ret = __rev0 + __noswap_vmovl_s16(__rev1);
69275		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	71793	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
69276	71794	return __ret;
69277	71795	}
69278	71796	#endif
69279	71797
69280	71798	#ifdef __LITTLE_ENDIAN__
69281		#define vget_lane_f16(__p0_831, __p1_831) __extension__ ({ \
69282		float16_t __ret_831; \
69283		float16x4_t __s0_831 = __p0_831; \
69284		__ret_831 = __builtin_bit_cast(float16_t, vget_lane_s16(__builtin_bit_cast(int16x4_t, __s0_831), __p1_831)); \
69285		__ret_831; \
	71799	#define vget_lane_f16(__p0_847, __p1_847) __extension__ ({ \
	71800	float16_t __ret_847; \
	71801	float16x4_t __s0_847 = __p0_847; \
	71802	__ret_847 = __builtin_bit_cast(float16_t, vget_lane_s16(__builtin_bit_cast(int16x4_t, __s0_847), __p1_847)); \
	71803	__ret_847; \
69286	71804	})
69287	71805	#else
69288		#define vget_lane_f16(__p0_832, __p1_832) __extension__ ({ \
69289		float16_t __ret_832; \
69290		float16x4_t __s0_832 = __p0_832; \
69291		float16x4_t __rev0_832; __rev0_832 = __builtin_shufflevector(__s0_832, __s0_832, 3, 2, 1, 0); \
69292		__ret_832 = __builtin_bit_cast(float16_t, __noswap_vget_lane_s16(__builtin_bit_cast(int16x4_t, __rev0_832), __p1_832)); \
69293		__ret_832; \
	71806	#define vget_lane_f16(__p0_848, __p1_848) __extension__ ({ \
	71807	float16_t __ret_848; \
	71808	float16x4_t __s0_848 = __p0_848; \
	71809	float16x4_t __rev0_848; __rev0_848 = __builtin_shufflevector(__s0_848, __s0_848, __lane_reverse_64_16); \
	71810	__ret_848 = __builtin_bit_cast(float16_t, __noswap_vget_lane_s16(__builtin_bit_cast(int16x4_t, __rev0_848), __p1_848)); \
	71811	__ret_848; \
69294	71812	})
69295		#define __noswap_vget_lane_f16(__p0_833, __p1_833) __extension__ ({ \
69296		float16_t __ret_833; \
69297		float16x4_t __s0_833 = __p0_833; \
69298		__ret_833 = __builtin_bit_cast(float16_t, __noswap_vget_lane_s16(__builtin_bit_cast(int16x4_t, __s0_833), __p1_833)); \
69299		__ret_833; \
	71813	#define __noswap_vget_lane_f16(__p0_849, __p1_849) __extension__ ({ \
	71814	float16_t __ret_849; \
	71815	float16x4_t __s0_849 = __p0_849; \
	71816	__ret_849 = __builtin_bit_cast(float16_t, __noswap_vget_lane_s16(__builtin_bit_cast(int16x4_t, __s0_849), __p1_849)); \
	71817	__ret_849; \
69300	71818	})
69301	71819	#endif
69302	71820
69303	71821	#ifdef __LITTLE_ENDIAN__
69304		#define vgetq_lane_f16(__p0_834, __p1_834) __extension__ ({ \
69305		float16_t __ret_834; \
69306		float16x8_t __s0_834 = __p0_834; \
69307		__ret_834 = __builtin_bit_cast(float16_t, vgetq_lane_s16(__builtin_bit_cast(int16x8_t, __s0_834), __p1_834)); \
69308		__ret_834; \
	71822	#define vgetq_lane_f16(__p0_850, __p1_850) __extension__ ({ \
	71823	float16_t __ret_850; \
	71824	float16x8_t __s0_850 = __p0_850; \
	71825	__ret_850 = __builtin_bit_cast(float16_t, vgetq_lane_s16(__builtin_bit_cast(int16x8_t, __s0_850), __p1_850)); \
	71826	__ret_850; \
69309	71827	})
69310	71828	#else
69311		#define vgetq_lane_f16(__p0_835, __p1_835) __extension__ ({ \
69312		float16_t __ret_835; \
69313		float16x8_t __s0_835 = __p0_835; \
69314		float16x8_t __rev0_835; __rev0_835 = __builtin_shufflevector(__s0_835, __s0_835, 7, 6, 5, 4, 3, 2, 1, 0); \
69315		__ret_835 = __builtin_bit_cast(float16_t, __noswap_vgetq_lane_s16(__builtin_bit_cast(int16x8_t, __rev0_835), __p1_835)); \
69316		__ret_835; \
	71829	#define vgetq_lane_f16(__p0_851, __p1_851) __extension__ ({ \
	71830	float16_t __ret_851; \
	71831	float16x8_t __s0_851 = __p0_851; \
	71832	float16x8_t __rev0_851; __rev0_851 = __builtin_shufflevector(__s0_851, __s0_851, __lane_reverse_128_16); \
	71833	__ret_851 = __builtin_bit_cast(float16_t, __noswap_vgetq_lane_s16(__builtin_bit_cast(int16x8_t, __rev0_851), __p1_851)); \
	71834	__ret_851; \
69317	71835	})
69318		#define __noswap_vgetq_lane_f16(__p0_836, __p1_836) __extension__ ({ \
69319		float16_t __ret_836; \
69320		float16x8_t __s0_836 = __p0_836; \
69321		__ret_836 = __builtin_bit_cast(float16_t, __noswap_vgetq_lane_s16(__builtin_bit_cast(int16x8_t, __s0_836), __p1_836)); \
69322		__ret_836; \
	71836	#define __noswap_vgetq_lane_f16(__p0_852, __p1_852) __extension__ ({ \
	71837	float16_t __ret_852; \
	71838	float16x8_t __s0_852 = __p0_852; \
	71839	__ret_852 = __builtin_bit_cast(float16_t, __noswap_vgetq_lane_s16(__builtin_bit_cast(int16x8_t, __s0_852), __p1_852)); \
	71840	__ret_852; \
69323	71841	})
69324	71842	#endif
69325	71843
...	...	@@ -69332,11 +71850,11 @@ __ai __attribute__((target("neon"))) uint16x8_t vmlal_u8(uint16x8_t __p0, uint8x
69332	71850	#else
69333	71851	__ai __attribute__((target("neon"))) uint16x8_t vmlal_u8(uint16x8_t __p0, uint8x8_t __p1, uint8x8_t __p2) {
69334	71852	uint16x8_t __ret;
69335		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
69336		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
69337		uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	71853	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	71854	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	71855	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
69338	71856	__ret = __rev0 + __noswap_vmull_u8(__rev1, __rev2);
69339		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	71857	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
69340	71858	return __ret;
69341	71859	}
69342	71860	__ai __attribute__((target("neon"))) uint16x8_t __noswap_vmlal_u8(uint16x8_t __p0, uint8x8_t __p1, uint8x8_t __p2) {
...	...	@@ -69355,11 +71873,11 @@ __ai __attribute__((target("neon"))) uint64x2_t vmlal_u32(uint64x2_t __p0, uint3
69355	71873	#else
69356	71874	__ai __attribute__((target("neon"))) uint64x2_t vmlal_u32(uint64x2_t __p0, uint32x2_t __p1, uint32x2_t __p2) {
69357	71875	uint64x2_t __ret;
69358		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
69359		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
69360		uint32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
	71876	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	71877	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	71878	uint32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_32);
69361	71879	__ret = __rev0 + __noswap_vmull_u32(__rev1, __rev2);
69362		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	71880	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
69363	71881	return __ret;
69364	71882	}
69365	71883	__ai __attribute__((target("neon"))) uint64x2_t __noswap_vmlal_u32(uint64x2_t __p0, uint32x2_t __p1, uint32x2_t __p2) {
...	...	@@ -69378,11 +71896,11 @@ __ai __attribute__((target("neon"))) uint32x4_t vmlal_u16(uint32x4_t __p0, uint1
69378	71896	#else
69379	71897	__ai __attribute__((target("neon"))) uint32x4_t vmlal_u16(uint32x4_t __p0, uint16x4_t __p1, uint16x4_t __p2) {
69380	71898	uint32x4_t __ret;
69381		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
69382		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
69383		uint16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	71899	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	71900	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	71901	uint16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
69384	71902	__ret = __rev0 + __noswap_vmull_u16(__rev1, __rev2);
69385		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	71903	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
69386	71904	return __ret;
69387	71905	}
69388	71906	__ai __attribute__((target("neon"))) uint32x4_t __noswap_vmlal_u16(uint32x4_t __p0, uint16x4_t __p1, uint16x4_t __p2) {
...	...	@@ -69401,11 +71919,11 @@ __ai __attribute__((target("neon"))) int16x8_t vmlal_s8(int16x8_t __p0, int8x8_t
69401	71919	#else
69402	71920	__ai __attribute__((target("neon"))) int16x8_t vmlal_s8(int16x8_t __p0, int8x8_t __p1, int8x8_t __p2) {
69403	71921	int16x8_t __ret;
69404		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
69405		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
69406		int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	71922	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	71923	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	71924	int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
69407	71925	__ret = __rev0 + __noswap_vmull_s8(__rev1, __rev2);
69408		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	71926	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
69409	71927	return __ret;
69410	71928	}
69411	71929	__ai __attribute__((target("neon"))) int16x8_t __noswap_vmlal_s8(int16x8_t __p0, int8x8_t __p1, int8x8_t __p2) {
...	...	@@ -69424,11 +71942,11 @@ __ai __attribute__((target("neon"))) int64x2_t vmlal_s32(int64x2_t __p0, int32x2
69424	71942	#else
69425	71943	__ai __attribute__((target("neon"))) int64x2_t vmlal_s32(int64x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
69426	71944	int64x2_t __ret;
69427		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
69428		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
69429		int32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
	71945	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	71946	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	71947	int32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_32);
69430	71948	__ret = __rev0 + __noswap_vmull_s32(__rev1, __rev2);
69431		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	71949	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
69432	71950	return __ret;
69433	71951	}
69434	71952	__ai __attribute__((target("neon"))) int64x2_t __noswap_vmlal_s32(int64x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
...	...	@@ -69447,11 +71965,11 @@ __ai __attribute__((target("neon"))) int32x4_t vmlal_s16(int32x4_t __p0, int16x4
69447	71965	#else
69448	71966	__ai __attribute__((target("neon"))) int32x4_t vmlal_s16(int32x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
69449	71967	int32x4_t __ret;
69450		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
69451		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
69452		int16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	71968	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	71969	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	71970	int16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
69453	71971	__ret = __rev0 + __noswap_vmull_s16(__rev1, __rev2);
69454		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	71972	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
69455	71973	return __ret;
69456	71974	}
69457	71975	__ai __attribute__((target("neon"))) int32x4_t __noswap_vmlal_s16(int32x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
...	...	@@ -69462,98 +71980,98 @@ __ai __attribute__((target("neon"))) int32x4_t __noswap_vmlal_s16(int32x4_t __p0
69462	71980	#endif
69463	71981
69464	71982	#ifdef __LITTLE_ENDIAN__
69465		#define vmlal_lane_u32(__p0_837, __p1_837, __p2_837, __p3_837) __extension__ ({ \
69466		uint64x2_t __ret_837; \
69467		uint64x2_t __s0_837 = __p0_837; \
69468		uint32x2_t __s1_837 = __p1_837; \
69469		uint32x2_t __s2_837 = __p2_837; \
69470		__ret_837 = __s0_837 + vmull_u32(__s1_837, splat_lane_u32(__s2_837, __p3_837)); \
69471		__ret_837; \
	71983	#define vmlal_lane_u32(__p0_853, __p1_853, __p2_853, __p3_853) __extension__ ({ \
	71984	uint64x2_t __ret_853; \
	71985	uint64x2_t __s0_853 = __p0_853; \
	71986	uint32x2_t __s1_853 = __p1_853; \
	71987	uint32x2_t __s2_853 = __p2_853; \
	71988	__ret_853 = __s0_853 + vmull_u32(__s1_853, splat_lane_u32(__s2_853, __p3_853)); \
	71989	__ret_853; \
69472	71990	})
69473	71991	#else
69474		#define vmlal_lane_u32(__p0_838, __p1_838, __p2_838, __p3_838) __extension__ ({ \
69475		uint64x2_t __ret_838; \
69476		uint64x2_t __s0_838 = __p0_838; \
69477		uint32x2_t __s1_838 = __p1_838; \
69478		uint32x2_t __s2_838 = __p2_838; \
69479		uint64x2_t __rev0_838; __rev0_838 = __builtin_shufflevector(__s0_838, __s0_838, 1, 0); \
69480		uint32x2_t __rev1_838; __rev1_838 = __builtin_shufflevector(__s1_838, __s1_838, 1, 0); \
69481		uint32x2_t __rev2_838; __rev2_838 = __builtin_shufflevector(__s2_838, __s2_838, 1, 0); \
69482		__ret_838 = __rev0_838 + __noswap_vmull_u32(__rev1_838, __noswap_splat_lane_u32(__rev2_838, __p3_838)); \
69483		__ret_838 = __builtin_shufflevector(__ret_838, __ret_838, 1, 0); \
69484		__ret_838; \
	71992	#define vmlal_lane_u32(__p0_854, __p1_854, __p2_854, __p3_854) __extension__ ({ \
	71993	uint64x2_t __ret_854; \
	71994	uint64x2_t __s0_854 = __p0_854; \
	71995	uint32x2_t __s1_854 = __p1_854; \
	71996	uint32x2_t __s2_854 = __p2_854; \
	71997	uint64x2_t __rev0_854; __rev0_854 = __builtin_shufflevector(__s0_854, __s0_854, __lane_reverse_128_64); \
	71998	uint32x2_t __rev1_854; __rev1_854 = __builtin_shufflevector(__s1_854, __s1_854, __lane_reverse_64_32); \
	71999	uint32x2_t __rev2_854; __rev2_854 = __builtin_shufflevector(__s2_854, __s2_854, __lane_reverse_64_32); \
	72000	__ret_854 = __rev0_854 + __noswap_vmull_u32(__rev1_854, __noswap_splat_lane_u32(__rev2_854, __p3_854)); \
	72001	__ret_854 = __builtin_shufflevector(__ret_854, __ret_854, __lane_reverse_128_64); \
	72002	__ret_854; \
69485	72003	})
69486	72004	#endif
69487	72005
69488	72006	#ifdef __LITTLE_ENDIAN__
69489		#define vmlal_lane_u16(__p0_839, __p1_839, __p2_839, __p3_839) __extension__ ({ \
69490		uint32x4_t __ret_839; \
69491		uint32x4_t __s0_839 = __p0_839; \
69492		uint16x4_t __s1_839 = __p1_839; \
69493		uint16x4_t __s2_839 = __p2_839; \
69494		__ret_839 = __s0_839 + vmull_u16(__s1_839, splat_lane_u16(__s2_839, __p3_839)); \
69495		__ret_839; \
	72007	#define vmlal_lane_u16(__p0_855, __p1_855, __p2_855, __p3_855) __extension__ ({ \
	72008	uint32x4_t __ret_855; \
	72009	uint32x4_t __s0_855 = __p0_855; \
	72010	uint16x4_t __s1_855 = __p1_855; \
	72011	uint16x4_t __s2_855 = __p2_855; \
	72012	__ret_855 = __s0_855 + vmull_u16(__s1_855, splat_lane_u16(__s2_855, __p3_855)); \
	72013	__ret_855; \
69496	72014	})
69497	72015	#else
69498		#define vmlal_lane_u16(__p0_840, __p1_840, __p2_840, __p3_840) __extension__ ({ \
69499		uint32x4_t __ret_840; \
69500		uint32x4_t __s0_840 = __p0_840; \
69501		uint16x4_t __s1_840 = __p1_840; \
69502		uint16x4_t __s2_840 = __p2_840; \
69503		uint32x4_t __rev0_840; __rev0_840 = __builtin_shufflevector(__s0_840, __s0_840, 3, 2, 1, 0); \
69504		uint16x4_t __rev1_840; __rev1_840 = __builtin_shufflevector(__s1_840, __s1_840, 3, 2, 1, 0); \
69505		uint16x4_t __rev2_840; __rev2_840 = __builtin_shufflevector(__s2_840, __s2_840, 3, 2, 1, 0); \
69506		__ret_840 = __rev0_840 + __noswap_vmull_u16(__rev1_840, __noswap_splat_lane_u16(__rev2_840, __p3_840)); \
69507		__ret_840 = __builtin_shufflevector(__ret_840, __ret_840, 3, 2, 1, 0); \
69508		__ret_840; \
	72016	#define vmlal_lane_u16(__p0_856, __p1_856, __p2_856, __p3_856) __extension__ ({ \
	72017	uint32x4_t __ret_856; \
	72018	uint32x4_t __s0_856 = __p0_856; \
	72019	uint16x4_t __s1_856 = __p1_856; \
	72020	uint16x4_t __s2_856 = __p2_856; \
	72021	uint32x4_t __rev0_856; __rev0_856 = __builtin_shufflevector(__s0_856, __s0_856, __lane_reverse_128_32); \
	72022	uint16x4_t __rev1_856; __rev1_856 = __builtin_shufflevector(__s1_856, __s1_856, __lane_reverse_64_16); \
	72023	uint16x4_t __rev2_856; __rev2_856 = __builtin_shufflevector(__s2_856, __s2_856, __lane_reverse_64_16); \
	72024	__ret_856 = __rev0_856 + __noswap_vmull_u16(__rev1_856, __noswap_splat_lane_u16(__rev2_856, __p3_856)); \
	72025	__ret_856 = __builtin_shufflevector(__ret_856, __ret_856, __lane_reverse_128_32); \
	72026	__ret_856; \
69509	72027	})
69510	72028	#endif
69511	72029
69512	72030	#ifdef __LITTLE_ENDIAN__
69513		#define vmlal_lane_s32(__p0_841, __p1_841, __p2_841, __p3_841) __extension__ ({ \
69514		int64x2_t __ret_841; \
69515		int64x2_t __s0_841 = __p0_841; \
69516		int32x2_t __s1_841 = __p1_841; \
69517		int32x2_t __s2_841 = __p2_841; \
69518		__ret_841 = __s0_841 + vmull_s32(__s1_841, splat_lane_s32(__s2_841, __p3_841)); \
69519		__ret_841; \
	72031	#define vmlal_lane_s32(__p0_857, __p1_857, __p2_857, __p3_857) __extension__ ({ \
	72032	int64x2_t __ret_857; \
	72033	int64x2_t __s0_857 = __p0_857; \
	72034	int32x2_t __s1_857 = __p1_857; \
	72035	int32x2_t __s2_857 = __p2_857; \
	72036	__ret_857 = __s0_857 + vmull_s32(__s1_857, splat_lane_s32(__s2_857, __p3_857)); \
	72037	__ret_857; \
69520	72038	})
69521	72039	#else
69522		#define vmlal_lane_s32(__p0_842, __p1_842, __p2_842, __p3_842) __extension__ ({ \
69523		int64x2_t __ret_842; \
69524		int64x2_t __s0_842 = __p0_842; \
69525		int32x2_t __s1_842 = __p1_842; \
69526		int32x2_t __s2_842 = __p2_842; \
69527		int64x2_t __rev0_842; __rev0_842 = __builtin_shufflevector(__s0_842, __s0_842, 1, 0); \
69528		int32x2_t __rev1_842; __rev1_842 = __builtin_shufflevector(__s1_842, __s1_842, 1, 0); \
69529		int32x2_t __rev2_842; __rev2_842 = __builtin_shufflevector(__s2_842, __s2_842, 1, 0); \
69530		__ret_842 = __rev0_842 + __noswap_vmull_s32(__rev1_842, __noswap_splat_lane_s32(__rev2_842, __p3_842)); \
69531		__ret_842 = __builtin_shufflevector(__ret_842, __ret_842, 1, 0); \
69532		__ret_842; \
	72040	#define vmlal_lane_s32(__p0_858, __p1_858, __p2_858, __p3_858) __extension__ ({ \
	72041	int64x2_t __ret_858; \
	72042	int64x2_t __s0_858 = __p0_858; \
	72043	int32x2_t __s1_858 = __p1_858; \
	72044	int32x2_t __s2_858 = __p2_858; \
	72045	int64x2_t __rev0_858; __rev0_858 = __builtin_shufflevector(__s0_858, __s0_858, __lane_reverse_128_64); \
	72046	int32x2_t __rev1_858; __rev1_858 = __builtin_shufflevector(__s1_858, __s1_858, __lane_reverse_64_32); \
	72047	int32x2_t __rev2_858; __rev2_858 = __builtin_shufflevector(__s2_858, __s2_858, __lane_reverse_64_32); \
	72048	__ret_858 = __rev0_858 + __noswap_vmull_s32(__rev1_858, __noswap_splat_lane_s32(__rev2_858, __p3_858)); \
	72049	__ret_858 = __builtin_shufflevector(__ret_858, __ret_858, __lane_reverse_128_64); \
	72050	__ret_858; \
69533	72051	})
69534	72052	#endif
69535	72053
69536	72054	#ifdef __LITTLE_ENDIAN__
69537		#define vmlal_lane_s16(__p0_843, __p1_843, __p2_843, __p3_843) __extension__ ({ \
69538		int32x4_t __ret_843; \
69539		int32x4_t __s0_843 = __p0_843; \
69540		int16x4_t __s1_843 = __p1_843; \
69541		int16x4_t __s2_843 = __p2_843; \
69542		__ret_843 = __s0_843 + vmull_s16(__s1_843, splat_lane_s16(__s2_843, __p3_843)); \
69543		__ret_843; \
	72055	#define vmlal_lane_s16(__p0_859, __p1_859, __p2_859, __p3_859) __extension__ ({ \
	72056	int32x4_t __ret_859; \
	72057	int32x4_t __s0_859 = __p0_859; \
	72058	int16x4_t __s1_859 = __p1_859; \
	72059	int16x4_t __s2_859 = __p2_859; \
	72060	__ret_859 = __s0_859 + vmull_s16(__s1_859, splat_lane_s16(__s2_859, __p3_859)); \
	72061	__ret_859; \
69544	72062	})
69545	72063	#else
69546		#define vmlal_lane_s16(__p0_844, __p1_844, __p2_844, __p3_844) __extension__ ({ \
69547		int32x4_t __ret_844; \
69548		int32x4_t __s0_844 = __p0_844; \
69549		int16x4_t __s1_844 = __p1_844; \
69550		int16x4_t __s2_844 = __p2_844; \
69551		int32x4_t __rev0_844; __rev0_844 = __builtin_shufflevector(__s0_844, __s0_844, 3, 2, 1, 0); \
69552		int16x4_t __rev1_844; __rev1_844 = __builtin_shufflevector(__s1_844, __s1_844, 3, 2, 1, 0); \
69553		int16x4_t __rev2_844; __rev2_844 = __builtin_shufflevector(__s2_844, __s2_844, 3, 2, 1, 0); \
69554		__ret_844 = __rev0_844 + __noswap_vmull_s16(__rev1_844, __noswap_splat_lane_s16(__rev2_844, __p3_844)); \
69555		__ret_844 = __builtin_shufflevector(__ret_844, __ret_844, 3, 2, 1, 0); \
69556		__ret_844; \
	72064	#define vmlal_lane_s16(__p0_860, __p1_860, __p2_860, __p3_860) __extension__ ({ \
	72065	int32x4_t __ret_860; \
	72066	int32x4_t __s0_860 = __p0_860; \
	72067	int16x4_t __s1_860 = __p1_860; \
	72068	int16x4_t __s2_860 = __p2_860; \
	72069	int32x4_t __rev0_860; __rev0_860 = __builtin_shufflevector(__s0_860, __s0_860, __lane_reverse_128_32); \
	72070	int16x4_t __rev1_860; __rev1_860 = __builtin_shufflevector(__s1_860, __s1_860, __lane_reverse_64_16); \
	72071	int16x4_t __rev2_860; __rev2_860 = __builtin_shufflevector(__s2_860, __s2_860, __lane_reverse_64_16); \
	72072	__ret_860 = __rev0_860 + __noswap_vmull_s16(__rev1_860, __noswap_splat_lane_s16(__rev2_860, __p3_860)); \
	72073	__ret_860 = __builtin_shufflevector(__ret_860, __ret_860, __lane_reverse_128_32); \
	72074	__ret_860; \
69557	72075	})
69558	72076	#endif
69559	72077
...	...	@@ -69566,10 +72084,10 @@ __ai __attribute__((target("neon"))) uint64x2_t vmlal_n_u32(uint64x2_t __p0, uin
69566	72084	#else
69567	72085	__ai __attribute__((target("neon"))) uint64x2_t vmlal_n_u32(uint64x2_t __p0, uint32x2_t __p1, uint32_t __p2) {
69568	72086	uint64x2_t __ret;
69569		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
69570		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	72087	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	72088	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
69571	72089	__ret = __rev0 + __noswap_vmull_u32(__rev1, (uint32x2_t) {__p2, __p2});
69572		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	72090	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
69573	72091	return __ret;
69574	72092	}
69575	72093	__ai __attribute__((target("neon"))) uint64x2_t __noswap_vmlal_n_u32(uint64x2_t __p0, uint32x2_t __p1, uint32_t __p2) {
...	...	@@ -69588,10 +72106,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vmlal_n_u16(uint32x4_t __p0, uin
69588	72106	#else
69589	72107	__ai __attribute__((target("neon"))) uint32x4_t vmlal_n_u16(uint32x4_t __p0, uint16x4_t __p1, uint16_t __p2) {
69590	72108	uint32x4_t __ret;
69591		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
69592		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	72109	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	72110	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
69593	72111	__ret = __rev0 + __noswap_vmull_u16(__rev1, (uint16x4_t) {__p2, __p2, __p2, __p2});
69594		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	72112	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
69595	72113	return __ret;
69596	72114	}
69597	72115	__ai __attribute__((target("neon"))) uint32x4_t __noswap_vmlal_n_u16(uint32x4_t __p0, uint16x4_t __p1, uint16_t __p2) {
...	...	@@ -69610,10 +72128,10 @@ __ai __attribute__((target("neon"))) int64x2_t vmlal_n_s32(int64x2_t __p0, int32
69610	72128	#else
69611	72129	__ai __attribute__((target("neon"))) int64x2_t vmlal_n_s32(int64x2_t __p0, int32x2_t __p1, int32_t __p2) {
69612	72130	int64x2_t __ret;
69613		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
69614		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	72131	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	72132	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
69615	72133	__ret = __rev0 + __noswap_vmull_s32(__rev1, (int32x2_t) {__p2, __p2});
69616		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	72134	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
69617	72135	return __ret;
69618	72136	}
69619	72137	__ai __attribute__((target("neon"))) int64x2_t __noswap_vmlal_n_s32(int64x2_t __p0, int32x2_t __p1, int32_t __p2) {
...	...	@@ -69632,10 +72150,10 @@ __ai __attribute__((target("neon"))) int32x4_t vmlal_n_s16(int32x4_t __p0, int16
69632	72150	#else
69633	72151	__ai __attribute__((target("neon"))) int32x4_t vmlal_n_s16(int32x4_t __p0, int16x4_t __p1, int16_t __p2) {
69634	72152	int32x4_t __ret;
69635		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
69636		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	72153	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	72154	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
69637	72155	__ret = __rev0 + __noswap_vmull_s16(__rev1, (int16x4_t) {__p2, __p2, __p2, __p2});
69638		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	72156	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
69639	72157	return __ret;
69640	72158	}
69641	72159	__ai __attribute__((target("neon"))) int32x4_t __noswap_vmlal_n_s16(int32x4_t __p0, int16x4_t __p1, int16_t __p2) {
...	...	@@ -69654,11 +72172,11 @@ __ai __attribute__((target("neon"))) uint16x8_t vmlsl_u8(uint16x8_t __p0, uint8x
69654	72172	#else
69655	72173	__ai __attribute__((target("neon"))) uint16x8_t vmlsl_u8(uint16x8_t __p0, uint8x8_t __p1, uint8x8_t __p2) {
69656	72174	uint16x8_t __ret;
69657		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
69658		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
69659		uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	72175	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	72176	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	72177	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
69660	72178	__ret = __rev0 - __noswap_vmull_u8(__rev1, __rev2);
69661		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	72179	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
69662	72180	return __ret;
69663	72181	}
69664	72182	__ai __attribute__((target("neon"))) uint16x8_t __noswap_vmlsl_u8(uint16x8_t __p0, uint8x8_t __p1, uint8x8_t __p2) {
...	...	@@ -69677,11 +72195,11 @@ __ai __attribute__((target("neon"))) uint64x2_t vmlsl_u32(uint64x2_t __p0, uint3
69677	72195	#else
69678	72196	__ai __attribute__((target("neon"))) uint64x2_t vmlsl_u32(uint64x2_t __p0, uint32x2_t __p1, uint32x2_t __p2) {
69679	72197	uint64x2_t __ret;
69680		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
69681		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
69682		uint32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
	72198	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	72199	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	72200	uint32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_32);
69683	72201	__ret = __rev0 - __noswap_vmull_u32(__rev1, __rev2);
69684		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	72202	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
69685	72203	return __ret;
69686	72204	}
69687	72205	__ai __attribute__((target("neon"))) uint64x2_t __noswap_vmlsl_u32(uint64x2_t __p0, uint32x2_t __p1, uint32x2_t __p2) {
...	...	@@ -69700,11 +72218,11 @@ __ai __attribute__((target("neon"))) uint32x4_t vmlsl_u16(uint32x4_t __p0, uint1
69700	72218	#else
69701	72219	__ai __attribute__((target("neon"))) uint32x4_t vmlsl_u16(uint32x4_t __p0, uint16x4_t __p1, uint16x4_t __p2) {
69702	72220	uint32x4_t __ret;
69703		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
69704		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
69705		uint16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	72221	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	72222	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	72223	uint16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
69706	72224	__ret = __rev0 - __noswap_vmull_u16(__rev1, __rev2);
69707		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	72225	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
69708	72226	return __ret;
69709	72227	}
69710	72228	__ai __attribute__((target("neon"))) uint32x4_t __noswap_vmlsl_u16(uint32x4_t __p0, uint16x4_t __p1, uint16x4_t __p2) {
...	...	@@ -69723,11 +72241,11 @@ __ai __attribute__((target("neon"))) int16x8_t vmlsl_s8(int16x8_t __p0, int8x8_t
69723	72241	#else
69724	72242	__ai __attribute__((target("neon"))) int16x8_t vmlsl_s8(int16x8_t __p0, int8x8_t __p1, int8x8_t __p2) {
69725	72243	int16x8_t __ret;
69726		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
69727		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
69728		int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	72244	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	72245	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	72246	int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
69729	72247	__ret = __rev0 - __noswap_vmull_s8(__rev1, __rev2);
69730		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	72248	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
69731	72249	return __ret;
69732	72250	}
69733	72251	__ai __attribute__((target("neon"))) int16x8_t __noswap_vmlsl_s8(int16x8_t __p0, int8x8_t __p1, int8x8_t __p2) {
...	...	@@ -69746,11 +72264,11 @@ __ai __attribute__((target("neon"))) int64x2_t vmlsl_s32(int64x2_t __p0, int32x2
69746	72264	#else
69747	72265	__ai __attribute__((target("neon"))) int64x2_t vmlsl_s32(int64x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
69748	72266	int64x2_t __ret;
69749		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
69750		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
69751		int32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
	72267	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	72268	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	72269	int32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_32);
69752	72270	__ret = __rev0 - __noswap_vmull_s32(__rev1, __rev2);
69753		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	72271	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
69754	72272	return __ret;
69755	72273	}
69756	72274	__ai __attribute__((target("neon"))) int64x2_t __noswap_vmlsl_s32(int64x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
...	...	@@ -69769,11 +72287,11 @@ __ai __attribute__((target("neon"))) int32x4_t vmlsl_s16(int32x4_t __p0, int16x4
69769	72287	#else
69770	72288	__ai __attribute__((target("neon"))) int32x4_t vmlsl_s16(int32x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
69771	72289	int32x4_t __ret;
69772		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
69773		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
69774		int16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	72290	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	72291	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	72292	int16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
69775	72293	__ret = __rev0 - __noswap_vmull_s16(__rev1, __rev2);
69776		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	72294	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
69777	72295	return __ret;
69778	72296	}
69779	72297	__ai __attribute__((target("neon"))) int32x4_t __noswap_vmlsl_s16(int32x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
...	...	@@ -69784,98 +72302,98 @@ __ai __attribute__((target("neon"))) int32x4_t __noswap_vmlsl_s16(int32x4_t __p0
69784	72302	#endif
69785	72303
69786	72304	#ifdef __LITTLE_ENDIAN__
69787		#define vmlsl_lane_u32(__p0_845, __p1_845, __p2_845, __p3_845) __extension__ ({ \
69788		uint64x2_t __ret_845; \
69789		uint64x2_t __s0_845 = __p0_845; \
69790		uint32x2_t __s1_845 = __p1_845; \
69791		uint32x2_t __s2_845 = __p2_845; \
69792		__ret_845 = __s0_845 - vmull_u32(__s1_845, splat_lane_u32(__s2_845, __p3_845)); \
69793		__ret_845; \
	72305	#define vmlsl_lane_u32(__p0_861, __p1_861, __p2_861, __p3_861) __extension__ ({ \
	72306	uint64x2_t __ret_861; \
	72307	uint64x2_t __s0_861 = __p0_861; \
	72308	uint32x2_t __s1_861 = __p1_861; \
	72309	uint32x2_t __s2_861 = __p2_861; \
	72310	__ret_861 = __s0_861 - vmull_u32(__s1_861, splat_lane_u32(__s2_861, __p3_861)); \
	72311	__ret_861; \
69794	72312	})
69795	72313	#else
69796		#define vmlsl_lane_u32(__p0_846, __p1_846, __p2_846, __p3_846) __extension__ ({ \
69797		uint64x2_t __ret_846; \
69798		uint64x2_t __s0_846 = __p0_846; \
69799		uint32x2_t __s1_846 = __p1_846; \
69800		uint32x2_t __s2_846 = __p2_846; \
69801		uint64x2_t __rev0_846; __rev0_846 = __builtin_shufflevector(__s0_846, __s0_846, 1, 0); \
69802		uint32x2_t __rev1_846; __rev1_846 = __builtin_shufflevector(__s1_846, __s1_846, 1, 0); \
69803		uint32x2_t __rev2_846; __rev2_846 = __builtin_shufflevector(__s2_846, __s2_846, 1, 0); \
69804		__ret_846 = __rev0_846 - __noswap_vmull_u32(__rev1_846, __noswap_splat_lane_u32(__rev2_846, __p3_846)); \
69805		__ret_846 = __builtin_shufflevector(__ret_846, __ret_846, 1, 0); \
69806		__ret_846; \
	72314	#define vmlsl_lane_u32(__p0_862, __p1_862, __p2_862, __p3_862) __extension__ ({ \
	72315	uint64x2_t __ret_862; \
	72316	uint64x2_t __s0_862 = __p0_862; \
	72317	uint32x2_t __s1_862 = __p1_862; \
	72318	uint32x2_t __s2_862 = __p2_862; \
	72319	uint64x2_t __rev0_862; __rev0_862 = __builtin_shufflevector(__s0_862, __s0_862, __lane_reverse_128_64); \
	72320	uint32x2_t __rev1_862; __rev1_862 = __builtin_shufflevector(__s1_862, __s1_862, __lane_reverse_64_32); \
	72321	uint32x2_t __rev2_862; __rev2_862 = __builtin_shufflevector(__s2_862, __s2_862, __lane_reverse_64_32); \
	72322	__ret_862 = __rev0_862 - __noswap_vmull_u32(__rev1_862, __noswap_splat_lane_u32(__rev2_862, __p3_862)); \
	72323	__ret_862 = __builtin_shufflevector(__ret_862, __ret_862, __lane_reverse_128_64); \
	72324	__ret_862; \
69807	72325	})
69808	72326	#endif
69809	72327
69810	72328	#ifdef __LITTLE_ENDIAN__
69811		#define vmlsl_lane_u16(__p0_847, __p1_847, __p2_847, __p3_847) __extension__ ({ \
69812		uint32x4_t __ret_847; \
69813		uint32x4_t __s0_847 = __p0_847; \
69814		uint16x4_t __s1_847 = __p1_847; \
69815		uint16x4_t __s2_847 = __p2_847; \
69816		__ret_847 = __s0_847 - vmull_u16(__s1_847, splat_lane_u16(__s2_847, __p3_847)); \
69817		__ret_847; \
	72329	#define vmlsl_lane_u16(__p0_863, __p1_863, __p2_863, __p3_863) __extension__ ({ \
	72330	uint32x4_t __ret_863; \
	72331	uint32x4_t __s0_863 = __p0_863; \
	72332	uint16x4_t __s1_863 = __p1_863; \
	72333	uint16x4_t __s2_863 = __p2_863; \
	72334	__ret_863 = __s0_863 - vmull_u16(__s1_863, splat_lane_u16(__s2_863, __p3_863)); \
	72335	__ret_863; \
69818	72336	})
69819	72337	#else
69820		#define vmlsl_lane_u16(__p0_848, __p1_848, __p2_848, __p3_848) __extension__ ({ \
69821		uint32x4_t __ret_848; \
69822		uint32x4_t __s0_848 = __p0_848; \
69823		uint16x4_t __s1_848 = __p1_848; \
69824		uint16x4_t __s2_848 = __p2_848; \
69825		uint32x4_t __rev0_848; __rev0_848 = __builtin_shufflevector(__s0_848, __s0_848, 3, 2, 1, 0); \
69826		uint16x4_t __rev1_848; __rev1_848 = __builtin_shufflevector(__s1_848, __s1_848, 3, 2, 1, 0); \
69827		uint16x4_t __rev2_848; __rev2_848 = __builtin_shufflevector(__s2_848, __s2_848, 3, 2, 1, 0); \
69828		__ret_848 = __rev0_848 - __noswap_vmull_u16(__rev1_848, __noswap_splat_lane_u16(__rev2_848, __p3_848)); \
69829		__ret_848 = __builtin_shufflevector(__ret_848, __ret_848, 3, 2, 1, 0); \
69830		__ret_848; \
	72338	#define vmlsl_lane_u16(__p0_864, __p1_864, __p2_864, __p3_864) __extension__ ({ \
	72339	uint32x4_t __ret_864; \
	72340	uint32x4_t __s0_864 = __p0_864; \
	72341	uint16x4_t __s1_864 = __p1_864; \
	72342	uint16x4_t __s2_864 = __p2_864; \
	72343	uint32x4_t __rev0_864; __rev0_864 = __builtin_shufflevector(__s0_864, __s0_864, __lane_reverse_128_32); \
	72344	uint16x4_t __rev1_864; __rev1_864 = __builtin_shufflevector(__s1_864, __s1_864, __lane_reverse_64_16); \
	72345	uint16x4_t __rev2_864; __rev2_864 = __builtin_shufflevector(__s2_864, __s2_864, __lane_reverse_64_16); \
	72346	__ret_864 = __rev0_864 - __noswap_vmull_u16(__rev1_864, __noswap_splat_lane_u16(__rev2_864, __p3_864)); \
	72347	__ret_864 = __builtin_shufflevector(__ret_864, __ret_864, __lane_reverse_128_32); \
	72348	__ret_864; \
69831	72349	})
69832	72350	#endif
69833	72351
69834	72352	#ifdef __LITTLE_ENDIAN__
69835		#define vmlsl_lane_s32(__p0_849, __p1_849, __p2_849, __p3_849) __extension__ ({ \
69836		int64x2_t __ret_849; \
69837		int64x2_t __s0_849 = __p0_849; \
69838		int32x2_t __s1_849 = __p1_849; \
69839		int32x2_t __s2_849 = __p2_849; \
69840		__ret_849 = __s0_849 - vmull_s32(__s1_849, splat_lane_s32(__s2_849, __p3_849)); \
69841		__ret_849; \
	72353	#define vmlsl_lane_s32(__p0_865, __p1_865, __p2_865, __p3_865) __extension__ ({ \
	72354	int64x2_t __ret_865; \
	72355	int64x2_t __s0_865 = __p0_865; \
	72356	int32x2_t __s1_865 = __p1_865; \
	72357	int32x2_t __s2_865 = __p2_865; \
	72358	__ret_865 = __s0_865 - vmull_s32(__s1_865, splat_lane_s32(__s2_865, __p3_865)); \
	72359	__ret_865; \
69842	72360	})
69843	72361	#else
69844		#define vmlsl_lane_s32(__p0_850, __p1_850, __p2_850, __p3_850) __extension__ ({ \
69845		int64x2_t __ret_850; \
69846		int64x2_t __s0_850 = __p0_850; \
69847		int32x2_t __s1_850 = __p1_850; \
69848		int32x2_t __s2_850 = __p2_850; \
69849		int64x2_t __rev0_850; __rev0_850 = __builtin_shufflevector(__s0_850, __s0_850, 1, 0); \
69850		int32x2_t __rev1_850; __rev1_850 = __builtin_shufflevector(__s1_850, __s1_850, 1, 0); \
69851		int32x2_t __rev2_850; __rev2_850 = __builtin_shufflevector(__s2_850, __s2_850, 1, 0); \
69852		__ret_850 = __rev0_850 - __noswap_vmull_s32(__rev1_850, __noswap_splat_lane_s32(__rev2_850, __p3_850)); \
69853		__ret_850 = __builtin_shufflevector(__ret_850, __ret_850, 1, 0); \
69854		__ret_850; \
	72362	#define vmlsl_lane_s32(__p0_866, __p1_866, __p2_866, __p3_866) __extension__ ({ \
	72363	int64x2_t __ret_866; \
	72364	int64x2_t __s0_866 = __p0_866; \
	72365	int32x2_t __s1_866 = __p1_866; \
	72366	int32x2_t __s2_866 = __p2_866; \
	72367	int64x2_t __rev0_866; __rev0_866 = __builtin_shufflevector(__s0_866, __s0_866, __lane_reverse_128_64); \
	72368	int32x2_t __rev1_866; __rev1_866 = __builtin_shufflevector(__s1_866, __s1_866, __lane_reverse_64_32); \
	72369	int32x2_t __rev2_866; __rev2_866 = __builtin_shufflevector(__s2_866, __s2_866, __lane_reverse_64_32); \
	72370	__ret_866 = __rev0_866 - __noswap_vmull_s32(__rev1_866, __noswap_splat_lane_s32(__rev2_866, __p3_866)); \
	72371	__ret_866 = __builtin_shufflevector(__ret_866, __ret_866, __lane_reverse_128_64); \
	72372	__ret_866; \
69855	72373	})
69856	72374	#endif
69857	72375
69858	72376	#ifdef __LITTLE_ENDIAN__
69859		#define vmlsl_lane_s16(__p0_851, __p1_851, __p2_851, __p3_851) __extension__ ({ \
69860		int32x4_t __ret_851; \
69861		int32x4_t __s0_851 = __p0_851; \
69862		int16x4_t __s1_851 = __p1_851; \
69863		int16x4_t __s2_851 = __p2_851; \
69864		__ret_851 = __s0_851 - vmull_s16(__s1_851, splat_lane_s16(__s2_851, __p3_851)); \
69865		__ret_851; \
	72377	#define vmlsl_lane_s16(__p0_867, __p1_867, __p2_867, __p3_867) __extension__ ({ \
	72378	int32x4_t __ret_867; \
	72379	int32x4_t __s0_867 = __p0_867; \
	72380	int16x4_t __s1_867 = __p1_867; \
	72381	int16x4_t __s2_867 = __p2_867; \
	72382	__ret_867 = __s0_867 - vmull_s16(__s1_867, splat_lane_s16(__s2_867, __p3_867)); \
	72383	__ret_867; \
69866	72384	})
69867	72385	#else
69868		#define vmlsl_lane_s16(__p0_852, __p1_852, __p2_852, __p3_852) __extension__ ({ \
69869		int32x4_t __ret_852; \
69870		int32x4_t __s0_852 = __p0_852; \
69871		int16x4_t __s1_852 = __p1_852; \
69872		int16x4_t __s2_852 = __p2_852; \
69873		int32x4_t __rev0_852; __rev0_852 = __builtin_shufflevector(__s0_852, __s0_852, 3, 2, 1, 0); \
69874		int16x4_t __rev1_852; __rev1_852 = __builtin_shufflevector(__s1_852, __s1_852, 3, 2, 1, 0); \
69875		int16x4_t __rev2_852; __rev2_852 = __builtin_shufflevector(__s2_852, __s2_852, 3, 2, 1, 0); \
69876		__ret_852 = __rev0_852 - __noswap_vmull_s16(__rev1_852, __noswap_splat_lane_s16(__rev2_852, __p3_852)); \
69877		__ret_852 = __builtin_shufflevector(__ret_852, __ret_852, 3, 2, 1, 0); \
69878		__ret_852; \
	72386	#define vmlsl_lane_s16(__p0_868, __p1_868, __p2_868, __p3_868) __extension__ ({ \
	72387	int32x4_t __ret_868; \
	72388	int32x4_t __s0_868 = __p0_868; \
	72389	int16x4_t __s1_868 = __p1_868; \
	72390	int16x4_t __s2_868 = __p2_868; \
	72391	int32x4_t __rev0_868; __rev0_868 = __builtin_shufflevector(__s0_868, __s0_868, __lane_reverse_128_32); \
	72392	int16x4_t __rev1_868; __rev1_868 = __builtin_shufflevector(__s1_868, __s1_868, __lane_reverse_64_16); \
	72393	int16x4_t __rev2_868; __rev2_868 = __builtin_shufflevector(__s2_868, __s2_868, __lane_reverse_64_16); \
	72394	__ret_868 = __rev0_868 - __noswap_vmull_s16(__rev1_868, __noswap_splat_lane_s16(__rev2_868, __p3_868)); \
	72395	__ret_868 = __builtin_shufflevector(__ret_868, __ret_868, __lane_reverse_128_32); \
	72396	__ret_868; \
69879	72397	})
69880	72398	#endif
69881	72399
...	...	@@ -69888,10 +72406,10 @@ __ai __attribute__((target("neon"))) uint64x2_t vmlsl_n_u32(uint64x2_t __p0, uin
69888	72406	#else
69889	72407	__ai __attribute__((target("neon"))) uint64x2_t vmlsl_n_u32(uint64x2_t __p0, uint32x2_t __p1, uint32_t __p2) {
69890	72408	uint64x2_t __ret;
69891		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
69892		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	72409	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	72410	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
69893	72411	__ret = __rev0 - __noswap_vmull_u32(__rev1, (uint32x2_t) {__p2, __p2});
69894		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	72412	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
69895	72413	return __ret;
69896	72414	}
69897	72415	__ai __attribute__((target("neon"))) uint64x2_t __noswap_vmlsl_n_u32(uint64x2_t __p0, uint32x2_t __p1, uint32_t __p2) {
...	...	@@ -69910,10 +72428,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vmlsl_n_u16(uint32x4_t __p0, uin
69910	72428	#else
69911	72429	__ai __attribute__((target("neon"))) uint32x4_t vmlsl_n_u16(uint32x4_t __p0, uint16x4_t __p1, uint16_t __p2) {
69912	72430	uint32x4_t __ret;
69913		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
69914		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	72431	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	72432	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
69915	72433	__ret = __rev0 - __noswap_vmull_u16(__rev1, (uint16x4_t) {__p2, __p2, __p2, __p2});
69916		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	72434	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
69917	72435	return __ret;
69918	72436	}
69919	72437	__ai __attribute__((target("neon"))) uint32x4_t __noswap_vmlsl_n_u16(uint32x4_t __p0, uint16x4_t __p1, uint16_t __p2) {
...	...	@@ -69932,10 +72450,10 @@ __ai __attribute__((target("neon"))) int64x2_t vmlsl_n_s32(int64x2_t __p0, int32
69932	72450	#else
69933	72451	__ai __attribute__((target("neon"))) int64x2_t vmlsl_n_s32(int64x2_t __p0, int32x2_t __p1, int32_t __p2) {
69934	72452	int64x2_t __ret;
69935		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
69936		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
	72453	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	72454	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
69937	72455	__ret = __rev0 - __noswap_vmull_s32(__rev1, (int32x2_t) {__p2, __p2});
69938		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	72456	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
69939	72457	return __ret;
69940	72458	}
69941	72459	__ai __attribute__((target("neon"))) int64x2_t __noswap_vmlsl_n_s32(int64x2_t __p0, int32x2_t __p1, int32_t __p2) {
...	...	@@ -69954,10 +72472,10 @@ __ai __attribute__((target("neon"))) int32x4_t vmlsl_n_s16(int32x4_t __p0, int16
69954	72472	#else
69955	72473	__ai __attribute__((target("neon"))) int32x4_t vmlsl_n_s16(int32x4_t __p0, int16x4_t __p1, int16_t __p2) {
69956	72474	int32x4_t __ret;
69957		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
69958		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	72475	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	72476	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
69959	72477	__ret = __rev0 - __noswap_vmull_s16(__rev1, (int16x4_t) {__p2, __p2, __p2, __p2});
69960		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	72478	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
69961	72479	return __ret;
69962	72480	}
69963	72481	__ai __attribute__((target("neon"))) int32x4_t __noswap_vmlsl_n_s16(int32x4_t __p0, int16x4_t __p1, int16_t __p2) {
...	...	@@ -69968,42 +72486,42 @@ __ai __attribute__((target("neon"))) int32x4_t __noswap_vmlsl_n_s16(int32x4_t __
69968	72486	#endif
69969	72487
69970	72488	#ifdef __LITTLE_ENDIAN__
69971		#define vset_lane_f16(__p0_853, __p1_853, __p2_853) __extension__ ({ \
69972		float16x4_t __ret_853; \
69973		float16_t __s0_853 = __p0_853; \
69974		float16x4_t __s1_853 = __p1_853; \
69975		__ret_853 = __builtin_bit_cast(float16x4_t, vset_lane_s16(__builtin_bit_cast(int16_t, __s0_853), __builtin_bit_cast(int16x4_t, __s1_853), __p2_853)); \
69976		__ret_853; \
	72489	#define vset_lane_f16(__p0_869, __p1_869, __p2_869) __extension__ ({ \
	72490	float16x4_t __ret_869; \
	72491	float16_t __s0_869 = __p0_869; \
	72492	float16x4_t __s1_869 = __p1_869; \
	72493	__ret_869 = __builtin_bit_cast(float16x4_t, vset_lane_s16(__builtin_bit_cast(int16_t, __s0_869), __builtin_bit_cast(int16x4_t, __s1_869), __p2_869)); \
	72494	__ret_869; \
69977	72495	})
69978	72496	#else
69979		#define vset_lane_f16(__p0_854, __p1_854, __p2_854) __extension__ ({ \
69980		float16x4_t __ret_854; \
69981		float16_t __s0_854 = __p0_854; \
69982		float16x4_t __s1_854 = __p1_854; \
69983		float16x4_t __rev1_854; __rev1_854 = __builtin_shufflevector(__s1_854, __s1_854, 3, 2, 1, 0); \
69984		__ret_854 = __builtin_bit_cast(float16x4_t, __noswap_vset_lane_s16(__builtin_bit_cast(int16_t, __s0_854), __builtin_bit_cast(int16x4_t, __rev1_854), __p2_854)); \
69985		__ret_854 = __builtin_shufflevector(__ret_854, __ret_854, 3, 2, 1, 0); \
69986		__ret_854; \
	72497	#define vset_lane_f16(__p0_870, __p1_870, __p2_870) __extension__ ({ \
	72498	float16x4_t __ret_870; \
	72499	float16_t __s0_870 = __p0_870; \
	72500	float16x4_t __s1_870 = __p1_870; \
	72501	float16x4_t __rev1_870; __rev1_870 = __builtin_shufflevector(__s1_870, __s1_870, __lane_reverse_64_16); \
	72502	__ret_870 = __builtin_bit_cast(float16x4_t, __noswap_vset_lane_s16(__builtin_bit_cast(int16_t, __s0_870), __builtin_bit_cast(int16x4_t, __rev1_870), __p2_870)); \
	72503	__ret_870 = __builtin_shufflevector(__ret_870, __ret_870, __lane_reverse_64_16); \
	72504	__ret_870; \
69987	72505	})
69988	72506	#endif
69989	72507
69990	72508	#ifdef __LITTLE_ENDIAN__
69991		#define vsetq_lane_f16(__p0_855, __p1_855, __p2_855) __extension__ ({ \
69992		float16x8_t __ret_855; \
69993		float16_t __s0_855 = __p0_855; \
69994		float16x8_t __s1_855 = __p1_855; \
69995		__ret_855 = __builtin_bit_cast(float16x8_t, vsetq_lane_s16(__builtin_bit_cast(int16_t, __s0_855), __builtin_bit_cast(int16x8_t, __s1_855), __p2_855)); \
69996		__ret_855; \
	72509	#define vsetq_lane_f16(__p0_871, __p1_871, __p2_871) __extension__ ({ \
	72510	float16x8_t __ret_871; \
	72511	float16_t __s0_871 = __p0_871; \
	72512	float16x8_t __s1_871 = __p1_871; \
	72513	__ret_871 = __builtin_bit_cast(float16x8_t, vsetq_lane_s16(__builtin_bit_cast(int16_t, __s0_871), __builtin_bit_cast(int16x8_t, __s1_871), __p2_871)); \
	72514	__ret_871; \
69997	72515	})
69998	72516	#else
69999		#define vsetq_lane_f16(__p0_856, __p1_856, __p2_856) __extension__ ({ \
70000		float16x8_t __ret_856; \
70001		float16_t __s0_856 = __p0_856; \
70002		float16x8_t __s1_856 = __p1_856; \
70003		float16x8_t __rev1_856; __rev1_856 = __builtin_shufflevector(__s1_856, __s1_856, 7, 6, 5, 4, 3, 2, 1, 0); \
70004		__ret_856 = __builtin_bit_cast(float16x8_t, __noswap_vsetq_lane_s16(__builtin_bit_cast(int16_t, __s0_856), __builtin_bit_cast(int16x8_t, __rev1_856), __p2_856)); \
70005		__ret_856 = __builtin_shufflevector(__ret_856, __ret_856, 7, 6, 5, 4, 3, 2, 1, 0); \
70006		__ret_856; \
	72517	#define vsetq_lane_f16(__p0_872, __p1_872, __p2_872) __extension__ ({ \
	72518	float16x8_t __ret_872; \
	72519	float16_t __s0_872 = __p0_872; \
	72520	float16x8_t __s1_872 = __p1_872; \
	72521	float16x8_t __rev1_872; __rev1_872 = __builtin_shufflevector(__s1_872, __s1_872, __lane_reverse_128_16); \
	72522	__ret_872 = __builtin_bit_cast(float16x8_t, __noswap_vsetq_lane_s16(__builtin_bit_cast(int16_t, __s0_872), __builtin_bit_cast(int16x8_t, __rev1_872), __p2_872)); \
	72523	__ret_872 = __builtin_shufflevector(__ret_872, __ret_872, __lane_reverse_128_16); \
	72524	__ret_872; \
70007	72525	})
70008	72526	#endif
70009	72527
...	...	@@ -70011,441 +72529,441 @@ __ai __attribute__((target("neon"))) int32x4_t __noswap_vmlsl_n_s16(int32x4_t __
70011	72529	#ifdef __LITTLE_ENDIAN__
70012	72530	__ai __attribute__((target("aes,neon"))) poly128_t vmull_high_p64(poly64x2_t __p0, poly64x2_t __p1) {
70013	72531	poly128_t __ret;
70014		__ret = vmull_p64((poly64_t)(vget_high_p64(__p0)), (poly64_t)(vget_high_p64(__p1)));
	72532	__ret = vmull_p64(__builtin_bit_cast(poly64_t, vget_high_p64(__p0)), __builtin_bit_cast(poly64_t, vget_high_p64(__p1)));
70015	72533	return __ret;
70016	72534	}
70017	72535	#else
70018	72536	__ai __attribute__((target("aes,neon"))) poly128_t vmull_high_p64(poly64x2_t __p0, poly64x2_t __p1) {
70019	72537	poly128_t __ret;
70020		poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
70021		poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
70022		__ret = vmull_p64((poly64_t)(__noswap_vget_high_p64(__rev0)), (poly64_t)(__noswap_vget_high_p64(__rev1)));
	72538	poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	72539	poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_64);
	72540	__ret = vmull_p64(__builtin_bit_cast(poly64_t, __noswap_vget_high_p64(__rev0)), __builtin_bit_cast(poly64_t, __noswap_vget_high_p64(__rev1)));
70023	72541	return __ret;
70024	72542	}
70025	72543	#endif
70026	72544
70027	72545	#ifdef __LITTLE_ENDIAN__
70028		#define vfmlalq_lane_high_f16(__p0_857, __p1_857, __p2_857, __p3_857) __extension__ ({ \
70029		float32x4_t __ret_857; \
70030		float32x4_t __s0_857 = __p0_857; \
70031		float16x8_t __s1_857 = __p1_857; \
70032		float16x4_t __s2_857 = __p2_857; \
70033		__ret_857 = vfmlalq_high_f16(__s0_857, __s1_857, (float16x8_t) {vget_lane_f16(__s2_857, __p3_857), vget_lane_f16(__s2_857, __p3_857), vget_lane_f16(__s2_857, __p3_857), vget_lane_f16(__s2_857, __p3_857), vget_lane_f16(__s2_857, __p3_857), vget_lane_f16(__s2_857, __p3_857), vget_lane_f16(__s2_857, __p3_857), vget_lane_f16(__s2_857, __p3_857)}); \
70034		__ret_857; \
70035		})
70036		#else
70037		#define vfmlalq_lane_high_f16(__p0_858, __p1_858, __p2_858, __p3_858) __extension__ ({ \
70038		float32x4_t __ret_858; \
70039		float32x4_t __s0_858 = __p0_858; \
70040		float16x8_t __s1_858 = __p1_858; \
70041		float16x4_t __s2_858 = __p2_858; \
70042		float32x4_t __rev0_858; __rev0_858 = __builtin_shufflevector(__s0_858, __s0_858, 3, 2, 1, 0); \
70043		float16x8_t __rev1_858; __rev1_858 = __builtin_shufflevector(__s1_858, __s1_858, 7, 6, 5, 4, 3, 2, 1, 0); \
70044		float16x4_t __rev2_858; __rev2_858 = __builtin_shufflevector(__s2_858, __s2_858, 3, 2, 1, 0); \
70045		__ret_858 = __noswap_vfmlalq_high_f16(__rev0_858, __rev1_858, (float16x8_t) {__noswap_vget_lane_f16(__rev2_858, __p3_858), __noswap_vget_lane_f16(__rev2_858, __p3_858), __noswap_vget_lane_f16(__rev2_858, __p3_858), __noswap_vget_lane_f16(__rev2_858, __p3_858), __noswap_vget_lane_f16(__rev2_858, __p3_858), __noswap_vget_lane_f16(__rev2_858, __p3_858), __noswap_vget_lane_f16(__rev2_858, __p3_858), __noswap_vget_lane_f16(__rev2_858, __p3_858)}); \
70046		__ret_858 = __builtin_shufflevector(__ret_858, __ret_858, 3, 2, 1, 0); \
70047		__ret_858; \
70048		})
70049		#endif
70050
70051		#ifdef __LITTLE_ENDIAN__
70052		#define vfmlal_lane_high_f16(__p0_859, __p1_859, __p2_859, __p3_859) __extension__ ({ \
70053		float32x2_t __ret_859; \
70054		float32x2_t __s0_859 = __p0_859; \
70055		float16x4_t __s1_859 = __p1_859; \
70056		float16x4_t __s2_859 = __p2_859; \
70057		__ret_859 = vfmlal_high_f16(__s0_859, __s1_859, (float16x4_t) {vget_lane_f16(__s2_859, __p3_859), vget_lane_f16(__s2_859, __p3_859), vget_lane_f16(__s2_859, __p3_859), vget_lane_f16(__s2_859, __p3_859)}); \
70058		__ret_859; \
70059		})
70060		#else
70061		#define vfmlal_lane_high_f16(__p0_860, __p1_860, __p2_860, __p3_860) __extension__ ({ \
70062		float32x2_t __ret_860; \
70063		float32x2_t __s0_860 = __p0_860; \
70064		float16x4_t __s1_860 = __p1_860; \
70065		float16x4_t __s2_860 = __p2_860; \
70066		float32x2_t __rev0_860; __rev0_860 = __builtin_shufflevector(__s0_860, __s0_860, 1, 0); \
70067		float16x4_t __rev1_860; __rev1_860 = __builtin_shufflevector(__s1_860, __s1_860, 3, 2, 1, 0); \
70068		float16x4_t __rev2_860; __rev2_860 = __builtin_shufflevector(__s2_860, __s2_860, 3, 2, 1, 0); \
70069		__ret_860 = __noswap_vfmlal_high_f16(__rev0_860, __rev1_860, (float16x4_t) {__noswap_vget_lane_f16(__rev2_860, __p3_860), __noswap_vget_lane_f16(__rev2_860, __p3_860), __noswap_vget_lane_f16(__rev2_860, __p3_860), __noswap_vget_lane_f16(__rev2_860, __p3_860)}); \
70070		__ret_860 = __builtin_shufflevector(__ret_860, __ret_860, 1, 0); \
70071		__ret_860; \
70072		})
70073		#endif
70074
70075		#ifdef __LITTLE_ENDIAN__
70076		#define vfmlalq_lane_low_f16(__p0_861, __p1_861, __p2_861, __p3_861) __extension__ ({ \
70077		float32x4_t __ret_861; \
70078		float32x4_t __s0_861 = __p0_861; \
70079		float16x8_t __s1_861 = __p1_861; \
70080		float16x4_t __s2_861 = __p2_861; \
70081		__ret_861 = vfmlalq_low_f16(__s0_861, __s1_861, (float16x8_t) {vget_lane_f16(__s2_861, __p3_861), vget_lane_f16(__s2_861, __p3_861), vget_lane_f16(__s2_861, __p3_861), vget_lane_f16(__s2_861, __p3_861), vget_lane_f16(__s2_861, __p3_861), vget_lane_f16(__s2_861, __p3_861), vget_lane_f16(__s2_861, __p3_861), vget_lane_f16(__s2_861, __p3_861)}); \
70082		__ret_861; \
70083		})
70084		#else
70085		#define vfmlalq_lane_low_f16(__p0_862, __p1_862, __p2_862, __p3_862) __extension__ ({ \
70086		float32x4_t __ret_862; \
70087		float32x4_t __s0_862 = __p0_862; \
70088		float16x8_t __s1_862 = __p1_862; \
70089		float16x4_t __s2_862 = __p2_862; \
70090		float32x4_t __rev0_862; __rev0_862 = __builtin_shufflevector(__s0_862, __s0_862, 3, 2, 1, 0); \
70091		float16x8_t __rev1_862; __rev1_862 = __builtin_shufflevector(__s1_862, __s1_862, 7, 6, 5, 4, 3, 2, 1, 0); \
70092		float16x4_t __rev2_862; __rev2_862 = __builtin_shufflevector(__s2_862, __s2_862, 3, 2, 1, 0); \
70093		__ret_862 = __noswap_vfmlalq_low_f16(__rev0_862, __rev1_862, (float16x8_t) {__noswap_vget_lane_f16(__rev2_862, __p3_862), __noswap_vget_lane_f16(__rev2_862, __p3_862), __noswap_vget_lane_f16(__rev2_862, __p3_862), __noswap_vget_lane_f16(__rev2_862, __p3_862), __noswap_vget_lane_f16(__rev2_862, __p3_862), __noswap_vget_lane_f16(__rev2_862, __p3_862), __noswap_vget_lane_f16(__rev2_862, __p3_862), __noswap_vget_lane_f16(__rev2_862, __p3_862)}); \
70094		__ret_862 = __builtin_shufflevector(__ret_862, __ret_862, 3, 2, 1, 0); \
70095		__ret_862; \
70096		})
70097		#endif
70098
70099		#ifdef __LITTLE_ENDIAN__
70100		#define vfmlal_lane_low_f16(__p0_863, __p1_863, __p2_863, __p3_863) __extension__ ({ \
70101		float32x2_t __ret_863; \
70102		float32x2_t __s0_863 = __p0_863; \
70103		float16x4_t __s1_863 = __p1_863; \
70104		float16x4_t __s2_863 = __p2_863; \
70105		__ret_863 = vfmlal_low_f16(__s0_863, __s1_863, (float16x4_t) {vget_lane_f16(__s2_863, __p3_863), vget_lane_f16(__s2_863, __p3_863), vget_lane_f16(__s2_863, __p3_863), vget_lane_f16(__s2_863, __p3_863)}); \
70106		__ret_863; \
70107		})
70108		#else
70109		#define vfmlal_lane_low_f16(__p0_864, __p1_864, __p2_864, __p3_864) __extension__ ({ \
70110		float32x2_t __ret_864; \
70111		float32x2_t __s0_864 = __p0_864; \
70112		float16x4_t __s1_864 = __p1_864; \
70113		float16x4_t __s2_864 = __p2_864; \
70114		float32x2_t __rev0_864; __rev0_864 = __builtin_shufflevector(__s0_864, __s0_864, 1, 0); \
70115		float16x4_t __rev1_864; __rev1_864 = __builtin_shufflevector(__s1_864, __s1_864, 3, 2, 1, 0); \
70116		float16x4_t __rev2_864; __rev2_864 = __builtin_shufflevector(__s2_864, __s2_864, 3, 2, 1, 0); \
70117		__ret_864 = __noswap_vfmlal_low_f16(__rev0_864, __rev1_864, (float16x4_t) {__noswap_vget_lane_f16(__rev2_864, __p3_864), __noswap_vget_lane_f16(__rev2_864, __p3_864), __noswap_vget_lane_f16(__rev2_864, __p3_864), __noswap_vget_lane_f16(__rev2_864, __p3_864)}); \
70118		__ret_864 = __builtin_shufflevector(__ret_864, __ret_864, 1, 0); \
70119		__ret_864; \
70120		})
70121		#endif
70122
70123		#ifdef __LITTLE_ENDIAN__
70124		#define vfmlalq_laneq_high_f16(__p0_865, __p1_865, __p2_865, __p3_865) __extension__ ({ \
70125		float32x4_t __ret_865; \
70126		float32x4_t __s0_865 = __p0_865; \
70127		float16x8_t __s1_865 = __p1_865; \
70128		float16x8_t __s2_865 = __p2_865; \
70129		__ret_865 = vfmlalq_high_f16(__s0_865, __s1_865, (float16x8_t) {vgetq_lane_f16(__s2_865, __p3_865), vgetq_lane_f16(__s2_865, __p3_865), vgetq_lane_f16(__s2_865, __p3_865), vgetq_lane_f16(__s2_865, __p3_865), vgetq_lane_f16(__s2_865, __p3_865), vgetq_lane_f16(__s2_865, __p3_865), vgetq_lane_f16(__s2_865, __p3_865), vgetq_lane_f16(__s2_865, __p3_865)}); \
70130		__ret_865; \
70131		})
70132		#else
70133		#define vfmlalq_laneq_high_f16(__p0_866, __p1_866, __p2_866, __p3_866) __extension__ ({ \
70134		float32x4_t __ret_866; \
70135		float32x4_t __s0_866 = __p0_866; \
70136		float16x8_t __s1_866 = __p1_866; \
70137		float16x8_t __s2_866 = __p2_866; \
70138		float32x4_t __rev0_866; __rev0_866 = __builtin_shufflevector(__s0_866, __s0_866, 3, 2, 1, 0); \
70139		float16x8_t __rev1_866; __rev1_866 = __builtin_shufflevector(__s1_866, __s1_866, 7, 6, 5, 4, 3, 2, 1, 0); \
70140		float16x8_t __rev2_866; __rev2_866 = __builtin_shufflevector(__s2_866, __s2_866, 7, 6, 5, 4, 3, 2, 1, 0); \
70141		__ret_866 = __noswap_vfmlalq_high_f16(__rev0_866, __rev1_866, (float16x8_t) {__noswap_vgetq_lane_f16(__rev2_866, __p3_866), __noswap_vgetq_lane_f16(__rev2_866, __p3_866), __noswap_vgetq_lane_f16(__rev2_866, __p3_866), __noswap_vgetq_lane_f16(__rev2_866, __p3_866), __noswap_vgetq_lane_f16(__rev2_866, __p3_866), __noswap_vgetq_lane_f16(__rev2_866, __p3_866), __noswap_vgetq_lane_f16(__rev2_866, __p3_866), __noswap_vgetq_lane_f16(__rev2_866, __p3_866)}); \
70142		__ret_866 = __builtin_shufflevector(__ret_866, __ret_866, 3, 2, 1, 0); \
70143		__ret_866; \
70144		})
70145		#endif
70146
70147		#ifdef __LITTLE_ENDIAN__
70148		#define vfmlal_laneq_high_f16(__p0_867, __p1_867, __p2_867, __p3_867) __extension__ ({ \
70149		float32x2_t __ret_867; \
70150		float32x2_t __s0_867 = __p0_867; \
70151		float16x4_t __s1_867 = __p1_867; \
70152		float16x8_t __s2_867 = __p2_867; \
70153		__ret_867 = vfmlal_high_f16(__s0_867, __s1_867, (float16x4_t) {vgetq_lane_f16(__s2_867, __p3_867), vgetq_lane_f16(__s2_867, __p3_867), vgetq_lane_f16(__s2_867, __p3_867), vgetq_lane_f16(__s2_867, __p3_867)}); \
70154		__ret_867; \
70155		})
70156		#else
70157		#define vfmlal_laneq_high_f16(__p0_868, __p1_868, __p2_868, __p3_868) __extension__ ({ \
70158		float32x2_t __ret_868; \
70159		float32x2_t __s0_868 = __p0_868; \
70160		float16x4_t __s1_868 = __p1_868; \
70161		float16x8_t __s2_868 = __p2_868; \
70162		float32x2_t __rev0_868; __rev0_868 = __builtin_shufflevector(__s0_868, __s0_868, 1, 0); \
70163		float16x4_t __rev1_868; __rev1_868 = __builtin_shufflevector(__s1_868, __s1_868, 3, 2, 1, 0); \
70164		float16x8_t __rev2_868; __rev2_868 = __builtin_shufflevector(__s2_868, __s2_868, 7, 6, 5, 4, 3, 2, 1, 0); \
70165		__ret_868 = __noswap_vfmlal_high_f16(__rev0_868, __rev1_868, (float16x4_t) {__noswap_vgetq_lane_f16(__rev2_868, __p3_868), __noswap_vgetq_lane_f16(__rev2_868, __p3_868), __noswap_vgetq_lane_f16(__rev2_868, __p3_868), __noswap_vgetq_lane_f16(__rev2_868, __p3_868)}); \
70166		__ret_868 = __builtin_shufflevector(__ret_868, __ret_868, 1, 0); \
70167		__ret_868; \
70168		})
70169		#endif
70170
70171		#ifdef __LITTLE_ENDIAN__
70172		#define vfmlalq_laneq_low_f16(__p0_869, __p1_869, __p2_869, __p3_869) __extension__ ({ \
70173		float32x4_t __ret_869; \
70174		float32x4_t __s0_869 = __p0_869; \
70175		float16x8_t __s1_869 = __p1_869; \
70176		float16x8_t __s2_869 = __p2_869; \
70177		__ret_869 = vfmlalq_low_f16(__s0_869, __s1_869, (float16x8_t) {vgetq_lane_f16(__s2_869, __p3_869), vgetq_lane_f16(__s2_869, __p3_869), vgetq_lane_f16(__s2_869, __p3_869), vgetq_lane_f16(__s2_869, __p3_869), vgetq_lane_f16(__s2_869, __p3_869), vgetq_lane_f16(__s2_869, __p3_869), vgetq_lane_f16(__s2_869, __p3_869), vgetq_lane_f16(__s2_869, __p3_869)}); \
70178		__ret_869; \
70179		})
70180		#else
70181		#define vfmlalq_laneq_low_f16(__p0_870, __p1_870, __p2_870, __p3_870) __extension__ ({ \
70182		float32x4_t __ret_870; \
70183		float32x4_t __s0_870 = __p0_870; \
70184		float16x8_t __s1_870 = __p1_870; \
70185		float16x8_t __s2_870 = __p2_870; \
70186		float32x4_t __rev0_870; __rev0_870 = __builtin_shufflevector(__s0_870, __s0_870, 3, 2, 1, 0); \
70187		float16x8_t __rev1_870; __rev1_870 = __builtin_shufflevector(__s1_870, __s1_870, 7, 6, 5, 4, 3, 2, 1, 0); \
70188		float16x8_t __rev2_870; __rev2_870 = __builtin_shufflevector(__s2_870, __s2_870, 7, 6, 5, 4, 3, 2, 1, 0); \
70189		__ret_870 = __noswap_vfmlalq_low_f16(__rev0_870, __rev1_870, (float16x8_t) {__noswap_vgetq_lane_f16(__rev2_870, __p3_870), __noswap_vgetq_lane_f16(__rev2_870, __p3_870), __noswap_vgetq_lane_f16(__rev2_870, __p3_870), __noswap_vgetq_lane_f16(__rev2_870, __p3_870), __noswap_vgetq_lane_f16(__rev2_870, __p3_870), __noswap_vgetq_lane_f16(__rev2_870, __p3_870), __noswap_vgetq_lane_f16(__rev2_870, __p3_870), __noswap_vgetq_lane_f16(__rev2_870, __p3_870)}); \
70190		__ret_870 = __builtin_shufflevector(__ret_870, __ret_870, 3, 2, 1, 0); \
70191		__ret_870; \
70192		})
70193		#endif
70194
70195		#ifdef __LITTLE_ENDIAN__
70196		#define vfmlal_laneq_low_f16(__p0_871, __p1_871, __p2_871, __p3_871) __extension__ ({ \
70197		float32x2_t __ret_871; \
70198		float32x2_t __s0_871 = __p0_871; \
70199		float16x4_t __s1_871 = __p1_871; \
70200		float16x8_t __s2_871 = __p2_871; \
70201		__ret_871 = vfmlal_low_f16(__s0_871, __s1_871, (float16x4_t) {vgetq_lane_f16(__s2_871, __p3_871), vgetq_lane_f16(__s2_871, __p3_871), vgetq_lane_f16(__s2_871, __p3_871), vgetq_lane_f16(__s2_871, __p3_871)}); \
70202		__ret_871; \
70203		})
70204		#else
70205		#define vfmlal_laneq_low_f16(__p0_872, __p1_872, __p2_872, __p3_872) __extension__ ({ \
70206		float32x2_t __ret_872; \
70207		float32x2_t __s0_872 = __p0_872; \
70208		float16x4_t __s1_872 = __p1_872; \
70209		float16x8_t __s2_872 = __p2_872; \
70210		float32x2_t __rev0_872; __rev0_872 = __builtin_shufflevector(__s0_872, __s0_872, 1, 0); \
70211		float16x4_t __rev1_872; __rev1_872 = __builtin_shufflevector(__s1_872, __s1_872, 3, 2, 1, 0); \
70212		float16x8_t __rev2_872; __rev2_872 = __builtin_shufflevector(__s2_872, __s2_872, 7, 6, 5, 4, 3, 2, 1, 0); \
70213		__ret_872 = __noswap_vfmlal_low_f16(__rev0_872, __rev1_872, (float16x4_t) {__noswap_vgetq_lane_f16(__rev2_872, __p3_872), __noswap_vgetq_lane_f16(__rev2_872, __p3_872), __noswap_vgetq_lane_f16(__rev2_872, __p3_872), __noswap_vgetq_lane_f16(__rev2_872, __p3_872)}); \
70214		__ret_872 = __builtin_shufflevector(__ret_872, __ret_872, 1, 0); \
70215		__ret_872; \
70216		})
70217		#endif
70218
70219		#ifdef __LITTLE_ENDIAN__
70220		#define vfmlslq_lane_high_f16(__p0_873, __p1_873, __p2_873, __p3_873) __extension__ ({ \
	72546	#define vfmlalq_lane_high_f16(__p0_873, __p1_873, __p2_873, __p3_873) __extension__ ({ \
70221	72547	float32x4_t __ret_873; \
70222	72548	float32x4_t __s0_873 = __p0_873; \
70223	72549	float16x8_t __s1_873 = __p1_873; \
70224	72550	float16x4_t __s2_873 = __p2_873; \
70225		__ret_873 = vfmlslq_high_f16(__s0_873, __s1_873, (float16x8_t) {vget_lane_f16(__s2_873, __p3_873), vget_lane_f16(__s2_873, __p3_873), vget_lane_f16(__s2_873, __p3_873), vget_lane_f16(__s2_873, __p3_873), vget_lane_f16(__s2_873, __p3_873), vget_lane_f16(__s2_873, __p3_873), vget_lane_f16(__s2_873, __p3_873), vget_lane_f16(__s2_873, __p3_873)}); \
	72551	__ret_873 = vfmlalq_high_f16(__s0_873, __s1_873, (float16x8_t) {vget_lane_f16(__s2_873, __p3_873), vget_lane_f16(__s2_873, __p3_873), vget_lane_f16(__s2_873, __p3_873), vget_lane_f16(__s2_873, __p3_873), vget_lane_f16(__s2_873, __p3_873), vget_lane_f16(__s2_873, __p3_873), vget_lane_f16(__s2_873, __p3_873), vget_lane_f16(__s2_873, __p3_873)}); \
70226	72552	__ret_873; \
70227	72553	})
70228	72554	#else
70229		#define vfmlslq_lane_high_f16(__p0_874, __p1_874, __p2_874, __p3_874) __extension__ ({ \
	72555	#define vfmlalq_lane_high_f16(__p0_874, __p1_874, __p2_874, __p3_874) __extension__ ({ \
70230	72556	float32x4_t __ret_874; \
70231	72557	float32x4_t __s0_874 = __p0_874; \
70232	72558	float16x8_t __s1_874 = __p1_874; \
70233	72559	float16x4_t __s2_874 = __p2_874; \
70234		float32x4_t __rev0_874; __rev0_874 = __builtin_shufflevector(__s0_874, __s0_874, 3, 2, 1, 0); \
70235		float16x8_t __rev1_874; __rev1_874 = __builtin_shufflevector(__s1_874, __s1_874, 7, 6, 5, 4, 3, 2, 1, 0); \
70236		float16x4_t __rev2_874; __rev2_874 = __builtin_shufflevector(__s2_874, __s2_874, 3, 2, 1, 0); \
70237		__ret_874 = __noswap_vfmlslq_high_f16(__rev0_874, __rev1_874, (float16x8_t) {__noswap_vget_lane_f16(__rev2_874, __p3_874), __noswap_vget_lane_f16(__rev2_874, __p3_874), __noswap_vget_lane_f16(__rev2_874, __p3_874), __noswap_vget_lane_f16(__rev2_874, __p3_874), __noswap_vget_lane_f16(__rev2_874, __p3_874), __noswap_vget_lane_f16(__rev2_874, __p3_874), __noswap_vget_lane_f16(__rev2_874, __p3_874), __noswap_vget_lane_f16(__rev2_874, __p3_874)}); \
70238		__ret_874 = __builtin_shufflevector(__ret_874, __ret_874, 3, 2, 1, 0); \
	72560	float32x4_t __rev0_874; __rev0_874 = __builtin_shufflevector(__s0_874, __s0_874, __lane_reverse_128_32); \
	72561	float16x8_t __rev1_874; __rev1_874 = __builtin_shufflevector(__s1_874, __s1_874, __lane_reverse_128_16); \
	72562	float16x4_t __rev2_874; __rev2_874 = __builtin_shufflevector(__s2_874, __s2_874, __lane_reverse_64_16); \
	72563	__ret_874 = __noswap_vfmlalq_high_f16(__rev0_874, __rev1_874, (float16x8_t) {__noswap_vget_lane_f16(__rev2_874, __p3_874), __noswap_vget_lane_f16(__rev2_874, __p3_874), __noswap_vget_lane_f16(__rev2_874, __p3_874), __noswap_vget_lane_f16(__rev2_874, __p3_874), __noswap_vget_lane_f16(__rev2_874, __p3_874), __noswap_vget_lane_f16(__rev2_874, __p3_874), __noswap_vget_lane_f16(__rev2_874, __p3_874), __noswap_vget_lane_f16(__rev2_874, __p3_874)}); \
	72564	__ret_874 = __builtin_shufflevector(__ret_874, __ret_874, __lane_reverse_128_32); \
70239	72565	__ret_874; \
70240	72566	})
70241	72567	#endif
70242	72568
70243	72569	#ifdef __LITTLE_ENDIAN__
70244		#define vfmlsl_lane_high_f16(__p0_875, __p1_875, __p2_875, __p3_875) __extension__ ({ \
	72570	#define vfmlal_lane_high_f16(__p0_875, __p1_875, __p2_875, __p3_875) __extension__ ({ \
70245	72571	float32x2_t __ret_875; \
70246	72572	float32x2_t __s0_875 = __p0_875; \
70247	72573	float16x4_t __s1_875 = __p1_875; \
70248	72574	float16x4_t __s2_875 = __p2_875; \
70249		__ret_875 = vfmlsl_high_f16(__s0_875, __s1_875, (float16x4_t) {vget_lane_f16(__s2_875, __p3_875), vget_lane_f16(__s2_875, __p3_875), vget_lane_f16(__s2_875, __p3_875), vget_lane_f16(__s2_875, __p3_875)}); \
	72575	__ret_875 = vfmlal_high_f16(__s0_875, __s1_875, (float16x4_t) {vget_lane_f16(__s2_875, __p3_875), vget_lane_f16(__s2_875, __p3_875), vget_lane_f16(__s2_875, __p3_875), vget_lane_f16(__s2_875, __p3_875)}); \
70250	72576	__ret_875; \
70251	72577	})
70252	72578	#else
70253		#define vfmlsl_lane_high_f16(__p0_876, __p1_876, __p2_876, __p3_876) __extension__ ({ \
	72579	#define vfmlal_lane_high_f16(__p0_876, __p1_876, __p2_876, __p3_876) __extension__ ({ \
70254	72580	float32x2_t __ret_876; \
70255	72581	float32x2_t __s0_876 = __p0_876; \
70256	72582	float16x4_t __s1_876 = __p1_876; \
70257	72583	float16x4_t __s2_876 = __p2_876; \
70258		float32x2_t __rev0_876; __rev0_876 = __builtin_shufflevector(__s0_876, __s0_876, 1, 0); \
70259		float16x4_t __rev1_876; __rev1_876 = __builtin_shufflevector(__s1_876, __s1_876, 3, 2, 1, 0); \
70260		float16x4_t __rev2_876; __rev2_876 = __builtin_shufflevector(__s2_876, __s2_876, 3, 2, 1, 0); \
70261		__ret_876 = __noswap_vfmlsl_high_f16(__rev0_876, __rev1_876, (float16x4_t) {__noswap_vget_lane_f16(__rev2_876, __p3_876), __noswap_vget_lane_f16(__rev2_876, __p3_876), __noswap_vget_lane_f16(__rev2_876, __p3_876), __noswap_vget_lane_f16(__rev2_876, __p3_876)}); \
70262		__ret_876 = __builtin_shufflevector(__ret_876, __ret_876, 1, 0); \
	72584	float32x2_t __rev0_876; __rev0_876 = __builtin_shufflevector(__s0_876, __s0_876, __lane_reverse_64_32); \
	72585	float16x4_t __rev1_876; __rev1_876 = __builtin_shufflevector(__s1_876, __s1_876, __lane_reverse_64_16); \
	72586	float16x4_t __rev2_876; __rev2_876 = __builtin_shufflevector(__s2_876, __s2_876, __lane_reverse_64_16); \
	72587	__ret_876 = __noswap_vfmlal_high_f16(__rev0_876, __rev1_876, (float16x4_t) {__noswap_vget_lane_f16(__rev2_876, __p3_876), __noswap_vget_lane_f16(__rev2_876, __p3_876), __noswap_vget_lane_f16(__rev2_876, __p3_876), __noswap_vget_lane_f16(__rev2_876, __p3_876)}); \
	72588	__ret_876 = __builtin_shufflevector(__ret_876, __ret_876, __lane_reverse_64_32); \
70263	72589	__ret_876; \
70264	72590	})
70265	72591	#endif
70266	72592
70267	72593	#ifdef __LITTLE_ENDIAN__
70268		#define vfmlslq_lane_low_f16(__p0_877, __p1_877, __p2_877, __p3_877) __extension__ ({ \
	72594	#define vfmlalq_lane_low_f16(__p0_877, __p1_877, __p2_877, __p3_877) __extension__ ({ \
70269	72595	float32x4_t __ret_877; \
70270	72596	float32x4_t __s0_877 = __p0_877; \
70271	72597	float16x8_t __s1_877 = __p1_877; \
70272	72598	float16x4_t __s2_877 = __p2_877; \
70273		__ret_877 = vfmlslq_low_f16(__s0_877, __s1_877, (float16x8_t) {vget_lane_f16(__s2_877, __p3_877), vget_lane_f16(__s2_877, __p3_877), vget_lane_f16(__s2_877, __p3_877), vget_lane_f16(__s2_877, __p3_877), vget_lane_f16(__s2_877, __p3_877), vget_lane_f16(__s2_877, __p3_877), vget_lane_f16(__s2_877, __p3_877), vget_lane_f16(__s2_877, __p3_877)}); \
	72599	__ret_877 = vfmlalq_low_f16(__s0_877, __s1_877, (float16x8_t) {vget_lane_f16(__s2_877, __p3_877), vget_lane_f16(__s2_877, __p3_877), vget_lane_f16(__s2_877, __p3_877), vget_lane_f16(__s2_877, __p3_877), vget_lane_f16(__s2_877, __p3_877), vget_lane_f16(__s2_877, __p3_877), vget_lane_f16(__s2_877, __p3_877), vget_lane_f16(__s2_877, __p3_877)}); \
70274	72600	__ret_877; \
70275	72601	})
70276	72602	#else
70277		#define vfmlslq_lane_low_f16(__p0_878, __p1_878, __p2_878, __p3_878) __extension__ ({ \
	72603	#define vfmlalq_lane_low_f16(__p0_878, __p1_878, __p2_878, __p3_878) __extension__ ({ \
70278	72604	float32x4_t __ret_878; \
70279	72605	float32x4_t __s0_878 = __p0_878; \
70280	72606	float16x8_t __s1_878 = __p1_878; \
70281	72607	float16x4_t __s2_878 = __p2_878; \
70282		float32x4_t __rev0_878; __rev0_878 = __builtin_shufflevector(__s0_878, __s0_878, 3, 2, 1, 0); \
70283		float16x8_t __rev1_878; __rev1_878 = __builtin_shufflevector(__s1_878, __s1_878, 7, 6, 5, 4, 3, 2, 1, 0); \
70284		float16x4_t __rev2_878; __rev2_878 = __builtin_shufflevector(__s2_878, __s2_878, 3, 2, 1, 0); \
70285		__ret_878 = __noswap_vfmlslq_low_f16(__rev0_878, __rev1_878, (float16x8_t) {__noswap_vget_lane_f16(__rev2_878, __p3_878), __noswap_vget_lane_f16(__rev2_878, __p3_878), __noswap_vget_lane_f16(__rev2_878, __p3_878), __noswap_vget_lane_f16(__rev2_878, __p3_878), __noswap_vget_lane_f16(__rev2_878, __p3_878), __noswap_vget_lane_f16(__rev2_878, __p3_878), __noswap_vget_lane_f16(__rev2_878, __p3_878), __noswap_vget_lane_f16(__rev2_878, __p3_878)}); \
70286		__ret_878 = __builtin_shufflevector(__ret_878, __ret_878, 3, 2, 1, 0); \
	72608	float32x4_t __rev0_878; __rev0_878 = __builtin_shufflevector(__s0_878, __s0_878, __lane_reverse_128_32); \
	72609	float16x8_t __rev1_878; __rev1_878 = __builtin_shufflevector(__s1_878, __s1_878, __lane_reverse_128_16); \
	72610	float16x4_t __rev2_878; __rev2_878 = __builtin_shufflevector(__s2_878, __s2_878, __lane_reverse_64_16); \
	72611	__ret_878 = __noswap_vfmlalq_low_f16(__rev0_878, __rev1_878, (float16x8_t) {__noswap_vget_lane_f16(__rev2_878, __p3_878), __noswap_vget_lane_f16(__rev2_878, __p3_878), __noswap_vget_lane_f16(__rev2_878, __p3_878), __noswap_vget_lane_f16(__rev2_878, __p3_878), __noswap_vget_lane_f16(__rev2_878, __p3_878), __noswap_vget_lane_f16(__rev2_878, __p3_878), __noswap_vget_lane_f16(__rev2_878, __p3_878), __noswap_vget_lane_f16(__rev2_878, __p3_878)}); \
	72612	__ret_878 = __builtin_shufflevector(__ret_878, __ret_878, __lane_reverse_128_32); \
70287	72613	__ret_878; \
70288	72614	})
70289	72615	#endif
70290	72616
70291	72617	#ifdef __LITTLE_ENDIAN__
70292		#define vfmlsl_lane_low_f16(__p0_879, __p1_879, __p2_879, __p3_879) __extension__ ({ \
	72618	#define vfmlal_lane_low_f16(__p0_879, __p1_879, __p2_879, __p3_879) __extension__ ({ \
70293	72619	float32x2_t __ret_879; \
70294	72620	float32x2_t __s0_879 = __p0_879; \
70295	72621	float16x4_t __s1_879 = __p1_879; \
70296	72622	float16x4_t __s2_879 = __p2_879; \
70297		__ret_879 = vfmlsl_low_f16(__s0_879, __s1_879, (float16x4_t) {vget_lane_f16(__s2_879, __p3_879), vget_lane_f16(__s2_879, __p3_879), vget_lane_f16(__s2_879, __p3_879), vget_lane_f16(__s2_879, __p3_879)}); \
	72623	__ret_879 = vfmlal_low_f16(__s0_879, __s1_879, (float16x4_t) {vget_lane_f16(__s2_879, __p3_879), vget_lane_f16(__s2_879, __p3_879), vget_lane_f16(__s2_879, __p3_879), vget_lane_f16(__s2_879, __p3_879)}); \
70298	72624	__ret_879; \
70299	72625	})
70300	72626	#else
70301		#define vfmlsl_lane_low_f16(__p0_880, __p1_880, __p2_880, __p3_880) __extension__ ({ \
	72627	#define vfmlal_lane_low_f16(__p0_880, __p1_880, __p2_880, __p3_880) __extension__ ({ \
70302	72628	float32x2_t __ret_880; \
70303	72629	float32x2_t __s0_880 = __p0_880; \
70304	72630	float16x4_t __s1_880 = __p1_880; \
70305	72631	float16x4_t __s2_880 = __p2_880; \
70306		float32x2_t __rev0_880; __rev0_880 = __builtin_shufflevector(__s0_880, __s0_880, 1, 0); \
70307		float16x4_t __rev1_880; __rev1_880 = __builtin_shufflevector(__s1_880, __s1_880, 3, 2, 1, 0); \
70308		float16x4_t __rev2_880; __rev2_880 = __builtin_shufflevector(__s2_880, __s2_880, 3, 2, 1, 0); \
70309		__ret_880 = __noswap_vfmlsl_low_f16(__rev0_880, __rev1_880, (float16x4_t) {__noswap_vget_lane_f16(__rev2_880, __p3_880), __noswap_vget_lane_f16(__rev2_880, __p3_880), __noswap_vget_lane_f16(__rev2_880, __p3_880), __noswap_vget_lane_f16(__rev2_880, __p3_880)}); \
70310		__ret_880 = __builtin_shufflevector(__ret_880, __ret_880, 1, 0); \
	72632	float32x2_t __rev0_880; __rev0_880 = __builtin_shufflevector(__s0_880, __s0_880, __lane_reverse_64_32); \
	72633	float16x4_t __rev1_880; __rev1_880 = __builtin_shufflevector(__s1_880, __s1_880, __lane_reverse_64_16); \
	72634	float16x4_t __rev2_880; __rev2_880 = __builtin_shufflevector(__s2_880, __s2_880, __lane_reverse_64_16); \
	72635	__ret_880 = __noswap_vfmlal_low_f16(__rev0_880, __rev1_880, (float16x4_t) {__noswap_vget_lane_f16(__rev2_880, __p3_880), __noswap_vget_lane_f16(__rev2_880, __p3_880), __noswap_vget_lane_f16(__rev2_880, __p3_880), __noswap_vget_lane_f16(__rev2_880, __p3_880)}); \
	72636	__ret_880 = __builtin_shufflevector(__ret_880, __ret_880, __lane_reverse_64_32); \
70311	72637	__ret_880; \
70312	72638	})
70313	72639	#endif
70314	72640
70315	72641	#ifdef __LITTLE_ENDIAN__
70316		#define vfmlslq_laneq_high_f16(__p0_881, __p1_881, __p2_881, __p3_881) __extension__ ({ \
	72642	#define vfmlalq_laneq_high_f16(__p0_881, __p1_881, __p2_881, __p3_881) __extension__ ({ \
70317	72643	float32x4_t __ret_881; \
70318	72644	float32x4_t __s0_881 = __p0_881; \
70319	72645	float16x8_t __s1_881 = __p1_881; \
70320	72646	float16x8_t __s2_881 = __p2_881; \
70321		__ret_881 = vfmlslq_high_f16(__s0_881, __s1_881, (float16x8_t) {vgetq_lane_f16(__s2_881, __p3_881), vgetq_lane_f16(__s2_881, __p3_881), vgetq_lane_f16(__s2_881, __p3_881), vgetq_lane_f16(__s2_881, __p3_881), vgetq_lane_f16(__s2_881, __p3_881), vgetq_lane_f16(__s2_881, __p3_881), vgetq_lane_f16(__s2_881, __p3_881), vgetq_lane_f16(__s2_881, __p3_881)}); \
	72647	__ret_881 = vfmlalq_high_f16(__s0_881, __s1_881, (float16x8_t) {vgetq_lane_f16(__s2_881, __p3_881), vgetq_lane_f16(__s2_881, __p3_881), vgetq_lane_f16(__s2_881, __p3_881), vgetq_lane_f16(__s2_881, __p3_881), vgetq_lane_f16(__s2_881, __p3_881), vgetq_lane_f16(__s2_881, __p3_881), vgetq_lane_f16(__s2_881, __p3_881), vgetq_lane_f16(__s2_881, __p3_881)}); \
70322	72648	__ret_881; \
70323	72649	})
70324	72650	#else
70325		#define vfmlslq_laneq_high_f16(__p0_882, __p1_882, __p2_882, __p3_882) __extension__ ({ \
	72651	#define vfmlalq_laneq_high_f16(__p0_882, __p1_882, __p2_882, __p3_882) __extension__ ({ \
70326	72652	float32x4_t __ret_882; \
70327	72653	float32x4_t __s0_882 = __p0_882; \
70328	72654	float16x8_t __s1_882 = __p1_882; \
70329	72655	float16x8_t __s2_882 = __p2_882; \
70330		float32x4_t __rev0_882; __rev0_882 = __builtin_shufflevector(__s0_882, __s0_882, 3, 2, 1, 0); \
70331		float16x8_t __rev1_882; __rev1_882 = __builtin_shufflevector(__s1_882, __s1_882, 7, 6, 5, 4, 3, 2, 1, 0); \
70332		float16x8_t __rev2_882; __rev2_882 = __builtin_shufflevector(__s2_882, __s2_882, 7, 6, 5, 4, 3, 2, 1, 0); \
70333		__ret_882 = __noswap_vfmlslq_high_f16(__rev0_882, __rev1_882, (float16x8_t) {__noswap_vgetq_lane_f16(__rev2_882, __p3_882), __noswap_vgetq_lane_f16(__rev2_882, __p3_882), __noswap_vgetq_lane_f16(__rev2_882, __p3_882), __noswap_vgetq_lane_f16(__rev2_882, __p3_882), __noswap_vgetq_lane_f16(__rev2_882, __p3_882), __noswap_vgetq_lane_f16(__rev2_882, __p3_882), __noswap_vgetq_lane_f16(__rev2_882, __p3_882), __noswap_vgetq_lane_f16(__rev2_882, __p3_882)}); \
70334		__ret_882 = __builtin_shufflevector(__ret_882, __ret_882, 3, 2, 1, 0); \
	72656	float32x4_t __rev0_882; __rev0_882 = __builtin_shufflevector(__s0_882, __s0_882, __lane_reverse_128_32); \
	72657	float16x8_t __rev1_882; __rev1_882 = __builtin_shufflevector(__s1_882, __s1_882, __lane_reverse_128_16); \
	72658	float16x8_t __rev2_882; __rev2_882 = __builtin_shufflevector(__s2_882, __s2_882, __lane_reverse_128_16); \
	72659	__ret_882 = __noswap_vfmlalq_high_f16(__rev0_882, __rev1_882, (float16x8_t) {__noswap_vgetq_lane_f16(__rev2_882, __p3_882), __noswap_vgetq_lane_f16(__rev2_882, __p3_882), __noswap_vgetq_lane_f16(__rev2_882, __p3_882), __noswap_vgetq_lane_f16(__rev2_882, __p3_882), __noswap_vgetq_lane_f16(__rev2_882, __p3_882), __noswap_vgetq_lane_f16(__rev2_882, __p3_882), __noswap_vgetq_lane_f16(__rev2_882, __p3_882), __noswap_vgetq_lane_f16(__rev2_882, __p3_882)}); \
	72660	__ret_882 = __builtin_shufflevector(__ret_882, __ret_882, __lane_reverse_128_32); \
70335	72661	__ret_882; \
70336	72662	})
70337	72663	#endif
70338	72664
70339	72665	#ifdef __LITTLE_ENDIAN__
70340		#define vfmlsl_laneq_high_f16(__p0_883, __p1_883, __p2_883, __p3_883) __extension__ ({ \
	72666	#define vfmlal_laneq_high_f16(__p0_883, __p1_883, __p2_883, __p3_883) __extension__ ({ \
70341	72667	float32x2_t __ret_883; \
70342	72668	float32x2_t __s0_883 = __p0_883; \
70343	72669	float16x4_t __s1_883 = __p1_883; \
70344	72670	float16x8_t __s2_883 = __p2_883; \
70345		__ret_883 = vfmlsl_high_f16(__s0_883, __s1_883, (float16x4_t) {vgetq_lane_f16(__s2_883, __p3_883), vgetq_lane_f16(__s2_883, __p3_883), vgetq_lane_f16(__s2_883, __p3_883), vgetq_lane_f16(__s2_883, __p3_883)}); \
	72671	__ret_883 = vfmlal_high_f16(__s0_883, __s1_883, (float16x4_t) {vgetq_lane_f16(__s2_883, __p3_883), vgetq_lane_f16(__s2_883, __p3_883), vgetq_lane_f16(__s2_883, __p3_883), vgetq_lane_f16(__s2_883, __p3_883)}); \
70346	72672	__ret_883; \
70347	72673	})
70348	72674	#else
70349		#define vfmlsl_laneq_high_f16(__p0_884, __p1_884, __p2_884, __p3_884) __extension__ ({ \
	72675	#define vfmlal_laneq_high_f16(__p0_884, __p1_884, __p2_884, __p3_884) __extension__ ({ \
70350	72676	float32x2_t __ret_884; \
70351	72677	float32x2_t __s0_884 = __p0_884; \
70352	72678	float16x4_t __s1_884 = __p1_884; \
70353	72679	float16x8_t __s2_884 = __p2_884; \
70354		float32x2_t __rev0_884; __rev0_884 = __builtin_shufflevector(__s0_884, __s0_884, 1, 0); \
70355		float16x4_t __rev1_884; __rev1_884 = __builtin_shufflevector(__s1_884, __s1_884, 3, 2, 1, 0); \
70356		float16x8_t __rev2_884; __rev2_884 = __builtin_shufflevector(__s2_884, __s2_884, 7, 6, 5, 4, 3, 2, 1, 0); \
70357		__ret_884 = __noswap_vfmlsl_high_f16(__rev0_884, __rev1_884, (float16x4_t) {__noswap_vgetq_lane_f16(__rev2_884, __p3_884), __noswap_vgetq_lane_f16(__rev2_884, __p3_884), __noswap_vgetq_lane_f16(__rev2_884, __p3_884), __noswap_vgetq_lane_f16(__rev2_884, __p3_884)}); \
70358		__ret_884 = __builtin_shufflevector(__ret_884, __ret_884, 1, 0); \
	72680	float32x2_t __rev0_884; __rev0_884 = __builtin_shufflevector(__s0_884, __s0_884, __lane_reverse_64_32); \
	72681	float16x4_t __rev1_884; __rev1_884 = __builtin_shufflevector(__s1_884, __s1_884, __lane_reverse_64_16); \
	72682	float16x8_t __rev2_884; __rev2_884 = __builtin_shufflevector(__s2_884, __s2_884, __lane_reverse_128_16); \
	72683	__ret_884 = __noswap_vfmlal_high_f16(__rev0_884, __rev1_884, (float16x4_t) {__noswap_vgetq_lane_f16(__rev2_884, __p3_884), __noswap_vgetq_lane_f16(__rev2_884, __p3_884), __noswap_vgetq_lane_f16(__rev2_884, __p3_884), __noswap_vgetq_lane_f16(__rev2_884, __p3_884)}); \
	72684	__ret_884 = __builtin_shufflevector(__ret_884, __ret_884, __lane_reverse_64_32); \
70359	72685	__ret_884; \
70360	72686	})
70361	72687	#endif
70362	72688
70363	72689	#ifdef __LITTLE_ENDIAN__
70364		#define vfmlslq_laneq_low_f16(__p0_885, __p1_885, __p2_885, __p3_885) __extension__ ({ \
	72690	#define vfmlalq_laneq_low_f16(__p0_885, __p1_885, __p2_885, __p3_885) __extension__ ({ \
70365	72691	float32x4_t __ret_885; \
70366	72692	float32x4_t __s0_885 = __p0_885; \
70367	72693	float16x8_t __s1_885 = __p1_885; \
70368	72694	float16x8_t __s2_885 = __p2_885; \
70369		__ret_885 = vfmlslq_low_f16(__s0_885, __s1_885, (float16x8_t) {vgetq_lane_f16(__s2_885, __p3_885), vgetq_lane_f16(__s2_885, __p3_885), vgetq_lane_f16(__s2_885, __p3_885), vgetq_lane_f16(__s2_885, __p3_885), vgetq_lane_f16(__s2_885, __p3_885), vgetq_lane_f16(__s2_885, __p3_885), vgetq_lane_f16(__s2_885, __p3_885), vgetq_lane_f16(__s2_885, __p3_885)}); \
	72695	__ret_885 = vfmlalq_low_f16(__s0_885, __s1_885, (float16x8_t) {vgetq_lane_f16(__s2_885, __p3_885), vgetq_lane_f16(__s2_885, __p3_885), vgetq_lane_f16(__s2_885, __p3_885), vgetq_lane_f16(__s2_885, __p3_885), vgetq_lane_f16(__s2_885, __p3_885), vgetq_lane_f16(__s2_885, __p3_885), vgetq_lane_f16(__s2_885, __p3_885), vgetq_lane_f16(__s2_885, __p3_885)}); \
70370	72696	__ret_885; \
70371	72697	})
70372	72698	#else
70373		#define vfmlslq_laneq_low_f16(__p0_886, __p1_886, __p2_886, __p3_886) __extension__ ({ \
	72699	#define vfmlalq_laneq_low_f16(__p0_886, __p1_886, __p2_886, __p3_886) __extension__ ({ \
70374	72700	float32x4_t __ret_886; \
70375	72701	float32x4_t __s0_886 = __p0_886; \
70376	72702	float16x8_t __s1_886 = __p1_886; \
70377	72703	float16x8_t __s2_886 = __p2_886; \
70378		float32x4_t __rev0_886; __rev0_886 = __builtin_shufflevector(__s0_886, __s0_886, 3, 2, 1, 0); \
70379		float16x8_t __rev1_886; __rev1_886 = __builtin_shufflevector(__s1_886, __s1_886, 7, 6, 5, 4, 3, 2, 1, 0); \
70380		float16x8_t __rev2_886; __rev2_886 = __builtin_shufflevector(__s2_886, __s2_886, 7, 6, 5, 4, 3, 2, 1, 0); \
70381		__ret_886 = __noswap_vfmlslq_low_f16(__rev0_886, __rev1_886, (float16x8_t) {__noswap_vgetq_lane_f16(__rev2_886, __p3_886), __noswap_vgetq_lane_f16(__rev2_886, __p3_886), __noswap_vgetq_lane_f16(__rev2_886, __p3_886), __noswap_vgetq_lane_f16(__rev2_886, __p3_886), __noswap_vgetq_lane_f16(__rev2_886, __p3_886), __noswap_vgetq_lane_f16(__rev2_886, __p3_886), __noswap_vgetq_lane_f16(__rev2_886, __p3_886), __noswap_vgetq_lane_f16(__rev2_886, __p3_886)}); \
70382		__ret_886 = __builtin_shufflevector(__ret_886, __ret_886, 3, 2, 1, 0); \
	72704	float32x4_t __rev0_886; __rev0_886 = __builtin_shufflevector(__s0_886, __s0_886, __lane_reverse_128_32); \
	72705	float16x8_t __rev1_886; __rev1_886 = __builtin_shufflevector(__s1_886, __s1_886, __lane_reverse_128_16); \
	72706	float16x8_t __rev2_886; __rev2_886 = __builtin_shufflevector(__s2_886, __s2_886, __lane_reverse_128_16); \
	72707	__ret_886 = __noswap_vfmlalq_low_f16(__rev0_886, __rev1_886, (float16x8_t) {__noswap_vgetq_lane_f16(__rev2_886, __p3_886), __noswap_vgetq_lane_f16(__rev2_886, __p3_886), __noswap_vgetq_lane_f16(__rev2_886, __p3_886), __noswap_vgetq_lane_f16(__rev2_886, __p3_886), __noswap_vgetq_lane_f16(__rev2_886, __p3_886), __noswap_vgetq_lane_f16(__rev2_886, __p3_886), __noswap_vgetq_lane_f16(__rev2_886, __p3_886), __noswap_vgetq_lane_f16(__rev2_886, __p3_886)}); \
	72708	__ret_886 = __builtin_shufflevector(__ret_886, __ret_886, __lane_reverse_128_32); \
70383	72709	__ret_886; \
70384	72710	})
70385	72711	#endif
70386	72712
70387	72713	#ifdef __LITTLE_ENDIAN__
70388		#define vfmlsl_laneq_low_f16(__p0_887, __p1_887, __p2_887, __p3_887) __extension__ ({ \
	72714	#define vfmlal_laneq_low_f16(__p0_887, __p1_887, __p2_887, __p3_887) __extension__ ({ \
70389	72715	float32x2_t __ret_887; \
70390	72716	float32x2_t __s0_887 = __p0_887; \
70391	72717	float16x4_t __s1_887 = __p1_887; \
70392	72718	float16x8_t __s2_887 = __p2_887; \
70393		__ret_887 = vfmlsl_low_f16(__s0_887, __s1_887, (float16x4_t) {vgetq_lane_f16(__s2_887, __p3_887), vgetq_lane_f16(__s2_887, __p3_887), vgetq_lane_f16(__s2_887, __p3_887), vgetq_lane_f16(__s2_887, __p3_887)}); \
	72719	__ret_887 = vfmlal_low_f16(__s0_887, __s1_887, (float16x4_t) {vgetq_lane_f16(__s2_887, __p3_887), vgetq_lane_f16(__s2_887, __p3_887), vgetq_lane_f16(__s2_887, __p3_887), vgetq_lane_f16(__s2_887, __p3_887)}); \
70394	72720	__ret_887; \
70395	72721	})
70396	72722	#else
70397		#define vfmlsl_laneq_low_f16(__p0_888, __p1_888, __p2_888, __p3_888) __extension__ ({ \
	72723	#define vfmlal_laneq_low_f16(__p0_888, __p1_888, __p2_888, __p3_888) __extension__ ({ \
70398	72724	float32x2_t __ret_888; \
70399	72725	float32x2_t __s0_888 = __p0_888; \
70400	72726	float16x4_t __s1_888 = __p1_888; \
70401	72727	float16x8_t __s2_888 = __p2_888; \
70402		float32x2_t __rev0_888; __rev0_888 = __builtin_shufflevector(__s0_888, __s0_888, 1, 0); \
70403		float16x4_t __rev1_888; __rev1_888 = __builtin_shufflevector(__s1_888, __s1_888, 3, 2, 1, 0); \
70404		float16x8_t __rev2_888; __rev2_888 = __builtin_shufflevector(__s2_888, __s2_888, 7, 6, 5, 4, 3, 2, 1, 0); \
70405		__ret_888 = __noswap_vfmlsl_low_f16(__rev0_888, __rev1_888, (float16x4_t) {__noswap_vgetq_lane_f16(__rev2_888, __p3_888), __noswap_vgetq_lane_f16(__rev2_888, __p3_888), __noswap_vgetq_lane_f16(__rev2_888, __p3_888), __noswap_vgetq_lane_f16(__rev2_888, __p3_888)}); \
70406		__ret_888 = __builtin_shufflevector(__ret_888, __ret_888, 1, 0); \
	72728	float32x2_t __rev0_888; __rev0_888 = __builtin_shufflevector(__s0_888, __s0_888, __lane_reverse_64_32); \
	72729	float16x4_t __rev1_888; __rev1_888 = __builtin_shufflevector(__s1_888, __s1_888, __lane_reverse_64_16); \
	72730	float16x8_t __rev2_888; __rev2_888 = __builtin_shufflevector(__s2_888, __s2_888, __lane_reverse_128_16); \
	72731	__ret_888 = __noswap_vfmlal_low_f16(__rev0_888, __rev1_888, (float16x4_t) {__noswap_vgetq_lane_f16(__rev2_888, __p3_888), __noswap_vgetq_lane_f16(__rev2_888, __p3_888), __noswap_vgetq_lane_f16(__rev2_888, __p3_888), __noswap_vgetq_lane_f16(__rev2_888, __p3_888)}); \
	72732	__ret_888 = __builtin_shufflevector(__ret_888, __ret_888, __lane_reverse_64_32); \
70407	72733	__ret_888; \
70408	72734	})
70409	72735	#endif
70410	72736
70411	72737	#ifdef __LITTLE_ENDIAN__
70412		#define vmulh_lane_f16(__p0_889, __p1_889, __p2_889) __extension__ ({ \
70413		float16_t __ret_889; \
70414		float16_t __s0_889 = __p0_889; \
70415		float16x4_t __s1_889 = __p1_889; \
70416		__ret_889 = __s0_889 * vget_lane_f16(__s1_889, __p2_889); \
	72738	#define vfmlslq_lane_high_f16(__p0_889, __p1_889, __p2_889, __p3_889) __extension__ ({ \
	72739	float32x4_t __ret_889; \
	72740	float32x4_t __s0_889 = __p0_889; \
	72741	float16x8_t __s1_889 = __p1_889; \
	72742	float16x4_t __s2_889 = __p2_889; \
	72743	__ret_889 = vfmlslq_high_f16(__s0_889, __s1_889, (float16x8_t) {vget_lane_f16(__s2_889, __p3_889), vget_lane_f16(__s2_889, __p3_889), vget_lane_f16(__s2_889, __p3_889), vget_lane_f16(__s2_889, __p3_889), vget_lane_f16(__s2_889, __p3_889), vget_lane_f16(__s2_889, __p3_889), vget_lane_f16(__s2_889, __p3_889), vget_lane_f16(__s2_889, __p3_889)}); \
70417	72744	__ret_889; \
70418	72745	})
70419	72746	#else
70420		#define vmulh_lane_f16(__p0_890, __p1_890, __p2_890) __extension__ ({ \
70421		float16_t __ret_890; \
70422		float16_t __s0_890 = __p0_890; \
70423		float16x4_t __s1_890 = __p1_890; \
70424		float16x4_t __rev1_890; __rev1_890 = __builtin_shufflevector(__s1_890, __s1_890, 3, 2, 1, 0); \
70425		__ret_890 = __s0_890 * __noswap_vget_lane_f16(__rev1_890, __p2_890); \
	72747	#define vfmlslq_lane_high_f16(__p0_890, __p1_890, __p2_890, __p3_890) __extension__ ({ \
	72748	float32x4_t __ret_890; \
	72749	float32x4_t __s0_890 = __p0_890; \
	72750	float16x8_t __s1_890 = __p1_890; \
	72751	float16x4_t __s2_890 = __p2_890; \
	72752	float32x4_t __rev0_890; __rev0_890 = __builtin_shufflevector(__s0_890, __s0_890, __lane_reverse_128_32); \
	72753	float16x8_t __rev1_890; __rev1_890 = __builtin_shufflevector(__s1_890, __s1_890, __lane_reverse_128_16); \
	72754	float16x4_t __rev2_890; __rev2_890 = __builtin_shufflevector(__s2_890, __s2_890, __lane_reverse_64_16); \
	72755	__ret_890 = __noswap_vfmlslq_high_f16(__rev0_890, __rev1_890, (float16x8_t) {__noswap_vget_lane_f16(__rev2_890, __p3_890), __noswap_vget_lane_f16(__rev2_890, __p3_890), __noswap_vget_lane_f16(__rev2_890, __p3_890), __noswap_vget_lane_f16(__rev2_890, __p3_890), __noswap_vget_lane_f16(__rev2_890, __p3_890), __noswap_vget_lane_f16(__rev2_890, __p3_890), __noswap_vget_lane_f16(__rev2_890, __p3_890), __noswap_vget_lane_f16(__rev2_890, __p3_890)}); \
	72756	__ret_890 = __builtin_shufflevector(__ret_890, __ret_890, __lane_reverse_128_32); \
70426	72757	__ret_890; \
70427	72758	})
70428	72759	#endif
70429	72760
70430	72761	#ifdef __LITTLE_ENDIAN__
70431		#define vmulh_laneq_f16(__p0_891, __p1_891, __p2_891) __extension__ ({ \
70432		float16_t __ret_891; \
70433		float16_t __s0_891 = __p0_891; \
70434		float16x8_t __s1_891 = __p1_891; \
70435		__ret_891 = __s0_891 * vgetq_lane_f16(__s1_891, __p2_891); \
	72762	#define vfmlsl_lane_high_f16(__p0_891, __p1_891, __p2_891, __p3_891) __extension__ ({ \
	72763	float32x2_t __ret_891; \
	72764	float32x2_t __s0_891 = __p0_891; \
	72765	float16x4_t __s1_891 = __p1_891; \
	72766	float16x4_t __s2_891 = __p2_891; \
	72767	__ret_891 = vfmlsl_high_f16(__s0_891, __s1_891, (float16x4_t) {vget_lane_f16(__s2_891, __p3_891), vget_lane_f16(__s2_891, __p3_891), vget_lane_f16(__s2_891, __p3_891), vget_lane_f16(__s2_891, __p3_891)}); \
70436	72768	__ret_891; \
70437	72769	})
70438	72770	#else
70439		#define vmulh_laneq_f16(__p0_892, __p1_892, __p2_892) __extension__ ({ \
70440		float16_t __ret_892; \
70441		float16_t __s0_892 = __p0_892; \
70442		float16x8_t __s1_892 = __p1_892; \
70443		float16x8_t __rev1_892; __rev1_892 = __builtin_shufflevector(__s1_892, __s1_892, 7, 6, 5, 4, 3, 2, 1, 0); \
70444		__ret_892 = __s0_892 * __noswap_vgetq_lane_f16(__rev1_892, __p2_892); \
	72771	#define vfmlsl_lane_high_f16(__p0_892, __p1_892, __p2_892, __p3_892) __extension__ ({ \
	72772	float32x2_t __ret_892; \
	72773	float32x2_t __s0_892 = __p0_892; \
	72774	float16x4_t __s1_892 = __p1_892; \
	72775	float16x4_t __s2_892 = __p2_892; \
	72776	float32x2_t __rev0_892; __rev0_892 = __builtin_shufflevector(__s0_892, __s0_892, __lane_reverse_64_32); \
	72777	float16x4_t __rev1_892; __rev1_892 = __builtin_shufflevector(__s1_892, __s1_892, __lane_reverse_64_16); \
	72778	float16x4_t __rev2_892; __rev2_892 = __builtin_shufflevector(__s2_892, __s2_892, __lane_reverse_64_16); \
	72779	__ret_892 = __noswap_vfmlsl_high_f16(__rev0_892, __rev1_892, (float16x4_t) {__noswap_vget_lane_f16(__rev2_892, __p3_892), __noswap_vget_lane_f16(__rev2_892, __p3_892), __noswap_vget_lane_f16(__rev2_892, __p3_892), __noswap_vget_lane_f16(__rev2_892, __p3_892)}); \
	72780	__ret_892 = __builtin_shufflevector(__ret_892, __ret_892, __lane_reverse_64_32); \
70445	72781	__ret_892; \
70446	72782	})
70447	72783	#endif
70448	72784
	72785	#ifdef __LITTLE_ENDIAN__
	72786	#define vfmlslq_lane_low_f16(__p0_893, __p1_893, __p2_893, __p3_893) __extension__ ({ \
	72787	float32x4_t __ret_893; \
	72788	float32x4_t __s0_893 = __p0_893; \
	72789	float16x8_t __s1_893 = __p1_893; \
	72790	float16x4_t __s2_893 = __p2_893; \
	72791	__ret_893 = vfmlslq_low_f16(__s0_893, __s1_893, (float16x8_t) {vget_lane_f16(__s2_893, __p3_893), vget_lane_f16(__s2_893, __p3_893), vget_lane_f16(__s2_893, __p3_893), vget_lane_f16(__s2_893, __p3_893), vget_lane_f16(__s2_893, __p3_893), vget_lane_f16(__s2_893, __p3_893), vget_lane_f16(__s2_893, __p3_893), vget_lane_f16(__s2_893, __p3_893)}); \
	72792	__ret_893; \
	72793	})
	72794	#else
	72795	#define vfmlslq_lane_low_f16(__p0_894, __p1_894, __p2_894, __p3_894) __extension__ ({ \
	72796	float32x4_t __ret_894; \
	72797	float32x4_t __s0_894 = __p0_894; \
	72798	float16x8_t __s1_894 = __p1_894; \
	72799	float16x4_t __s2_894 = __p2_894; \
	72800	float32x4_t __rev0_894; __rev0_894 = __builtin_shufflevector(__s0_894, __s0_894, __lane_reverse_128_32); \
	72801	float16x8_t __rev1_894; __rev1_894 = __builtin_shufflevector(__s1_894, __s1_894, __lane_reverse_128_16); \
	72802	float16x4_t __rev2_894; __rev2_894 = __builtin_shufflevector(__s2_894, __s2_894, __lane_reverse_64_16); \
	72803	__ret_894 = __noswap_vfmlslq_low_f16(__rev0_894, __rev1_894, (float16x8_t) {__noswap_vget_lane_f16(__rev2_894, __p3_894), __noswap_vget_lane_f16(__rev2_894, __p3_894), __noswap_vget_lane_f16(__rev2_894, __p3_894), __noswap_vget_lane_f16(__rev2_894, __p3_894), __noswap_vget_lane_f16(__rev2_894, __p3_894), __noswap_vget_lane_f16(__rev2_894, __p3_894), __noswap_vget_lane_f16(__rev2_894, __p3_894), __noswap_vget_lane_f16(__rev2_894, __p3_894)}); \
	72804	__ret_894 = __builtin_shufflevector(__ret_894, __ret_894, __lane_reverse_128_32); \
	72805	__ret_894; \
	72806	})
	72807	#endif
	72808
	72809	#ifdef __LITTLE_ENDIAN__
	72810	#define vfmlsl_lane_low_f16(__p0_895, __p1_895, __p2_895, __p3_895) __extension__ ({ \
	72811	float32x2_t __ret_895; \
	72812	float32x2_t __s0_895 = __p0_895; \
	72813	float16x4_t __s1_895 = __p1_895; \
	72814	float16x4_t __s2_895 = __p2_895; \
	72815	__ret_895 = vfmlsl_low_f16(__s0_895, __s1_895, (float16x4_t) {vget_lane_f16(__s2_895, __p3_895), vget_lane_f16(__s2_895, __p3_895), vget_lane_f16(__s2_895, __p3_895), vget_lane_f16(__s2_895, __p3_895)}); \
	72816	__ret_895; \
	72817	})
	72818	#else
	72819	#define vfmlsl_lane_low_f16(__p0_896, __p1_896, __p2_896, __p3_896) __extension__ ({ \
	72820	float32x2_t __ret_896; \
	72821	float32x2_t __s0_896 = __p0_896; \
	72822	float16x4_t __s1_896 = __p1_896; \
	72823	float16x4_t __s2_896 = __p2_896; \
	72824	float32x2_t __rev0_896; __rev0_896 = __builtin_shufflevector(__s0_896, __s0_896, __lane_reverse_64_32); \
	72825	float16x4_t __rev1_896; __rev1_896 = __builtin_shufflevector(__s1_896, __s1_896, __lane_reverse_64_16); \
	72826	float16x4_t __rev2_896; __rev2_896 = __builtin_shufflevector(__s2_896, __s2_896, __lane_reverse_64_16); \
	72827	__ret_896 = __noswap_vfmlsl_low_f16(__rev0_896, __rev1_896, (float16x4_t) {__noswap_vget_lane_f16(__rev2_896, __p3_896), __noswap_vget_lane_f16(__rev2_896, __p3_896), __noswap_vget_lane_f16(__rev2_896, __p3_896), __noswap_vget_lane_f16(__rev2_896, __p3_896)}); \
	72828	__ret_896 = __builtin_shufflevector(__ret_896, __ret_896, __lane_reverse_64_32); \
	72829	__ret_896; \
	72830	})
	72831	#endif
	72832
	72833	#ifdef __LITTLE_ENDIAN__
	72834	#define vfmlslq_laneq_high_f16(__p0_897, __p1_897, __p2_897, __p3_897) __extension__ ({ \
	72835	float32x4_t __ret_897; \
	72836	float32x4_t __s0_897 = __p0_897; \
	72837	float16x8_t __s1_897 = __p1_897; \
	72838	float16x8_t __s2_897 = __p2_897; \
	72839	__ret_897 = vfmlslq_high_f16(__s0_897, __s1_897, (float16x8_t) {vgetq_lane_f16(__s2_897, __p3_897), vgetq_lane_f16(__s2_897, __p3_897), vgetq_lane_f16(__s2_897, __p3_897), vgetq_lane_f16(__s2_897, __p3_897), vgetq_lane_f16(__s2_897, __p3_897), vgetq_lane_f16(__s2_897, __p3_897), vgetq_lane_f16(__s2_897, __p3_897), vgetq_lane_f16(__s2_897, __p3_897)}); \
	72840	__ret_897; \
	72841	})
	72842	#else
	72843	#define vfmlslq_laneq_high_f16(__p0_898, __p1_898, __p2_898, __p3_898) __extension__ ({ \
	72844	float32x4_t __ret_898; \
	72845	float32x4_t __s0_898 = __p0_898; \
	72846	float16x8_t __s1_898 = __p1_898; \
	72847	float16x8_t __s2_898 = __p2_898; \
	72848	float32x4_t __rev0_898; __rev0_898 = __builtin_shufflevector(__s0_898, __s0_898, __lane_reverse_128_32); \
	72849	float16x8_t __rev1_898; __rev1_898 = __builtin_shufflevector(__s1_898, __s1_898, __lane_reverse_128_16); \
	72850	float16x8_t __rev2_898; __rev2_898 = __builtin_shufflevector(__s2_898, __s2_898, __lane_reverse_128_16); \
	72851	__ret_898 = __noswap_vfmlslq_high_f16(__rev0_898, __rev1_898, (float16x8_t) {__noswap_vgetq_lane_f16(__rev2_898, __p3_898), __noswap_vgetq_lane_f16(__rev2_898, __p3_898), __noswap_vgetq_lane_f16(__rev2_898, __p3_898), __noswap_vgetq_lane_f16(__rev2_898, __p3_898), __noswap_vgetq_lane_f16(__rev2_898, __p3_898), __noswap_vgetq_lane_f16(__rev2_898, __p3_898), __noswap_vgetq_lane_f16(__rev2_898, __p3_898), __noswap_vgetq_lane_f16(__rev2_898, __p3_898)}); \
	72852	__ret_898 = __builtin_shufflevector(__ret_898, __ret_898, __lane_reverse_128_32); \
	72853	__ret_898; \
	72854	})
	72855	#endif
	72856
	72857	#ifdef __LITTLE_ENDIAN__
	72858	#define vfmlsl_laneq_high_f16(__p0_899, __p1_899, __p2_899, __p3_899) __extension__ ({ \
	72859	float32x2_t __ret_899; \
	72860	float32x2_t __s0_899 = __p0_899; \
	72861	float16x4_t __s1_899 = __p1_899; \
	72862	float16x8_t __s2_899 = __p2_899; \
	72863	__ret_899 = vfmlsl_high_f16(__s0_899, __s1_899, (float16x4_t) {vgetq_lane_f16(__s2_899, __p3_899), vgetq_lane_f16(__s2_899, __p3_899), vgetq_lane_f16(__s2_899, __p3_899), vgetq_lane_f16(__s2_899, __p3_899)}); \
	72864	__ret_899; \
	72865	})
	72866	#else
	72867	#define vfmlsl_laneq_high_f16(__p0_900, __p1_900, __p2_900, __p3_900) __extension__ ({ \
	72868	float32x2_t __ret_900; \
	72869	float32x2_t __s0_900 = __p0_900; \
	72870	float16x4_t __s1_900 = __p1_900; \
	72871	float16x8_t __s2_900 = __p2_900; \
	72872	float32x2_t __rev0_900; __rev0_900 = __builtin_shufflevector(__s0_900, __s0_900, __lane_reverse_64_32); \
	72873	float16x4_t __rev1_900; __rev1_900 = __builtin_shufflevector(__s1_900, __s1_900, __lane_reverse_64_16); \
	72874	float16x8_t __rev2_900; __rev2_900 = __builtin_shufflevector(__s2_900, __s2_900, __lane_reverse_128_16); \
	72875	__ret_900 = __noswap_vfmlsl_high_f16(__rev0_900, __rev1_900, (float16x4_t) {__noswap_vgetq_lane_f16(__rev2_900, __p3_900), __noswap_vgetq_lane_f16(__rev2_900, __p3_900), __noswap_vgetq_lane_f16(__rev2_900, __p3_900), __noswap_vgetq_lane_f16(__rev2_900, __p3_900)}); \
	72876	__ret_900 = __builtin_shufflevector(__ret_900, __ret_900, __lane_reverse_64_32); \
	72877	__ret_900; \
	72878	})
	72879	#endif
	72880
	72881	#ifdef __LITTLE_ENDIAN__
	72882	#define vfmlslq_laneq_low_f16(__p0_901, __p1_901, __p2_901, __p3_901) __extension__ ({ \
	72883	float32x4_t __ret_901; \
	72884	float32x4_t __s0_901 = __p0_901; \
	72885	float16x8_t __s1_901 = __p1_901; \
	72886	float16x8_t __s2_901 = __p2_901; \
	72887	__ret_901 = vfmlslq_low_f16(__s0_901, __s1_901, (float16x8_t) {vgetq_lane_f16(__s2_901, __p3_901), vgetq_lane_f16(__s2_901, __p3_901), vgetq_lane_f16(__s2_901, __p3_901), vgetq_lane_f16(__s2_901, __p3_901), vgetq_lane_f16(__s2_901, __p3_901), vgetq_lane_f16(__s2_901, __p3_901), vgetq_lane_f16(__s2_901, __p3_901), vgetq_lane_f16(__s2_901, __p3_901)}); \
	72888	__ret_901; \
	72889	})
	72890	#else
	72891	#define vfmlslq_laneq_low_f16(__p0_902, __p1_902, __p2_902, __p3_902) __extension__ ({ \
	72892	float32x4_t __ret_902; \
	72893	float32x4_t __s0_902 = __p0_902; \
	72894	float16x8_t __s1_902 = __p1_902; \
	72895	float16x8_t __s2_902 = __p2_902; \
	72896	float32x4_t __rev0_902; __rev0_902 = __builtin_shufflevector(__s0_902, __s0_902, __lane_reverse_128_32); \
	72897	float16x8_t __rev1_902; __rev1_902 = __builtin_shufflevector(__s1_902, __s1_902, __lane_reverse_128_16); \
	72898	float16x8_t __rev2_902; __rev2_902 = __builtin_shufflevector(__s2_902, __s2_902, __lane_reverse_128_16); \
	72899	__ret_902 = __noswap_vfmlslq_low_f16(__rev0_902, __rev1_902, (float16x8_t) {__noswap_vgetq_lane_f16(__rev2_902, __p3_902), __noswap_vgetq_lane_f16(__rev2_902, __p3_902), __noswap_vgetq_lane_f16(__rev2_902, __p3_902), __noswap_vgetq_lane_f16(__rev2_902, __p3_902), __noswap_vgetq_lane_f16(__rev2_902, __p3_902), __noswap_vgetq_lane_f16(__rev2_902, __p3_902), __noswap_vgetq_lane_f16(__rev2_902, __p3_902), __noswap_vgetq_lane_f16(__rev2_902, __p3_902)}); \
	72900	__ret_902 = __builtin_shufflevector(__ret_902, __ret_902, __lane_reverse_128_32); \
	72901	__ret_902; \
	72902	})
	72903	#endif
	72904
	72905	#ifdef __LITTLE_ENDIAN__
	72906	#define vfmlsl_laneq_low_f16(__p0_903, __p1_903, __p2_903, __p3_903) __extension__ ({ \
	72907	float32x2_t __ret_903; \
	72908	float32x2_t __s0_903 = __p0_903; \
	72909	float16x4_t __s1_903 = __p1_903; \
	72910	float16x8_t __s2_903 = __p2_903; \
	72911	__ret_903 = vfmlsl_low_f16(__s0_903, __s1_903, (float16x4_t) {vgetq_lane_f16(__s2_903, __p3_903), vgetq_lane_f16(__s2_903, __p3_903), vgetq_lane_f16(__s2_903, __p3_903), vgetq_lane_f16(__s2_903, __p3_903)}); \
	72912	__ret_903; \
	72913	})
	72914	#else
	72915	#define vfmlsl_laneq_low_f16(__p0_904, __p1_904, __p2_904, __p3_904) __extension__ ({ \
	72916	float32x2_t __ret_904; \
	72917	float32x2_t __s0_904 = __p0_904; \
	72918	float16x4_t __s1_904 = __p1_904; \
	72919	float16x8_t __s2_904 = __p2_904; \
	72920	float32x2_t __rev0_904; __rev0_904 = __builtin_shufflevector(__s0_904, __s0_904, __lane_reverse_64_32); \
	72921	float16x4_t __rev1_904; __rev1_904 = __builtin_shufflevector(__s1_904, __s1_904, __lane_reverse_64_16); \
	72922	float16x8_t __rev2_904; __rev2_904 = __builtin_shufflevector(__s2_904, __s2_904, __lane_reverse_128_16); \
	72923	__ret_904 = __noswap_vfmlsl_low_f16(__rev0_904, __rev1_904, (float16x4_t) {__noswap_vgetq_lane_f16(__rev2_904, __p3_904), __noswap_vgetq_lane_f16(__rev2_904, __p3_904), __noswap_vgetq_lane_f16(__rev2_904, __p3_904), __noswap_vgetq_lane_f16(__rev2_904, __p3_904)}); \
	72924	__ret_904 = __builtin_shufflevector(__ret_904, __ret_904, __lane_reverse_64_32); \
	72925	__ret_904; \
	72926	})
	72927	#endif
	72928
	72929	#ifdef __LITTLE_ENDIAN__
	72930	#define vmulh_lane_f16(__p0_905, __p1_905, __p2_905) __extension__ ({ \
	72931	float16_t __ret_905; \
	72932	float16_t __s0_905 = __p0_905; \
	72933	float16x4_t __s1_905 = __p1_905; \
	72934	__ret_905 = __s0_905 * vget_lane_f16(__s1_905, __p2_905); \
	72935	__ret_905; \
	72936	})
	72937	#else
	72938	#define vmulh_lane_f16(__p0_906, __p1_906, __p2_906) __extension__ ({ \
	72939	float16_t __ret_906; \
	72940	float16_t __s0_906 = __p0_906; \
	72941	float16x4_t __s1_906 = __p1_906; \
	72942	float16x4_t __rev1_906; __rev1_906 = __builtin_shufflevector(__s1_906, __s1_906, __lane_reverse_64_16); \
	72943	__ret_906 = __s0_906 * __noswap_vget_lane_f16(__rev1_906, __p2_906); \
	72944	__ret_906; \
	72945	})
	72946	#endif
	72947
	72948	#ifdef __LITTLE_ENDIAN__
	72949	#define vmulh_laneq_f16(__p0_907, __p1_907, __p2_907) __extension__ ({ \
	72950	float16_t __ret_907; \
	72951	float16_t __s0_907 = __p0_907; \
	72952	float16x8_t __s1_907 = __p1_907; \
	72953	__ret_907 = __s0_907 * vgetq_lane_f16(__s1_907, __p2_907); \
	72954	__ret_907; \
	72955	})
	72956	#else
	72957	#define vmulh_laneq_f16(__p0_908, __p1_908, __p2_908) __extension__ ({ \
	72958	float16_t __ret_908; \
	72959	float16_t __s0_908 = __p0_908; \
	72960	float16x8_t __s1_908 = __p1_908; \
	72961	float16x8_t __rev1_908; __rev1_908 = __builtin_shufflevector(__s1_908, __s1_908, __lane_reverse_128_16); \
	72962	__ret_908 = __s0_908 * __noswap_vgetq_lane_f16(__rev1_908, __p2_908); \
	72963	__ret_908; \
	72964	})
	72965	#endif
	72966
70449	72967	#ifdef __LITTLE_ENDIAN__
70450	72968	__ai __attribute__((target("neon"))) uint16x8_t vabdl_high_u8(uint8x16_t __p0, uint8x16_t __p1) {
70451	72969	uint16x8_t __ret;
...	...	@@ -70455,10 +72973,10 @@ __ai __attribute__((target("neon"))) uint16x8_t vabdl_high_u8(uint8x16_t __p0, u
70455	72973	#else
70456	72974	__ai __attribute__((target("neon"))) uint16x8_t vabdl_high_u8(uint8x16_t __p0, uint8x16_t __p1) {
70457	72975	uint16x8_t __ret;
70458		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
70459		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	72976	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	72977	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
70460	72978	__ret = __noswap_vabdl_u8(__noswap_vget_high_u8(__rev0), __noswap_vget_high_u8(__rev1));
70461		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	72979	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
70462	72980	return __ret;
70463	72981	}
70464	72982	#endif
...	...	@@ -70472,10 +72990,10 @@ __ai __attribute__((target("neon"))) uint64x2_t vabdl_high_u32(uint32x4_t __p0,
70472	72990	#else
70473	72991	__ai __attribute__((target("neon"))) uint64x2_t vabdl_high_u32(uint32x4_t __p0, uint32x4_t __p1) {
70474	72992	uint64x2_t __ret;
70475		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
70476		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	72993	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	72994	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
70477	72995	__ret = __noswap_vabdl_u32(__noswap_vget_high_u32(__rev0), __noswap_vget_high_u32(__rev1));
70478		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	72996	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
70479	72997	return __ret;
70480	72998	}
70481	72999	#endif
...	...	@@ -70489,10 +73007,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vabdl_high_u16(uint16x8_t __p0,
70489	73007	#else
70490	73008	__ai __attribute__((target("neon"))) uint32x4_t vabdl_high_u16(uint16x8_t __p0, uint16x8_t __p1) {
70491	73009	uint32x4_t __ret;
70492		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
70493		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	73010	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	73011	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
70494	73012	__ret = __noswap_vabdl_u16(__noswap_vget_high_u16(__rev0), __noswap_vget_high_u16(__rev1));
70495		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	73013	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
70496	73014	return __ret;
70497	73015	}
70498	73016	#endif
...	...	@@ -70506,10 +73024,10 @@ __ai __attribute__((target("neon"))) int16x8_t vabdl_high_s8(int8x16_t __p0, int
70506	73024	#else
70507	73025	__ai __attribute__((target("neon"))) int16x8_t vabdl_high_s8(int8x16_t __p0, int8x16_t __p1) {
70508	73026	int16x8_t __ret;
70509		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
70510		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	73027	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	73028	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
70511	73029	__ret = __noswap_vabdl_s8(__noswap_vget_high_s8(__rev0), __noswap_vget_high_s8(__rev1));
70512		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	73030	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
70513	73031	return __ret;
70514	73032	}
70515	73033	#endif
...	...	@@ -70523,10 +73041,10 @@ __ai __attribute__((target("neon"))) int64x2_t vabdl_high_s32(int32x4_t __p0, in
70523	73041	#else
70524	73042	__ai __attribute__((target("neon"))) int64x2_t vabdl_high_s32(int32x4_t __p0, int32x4_t __p1) {
70525	73043	int64x2_t __ret;
70526		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
70527		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	73044	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	73045	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
70528	73046	__ret = __noswap_vabdl_s32(__noswap_vget_high_s32(__rev0), __noswap_vget_high_s32(__rev1));
70529		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	73047	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
70530	73048	return __ret;
70531	73049	}
70532	73050	#endif
...	...	@@ -70540,10 +73058,10 @@ __ai __attribute__((target("neon"))) int32x4_t vabdl_high_s16(int16x8_t __p0, in
70540	73058	#else
70541	73059	__ai __attribute__((target("neon"))) int32x4_t vabdl_high_s16(int16x8_t __p0, int16x8_t __p1) {
70542	73060	int32x4_t __ret;
70543		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
70544		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	73061	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	73062	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
70545	73063	__ret = __noswap_vabdl_s16(__noswap_vget_high_s16(__rev0), __noswap_vget_high_s16(__rev1));
70546		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	73064	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
70547	73065	return __ret;
70548	73066	}
70549	73067	#endif
...	...	@@ -70557,10 +73075,10 @@ __ai __attribute__((target("neon"))) uint16x8_t vaddl_high_u8(uint8x16_t __p0, u
70557	73075	#else
70558	73076	__ai __attribute__((target("neon"))) uint16x8_t vaddl_high_u8(uint8x16_t __p0, uint8x16_t __p1) {
70559	73077	uint16x8_t __ret;
70560		uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
70561		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	73078	uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	73079	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
70562	73080	__ret = __noswap_vmovl_high_u8(__rev0) + __noswap_vmovl_high_u8(__rev1);
70563		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	73081	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
70564	73082	return __ret;
70565	73083	}
70566	73084	#endif
...	...	@@ -70574,10 +73092,10 @@ __ai __attribute__((target("neon"))) uint64x2_t vaddl_high_u32(uint32x4_t __p0,
70574	73092	#else
70575	73093	__ai __attribute__((target("neon"))) uint64x2_t vaddl_high_u32(uint32x4_t __p0, uint32x4_t __p1) {
70576	73094	uint64x2_t __ret;
70577		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
70578		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	73095	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	73096	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
70579	73097	__ret = __noswap_vmovl_high_u32(__rev0) + __noswap_vmovl_high_u32(__rev1);
70580		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	73098	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
70581	73099	return __ret;
70582	73100	}
70583	73101	#endif
...	...	@@ -70591,10 +73109,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vaddl_high_u16(uint16x8_t __p0,
70591	73109	#else
70592	73110	__ai __attribute__((target("neon"))) uint32x4_t vaddl_high_u16(uint16x8_t __p0, uint16x8_t __p1) {
70593	73111	uint32x4_t __ret;
70594		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
70595		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	73112	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	73113	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
70596	73114	__ret = __noswap_vmovl_high_u16(__rev0) + __noswap_vmovl_high_u16(__rev1);
70597		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	73115	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
70598	73116	return __ret;
70599	73117	}
70600	73118	#endif
...	...	@@ -70608,10 +73126,10 @@ __ai __attribute__((target("neon"))) int16x8_t vaddl_high_s8(int8x16_t __p0, int
70608	73126	#else
70609	73127	__ai __attribute__((target("neon"))) int16x8_t vaddl_high_s8(int8x16_t __p0, int8x16_t __p1) {
70610	73128	int16x8_t __ret;
70611		int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
70612		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	73129	int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_8);
	73130	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
70613	73131	__ret = __noswap_vmovl_high_s8(__rev0) + __noswap_vmovl_high_s8(__rev1);
70614		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	73132	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
70615	73133	return __ret;
70616	73134	}
70617	73135	#endif
...	...	@@ -70625,10 +73143,10 @@ __ai __attribute__((target("neon"))) int64x2_t vaddl_high_s32(int32x4_t __p0, in
70625	73143	#else
70626	73144	__ai __attribute__((target("neon"))) int64x2_t vaddl_high_s32(int32x4_t __p0, int32x4_t __p1) {
70627	73145	int64x2_t __ret;
70628		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
70629		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	73146	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	73147	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
70630	73148	__ret = __noswap_vmovl_high_s32(__rev0) + __noswap_vmovl_high_s32(__rev1);
70631		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	73149	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
70632	73150	return __ret;
70633	73151	}
70634	73152	#endif
...	...	@@ -70642,10 +73160,10 @@ __ai __attribute__((target("neon"))) int32x4_t vaddl_high_s16(int16x8_t __p0, in
70642	73160	#else
70643	73161	__ai __attribute__((target("neon"))) int32x4_t vaddl_high_s16(int16x8_t __p0, int16x8_t __p1) {
70644	73162	int32x4_t __ret;
70645		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
70646		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	73163	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	73164	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
70647	73165	__ret = __noswap_vmovl_high_s16(__rev0) + __noswap_vmovl_high_s16(__rev1);
70648		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	73166	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
70649	73167	return __ret;
70650	73168	}
70651	73169	#endif
...	...	@@ -70659,10 +73177,10 @@ __ai __attribute__((target("neon"))) uint16x8_t vaddw_high_u8(uint16x8_t __p0, u
70659	73177	#else
70660	73178	__ai __attribute__((target("neon"))) uint16x8_t vaddw_high_u8(uint16x8_t __p0, uint8x16_t __p1) {
70661	73179	uint16x8_t __ret;
70662		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
70663		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	73180	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	73181	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
70664	73182	__ret = __rev0 + __noswap_vmovl_high_u8(__rev1);
70665		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	73183	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
70666	73184	return __ret;
70667	73185	}
70668	73186	#endif
...	...	@@ -70676,10 +73194,10 @@ __ai __attribute__((target("neon"))) uint64x2_t vaddw_high_u32(uint64x2_t __p0,
70676	73194	#else
70677	73195	__ai __attribute__((target("neon"))) uint64x2_t vaddw_high_u32(uint64x2_t __p0, uint32x4_t __p1) {
70678	73196	uint64x2_t __ret;
70679		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
70680		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	73197	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	73198	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
70681	73199	__ret = __rev0 + __noswap_vmovl_high_u32(__rev1);
70682		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	73200	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
70683	73201	return __ret;
70684	73202	}
70685	73203	#endif
...	...	@@ -70693,10 +73211,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vaddw_high_u16(uint32x4_t __p0,
70693	73211	#else
70694	73212	__ai __attribute__((target("neon"))) uint32x4_t vaddw_high_u16(uint32x4_t __p0, uint16x8_t __p1) {
70695	73213	uint32x4_t __ret;
70696		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
70697		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	73214	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	73215	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
70698	73216	__ret = __rev0 + __noswap_vmovl_high_u16(__rev1);
70699		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	73217	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
70700	73218	return __ret;
70701	73219	}
70702	73220	#endif
...	...	@@ -70710,10 +73228,10 @@ __ai __attribute__((target("neon"))) int16x8_t vaddw_high_s8(int16x8_t __p0, int
70710	73228	#else
70711	73229	__ai __attribute__((target("neon"))) int16x8_t vaddw_high_s8(int16x8_t __p0, int8x16_t __p1) {
70712	73230	int16x8_t __ret;
70713		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
70714		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	73231	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	73232	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
70715	73233	__ret = __rev0 + __noswap_vmovl_high_s8(__rev1);
70716		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	73234	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
70717	73235	return __ret;
70718	73236	}
70719	73237	#endif
...	...	@@ -70727,10 +73245,10 @@ __ai __attribute__((target("neon"))) int64x2_t vaddw_high_s32(int64x2_t __p0, in
70727	73245	#else
70728	73246	__ai __attribute__((target("neon"))) int64x2_t vaddw_high_s32(int64x2_t __p0, int32x4_t __p1) {
70729	73247	int64x2_t __ret;
70730		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
70731		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	73248	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	73249	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
70732	73250	__ret = __rev0 + __noswap_vmovl_high_s32(__rev1);
70733		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	73251	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
70734	73252	return __ret;
70735	73253	}
70736	73254	#endif
...	...	@@ -70744,145 +73262,145 @@ __ai __attribute__((target("neon"))) int32x4_t vaddw_high_s16(int32x4_t __p0, in
70744	73262	#else
70745	73263	__ai __attribute__((target("neon"))) int32x4_t vaddw_high_s16(int32x4_t __p0, int16x8_t __p1) {
70746	73264	int32x4_t __ret;
70747		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
70748		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	73265	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	73266	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
70749	73267	__ret = __rev0 + __noswap_vmovl_high_s16(__rev1);
70750		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	73268	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
70751	73269	return __ret;
70752	73270	}
70753	73271	#endif
70754	73272
70755	73273	#ifdef __LITTLE_ENDIAN__
70756		#define vcopyq_lane_p64(__p0_893, __p1_893, __p2_893, __p3_893) __extension__ ({ \
70757		poly64x2_t __ret_893; \
70758		poly64x2_t __s0_893 = __p0_893; \
70759		poly64x1_t __s2_893 = __p2_893; \
70760		__ret_893 = vsetq_lane_p64(vget_lane_p64(__s2_893, __p3_893), __s0_893, __p1_893); \
70761		__ret_893; \
	73274	#define vcopyq_lane_p64(__p0_909, __p1_909, __p2_909, __p3_909) __extension__ ({ \
	73275	poly64x2_t __ret_909; \
	73276	poly64x2_t __s0_909 = __p0_909; \
	73277	poly64x1_t __s2_909 = __p2_909; \
	73278	__ret_909 = vsetq_lane_p64(vget_lane_p64(__s2_909, __p3_909), __s0_909, __p1_909); \
	73279	__ret_909; \
70762	73280	})
70763	73281	#else
70764		#define vcopyq_lane_p64(__p0_894, __p1_894, __p2_894, __p3_894) __extension__ ({ \
70765		poly64x2_t __ret_894; \
70766		poly64x2_t __s0_894 = __p0_894; \
70767		poly64x1_t __s2_894 = __p2_894; \
70768		poly64x2_t __rev0_894; __rev0_894 = __builtin_shufflevector(__s0_894, __s0_894, 1, 0); \
70769		__ret_894 = __noswap_vsetq_lane_p64(vget_lane_p64(__s2_894, __p3_894), __rev0_894, __p1_894); \
70770		__ret_894 = __builtin_shufflevector(__ret_894, __ret_894, 1, 0); \
70771		__ret_894; \
	73282	#define vcopyq_lane_p64(__p0_910, __p1_910, __p2_910, __p3_910) __extension__ ({ \
	73283	poly64x2_t __ret_910; \
	73284	poly64x2_t __s0_910 = __p0_910; \
	73285	poly64x1_t __s2_910 = __p2_910; \
	73286	poly64x2_t __rev0_910; __rev0_910 = __builtin_shufflevector(__s0_910, __s0_910, __lane_reverse_128_64); \
	73287	__ret_910 = __noswap_vsetq_lane_p64(vget_lane_p64(__s2_910, __p3_910), __rev0_910, __p1_910); \
	73288	__ret_910 = __builtin_shufflevector(__ret_910, __ret_910, __lane_reverse_128_64); \
	73289	__ret_910; \
70772	73290	})
70773	73291	#endif
70774	73292
70775	73293	#ifdef __LITTLE_ENDIAN__
70776		#define vcopyq_lane_f64(__p0_895, __p1_895, __p2_895, __p3_895) __extension__ ({ \
70777		float64x2_t __ret_895; \
70778		float64x2_t __s0_895 = __p0_895; \
70779		float64x1_t __s2_895 = __p2_895; \
70780		__ret_895 = vsetq_lane_f64(vget_lane_f64(__s2_895, __p3_895), __s0_895, __p1_895); \
70781		__ret_895; \
	73294	#define vcopyq_lane_f64(__p0_911, __p1_911, __p2_911, __p3_911) __extension__ ({ \
	73295	float64x2_t __ret_911; \
	73296	float64x2_t __s0_911 = __p0_911; \
	73297	float64x1_t __s2_911 = __p2_911; \
	73298	__ret_911 = vsetq_lane_f64(vget_lane_f64(__s2_911, __p3_911), __s0_911, __p1_911); \
	73299	__ret_911; \
70782	73300	})
70783	73301	#else
70784		#define vcopyq_lane_f64(__p0_896, __p1_896, __p2_896, __p3_896) __extension__ ({ \
70785		float64x2_t __ret_896; \
70786		float64x2_t __s0_896 = __p0_896; \
70787		float64x1_t __s2_896 = __p2_896; \
70788		float64x2_t __rev0_896; __rev0_896 = __builtin_shufflevector(__s0_896, __s0_896, 1, 0); \
70789		__ret_896 = __noswap_vsetq_lane_f64(vget_lane_f64(__s2_896, __p3_896), __rev0_896, __p1_896); \
70790		__ret_896 = __builtin_shufflevector(__ret_896, __ret_896, 1, 0); \
70791		__ret_896; \
	73302	#define vcopyq_lane_f64(__p0_912, __p1_912, __p2_912, __p3_912) __extension__ ({ \
	73303	float64x2_t __ret_912; \
	73304	float64x2_t __s0_912 = __p0_912; \
	73305	float64x1_t __s2_912 = __p2_912; \
	73306	float64x2_t __rev0_912; __rev0_912 = __builtin_shufflevector(__s0_912, __s0_912, __lane_reverse_128_64); \
	73307	__ret_912 = __noswap_vsetq_lane_f64(vget_lane_f64(__s2_912, __p3_912), __rev0_912, __p1_912); \
	73308	__ret_912 = __builtin_shufflevector(__ret_912, __ret_912, __lane_reverse_128_64); \
	73309	__ret_912; \
70792	73310	})
70793	73311	#endif
70794	73312
70795		#define vcopy_lane_p64(__p0_897, __p1_897, __p2_897, __p3_897) __extension__ ({ \
70796		poly64x1_t __ret_897; \
70797		poly64x1_t __s0_897 = __p0_897; \
70798		poly64x1_t __s2_897 = __p2_897; \
70799		__ret_897 = vset_lane_p64(vget_lane_p64(__s2_897, __p3_897), __s0_897, __p1_897); \
70800		__ret_897; \
	73313	#define vcopy_lane_p64(__p0_913, __p1_913, __p2_913, __p3_913) __extension__ ({ \
	73314	poly64x1_t __ret_913; \
	73315	poly64x1_t __s0_913 = __p0_913; \
	73316	poly64x1_t __s2_913 = __p2_913; \
	73317	__ret_913 = vset_lane_p64(vget_lane_p64(__s2_913, __p3_913), __s0_913, __p1_913); \
	73318	__ret_913; \
70801	73319	})
70802		#define vcopy_lane_f64(__p0_898, __p1_898, __p2_898, __p3_898) __extension__ ({ \
70803		float64x1_t __ret_898; \
70804		float64x1_t __s0_898 = __p0_898; \
70805		float64x1_t __s2_898 = __p2_898; \
70806		__ret_898 = vset_lane_f64(vget_lane_f64(__s2_898, __p3_898), __s0_898, __p1_898); \
70807		__ret_898; \
	73320	#define vcopy_lane_f64(__p0_914, __p1_914, __p2_914, __p3_914) __extension__ ({ \
	73321	float64x1_t __ret_914; \
	73322	float64x1_t __s0_914 = __p0_914; \
	73323	float64x1_t __s2_914 = __p2_914; \
	73324	__ret_914 = vset_lane_f64(vget_lane_f64(__s2_914, __p3_914), __s0_914, __p1_914); \
	73325	__ret_914; \
70808	73326	})
70809	73327	#ifdef __LITTLE_ENDIAN__
70810		#define vcopyq_laneq_p64(__p0_899, __p1_899, __p2_899, __p3_899) __extension__ ({ \
70811		poly64x2_t __ret_899; \
70812		poly64x2_t __s0_899 = __p0_899; \
70813		poly64x2_t __s2_899 = __p2_899; \
70814		__ret_899 = vsetq_lane_p64(vgetq_lane_p64(__s2_899, __p3_899), __s0_899, __p1_899); \
70815		__ret_899; \
	73328	#define vcopyq_laneq_p64(__p0_915, __p1_915, __p2_915, __p3_915) __extension__ ({ \
	73329	poly64x2_t __ret_915; \
	73330	poly64x2_t __s0_915 = __p0_915; \
	73331	poly64x2_t __s2_915 = __p2_915; \
	73332	__ret_915 = vsetq_lane_p64(vgetq_lane_p64(__s2_915, __p3_915), __s0_915, __p1_915); \
	73333	__ret_915; \
70816	73334	})
70817	73335	#else
70818		#define vcopyq_laneq_p64(__p0_900, __p1_900, __p2_900, __p3_900) __extension__ ({ \
70819		poly64x2_t __ret_900; \
70820		poly64x2_t __s0_900 = __p0_900; \
70821		poly64x2_t __s2_900 = __p2_900; \
70822		poly64x2_t __rev0_900; __rev0_900 = __builtin_shufflevector(__s0_900, __s0_900, 1, 0); \
70823		poly64x2_t __rev2_900; __rev2_900 = __builtin_shufflevector(__s2_900, __s2_900, 1, 0); \
70824		__ret_900 = __noswap_vsetq_lane_p64(__noswap_vgetq_lane_p64(__rev2_900, __p3_900), __rev0_900, __p1_900); \
70825		__ret_900 = __builtin_shufflevector(__ret_900, __ret_900, 1, 0); \
70826		__ret_900; \
	73336	#define vcopyq_laneq_p64(__p0_916, __p1_916, __p2_916, __p3_916) __extension__ ({ \
	73337	poly64x2_t __ret_916; \
	73338	poly64x2_t __s0_916 = __p0_916; \
	73339	poly64x2_t __s2_916 = __p2_916; \
	73340	poly64x2_t __rev0_916; __rev0_916 = __builtin_shufflevector(__s0_916, __s0_916, __lane_reverse_128_64); \
	73341	poly64x2_t __rev2_916; __rev2_916 = __builtin_shufflevector(__s2_916, __s2_916, __lane_reverse_128_64); \
	73342	__ret_916 = __noswap_vsetq_lane_p64(__noswap_vgetq_lane_p64(__rev2_916, __p3_916), __rev0_916, __p1_916); \
	73343	__ret_916 = __builtin_shufflevector(__ret_916, __ret_916, __lane_reverse_128_64); \
	73344	__ret_916; \
70827	73345	})
70828	73346	#endif
70829	73347
70830	73348	#ifdef __LITTLE_ENDIAN__
70831		#define vcopyq_laneq_f64(__p0_901, __p1_901, __p2_901, __p3_901) __extension__ ({ \
70832		float64x2_t __ret_901; \
70833		float64x2_t __s0_901 = __p0_901; \
70834		float64x2_t __s2_901 = __p2_901; \
70835		__ret_901 = vsetq_lane_f64(vgetq_lane_f64(__s2_901, __p3_901), __s0_901, __p1_901); \
70836		__ret_901; \
	73349	#define vcopyq_laneq_f64(__p0_917, __p1_917, __p2_917, __p3_917) __extension__ ({ \
	73350	float64x2_t __ret_917; \
	73351	float64x2_t __s0_917 = __p0_917; \
	73352	float64x2_t __s2_917 = __p2_917; \
	73353	__ret_917 = vsetq_lane_f64(vgetq_lane_f64(__s2_917, __p3_917), __s0_917, __p1_917); \
	73354	__ret_917; \
70837	73355	})
70838	73356	#else
70839		#define vcopyq_laneq_f64(__p0_902, __p1_902, __p2_902, __p3_902) __extension__ ({ \
70840		float64x2_t __ret_902; \
70841		float64x2_t __s0_902 = __p0_902; \
70842		float64x2_t __s2_902 = __p2_902; \
70843		float64x2_t __rev0_902; __rev0_902 = __builtin_shufflevector(__s0_902, __s0_902, 1, 0); \
70844		float64x2_t __rev2_902; __rev2_902 = __builtin_shufflevector(__s2_902, __s2_902, 1, 0); \
70845		__ret_902 = __noswap_vsetq_lane_f64(__noswap_vgetq_lane_f64(__rev2_902, __p3_902), __rev0_902, __p1_902); \
70846		__ret_902 = __builtin_shufflevector(__ret_902, __ret_902, 1, 0); \
70847		__ret_902; \
	73357	#define vcopyq_laneq_f64(__p0_918, __p1_918, __p2_918, __p3_918) __extension__ ({ \
	73358	float64x2_t __ret_918; \
	73359	float64x2_t __s0_918 = __p0_918; \
	73360	float64x2_t __s2_918 = __p2_918; \
	73361	float64x2_t __rev0_918; __rev0_918 = __builtin_shufflevector(__s0_918, __s0_918, __lane_reverse_128_64); \
	73362	float64x2_t __rev2_918; __rev2_918 = __builtin_shufflevector(__s2_918, __s2_918, __lane_reverse_128_64); \
	73363	__ret_918 = __noswap_vsetq_lane_f64(__noswap_vgetq_lane_f64(__rev2_918, __p3_918), __rev0_918, __p1_918); \
	73364	__ret_918 = __builtin_shufflevector(__ret_918, __ret_918, __lane_reverse_128_64); \
	73365	__ret_918; \
70848	73366	})
70849	73367	#endif
70850	73368
70851	73369	#ifdef __LITTLE_ENDIAN__
70852		#define vcopy_laneq_p64(__p0_903, __p1_903, __p2_903, __p3_903) __extension__ ({ \
70853		poly64x1_t __ret_903; \
70854		poly64x1_t __s0_903 = __p0_903; \
70855		poly64x2_t __s2_903 = __p2_903; \
70856		__ret_903 = vset_lane_p64(vgetq_lane_p64(__s2_903, __p3_903), __s0_903, __p1_903); \
70857		__ret_903; \
	73370	#define vcopy_laneq_p64(__p0_919, __p1_919, __p2_919, __p3_919) __extension__ ({ \
	73371	poly64x1_t __ret_919; \
	73372	poly64x1_t __s0_919 = __p0_919; \
	73373	poly64x2_t __s2_919 = __p2_919; \
	73374	__ret_919 = vset_lane_p64(vgetq_lane_p64(__s2_919, __p3_919), __s0_919, __p1_919); \
	73375	__ret_919; \
70858	73376	})
70859	73377	#else
70860		#define vcopy_laneq_p64(__p0_904, __p1_904, __p2_904, __p3_904) __extension__ ({ \
70861		poly64x1_t __ret_904; \
70862		poly64x1_t __s0_904 = __p0_904; \
70863		poly64x2_t __s2_904 = __p2_904; \
70864		poly64x2_t __rev2_904; __rev2_904 = __builtin_shufflevector(__s2_904, __s2_904, 1, 0); \
70865		__ret_904 = vset_lane_p64(__noswap_vgetq_lane_p64(__rev2_904, __p3_904), __s0_904, __p1_904); \
70866		__ret_904; \
	73378	#define vcopy_laneq_p64(__p0_920, __p1_920, __p2_920, __p3_920) __extension__ ({ \
	73379	poly64x1_t __ret_920; \
	73380	poly64x1_t __s0_920 = __p0_920; \
	73381	poly64x2_t __s2_920 = __p2_920; \
	73382	poly64x2_t __rev2_920; __rev2_920 = __builtin_shufflevector(__s2_920, __s2_920, __lane_reverse_128_64); \
	73383	__ret_920 = vset_lane_p64(__noswap_vgetq_lane_p64(__rev2_920, __p3_920), __s0_920, __p1_920); \
	73384	__ret_920; \
70867	73385	})
70868	73386	#endif
70869	73387
70870	73388	#ifdef __LITTLE_ENDIAN__
70871		#define vcopy_laneq_f64(__p0_905, __p1_905, __p2_905, __p3_905) __extension__ ({ \
70872		float64x1_t __ret_905; \
70873		float64x1_t __s0_905 = __p0_905; \
70874		float64x2_t __s2_905 = __p2_905; \
70875		__ret_905 = vset_lane_f64(vgetq_lane_f64(__s2_905, __p3_905), __s0_905, __p1_905); \
70876		__ret_905; \
	73389	#define vcopy_laneq_f64(__p0_921, __p1_921, __p2_921, __p3_921) __extension__ ({ \
	73390	float64x1_t __ret_921; \
	73391	float64x1_t __s0_921 = __p0_921; \
	73392	float64x2_t __s2_921 = __p2_921; \
	73393	__ret_921 = vset_lane_f64(vgetq_lane_f64(__s2_921, __p3_921), __s0_921, __p1_921); \
	73394	__ret_921; \
70877	73395	})
70878	73396	#else
70879		#define vcopy_laneq_f64(__p0_906, __p1_906, __p2_906, __p3_906) __extension__ ({ \
70880		float64x1_t __ret_906; \
70881		float64x1_t __s0_906 = __p0_906; \
70882		float64x2_t __s2_906 = __p2_906; \
70883		float64x2_t __rev2_906; __rev2_906 = __builtin_shufflevector(__s2_906, __s2_906, 1, 0); \
70884		__ret_906 = vset_lane_f64(__noswap_vgetq_lane_f64(__rev2_906, __p3_906), __s0_906, __p1_906); \
70885		__ret_906; \
	73397	#define vcopy_laneq_f64(__p0_922, __p1_922, __p2_922, __p3_922) __extension__ ({ \
	73398	float64x1_t __ret_922; \
	73399	float64x1_t __s0_922 = __p0_922; \
	73400	float64x2_t __s2_922 = __p2_922; \
	73401	float64x2_t __rev2_922; __rev2_922 = __builtin_shufflevector(__s2_922, __s2_922, __lane_reverse_128_64); \
	73402	__ret_922 = vset_lane_f64(__noswap_vgetq_lane_f64(__rev2_922, __p3_922), __s0_922, __p1_922); \
	73403	__ret_922; \
70886	73404	})
70887	73405	#endif
70888	73406
...	...	@@ -70895,11 +73413,11 @@ __ai __attribute__((target("neon"))) uint16x8_t vmlal_high_u8(uint16x8_t __p0, u
70895	73413	#else
70896	73414	__ai __attribute__((target("neon"))) uint16x8_t vmlal_high_u8(uint16x8_t __p0, uint8x16_t __p1, uint8x16_t __p2) {
70897	73415	uint16x8_t __ret;
70898		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
70899		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
70900		uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	73416	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	73417	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	73418	uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
70901	73419	__ret = __noswap_vmlal_u8(__rev0, __noswap_vget_high_u8(__rev1), __noswap_vget_high_u8(__rev2));
70902		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	73420	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
70903	73421	return __ret;
70904	73422	}
70905	73423	#endif
...	...	@@ -70913,11 +73431,11 @@ __ai __attribute__((target("neon"))) uint64x2_t vmlal_high_u32(uint64x2_t __p0,
70913	73431	#else
70914	73432	__ai __attribute__((target("neon"))) uint64x2_t vmlal_high_u32(uint64x2_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
70915	73433	uint64x2_t __ret;
70916		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
70917		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
70918		uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	73434	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	73435	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	73436	uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
70919	73437	__ret = __noswap_vmlal_u32(__rev0, __noswap_vget_high_u32(__rev1), __noswap_vget_high_u32(__rev2));
70920		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	73438	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
70921	73439	return __ret;
70922	73440	}
70923	73441	#endif
...	...	@@ -70931,11 +73449,11 @@ __ai __attribute__((target("neon"))) uint32x4_t vmlal_high_u16(uint32x4_t __p0,
70931	73449	#else
70932	73450	__ai __attribute__((target("neon"))) uint32x4_t vmlal_high_u16(uint32x4_t __p0, uint16x8_t __p1, uint16x8_t __p2) {
70933	73451	uint32x4_t __ret;
70934		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
70935		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
70936		uint16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	73452	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	73453	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	73454	uint16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
70937	73455	__ret = __noswap_vmlal_u16(__rev0, __noswap_vget_high_u16(__rev1), __noswap_vget_high_u16(__rev2));
70938		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	73456	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
70939	73457	return __ret;
70940	73458	}
70941	73459	#endif
...	...	@@ -70949,11 +73467,11 @@ __ai __attribute__((target("neon"))) int16x8_t vmlal_high_s8(int16x8_t __p0, int
70949	73467	#else
70950	73468	__ai __attribute__((target("neon"))) int16x8_t vmlal_high_s8(int16x8_t __p0, int8x16_t __p1, int8x16_t __p2) {
70951	73469	int16x8_t __ret;
70952		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
70953		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
70954		int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	73470	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	73471	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	73472	int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
70955	73473	__ret = __noswap_vmlal_s8(__rev0, __noswap_vget_high_s8(__rev1), __noswap_vget_high_s8(__rev2));
70956		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	73474	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
70957	73475	return __ret;
70958	73476	}
70959	73477	#endif
...	...	@@ -70967,11 +73485,11 @@ __ai __attribute__((target("neon"))) int64x2_t vmlal_high_s32(int64x2_t __p0, in
70967	73485	#else
70968	73486	__ai __attribute__((target("neon"))) int64x2_t vmlal_high_s32(int64x2_t __p0, int32x4_t __p1, int32x4_t __p2) {
70969	73487	int64x2_t __ret;
70970		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
70971		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
70972		int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	73488	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	73489	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	73490	int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
70973	73491	__ret = __noswap_vmlal_s32(__rev0, __noswap_vget_high_s32(__rev1), __noswap_vget_high_s32(__rev2));
70974		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	73492	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
70975	73493	return __ret;
70976	73494	}
70977	73495	#endif
...	...	@@ -70985,11 +73503,11 @@ __ai __attribute__((target("neon"))) int32x4_t vmlal_high_s16(int32x4_t __p0, in
70985	73503	#else
70986	73504	__ai __attribute__((target("neon"))) int32x4_t vmlal_high_s16(int32x4_t __p0, int16x8_t __p1, int16x8_t __p2) {
70987	73505	int32x4_t __ret;
70988		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
70989		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
70990		int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	73506	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	73507	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	73508	int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
70991	73509	__ret = __noswap_vmlal_s16(__rev0, __noswap_vget_high_s16(__rev1), __noswap_vget_high_s16(__rev2));
70992		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	73510	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
70993	73511	return __ret;
70994	73512	}
70995	73513	#endif
...	...	@@ -71003,10 +73521,10 @@ __ai __attribute__((target("neon"))) uint64x2_t vmlal_high_n_u32(uint64x2_t __p0
71003	73521	#else
71004	73522	__ai __attribute__((target("neon"))) uint64x2_t vmlal_high_n_u32(uint64x2_t __p0, uint32x4_t __p1, uint32_t __p2) {
71005	73523	uint64x2_t __ret;
71006		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
71007		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	73524	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	73525	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
71008	73526	__ret = __noswap_vmlal_n_u32(__rev0, __noswap_vget_high_u32(__rev1), __p2);
71009		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	73527	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
71010	73528	return __ret;
71011	73529	}
71012	73530	#endif
...	...	@@ -71020,10 +73538,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vmlal_high_n_u16(uint32x4_t __p0
71020	73538	#else
71021	73539	__ai __attribute__((target("neon"))) uint32x4_t vmlal_high_n_u16(uint32x4_t __p0, uint16x8_t __p1, uint16_t __p2) {
71022	73540	uint32x4_t __ret;
71023		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
71024		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	73541	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	73542	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
71025	73543	__ret = __noswap_vmlal_n_u16(__rev0, __noswap_vget_high_u16(__rev1), __p2);
71026		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	73544	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
71027	73545	return __ret;
71028	73546	}
71029	73547	#endif
...	...	@@ -71037,10 +73555,10 @@ __ai __attribute__((target("neon"))) int64x2_t vmlal_high_n_s32(int64x2_t __p0,
71037	73555	#else
71038	73556	__ai __attribute__((target("neon"))) int64x2_t vmlal_high_n_s32(int64x2_t __p0, int32x4_t __p1, int32_t __p2) {
71039	73557	int64x2_t __ret;
71040		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
71041		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	73558	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	73559	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
71042	73560	__ret = __noswap_vmlal_n_s32(__rev0, __noswap_vget_high_s32(__rev1), __p2);
71043		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	73561	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
71044	73562	return __ret;
71045	73563	}
71046	73564	#endif
...	...	@@ -71054,10 +73572,10 @@ __ai __attribute__((target("neon"))) int32x4_t vmlal_high_n_s16(int32x4_t __p0,
71054	73572	#else
71055	73573	__ai __attribute__((target("neon"))) int32x4_t vmlal_high_n_s16(int32x4_t __p0, int16x8_t __p1, int16_t __p2) {
71056	73574	int32x4_t __ret;
71057		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
71058		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	73575	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	73576	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
71059	73577	__ret = __noswap_vmlal_n_s16(__rev0, __noswap_vget_high_s16(__rev1), __p2);
71060		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	73578	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
71061	73579	return __ret;
71062	73580	}
71063	73581	#endif
...	...	@@ -71071,11 +73589,11 @@ __ai __attribute__((target("neon"))) uint16x8_t vmlsl_high_u8(uint16x8_t __p0, u
71071	73589	#else
71072	73590	__ai __attribute__((target("neon"))) uint16x8_t vmlsl_high_u8(uint16x8_t __p0, uint8x16_t __p1, uint8x16_t __p2) {
71073	73591	uint16x8_t __ret;
71074		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
71075		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
71076		uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	73592	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	73593	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	73594	uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
71077	73595	__ret = __noswap_vmlsl_u8(__rev0, __noswap_vget_high_u8(__rev1), __noswap_vget_high_u8(__rev2));
71078		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	73596	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
71079	73597	return __ret;
71080	73598	}
71081	73599	#endif
...	...	@@ -71089,11 +73607,11 @@ __ai __attribute__((target("neon"))) uint64x2_t vmlsl_high_u32(uint64x2_t __p0,
71089	73607	#else
71090	73608	__ai __attribute__((target("neon"))) uint64x2_t vmlsl_high_u32(uint64x2_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
71091	73609	uint64x2_t __ret;
71092		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
71093		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
71094		uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	73610	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	73611	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	73612	uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
71095	73613	__ret = __noswap_vmlsl_u32(__rev0, __noswap_vget_high_u32(__rev1), __noswap_vget_high_u32(__rev2));
71096		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	73614	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
71097	73615	return __ret;
71098	73616	}
71099	73617	#endif
...	...	@@ -71107,11 +73625,11 @@ __ai __attribute__((target("neon"))) uint32x4_t vmlsl_high_u16(uint32x4_t __p0,
71107	73625	#else
71108	73626	__ai __attribute__((target("neon"))) uint32x4_t vmlsl_high_u16(uint32x4_t __p0, uint16x8_t __p1, uint16x8_t __p2) {
71109	73627	uint32x4_t __ret;
71110		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
71111		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
71112		uint16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	73628	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	73629	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	73630	uint16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
71113	73631	__ret = __noswap_vmlsl_u16(__rev0, __noswap_vget_high_u16(__rev1), __noswap_vget_high_u16(__rev2));
71114		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	73632	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
71115	73633	return __ret;
71116	73634	}
71117	73635	#endif
...	...	@@ -71125,11 +73643,11 @@ __ai __attribute__((target("neon"))) int16x8_t vmlsl_high_s8(int16x8_t __p0, int
71125	73643	#else
71126	73644	__ai __attribute__((target("neon"))) int16x8_t vmlsl_high_s8(int16x8_t __p0, int8x16_t __p1, int8x16_t __p2) {
71127	73645	int16x8_t __ret;
71128		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
71129		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
71130		int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	73646	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	73647	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	73648	int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
71131	73649	__ret = __noswap_vmlsl_s8(__rev0, __noswap_vget_high_s8(__rev1), __noswap_vget_high_s8(__rev2));
71132		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	73650	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
71133	73651	return __ret;
71134	73652	}
71135	73653	#endif
...	...	@@ -71143,11 +73661,11 @@ __ai __attribute__((target("neon"))) int64x2_t vmlsl_high_s32(int64x2_t __p0, in
71143	73661	#else
71144	73662	__ai __attribute__((target("neon"))) int64x2_t vmlsl_high_s32(int64x2_t __p0, int32x4_t __p1, int32x4_t __p2) {
71145	73663	int64x2_t __ret;
71146		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
71147		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
71148		int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	73664	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	73665	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	73666	int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
71149	73667	__ret = __noswap_vmlsl_s32(__rev0, __noswap_vget_high_s32(__rev1), __noswap_vget_high_s32(__rev2));
71150		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	73668	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
71151	73669	return __ret;
71152	73670	}
71153	73671	#endif
...	...	@@ -71161,11 +73679,11 @@ __ai __attribute__((target("neon"))) int32x4_t vmlsl_high_s16(int32x4_t __p0, in
71161	73679	#else
71162	73680	__ai __attribute__((target("neon"))) int32x4_t vmlsl_high_s16(int32x4_t __p0, int16x8_t __p1, int16x8_t __p2) {
71163	73681	int32x4_t __ret;
71164		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
71165		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
71166		int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	73682	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	73683	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	73684	int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
71167	73685	__ret = __noswap_vmlsl_s16(__rev0, __noswap_vget_high_s16(__rev1), __noswap_vget_high_s16(__rev2));
71168		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	73686	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
71169	73687	return __ret;
71170	73688	}
71171	73689	#endif
...	...	@@ -71179,10 +73697,10 @@ __ai __attribute__((target("neon"))) uint64x2_t vmlsl_high_n_u32(uint64x2_t __p0
71179	73697	#else
71180	73698	__ai __attribute__((target("neon"))) uint64x2_t vmlsl_high_n_u32(uint64x2_t __p0, uint32x4_t __p1, uint32_t __p2) {
71181	73699	uint64x2_t __ret;
71182		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
71183		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	73700	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	73701	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
71184	73702	__ret = __noswap_vmlsl_n_u32(__rev0, __noswap_vget_high_u32(__rev1), __p2);
71185		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	73703	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
71186	73704	return __ret;
71187	73705	}
71188	73706	#endif
...	...	@@ -71196,10 +73714,10 @@ __ai __attribute__((target("neon"))) uint32x4_t vmlsl_high_n_u16(uint32x4_t __p0
71196	73714	#else
71197	73715	__ai __attribute__((target("neon"))) uint32x4_t vmlsl_high_n_u16(uint32x4_t __p0, uint16x8_t __p1, uint16_t __p2) {
71198	73716	uint32x4_t __ret;
71199		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
71200		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	73717	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	73718	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
71201	73719	__ret = __noswap_vmlsl_n_u16(__rev0, __noswap_vget_high_u16(__rev1), __p2);
71202		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	73720	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
71203	73721	return __ret;
71204	73722	}
71205	73723	#endif
...	...	@@ -71213,10 +73731,10 @@ __ai __attribute__((target("neon"))) int64x2_t vmlsl_high_n_s32(int64x2_t __p0,
71213	73731	#else
71214	73732	__ai __attribute__((target("neon"))) int64x2_t vmlsl_high_n_s32(int64x2_t __p0, int32x4_t __p1, int32_t __p2) {
71215	73733	int64x2_t __ret;
71216		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
71217		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
	73734	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	73735	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
71218	73736	__ret = __noswap_vmlsl_n_s32(__rev0, __noswap_vget_high_s32(__rev1), __p2);
71219		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	73737	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
71220	73738	return __ret;
71221	73739	}
71222	73740	#endif
...	...	@@ -71230,46 +73748,46 @@ __ai __attribute__((target("neon"))) int32x4_t vmlsl_high_n_s16(int32x4_t __p0,
71230	73748	#else
71231	73749	__ai __attribute__((target("neon"))) int32x4_t vmlsl_high_n_s16(int32x4_t __p0, int16x8_t __p1, int16_t __p2) {
71232	73750	int32x4_t __ret;
71233		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
71234		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
	73751	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	73752	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
71235	73753	__ret = __noswap_vmlsl_n_s16(__rev0, __noswap_vget_high_s16(__rev1), __p2);
71236		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	73754	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
71237	73755	return __ret;
71238	73756	}
71239	73757	#endif
71240	73758
71241		#define vmulx_lane_f64(__p0_907, __p1_907, __p2_907) __extension__ ({ \
71242		float64x1_t __ret_907; \
71243		float64x1_t __s0_907 = __p0_907; \
71244		float64x1_t __s1_907 = __p1_907; \
71245		float64_t __x_907 = vget_lane_f64(__s0_907, 0); \
71246		float64_t __y_907 = vget_lane_f64(__s1_907, __p2_907); \
71247		float64_t __z_907 = vmulxd_f64(__x_907, __y_907); \
71248		__ret_907 = vset_lane_f64(__z_907, __s0_907, __p2_907); \
71249		__ret_907; \
	73759	#define vmulx_lane_f64(__p0_923, __p1_923, __p2_923) __extension__ ({ \
	73760	float64x1_t __ret_923; \
	73761	float64x1_t __s0_923 = __p0_923; \
	73762	float64x1_t __s1_923 = __p1_923; \
	73763	float64_t __x_923 = vget_lane_f64(__s0_923, 0); \
	73764	float64_t __y_923 = vget_lane_f64(__s1_923, __p2_923); \
	73765	float64_t __z_923 = vmulxd_f64(__x_923, __y_923); \
	73766	__ret_923 = vset_lane_f64(__z_923, __s0_923, __p2_923); \
	73767	__ret_923; \
71250	73768	})
71251	73769	#ifdef __LITTLE_ENDIAN__
71252		#define vmulx_laneq_f64(__p0_908, __p1_908, __p2_908) __extension__ ({ \
71253		float64x1_t __ret_908; \
71254		float64x1_t __s0_908 = __p0_908; \
71255		float64x2_t __s1_908 = __p1_908; \
71256		float64_t __x_908 = vget_lane_f64(__s0_908, 0); \
71257		float64_t __y_908 = vgetq_lane_f64(__s1_908, __p2_908); \
71258		float64_t __z_908 = vmulxd_f64(__x_908, __y_908); \
71259		__ret_908 = vset_lane_f64(__z_908, __s0_908, 0); \
71260		__ret_908; \
	73770	#define vmulx_laneq_f64(__p0_924, __p1_924, __p2_924) __extension__ ({ \
	73771	float64x1_t __ret_924; \
	73772	float64x1_t __s0_924 = __p0_924; \
	73773	float64x2_t __s1_924 = __p1_924; \
	73774	float64_t __x_924 = vget_lane_f64(__s0_924, 0); \
	73775	float64_t __y_924 = vgetq_lane_f64(__s1_924, __p2_924); \
	73776	float64_t __z_924 = vmulxd_f64(__x_924, __y_924); \
	73777	__ret_924 = vset_lane_f64(__z_924, __s0_924, 0); \
	73778	__ret_924; \
71261	73779	})
71262	73780	#else
71263		#define vmulx_laneq_f64(__p0_909, __p1_909, __p2_909) __extension__ ({ \
71264		float64x1_t __ret_909; \
71265		float64x1_t __s0_909 = __p0_909; \
71266		float64x2_t __s1_909 = __p1_909; \
71267		float64x2_t __rev1_909; __rev1_909 = __builtin_shufflevector(__s1_909, __s1_909, 1, 0); \
71268		float64_t __x_909 = vget_lane_f64(__s0_909, 0); \
71269		float64_t __y_909 = __noswap_vgetq_lane_f64(__rev1_909, __p2_909); \
71270		float64_t __z_909 = vmulxd_f64(__x_909, __y_909); \
71271		__ret_909 = vset_lane_f64(__z_909, __s0_909, 0); \
71272		__ret_909; \
	73781	#define vmulx_laneq_f64(__p0_925, __p1_925, __p2_925) __extension__ ({ \
	73782	float64x1_t __ret_925; \
	73783	float64x1_t __s0_925 = __p0_925; \
	73784	float64x2_t __s1_925 = __p1_925; \
	73785	float64x2_t __rev1_925; __rev1_925 = __builtin_shufflevector(__s1_925, __s1_925, __lane_reverse_128_64); \
	73786	float64_t __x_925 = vget_lane_f64(__s0_925, 0); \
	73787	float64_t __y_925 = __noswap_vgetq_lane_f64(__rev1_925, __p2_925); \
	73788	float64_t __z_925 = vmulxd_f64(__x_925, __y_925); \
	73789	__ret_925 = vset_lane_f64(__z_925, __s0_925, 0); \
	73790	__ret_925; \
71273	73791	})
71274	73792	#endif
71275	73793
...	...	@@ -71283,11 +73801,11 @@ __ai __attribute__((target("neon"))) uint16x8_t vabal_u8(uint16x8_t __p0, uint8x
71283	73801	#else
71284	73802	__ai __attribute__((target("neon"))) uint16x8_t vabal_u8(uint16x8_t __p0, uint8x8_t __p1, uint8x8_t __p2) {
71285	73803	uint16x8_t __ret;
71286		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
71287		uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
71288		uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	73804	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	73805	uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	73806	uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
71289	73807	__ret = __rev0 + __noswap_vabdl_u8(__rev1, __rev2);
71290		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	73808	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
71291	73809	return __ret;
71292	73810	}
71293	73811	__ai __attribute__((target("neon"))) uint16x8_t __noswap_vabal_u8(uint16x8_t __p0, uint8x8_t __p1, uint8x8_t __p2) {
...	...	@@ -71306,11 +73824,11 @@ __ai __attribute__((target("neon"))) uint64x2_t vabal_u32(uint64x2_t __p0, uint3
71306	73824	#else
71307	73825	__ai __attribute__((target("neon"))) uint64x2_t vabal_u32(uint64x2_t __p0, uint32x2_t __p1, uint32x2_t __p2) {
71308	73826	uint64x2_t __ret;
71309		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
71310		uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
71311		uint32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
	73827	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	73828	uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	73829	uint32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_32);
71312	73830	__ret = __rev0 + __noswap_vabdl_u32(__rev1, __rev2);
71313		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	73831	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
71314	73832	return __ret;
71315	73833	}
71316	73834	__ai __attribute__((target("neon"))) uint64x2_t __noswap_vabal_u32(uint64x2_t __p0, uint32x2_t __p1, uint32x2_t __p2) {
...	...	@@ -71329,11 +73847,11 @@ __ai __attribute__((target("neon"))) uint32x4_t vabal_u16(uint32x4_t __p0, uint1
71329	73847	#else
71330	73848	__ai __attribute__((target("neon"))) uint32x4_t vabal_u16(uint32x4_t __p0, uint16x4_t __p1, uint16x4_t __p2) {
71331	73849	uint32x4_t __ret;
71332		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
71333		uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
71334		uint16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	73850	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	73851	uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	73852	uint16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
71335	73853	__ret = __rev0 + __noswap_vabdl_u16(__rev1, __rev2);
71336		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	73854	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
71337	73855	return __ret;
71338	73856	}
71339	73857	__ai __attribute__((target("neon"))) uint32x4_t __noswap_vabal_u16(uint32x4_t __p0, uint16x4_t __p1, uint16x4_t __p2) {
...	...	@@ -71352,11 +73870,11 @@ __ai __attribute__((target("neon"))) int16x8_t vabal_s8(int16x8_t __p0, int8x8_t
71352	73870	#else
71353	73871	__ai __attribute__((target("neon"))) int16x8_t vabal_s8(int16x8_t __p0, int8x8_t __p1, int8x8_t __p2) {
71354	73872	int16x8_t __ret;
71355		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
71356		int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
71357		int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	73873	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	73874	int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_8);
	73875	int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_8);
71358	73876	__ret = __rev0 + __noswap_vabdl_s8(__rev1, __rev2);
71359		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	73877	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
71360	73878	return __ret;
71361	73879	}
71362	73880	__ai __attribute__((target("neon"))) int16x8_t __noswap_vabal_s8(int16x8_t __p0, int8x8_t __p1, int8x8_t __p2) {
...	...	@@ -71375,11 +73893,11 @@ __ai __attribute__((target("neon"))) int64x2_t vabal_s32(int64x2_t __p0, int32x2
71375	73893	#else
71376	73894	__ai __attribute__((target("neon"))) int64x2_t vabal_s32(int64x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
71377	73895	int64x2_t __ret;
71378		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
71379		int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
71380		int32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
	73896	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	73897	int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_32);
	73898	int32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_32);
71381	73899	__ret = __rev0 + __noswap_vabdl_s32(__rev1, __rev2);
71382		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	73900	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
71383	73901	return __ret;
71384	73902	}
71385	73903	__ai __attribute__((target("neon"))) int64x2_t __noswap_vabal_s32(int64x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
...	...	@@ -71398,11 +73916,11 @@ __ai __attribute__((target("neon"))) int32x4_t vabal_s16(int32x4_t __p0, int16x4
71398	73916	#else
71399	73917	__ai __attribute__((target("neon"))) int32x4_t vabal_s16(int32x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
71400	73918	int32x4_t __ret;
71401		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
71402		int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
71403		int16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	73919	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	73920	int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_64_16);
	73921	int16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_64_16);
71404	73922	__ret = __rev0 + __noswap_vabdl_s16(__rev1, __rev2);
71405		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	73923	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
71406	73924	return __ret;
71407	73925	}
71408	73926	__ai __attribute__((target("neon"))) int32x4_t __noswap_vabal_s16(int32x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
...	...	@@ -71422,11 +73940,11 @@ __ai __attribute__((target("neon"))) uint16x8_t vabal_high_u8(uint16x8_t __p0, u
71422	73940	#else
71423	73941	__ai __attribute__((target("neon"))) uint16x8_t vabal_high_u8(uint16x8_t __p0, uint8x16_t __p1, uint8x16_t __p2) {
71424	73942	uint16x8_t __ret;
71425		uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
71426		uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
71427		uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	73943	uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	73944	uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	73945	uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
71428	73946	__ret = __noswap_vabal_u8(__rev0, __noswap_vget_high_u8(__rev1), __noswap_vget_high_u8(__rev2));
71429		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	73947	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
71430	73948	return __ret;
71431	73949	}
71432	73950	#endif
...	...	@@ -71440,11 +73958,11 @@ __ai __attribute__((target("neon"))) uint64x2_t vabal_high_u32(uint64x2_t __p0,
71440	73958	#else
71441	73959	__ai __attribute__((target("neon"))) uint64x2_t vabal_high_u32(uint64x2_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
71442	73960	uint64x2_t __ret;
71443		uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
71444		uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
71445		uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	73961	uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	73962	uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	73963	uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
71446	73964	__ret = __noswap_vabal_u32(__rev0, __noswap_vget_high_u32(__rev1), __noswap_vget_high_u32(__rev2));
71447		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	73965	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
71448	73966	return __ret;
71449	73967	}
71450	73968	#endif
...	...	@@ -71458,11 +73976,11 @@ __ai __attribute__((target("neon"))) uint32x4_t vabal_high_u16(uint32x4_t __p0,
71458	73976	#else
71459	73977	__ai __attribute__((target("neon"))) uint32x4_t vabal_high_u16(uint32x4_t __p0, uint16x8_t __p1, uint16x8_t __p2) {
71460	73978	uint32x4_t __ret;
71461		uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
71462		uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
71463		uint16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	73979	uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	73980	uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	73981	uint16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
71464	73982	__ret = __noswap_vabal_u16(__rev0, __noswap_vget_high_u16(__rev1), __noswap_vget_high_u16(__rev2));
71465		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	73983	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
71466	73984	return __ret;
71467	73985	}
71468	73986	#endif
...	...	@@ -71476,11 +73994,11 @@ __ai __attribute__((target("neon"))) int16x8_t vabal_high_s8(int16x8_t __p0, int
71476	73994	#else
71477	73995	__ai __attribute__((target("neon"))) int16x8_t vabal_high_s8(int16x8_t __p0, int8x16_t __p1, int8x16_t __p2) {
71478	73996	int16x8_t __ret;
71479		int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
71480		int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
71481		int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
	73997	int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_16);
	73998	int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_8);
	73999	int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_8);
71482	74000	__ret = __noswap_vabal_s8(__rev0, __noswap_vget_high_s8(__rev1), __noswap_vget_high_s8(__rev2));
71483		__ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
	74001	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_16);
71484	74002	return __ret;
71485	74003	}
71486	74004	#endif
...	...	@@ -71494,11 +74012,11 @@ __ai __attribute__((target("neon"))) int64x2_t vabal_high_s32(int64x2_t __p0, in
71494	74012	#else
71495	74013	__ai __attribute__((target("neon"))) int64x2_t vabal_high_s32(int64x2_t __p0, int32x4_t __p1, int32x4_t __p2) {
71496	74014	int64x2_t __ret;
71497		int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
71498		int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
71499		int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
	74015	int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_64);
	74016	int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_32);
	74017	int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_32);
71500	74018	__ret = __noswap_vabal_s32(__rev0, __noswap_vget_high_s32(__rev1), __noswap_vget_high_s32(__rev2));
71501		__ret = __builtin_shufflevector(__ret, __ret, 1, 0);
	74019	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_64);
71502	74020	return __ret;
71503	74021	}
71504	74022	#endif
...	...	@@ -71512,11 +74030,11 @@ __ai __attribute__((target("neon"))) int32x4_t vabal_high_s16(int32x4_t __p0, in
71512	74030	#else
71513	74031	__ai __attribute__((target("neon"))) int32x4_t vabal_high_s16(int32x4_t __p0, int16x8_t __p1, int16x8_t __p2) {
71514	74032	int32x4_t __ret;
71515		int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
71516		int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
71517		int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
	74033	int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, __lane_reverse_128_32);
	74034	int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, __lane_reverse_128_16);
	74035	int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, __lane_reverse_128_16);
71518	74036	__ret = __noswap_vabal_s16(__rev0, __noswap_vget_high_s16(__rev1), __noswap_vget_high_s16(__rev2));
71519		__ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
	74037	__ret = __builtin_shufflevector(__ret, __ret, __lane_reverse_128_32);
71520	74038	return __ret;
71521	74039	}
71522	74040	#endif

lib/include/arm_sme.h+694

...	...	@@ -146,6 +146,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_hor_za128_s32
146	146	svint32_t svread_hor_za128_s32_m(svint32_t, svbool_t, uint64_t, uint32_t);
147	147	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_hor_za128_s64_m)))
148	148	svint64_t svread_hor_za128_s64_m(svint64_t, svbool_t, uint64_t, uint32_t);
	149	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_hor_za128_mf8_m)))
	150	svmfloat8_t svread_hor_za128_mf8_m(svmfloat8_t, svbool_t, uint64_t, uint32_t);
149	151	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_hor_za128_s16_m)))
150	152	svint16_t svread_hor_za128_s16_m(svint16_t, svbool_t, uint64_t, uint32_t);
151	153	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_hor_za16_u16_m)))
...	...	@@ -172,6 +174,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_hor_za8_u8_m)
172	174	svuint8_t svread_hor_za8_u8_m(svuint8_t, svbool_t, uint64_t, uint32_t);
173	175	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_hor_za8_s8_m)))
174	176	svint8_t svread_hor_za8_s8_m(svint8_t, svbool_t, uint64_t, uint32_t);
	177	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_hor_za8_mf8_m)))
	178	svmfloat8_t svread_hor_za8_mf8_m(svmfloat8_t, svbool_t, uint64_t, uint32_t);
175	179	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_ver_za128_u8_m)))
176	180	svuint8_t svread_ver_za128_u8_m(svuint8_t, svbool_t, uint64_t, uint32_t);
177	181	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_ver_za128_u32_m)))
...	...	@@ -194,6 +198,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_ver_za128_s32
194	198	svint32_t svread_ver_za128_s32_m(svint32_t, svbool_t, uint64_t, uint32_t);
195	199	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_ver_za128_s64_m)))
196	200	svint64_t svread_ver_za128_s64_m(svint64_t, svbool_t, uint64_t, uint32_t);
	201	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_ver_za128_mf8_m)))
	202	svmfloat8_t svread_ver_za128_mf8_m(svmfloat8_t, svbool_t, uint64_t, uint32_t);
197	203	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_ver_za128_s16_m)))
198	204	svint16_t svread_ver_za128_s16_m(svint16_t, svbool_t, uint64_t, uint32_t);
199	205	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_ver_za16_u16_m)))
...	...	@@ -220,6 +226,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_ver_za8_u8_m)
220	226	svuint8_t svread_ver_za8_u8_m(svuint8_t, svbool_t, uint64_t, uint32_t);
221	227	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_ver_za8_s8_m)))
222	228	svint8_t svread_ver_za8_s8_m(svint8_t, svbool_t, uint64_t, uint32_t);
	229	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_ver_za8_mf8_m)))
	230	svmfloat8_t svread_ver_za8_mf8_m(svmfloat8_t, svbool_t, uint64_t, uint32_t);
223	231	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svst1_hor_vnum_za128)))
224	232	void svst1_hor_vnum_za128(uint64_t, uint32_t, svbool_t, void *, int64_t);
225	233	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svst1_hor_vnum_za16)))
...	...	@@ -294,6 +302,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_hor_za128_s3
294	302	void svwrite_hor_za128_s32_m(uint64_t, uint32_t, svbool_t, svint32_t);
295	303	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_hor_za128_s64_m)))
296	304	void svwrite_hor_za128_s64_m(uint64_t, uint32_t, svbool_t, svint64_t);
	305	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_hor_za128_mf8_m)))
	306	void svwrite_hor_za128_mf8_m(uint64_t, uint32_t, svbool_t, svmfloat8_t);
297	307	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_hor_za128_s16_m)))
298	308	void svwrite_hor_za128_s16_m(uint64_t, uint32_t, svbool_t, svint16_t);
299	309	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_hor_za16_u16_m)))
...	...	@@ -320,6 +330,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_hor_za8_u8_m
320	330	void svwrite_hor_za8_u8_m(uint64_t, uint32_t, svbool_t, svuint8_t);
321	331	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_hor_za8_s8_m)))
322	332	void svwrite_hor_za8_s8_m(uint64_t, uint32_t, svbool_t, svint8_t);
	333	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_hor_za8_mf8_m)))
	334	void svwrite_hor_za8_mf8_m(uint64_t, uint32_t, svbool_t, svmfloat8_t);
323	335	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za128_u8_m)))
324	336	void svwrite_ver_za128_u8_m(uint64_t, uint32_t, svbool_t, svuint8_t);
325	337	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za128_u32_m)))
...	...	@@ -342,6 +354,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za128_s3
342	354	void svwrite_ver_za128_s32_m(uint64_t, uint32_t, svbool_t, svint32_t);
343	355	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za128_s64_m)))
344	356	void svwrite_ver_za128_s64_m(uint64_t, uint32_t, svbool_t, svint64_t);
	357	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za128_mf8_m)))
	358	void svwrite_ver_za128_mf8_m(uint64_t, uint32_t, svbool_t, svmfloat8_t);
345	359	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za128_s16_m)))
346	360	void svwrite_ver_za128_s16_m(uint64_t, uint32_t, svbool_t, svint16_t);
347	361	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za16_u16_m)))
...	...	@@ -368,6 +382,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za8_u8_m
368	382	void svwrite_ver_za8_u8_m(uint64_t, uint32_t, svbool_t, svuint8_t);
369	383	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za8_s8_m)))
370	384	void svwrite_ver_za8_s8_m(uint64_t, uint32_t, svbool_t, svint8_t);
	385	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za8_mf8_m)))
	386	void svwrite_ver_za8_mf8_m(uint64_t, uint32_t, svbool_t, svmfloat8_t);
371	387	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svzero_mask_za)))
372	388	void svzero_mask_za(uint64_t);
373	389	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svzero_za)))
...	...	@@ -422,6 +438,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_hor_za128_s3
422	438	svint32_t svread_hor_za128_m(svint32_t, svbool_t, uint64_t, uint32_t);
423	439	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_hor_za128_s64_m)))
424	440	svint64_t svread_hor_za128_m(svint64_t, svbool_t, uint64_t, uint32_t);
	441	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_hor_za128_mf8_m)))
	442	svmfloat8_t svread_hor_za128_m(svmfloat8_t, svbool_t, uint64_t, uint32_t);
425	443	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_hor_za128_s16_m)))
426	444	svint16_t svread_hor_za128_m(svint16_t, svbool_t, uint64_t, uint32_t);
427	445	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_hor_za16_u16_m)))
...	...	@@ -448,6 +466,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_hor_za8_u8_m
448	466	svuint8_t svread_hor_za8_m(svuint8_t, svbool_t, uint64_t, uint32_t);
449	467	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_hor_za8_s8_m)))
450	468	svint8_t svread_hor_za8_m(svint8_t, svbool_t, uint64_t, uint32_t);
	469	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_hor_za8_mf8_m)))
	470	svmfloat8_t svread_hor_za8_m(svmfloat8_t, svbool_t, uint64_t, uint32_t);
451	471	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_ver_za128_u8_m)))
452	472	svuint8_t svread_ver_za128_m(svuint8_t, svbool_t, uint64_t, uint32_t);
453	473	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_ver_za128_u32_m)))
...	...	@@ -470,6 +490,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_ver_za128_s3
470	490	svint32_t svread_ver_za128_m(svint32_t, svbool_t, uint64_t, uint32_t);
471	491	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_ver_za128_s64_m)))
472	492	svint64_t svread_ver_za128_m(svint64_t, svbool_t, uint64_t, uint32_t);
	493	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_ver_za128_mf8_m)))
	494	svmfloat8_t svread_ver_za128_m(svmfloat8_t, svbool_t, uint64_t, uint32_t);
473	495	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_ver_za128_s16_m)))
474	496	svint16_t svread_ver_za128_m(svint16_t, svbool_t, uint64_t, uint32_t);
475	497	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_ver_za16_u16_m)))
...	...	@@ -496,6 +518,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_ver_za8_u8_m
496	518	svuint8_t svread_ver_za8_m(svuint8_t, svbool_t, uint64_t, uint32_t);
497	519	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_ver_za8_s8_m)))
498	520	svint8_t svread_ver_za8_m(svint8_t, svbool_t, uint64_t, uint32_t);
	521	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_ver_za8_mf8_m)))
	522	svmfloat8_t svread_ver_za8_m(svmfloat8_t, svbool_t, uint64_t, uint32_t);
499	523	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svsumopa_za32_s8_m)))
500	524	void svsumopa_za32_m(uint64_t, svbool_t, svbool_t, svint8_t, svuint8_t);
501	525	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svsumops_za32_s8_m)))
...	...	@@ -526,6 +550,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_hor_za128_s
526	550	void svwrite_hor_za128_m(uint64_t, uint32_t, svbool_t, svint32_t);
527	551	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_hor_za128_s64_m)))
528	552	void svwrite_hor_za128_m(uint64_t, uint32_t, svbool_t, svint64_t);
	553	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_hor_za128_mf8_m)))
	554	void svwrite_hor_za128_m(uint64_t, uint32_t, svbool_t, svmfloat8_t);
529	555	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_hor_za128_s16_m)))
530	556	void svwrite_hor_za128_m(uint64_t, uint32_t, svbool_t, svint16_t);
531	557	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_hor_za16_u16_m)))
...	...	@@ -552,6 +578,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_hor_za8_u8_
552	578	void svwrite_hor_za8_m(uint64_t, uint32_t, svbool_t, svuint8_t);
553	579	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_hor_za8_s8_m)))
554	580	void svwrite_hor_za8_m(uint64_t, uint32_t, svbool_t, svint8_t);
	581	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_hor_za8_mf8_m)))
	582	void svwrite_hor_za8_m(uint64_t, uint32_t, svbool_t, svmfloat8_t);
555	583	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za128_u8_m)))
556	584	void svwrite_ver_za128_m(uint64_t, uint32_t, svbool_t, svuint8_t);
557	585	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za128_u32_m)))
...	...	@@ -574,6 +602,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za128_s
574	602	void svwrite_ver_za128_m(uint64_t, uint32_t, svbool_t, svint32_t);
575	603	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za128_s64_m)))
576	604	void svwrite_ver_za128_m(uint64_t, uint32_t, svbool_t, svint64_t);
	605	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za128_mf8_m)))
	606	void svwrite_ver_za128_m(uint64_t, uint32_t, svbool_t, svmfloat8_t);
577	607	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za128_s16_m)))
578	608	void svwrite_ver_za128_m(uint64_t, uint32_t, svbool_t, svint16_t);
579	609	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za16_u16_m)))
...	...	@@ -600,6 +630,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za8_u8_
600	630	void svwrite_ver_za8_m(uint64_t, uint32_t, svbool_t, svuint8_t);
601	631	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za8_s8_m)))
602	632	void svwrite_ver_za8_m(uint64_t, uint32_t, svbool_t, svint8_t);
	633	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za8_mf8_m)))
	634	void svwrite_ver_za8_m(uint64_t, uint32_t, svbool_t, svmfloat8_t);
603	635	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svadd_za16_f16_vg1x2)))
604	636	void svadd_za16_f16_vg1x2(uint32_t, svfloat16x2_t);
605	637	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svadd_za16_f16_vg1x4)))
...	...	@@ -1158,6 +1190,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svluti2_lane_zt_f16)
1158	1190	svfloat16_t svluti2_lane_zt_f16(uint64_t, svuint8_t, uint64_t);
1159	1191	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svluti2_lane_zt_s32)))
1160	1192	svint32_t svluti2_lane_zt_s32(uint64_t, svuint8_t, uint64_t);
	1193	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svluti2_lane_zt_mf8)))
	1194	svmfloat8_t svluti2_lane_zt_mf8(uint64_t, svuint8_t, uint64_t);
1161	1195	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svluti2_lane_zt_s16)))
1162	1196	svint16_t svluti2_lane_zt_s16(uint64_t, svuint8_t, uint64_t);
1163	1197	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svluti2_lane_zt_u8_x2)))
...	...	@@ -1176,6 +1210,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svluti2_lane_zt_f16_
1176	1210	svfloat16x2_t svluti2_lane_zt_f16_x2(uint64_t, svuint8_t, uint64_t);
1177	1211	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svluti2_lane_zt_s32_x2)))
1178	1212	svint32x2_t svluti2_lane_zt_s32_x2(uint64_t, svuint8_t, uint64_t);
	1213	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svluti2_lane_zt_mf8_x2)))
	1214	svmfloat8x2_t svluti2_lane_zt_mf8_x2(uint64_t, svuint8_t, uint64_t);
1179	1215	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svluti2_lane_zt_s16_x2)))
1180	1216	svint16x2_t svluti2_lane_zt_s16_x2(uint64_t, svuint8_t, uint64_t);
1181	1217	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svluti2_lane_zt_u8_x4)))
...	...	@@ -1194,6 +1230,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svluti2_lane_zt_f16_
1194	1230	svfloat16x4_t svluti2_lane_zt_f16_x4(uint64_t, svuint8_t, uint64_t);
1195	1231	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svluti2_lane_zt_s32_x4)))
1196	1232	svint32x4_t svluti2_lane_zt_s32_x4(uint64_t, svuint8_t, uint64_t);
	1233	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svluti2_lane_zt_mf8_x4)))
	1234	svmfloat8x4_t svluti2_lane_zt_mf8_x4(uint64_t, svuint8_t, uint64_t);
1197	1235	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svluti2_lane_zt_s16_x4)))
1198	1236	svint16x4_t svluti2_lane_zt_s16_x4(uint64_t, svuint8_t, uint64_t);
1199	1237	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svluti4_lane_zt_u8)))
...	...	@@ -1212,6 +1250,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svluti4_lane_zt_f16)
1212	1250	svfloat16_t svluti4_lane_zt_f16(uint64_t, svuint8_t, uint64_t);
1213	1251	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svluti4_lane_zt_s32)))
1214	1252	svint32_t svluti4_lane_zt_s32(uint64_t, svuint8_t, uint64_t);
	1253	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svluti4_lane_zt_mf8)))
	1254	svmfloat8_t svluti4_lane_zt_mf8(uint64_t, svuint8_t, uint64_t);
1215	1255	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svluti4_lane_zt_s16)))
1216	1256	svint16_t svluti4_lane_zt_s16(uint64_t, svuint8_t, uint64_t);
1217	1257	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svluti4_lane_zt_u8_x2)))
...	...	@@ -1230,6 +1270,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svluti4_lane_zt_f16_
1230	1270	svfloat16x2_t svluti4_lane_zt_f16_x2(uint64_t, svuint8_t, uint64_t);
1231	1271	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svluti4_lane_zt_s32_x2)))
1232	1272	svint32x2_t svluti4_lane_zt_s32_x2(uint64_t, svuint8_t, uint64_t);
	1273	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svluti4_lane_zt_mf8_x2)))
	1274	svmfloat8x2_t svluti4_lane_zt_mf8_x2(uint64_t, svuint8_t, uint64_t);
1233	1275	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svluti4_lane_zt_s16_x2)))
1234	1276	svint16x2_t svluti4_lane_zt_s16_x2(uint64_t, svuint8_t, uint64_t);
1235	1277	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svluti4_lane_zt_u32_x4)))
...	...	@@ -1514,10 +1556,14 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_hor_za8_u8_vg
1514	1556	svuint8x2_t svread_hor_za8_u8_vg2(uint64_t, uint32_t);
1515	1557	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_hor_za8_s8_vg2)))
1516	1558	svint8x2_t svread_hor_za8_s8_vg2(uint64_t, uint32_t);
	1559	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_hor_za8_mf8_vg2)))
	1560	svmfloat8x2_t svread_hor_za8_mf8_vg2(uint64_t, uint32_t);
1517	1561	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_hor_za8_u8_vg4)))
1518	1562	svuint8x4_t svread_hor_za8_u8_vg4(uint64_t, uint32_t);
1519	1563	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_hor_za8_s8_vg4)))
1520	1564	svint8x4_t svread_hor_za8_s8_vg4(uint64_t, uint32_t);
	1565	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_hor_za8_mf8_vg4)))
	1566	svmfloat8x4_t svread_hor_za8_mf8_vg4(uint64_t, uint32_t);
1521	1567	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_ver_za16_u16_vg2)))
1522	1568	svuint16x2_t svread_ver_za16_u16_vg2(uint64_t, uint32_t);
1523	1569	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_ver_za16_bf16_vg2)))
...	...	@@ -1562,10 +1608,14 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_ver_za8_u8_vg
1562	1608	svuint8x2_t svread_ver_za8_u8_vg2(uint64_t, uint32_t);
1563	1609	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_ver_za8_s8_vg2)))
1564	1610	svint8x2_t svread_ver_za8_s8_vg2(uint64_t, uint32_t);
	1611	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_ver_za8_mf8_vg2)))
	1612	svmfloat8x2_t svread_ver_za8_mf8_vg2(uint64_t, uint32_t);
1565	1613	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_ver_za8_u8_vg4)))
1566	1614	svuint8x4_t svread_ver_za8_u8_vg4(uint64_t, uint32_t);
1567	1615	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_ver_za8_s8_vg4)))
1568	1616	svint8x4_t svread_ver_za8_s8_vg4(uint64_t, uint32_t);
	1617	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_ver_za8_mf8_vg4)))
	1618	svmfloat8x4_t svread_ver_za8_mf8_vg4(uint64_t, uint32_t);
1569	1619	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_za16_u16_vg1x2)))
1570	1620	svuint16x2_t svread_za16_u16_vg1x2(uint32_t);
1571	1621	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_za16_bf16_vg1x2)))
...	...	@@ -1610,10 +1660,14 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_za8_u8_vg1x2)
1610	1660	svuint8x2_t svread_za8_u8_vg1x2(uint32_t);
1611	1661	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_za8_s8_vg1x2)))
1612	1662	svint8x2_t svread_za8_s8_vg1x2(uint32_t);
	1663	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_za8_mf8_vg1x2)))
	1664	svmfloat8x2_t svread_za8_mf8_vg1x2(uint32_t);
1613	1665	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_za8_u8_vg1x4)))
1614	1666	svuint8x4_t svread_za8_u8_vg1x4(uint32_t);
1615	1667	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_za8_s8_vg1x4)))
1616	1668	svint8x4_t svread_za8_s8_vg1x4(uint32_t);
	1669	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svread_za8_mf8_vg1x4)))
	1670	svmfloat8x4_t svread_za8_mf8_vg1x4(uint32_t);
1617	1671	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svstr_zt)))
1618	1672	void svstr_zt(uint64_t, void *);
1619	1673	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svsub_write_single_za32_u32_vg1x2)))
...	...	@@ -1760,10 +1814,14 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_hor_za8_u8_v
1760	1814	void svwrite_hor_za8_u8_vg2(uint64_t, uint32_t, svuint8x2_t);
1761	1815	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_hor_za8_s8_vg2)))
1762	1816	void svwrite_hor_za8_s8_vg2(uint64_t, uint32_t, svint8x2_t);
	1817	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_hor_za8_mf8_vg2)))
	1818	void svwrite_hor_za8_mf8_vg2(uint64_t, uint32_t, svmfloat8x2_t);
1763	1819	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_hor_za8_u8_vg4)))
1764	1820	void svwrite_hor_za8_u8_vg4(uint64_t, uint32_t, svuint8x4_t);
1765	1821	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_hor_za8_s8_vg4)))
1766	1822	void svwrite_hor_za8_s8_vg4(uint64_t, uint32_t, svint8x4_t);
	1823	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_hor_za8_mf8_vg4)))
	1824	void svwrite_hor_za8_mf8_vg4(uint64_t, uint32_t, svmfloat8x4_t);
1767	1825	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za16_u16_vg2)))
1768	1826	void svwrite_ver_za16_u16_vg2(uint64_t, uint32_t, svuint16x2_t);
1769	1827	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za16_bf16_vg2)))
...	...	@@ -1808,10 +1866,14 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za8_u8_v
1808	1866	void svwrite_ver_za8_u8_vg2(uint64_t, uint32_t, svuint8x2_t);
1809	1867	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za8_s8_vg2)))
1810	1868	void svwrite_ver_za8_s8_vg2(uint64_t, uint32_t, svint8x2_t);
	1869	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za8_mf8_vg2)))
	1870	void svwrite_ver_za8_mf8_vg2(uint64_t, uint32_t, svmfloat8x2_t);
1811	1871	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za8_u8_vg4)))
1812	1872	void svwrite_ver_za8_u8_vg4(uint64_t, uint32_t, svuint8x4_t);
1813	1873	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za8_s8_vg4)))
1814	1874	void svwrite_ver_za8_s8_vg4(uint64_t, uint32_t, svint8x4_t);
	1875	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za8_mf8_vg4)))
	1876	void svwrite_ver_za8_mf8_vg4(uint64_t, uint32_t, svmfloat8x4_t);
1815	1877	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_za16_u16_vg1x2)))
1816	1878	void svwrite_za16_u16_vg1x2(uint32_t, svuint16x2_t);
1817	1879	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_za16_bf16_vg1x2)))
...	...	@@ -1856,10 +1918,14 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_za8_u8_vg1x2
1856	1918	void svwrite_za8_u8_vg1x2(uint32_t, svuint8x2_t);
1857	1919	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_za8_s8_vg1x2)))
1858	1920	void svwrite_za8_s8_vg1x2(uint32_t, svint8x2_t);
	1921	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_za8_mf8_vg1x2)))
	1922	void svwrite_za8_mf8_vg1x2(uint32_t, svmfloat8x2_t);
1859	1923	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_za8_u8_vg1x4)))
1860	1924	void svwrite_za8_u8_vg1x4(uint32_t, svuint8x4_t);
1861	1925	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_za8_s8_vg1x4)))
1862	1926	void svwrite_za8_s8_vg1x4(uint32_t, svint8x4_t);
	1927	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_za8_mf8_vg1x4)))
	1928	void svwrite_za8_mf8_vg1x4(uint32_t, svmfloat8x4_t);
1863	1929	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svzero_zt)))
1864	1930	void svzero_zt(uint64_t);
1865	1931	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svadd_write_single_za32_u32_vg1x2)))
...	...	@@ -2338,10 +2404,14 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_hor_za8_u8_
2338	2404	void svwrite_hor_za8_vg2(uint64_t, uint32_t, svuint8x2_t);
2339	2405	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_hor_za8_s8_vg2)))
2340	2406	void svwrite_hor_za8_vg2(uint64_t, uint32_t, svint8x2_t);
	2407	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_hor_za8_mf8_vg2)))
	2408	void svwrite_hor_za8_vg2(uint64_t, uint32_t, svmfloat8x2_t);
2341	2409	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_hor_za8_u8_vg4)))
2342	2410	void svwrite_hor_za8_vg4(uint64_t, uint32_t, svuint8x4_t);
2343	2411	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_hor_za8_s8_vg4)))
2344	2412	void svwrite_hor_za8_vg4(uint64_t, uint32_t, svint8x4_t);
	2413	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_hor_za8_mf8_vg4)))
	2414	void svwrite_hor_za8_vg4(uint64_t, uint32_t, svmfloat8x4_t);
2345	2415	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za16_u16_vg2)))
2346	2416	void svwrite_ver_za16_vg2(uint64_t, uint32_t, svuint16x2_t);
2347	2417	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za16_bf16_vg2)))
...	...	@@ -2386,10 +2456,14 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za8_u8_
2386	2456	void svwrite_ver_za8_vg2(uint64_t, uint32_t, svuint8x2_t);
2387	2457	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za8_s8_vg2)))
2388	2458	void svwrite_ver_za8_vg2(uint64_t, uint32_t, svint8x2_t);
	2459	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za8_mf8_vg2)))
	2460	void svwrite_ver_za8_vg2(uint64_t, uint32_t, svmfloat8x2_t);
2389	2461	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za8_u8_vg4)))
2390	2462	void svwrite_ver_za8_vg4(uint64_t, uint32_t, svuint8x4_t);
2391	2463	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za8_s8_vg4)))
2392	2464	void svwrite_ver_za8_vg4(uint64_t, uint32_t, svint8x4_t);
	2465	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_ver_za8_mf8_vg4)))
	2466	void svwrite_ver_za8_vg4(uint64_t, uint32_t, svmfloat8x4_t);
2393	2467	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_za16_u16_vg1x2)))
2394	2468	void svwrite_za16_vg1x2(uint32_t, svuint16x2_t);
2395	2469	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_za16_bf16_vg1x2)))
...	...	@@ -2434,10 +2508,14 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_za8_u8_vg1x
2434	2508	void svwrite_za8_vg1x2(uint32_t, svuint8x2_t);
2435	2509	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_za8_s8_vg1x2)))
2436	2510	void svwrite_za8_vg1x2(uint32_t, svint8x2_t);
	2511	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_za8_mf8_vg1x2)))
	2512	void svwrite_za8_vg1x2(uint32_t, svmfloat8x2_t);
2437	2513	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_za8_u8_vg1x4)))
2438	2514	void svwrite_za8_vg1x4(uint32_t, svuint8x4_t);
2439	2515	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_za8_s8_vg1x4)))
2440	2516	void svwrite_za8_vg1x4(uint32_t, svint8x4_t);
	2517	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svwrite_za8_mf8_vg1x4)))
	2518	void svwrite_za8_vg1x4(uint32_t, svmfloat8x4_t);
2441	2519	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svadd_za64_f64_vg1x2)))
2442	2520	void svadd_za64_f64_vg1x2(uint32_t, svfloat64x2_t);
2443	2521	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svadd_za64_f64_vg1x4)))
...	...	@@ -2782,6 +2860,602 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svvdot_lane_za64_s1
2782	2860	void svvdot_lane_za64_vg1x4(uint32_t, svint16x4_t, svint16_t, uint64_t);
2783	2861	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svvdot_lane_za64_u16_vg1x4)))
2784	2862	void svvdot_lane_za64_vg1x4(uint32_t, svuint16x4_t, svuint16_t, uint64_t);
	2863	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za32_s8_u8)))
	2864	void svmop4a_1x1_za32_s8_u8(uint64_t, svint8_t, svuint8_t);
	2865	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za32_u8_s8)))
	2866	void svmop4a_1x1_za32_u8_s8(uint64_t, svuint8_t, svint8_t);
	2867	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za32_bf16_bf16)))
	2868	void svmop4a_1x1_za32_bf16_bf16(uint64_t, svbfloat16_t, svbfloat16_t);
	2869	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za32_f16_f16)))
	2870	void svmop4a_1x1_za32_f16_f16(uint64_t, svfloat16_t, svfloat16_t);
	2871	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za32_f32_f32)))
	2872	void svmop4a_1x1_za32_f32_f32(uint64_t, svfloat32_t, svfloat32_t);
	2873	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za32_s8_s8)))
	2874	void svmop4a_1x1_za32_s8_s8(uint64_t, svint8_t, svint8_t);
	2875	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za32_s16_s16)))
	2876	void svmop4a_1x1_za32_s16_s16(uint64_t, svint16_t, svint16_t);
	2877	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za32_u8_u8)))
	2878	void svmop4a_1x1_za32_u8_u8(uint64_t, svuint8_t, svuint8_t);
	2879	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za32_u16_u16)))
	2880	void svmop4a_1x1_za32_u16_u16(uint64_t, svuint16_t, svuint16_t);
	2881	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za32_s8_u8)))
	2882	void svmop4a_1x2_za32_s8_u8(uint64_t, svint8_t, svuint8x2_t);
	2883	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za32_u8_s8)))
	2884	void svmop4a_1x2_za32_u8_s8(uint64_t, svuint8_t, svint8x2_t);
	2885	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za32_bf16_bf16)))
	2886	void svmop4a_1x2_za32_bf16_bf16(uint64_t, svbfloat16_t, svbfloat16x2_t);
	2887	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za32_f16_f16)))
	2888	void svmop4a_1x2_za32_f16_f16(uint64_t, svfloat16_t, svfloat16x2_t);
	2889	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za32_f32_f32)))
	2890	void svmop4a_1x2_za32_f32_f32(uint64_t, svfloat32_t, svfloat32x2_t);
	2891	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za32_s8_s8)))
	2892	void svmop4a_1x2_za32_s8_s8(uint64_t, svint8_t, svint8x2_t);
	2893	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za32_s16_s16)))
	2894	void svmop4a_1x2_za32_s16_s16(uint64_t, svint16_t, svint16x2_t);
	2895	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za32_u8_u8)))
	2896	void svmop4a_1x2_za32_u8_u8(uint64_t, svuint8_t, svuint8x2_t);
	2897	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za32_u16_u16)))
	2898	void svmop4a_1x2_za32_u16_u16(uint64_t, svuint16_t, svuint16x2_t);
	2899	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za32_s8_u8)))
	2900	void svmop4a_2x1_za32_s8_u8(uint64_t, svint8x2_t, svuint8_t);
	2901	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za32_u8_s8)))
	2902	void svmop4a_2x1_za32_u8_s8(uint64_t, svuint8x2_t, svint8_t);
	2903	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za32_bf16_bf16)))
	2904	void svmop4a_2x1_za32_bf16_bf16(uint64_t, svbfloat16x2_t, svbfloat16_t);
	2905	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za32_f16_f16)))
	2906	void svmop4a_2x1_za32_f16_f16(uint64_t, svfloat16x2_t, svfloat16_t);
	2907	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za32_f32_f32)))
	2908	void svmop4a_2x1_za32_f32_f32(uint64_t, svfloat32x2_t, svfloat32_t);
	2909	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za32_s8_s8)))
	2910	void svmop4a_2x1_za32_s8_s8(uint64_t, svint8x2_t, svint8_t);
	2911	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za32_s16_s16)))
	2912	void svmop4a_2x1_za32_s16_s16(uint64_t, svint16x2_t, svint16_t);
	2913	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za32_u8_u8)))
	2914	void svmop4a_2x1_za32_u8_u8(uint64_t, svuint8x2_t, svuint8_t);
	2915	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za32_u16_u16)))
	2916	void svmop4a_2x1_za32_u16_u16(uint64_t, svuint16x2_t, svuint16_t);
	2917	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za32_s8_u8)))
	2918	void svmop4a_2x2_za32_s8_u8(uint64_t, svint8x2_t, svuint8x2_t);
	2919	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za32_u8_s8)))
	2920	void svmop4a_2x2_za32_u8_s8(uint64_t, svuint8x2_t, svint8x2_t);
	2921	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za32_bf16_bf16)))
	2922	void svmop4a_2x2_za32_bf16_bf16(uint64_t, svbfloat16x2_t, svbfloat16x2_t);
	2923	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za32_f16_f16)))
	2924	void svmop4a_2x2_za32_f16_f16(uint64_t, svfloat16x2_t, svfloat16x2_t);
	2925	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za32_f32_f32)))
	2926	void svmop4a_2x2_za32_f32_f32(uint64_t, svfloat32x2_t, svfloat32x2_t);
	2927	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za32_s8_s8)))
	2928	void svmop4a_2x2_za32_s8_s8(uint64_t, svint8x2_t, svint8x2_t);
	2929	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za32_s16_s16)))
	2930	void svmop4a_2x2_za32_s16_s16(uint64_t, svint16x2_t, svint16x2_t);
	2931	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za32_u8_u8)))
	2932	void svmop4a_2x2_za32_u8_u8(uint64_t, svuint8x2_t, svuint8x2_t);
	2933	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za32_u16_u16)))
	2934	void svmop4a_2x2_za32_u16_u16(uint64_t, svuint16x2_t, svuint16x2_t);
	2935	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za32_s8_u8)))
	2936	void svmop4s_1x1_za32_s8_u8(uint64_t, svint8_t, svuint8_t);
	2937	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za32_u8_s8)))
	2938	void svmop4s_1x1_za32_u8_s8(uint64_t, svuint8_t, svint8_t);
	2939	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za32_bf16_bf16)))
	2940	void svmop4s_1x1_za32_bf16_bf16(uint64_t, svbfloat16_t, svbfloat16_t);
	2941	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za32_f16_f16)))
	2942	void svmop4s_1x1_za32_f16_f16(uint64_t, svfloat16_t, svfloat16_t);
	2943	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za32_f32_f32)))
	2944	void svmop4s_1x1_za32_f32_f32(uint64_t, svfloat32_t, svfloat32_t);
	2945	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za32_s8_s8)))
	2946	void svmop4s_1x1_za32_s8_s8(uint64_t, svint8_t, svint8_t);
	2947	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za32_s16_s16)))
	2948	void svmop4s_1x1_za32_s16_s16(uint64_t, svint16_t, svint16_t);
	2949	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za32_u8_u8)))
	2950	void svmop4s_1x1_za32_u8_u8(uint64_t, svuint8_t, svuint8_t);
	2951	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za32_u16_u16)))
	2952	void svmop4s_1x1_za32_u16_u16(uint64_t, svuint16_t, svuint16_t);
	2953	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za32_s8_u8)))
	2954	void svmop4s_1x2_za32_s8_u8(uint64_t, svint8_t, svuint8x2_t);
	2955	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za32_u8_s8)))
	2956	void svmop4s_1x2_za32_u8_s8(uint64_t, svuint8_t, svint8x2_t);
	2957	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za32_bf16_bf16)))
	2958	void svmop4s_1x2_za32_bf16_bf16(uint64_t, svbfloat16_t, svbfloat16x2_t);
	2959	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za32_f16_f16)))
	2960	void svmop4s_1x2_za32_f16_f16(uint64_t, svfloat16_t, svfloat16x2_t);
	2961	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za32_f32_f32)))
	2962	void svmop4s_1x2_za32_f32_f32(uint64_t, svfloat32_t, svfloat32x2_t);
	2963	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za32_s8_s8)))
	2964	void svmop4s_1x2_za32_s8_s8(uint64_t, svint8_t, svint8x2_t);
	2965	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za32_s16_s16)))
	2966	void svmop4s_1x2_za32_s16_s16(uint64_t, svint16_t, svint16x2_t);
	2967	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za32_u8_u8)))
	2968	void svmop4s_1x2_za32_u8_u8(uint64_t, svuint8_t, svuint8x2_t);
	2969	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za32_u16_u16)))
	2970	void svmop4s_1x2_za32_u16_u16(uint64_t, svuint16_t, svuint16x2_t);
	2971	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za32_s8_u8)))
	2972	void svmop4s_2x1_za32_s8_u8(uint64_t, svint8x2_t, svuint8_t);
	2973	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za32_u8_s8)))
	2974	void svmop4s_2x1_za32_u8_s8(uint64_t, svuint8x2_t, svint8_t);
	2975	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za32_bf16_bf16)))
	2976	void svmop4s_2x1_za32_bf16_bf16(uint64_t, svbfloat16x2_t, svbfloat16_t);
	2977	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za32_f16_f16)))
	2978	void svmop4s_2x1_za32_f16_f16(uint64_t, svfloat16x2_t, svfloat16_t);
	2979	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za32_f32_f32)))
	2980	void svmop4s_2x1_za32_f32_f32(uint64_t, svfloat32x2_t, svfloat32_t);
	2981	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za32_s8_s8)))
	2982	void svmop4s_2x1_za32_s8_s8(uint64_t, svint8x2_t, svint8_t);
	2983	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za32_s16_s16)))
	2984	void svmop4s_2x1_za32_s16_s16(uint64_t, svint16x2_t, svint16_t);
	2985	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za32_u8_u8)))
	2986	void svmop4s_2x1_za32_u8_u8(uint64_t, svuint8x2_t, svuint8_t);
	2987	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za32_u16_u16)))
	2988	void svmop4s_2x1_za32_u16_u16(uint64_t, svuint16x2_t, svuint16_t);
	2989	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za32_s8_u8)))
	2990	void svmop4s_2x2_za32_s8_u8(uint64_t, svint8x2_t, svuint8x2_t);
	2991	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za32_u8_s8)))
	2992	void svmop4s_2x2_za32_u8_s8(uint64_t, svuint8x2_t, svint8x2_t);
	2993	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za32_bf16_bf16)))
	2994	void svmop4s_2x2_za32_bf16_bf16(uint64_t, svbfloat16x2_t, svbfloat16x2_t);
	2995	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za32_f16_f16)))
	2996	void svmop4s_2x2_za32_f16_f16(uint64_t, svfloat16x2_t, svfloat16x2_t);
	2997	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za32_f32_f32)))
	2998	void svmop4s_2x2_za32_f32_f32(uint64_t, svfloat32x2_t, svfloat32x2_t);
	2999	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za32_s8_s8)))
	3000	void svmop4s_2x2_za32_s8_s8(uint64_t, svint8x2_t, svint8x2_t);
	3001	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za32_s16_s16)))
	3002	void svmop4s_2x2_za32_s16_s16(uint64_t, svint16x2_t, svint16x2_t);
	3003	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za32_u8_u8)))
	3004	void svmop4s_2x2_za32_u8_u8(uint64_t, svuint8x2_t, svuint8x2_t);
	3005	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za32_u16_u16)))
	3006	void svmop4s_2x2_za32_u16_u16(uint64_t, svuint16x2_t, svuint16x2_t);
	3007	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za32_s8_u8)))
	3008	void svmop4a_za32(uint64_t, svint8_t, svuint8_t);
	3009	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za32_u8_s8)))
	3010	void svmop4a_za32(uint64_t, svuint8_t, svint8_t);
	3011	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za32_bf16_bf16)))
	3012	void svmop4a_za32(uint64_t, svbfloat16_t, svbfloat16_t);
	3013	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za32_f16_f16)))
	3014	void svmop4a_za32(uint64_t, svfloat16_t, svfloat16_t);
	3015	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za32_f32_f32)))
	3016	void svmop4a_za32(uint64_t, svfloat32_t, svfloat32_t);
	3017	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za32_s8_s8)))
	3018	void svmop4a_za32(uint64_t, svint8_t, svint8_t);
	3019	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za32_s16_s16)))
	3020	void svmop4a_za32(uint64_t, svint16_t, svint16_t);
	3021	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za32_u8_u8)))
	3022	void svmop4a_za32(uint64_t, svuint8_t, svuint8_t);
	3023	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za32_u16_u16)))
	3024	void svmop4a_za32(uint64_t, svuint16_t, svuint16_t);
	3025	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za32_s8_u8)))
	3026	void svmop4a_za32(uint64_t, svint8_t, svuint8x2_t);
	3027	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za32_u8_s8)))
	3028	void svmop4a_za32(uint64_t, svuint8_t, svint8x2_t);
	3029	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za32_bf16_bf16)))
	3030	void svmop4a_za32(uint64_t, svbfloat16_t, svbfloat16x2_t);
	3031	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za32_f16_f16)))
	3032	void svmop4a_za32(uint64_t, svfloat16_t, svfloat16x2_t);
	3033	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za32_f32_f32)))
	3034	void svmop4a_za32(uint64_t, svfloat32_t, svfloat32x2_t);
	3035	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za32_s8_s8)))
	3036	void svmop4a_za32(uint64_t, svint8_t, svint8x2_t);
	3037	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za32_s16_s16)))
	3038	void svmop4a_za32(uint64_t, svint16_t, svint16x2_t);
	3039	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za32_u8_u8)))
	3040	void svmop4a_za32(uint64_t, svuint8_t, svuint8x2_t);
	3041	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za32_u16_u16)))
	3042	void svmop4a_za32(uint64_t, svuint16_t, svuint16x2_t);
	3043	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za32_s8_u8)))
	3044	void svmop4a_za32(uint64_t, svint8x2_t, svuint8_t);
	3045	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za32_u8_s8)))
	3046	void svmop4a_za32(uint64_t, svuint8x2_t, svint8_t);
	3047	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za32_bf16_bf16)))
	3048	void svmop4a_za32(uint64_t, svbfloat16x2_t, svbfloat16_t);
	3049	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za32_f16_f16)))
	3050	void svmop4a_za32(uint64_t, svfloat16x2_t, svfloat16_t);
	3051	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za32_f32_f32)))
	3052	void svmop4a_za32(uint64_t, svfloat32x2_t, svfloat32_t);
	3053	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za32_s8_s8)))
	3054	void svmop4a_za32(uint64_t, svint8x2_t, svint8_t);
	3055	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za32_s16_s16)))
	3056	void svmop4a_za32(uint64_t, svint16x2_t, svint16_t);
	3057	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za32_u8_u8)))
	3058	void svmop4a_za32(uint64_t, svuint8x2_t, svuint8_t);
	3059	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za32_u16_u16)))
	3060	void svmop4a_za32(uint64_t, svuint16x2_t, svuint16_t);
	3061	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za32_s8_u8)))
	3062	void svmop4a_za32(uint64_t, svint8x2_t, svuint8x2_t);
	3063	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za32_u8_s8)))
	3064	void svmop4a_za32(uint64_t, svuint8x2_t, svint8x2_t);
	3065	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za32_bf16_bf16)))
	3066	void svmop4a_za32(uint64_t, svbfloat16x2_t, svbfloat16x2_t);
	3067	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za32_f16_f16)))
	3068	void svmop4a_za32(uint64_t, svfloat16x2_t, svfloat16x2_t);
	3069	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za32_f32_f32)))
	3070	void svmop4a_za32(uint64_t, svfloat32x2_t, svfloat32x2_t);
	3071	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za32_s8_s8)))
	3072	void svmop4a_za32(uint64_t, svint8x2_t, svint8x2_t);
	3073	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za32_s16_s16)))
	3074	void svmop4a_za32(uint64_t, svint16x2_t, svint16x2_t);
	3075	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za32_u8_u8)))
	3076	void svmop4a_za32(uint64_t, svuint8x2_t, svuint8x2_t);
	3077	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za32_u16_u16)))
	3078	void svmop4a_za32(uint64_t, svuint16x2_t, svuint16x2_t);
	3079	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za32_s8_u8)))
	3080	void svmop4s_za32(uint64_t, svint8_t, svuint8_t);
	3081	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za32_u8_s8)))
	3082	void svmop4s_za32(uint64_t, svuint8_t, svint8_t);
	3083	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za32_bf16_bf16)))
	3084	void svmop4s_za32(uint64_t, svbfloat16_t, svbfloat16_t);
	3085	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za32_f16_f16)))
	3086	void svmop4s_za32(uint64_t, svfloat16_t, svfloat16_t);
	3087	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za32_f32_f32)))
	3088	void svmop4s_za32(uint64_t, svfloat32_t, svfloat32_t);
	3089	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za32_s8_s8)))
	3090	void svmop4s_za32(uint64_t, svint8_t, svint8_t);
	3091	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za32_s16_s16)))
	3092	void svmop4s_za32(uint64_t, svint16_t, svint16_t);
	3093	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za32_u8_u8)))
	3094	void svmop4s_za32(uint64_t, svuint8_t, svuint8_t);
	3095	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za32_u16_u16)))
	3096	void svmop4s_za32(uint64_t, svuint16_t, svuint16_t);
	3097	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za32_s8_u8)))
	3098	void svmop4s_za32(uint64_t, svint8_t, svuint8x2_t);
	3099	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za32_u8_s8)))
	3100	void svmop4s_za32(uint64_t, svuint8_t, svint8x2_t);
	3101	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za32_bf16_bf16)))
	3102	void svmop4s_za32(uint64_t, svbfloat16_t, svbfloat16x2_t);
	3103	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za32_f16_f16)))
	3104	void svmop4s_za32(uint64_t, svfloat16_t, svfloat16x2_t);
	3105	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za32_f32_f32)))
	3106	void svmop4s_za32(uint64_t, svfloat32_t, svfloat32x2_t);
	3107	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za32_s8_s8)))
	3108	void svmop4s_za32(uint64_t, svint8_t, svint8x2_t);
	3109	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za32_s16_s16)))
	3110	void svmop4s_za32(uint64_t, svint16_t, svint16x2_t);
	3111	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za32_u8_u8)))
	3112	void svmop4s_za32(uint64_t, svuint8_t, svuint8x2_t);
	3113	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za32_u16_u16)))
	3114	void svmop4s_za32(uint64_t, svuint16_t, svuint16x2_t);
	3115	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za32_s8_u8)))
	3116	void svmop4s_za32(uint64_t, svint8x2_t, svuint8_t);
	3117	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za32_u8_s8)))
	3118	void svmop4s_za32(uint64_t, svuint8x2_t, svint8_t);
	3119	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za32_bf16_bf16)))
	3120	void svmop4s_za32(uint64_t, svbfloat16x2_t, svbfloat16_t);
	3121	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za32_f16_f16)))
	3122	void svmop4s_za32(uint64_t, svfloat16x2_t, svfloat16_t);
	3123	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za32_f32_f32)))
	3124	void svmop4s_za32(uint64_t, svfloat32x2_t, svfloat32_t);
	3125	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za32_s8_s8)))
	3126	void svmop4s_za32(uint64_t, svint8x2_t, svint8_t);
	3127	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za32_s16_s16)))
	3128	void svmop4s_za32(uint64_t, svint16x2_t, svint16_t);
	3129	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za32_u8_u8)))
	3130	void svmop4s_za32(uint64_t, svuint8x2_t, svuint8_t);
	3131	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za32_u16_u16)))
	3132	void svmop4s_za32(uint64_t, svuint16x2_t, svuint16_t);
	3133	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za32_s8_u8)))
	3134	void svmop4s_za32(uint64_t, svint8x2_t, svuint8x2_t);
	3135	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za32_u8_s8)))
	3136	void svmop4s_za32(uint64_t, svuint8x2_t, svint8x2_t);
	3137	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za32_bf16_bf16)))
	3138	void svmop4s_za32(uint64_t, svbfloat16x2_t, svbfloat16x2_t);
	3139	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za32_f16_f16)))
	3140	void svmop4s_za32(uint64_t, svfloat16x2_t, svfloat16x2_t);
	3141	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za32_f32_f32)))
	3142	void svmop4s_za32(uint64_t, svfloat32x2_t, svfloat32x2_t);
	3143	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za32_s8_s8)))
	3144	void svmop4s_za32(uint64_t, svint8x2_t, svint8x2_t);
	3145	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za32_s16_s16)))
	3146	void svmop4s_za32(uint64_t, svint16x2_t, svint16x2_t);
	3147	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za32_u8_u8)))
	3148	void svmop4s_za32(uint64_t, svuint8x2_t, svuint8x2_t);
	3149	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za32_u16_u16)))
	3150	void svmop4s_za32(uint64_t, svuint16x2_t, svuint16x2_t);
	3151	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za16_bf16_bf16)))
	3152	void svmop4a_1x1_za16_bf16_bf16(uint64_t, svbfloat16_t, svbfloat16_t);
	3153	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za16_bf16_bf16)))
	3154	void svmop4a_1x2_za16_bf16_bf16(uint64_t, svbfloat16_t, svbfloat16x2_t);
	3155	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za16_bf16_bf16)))
	3156	void svmop4a_2x1_za16_bf16_bf16(uint64_t, svbfloat16x2_t, svbfloat16_t);
	3157	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za16_bf16_bf16)))
	3158	void svmop4a_2x2_za16_bf16_bf16(uint64_t, svbfloat16x2_t, svbfloat16x2_t);
	3159	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za16_bf16_bf16)))
	3160	void svmop4s_1x1_za16_bf16_bf16(uint64_t, svbfloat16_t, svbfloat16_t);
	3161	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za16_bf16_bf16)))
	3162	void svmop4s_1x2_za16_bf16_bf16(uint64_t, svbfloat16_t, svbfloat16x2_t);
	3163	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za16_bf16_bf16)))
	3164	void svmop4s_2x1_za16_bf16_bf16(uint64_t, svbfloat16x2_t, svbfloat16_t);
	3165	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za16_bf16_bf16)))
	3166	void svmop4s_2x2_za16_bf16_bf16(uint64_t, svbfloat16x2_t, svbfloat16x2_t);
	3167	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za16_bf16_bf16)))
	3168	void svmop4a_za16(uint64_t, svbfloat16_t, svbfloat16_t);
	3169	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za16_bf16_bf16)))
	3170	void svmop4a_za16(uint64_t, svbfloat16_t, svbfloat16x2_t);
	3171	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za16_bf16_bf16)))
	3172	void svmop4a_za16(uint64_t, svbfloat16x2_t, svbfloat16_t);
	3173	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za16_bf16_bf16)))
	3174	void svmop4a_za16(uint64_t, svbfloat16x2_t, svbfloat16x2_t);
	3175	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za16_bf16_bf16)))
	3176	void svmop4s_za16(uint64_t, svbfloat16_t, svbfloat16_t);
	3177	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za16_bf16_bf16)))
	3178	void svmop4s_za16(uint64_t, svbfloat16_t, svbfloat16x2_t);
	3179	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za16_bf16_bf16)))
	3180	void svmop4s_za16(uint64_t, svbfloat16x2_t, svbfloat16_t);
	3181	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za16_bf16_bf16)))
	3182	void svmop4s_za16(uint64_t, svbfloat16x2_t, svbfloat16x2_t);
	3183	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za16_f16_f16)))
	3184	void svmop4a_1x1_za16_f16_f16(uint64_t, svfloat16_t, svfloat16_t);
	3185	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za16_f16_f16)))
	3186	void svmop4a_1x2_za16_f16_f16(uint64_t, svfloat16_t, svfloat16x2_t);
	3187	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za16_f16_f16)))
	3188	void svmop4a_2x1_za16_f16_f16(uint64_t, svfloat16x2_t, svfloat16_t);
	3189	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za16_f16_f16)))
	3190	void svmop4a_2x2_za16_f16_f16(uint64_t, svfloat16x2_t, svfloat16x2_t);
	3191	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za16_f16_f16)))
	3192	void svmop4s_1x1_za16_f16_f16(uint64_t, svfloat16_t, svfloat16_t);
	3193	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za16_f16_f16)))
	3194	void svmop4s_1x2_za16_f16_f16(uint64_t, svfloat16_t, svfloat16x2_t);
	3195	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za16_f16_f16)))
	3196	void svmop4s_2x1_za16_f16_f16(uint64_t, svfloat16x2_t, svfloat16_t);
	3197	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za16_f16_f16)))
	3198	void svmop4s_2x2_za16_f16_f16(uint64_t, svfloat16x2_t, svfloat16x2_t);
	3199	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za16_f16_f16)))
	3200	void svmop4a_za16(uint64_t, svfloat16_t, svfloat16_t);
	3201	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za16_f16_f16)))
	3202	void svmop4a_za16(uint64_t, svfloat16_t, svfloat16x2_t);
	3203	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za16_f16_f16)))
	3204	void svmop4a_za16(uint64_t, svfloat16x2_t, svfloat16_t);
	3205	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za16_f16_f16)))
	3206	void svmop4a_za16(uint64_t, svfloat16x2_t, svfloat16x2_t);
	3207	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za16_f16_f16)))
	3208	void svmop4s_za16(uint64_t, svfloat16_t, svfloat16_t);
	3209	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za16_f16_f16)))
	3210	void svmop4s_za16(uint64_t, svfloat16_t, svfloat16x2_t);
	3211	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za16_f16_f16)))
	3212	void svmop4s_za16(uint64_t, svfloat16x2_t, svfloat16_t);
	3213	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za16_f16_f16)))
	3214	void svmop4s_za16(uint64_t, svfloat16x2_t, svfloat16x2_t);
	3215	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za64_f64_f64)))
	3216	void svmop4a_1x1_za64_f64_f64(uint64_t, svfloat64_t, svfloat64_t);
	3217	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za64_f64_f64)))
	3218	void svmop4a_1x2_za64_f64_f64(uint64_t, svfloat64_t, svfloat64x2_t);
	3219	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za64_f64_f64)))
	3220	void svmop4a_2x1_za64_f64_f64(uint64_t, svfloat64x2_t, svfloat64_t);
	3221	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za64_f64_f64)))
	3222	void svmop4a_2x2_za64_f64_f64(uint64_t, svfloat64x2_t, svfloat64x2_t);
	3223	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za64_f64_f64)))
	3224	void svmop4s_1x1_za64_f64_f64(uint64_t, svfloat64_t, svfloat64_t);
	3225	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za64_f64_f64)))
	3226	void svmop4s_1x2_za64_f64_f64(uint64_t, svfloat64_t, svfloat64x2_t);
	3227	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za64_f64_f64)))
	3228	void svmop4s_2x1_za64_f64_f64(uint64_t, svfloat64x2_t, svfloat64_t);
	3229	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za64_f64_f64)))
	3230	void svmop4s_2x2_za64_f64_f64(uint64_t, svfloat64x2_t, svfloat64x2_t);
	3231	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za64_f64_f64)))
	3232	void svmop4a_za64(uint64_t, svfloat64_t, svfloat64_t);
	3233	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za64_f64_f64)))
	3234	void svmop4a_za64(uint64_t, svfloat64_t, svfloat64x2_t);
	3235	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za64_f64_f64)))
	3236	void svmop4a_za64(uint64_t, svfloat64x2_t, svfloat64_t);
	3237	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za64_f64_f64)))
	3238	void svmop4a_za64(uint64_t, svfloat64x2_t, svfloat64x2_t);
	3239	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za64_f64_f64)))
	3240	void svmop4s_za64(uint64_t, svfloat64_t, svfloat64_t);
	3241	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za64_f64_f64)))
	3242	void svmop4s_za64(uint64_t, svfloat64_t, svfloat64x2_t);
	3243	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za64_f64_f64)))
	3244	void svmop4s_za64(uint64_t, svfloat64x2_t, svfloat64_t);
	3245	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za64_f64_f64)))
	3246	void svmop4s_za64(uint64_t, svfloat64x2_t, svfloat64x2_t);
	3247	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za16_mf8_mf8_fpm)))
	3248	void svmop4a_1x1_za16_mf8_mf8_fpm(uint64_t, svmfloat8_t, svmfloat8_t, fpm_t);
	3249	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za16_mf8_mf8_fpm)))
	3250	void svmop4a_1x2_za16_mf8_mf8_fpm(uint64_t, svmfloat8_t, svmfloat8x2_t, fpm_t);
	3251	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za16_mf8_mf8_fpm)))
	3252	void svmop4a_2x1_za16_mf8_mf8_fpm(uint64_t, svmfloat8x2_t, svmfloat8_t, fpm_t);
	3253	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za16_mf8_mf8_fpm)))
	3254	void svmop4a_2x2_za16_mf8_mf8_fpm(uint64_t, svmfloat8x2_t, svmfloat8x2_t, fpm_t);
	3255	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za16_mf8_mf8_fpm)))
	3256	void svmop4a_za16_fpm(uint64_t, svmfloat8_t, svmfloat8_t, fpm_t);
	3257	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za16_mf8_mf8_fpm)))
	3258	void svmop4a_za16_fpm(uint64_t, svmfloat8_t, svmfloat8x2_t, fpm_t);
	3259	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za16_mf8_mf8_fpm)))
	3260	void svmop4a_za16_fpm(uint64_t, svmfloat8x2_t, svmfloat8_t, fpm_t);
	3261	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za16_mf8_mf8_fpm)))
	3262	void svmop4a_za16_fpm(uint64_t, svmfloat8x2_t, svmfloat8x2_t, fpm_t);
	3263	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za32_mf8_mf8_fpm)))
	3264	void svmop4a_1x1_za32_mf8_mf8_fpm(uint64_t, svmfloat8_t, svmfloat8_t, fpm_t);
	3265	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za32_mf8_mf8_fpm)))
	3266	void svmop4a_1x2_za32_mf8_mf8_fpm(uint64_t, svmfloat8_t, svmfloat8x2_t, fpm_t);
	3267	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za32_mf8_mf8_fpm)))
	3268	void svmop4a_2x1_za32_mf8_mf8_fpm(uint64_t, svmfloat8x2_t, svmfloat8_t, fpm_t);
	3269	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za32_mf8_mf8_fpm)))
	3270	void svmop4a_2x2_za32_mf8_mf8_fpm(uint64_t, svmfloat8x2_t, svmfloat8x2_t, fpm_t);
	3271	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za32_mf8_mf8_fpm)))
	3272	void svmop4a_za32_fpm(uint64_t, svmfloat8_t, svmfloat8_t, fpm_t);
	3273	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za32_mf8_mf8_fpm)))
	3274	void svmop4a_za32_fpm(uint64_t, svmfloat8_t, svmfloat8x2_t, fpm_t);
	3275	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za32_mf8_mf8_fpm)))
	3276	void svmop4a_za32_fpm(uint64_t, svmfloat8x2_t, svmfloat8_t, fpm_t);
	3277	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za32_mf8_mf8_fpm)))
	3278	void svmop4a_za32_fpm(uint64_t, svmfloat8x2_t, svmfloat8x2_t, fpm_t);
	3279	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za64_s16_u16)))
	3280	void svmop4a_1x1_za64_s16_u16(uint64_t, svint16_t, svuint16_t);
	3281	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za64_u16_s16)))
	3282	void svmop4a_1x1_za64_u16_s16(uint64_t, svuint16_t, svint16_t);
	3283	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za64_s16_s16)))
	3284	void svmop4a_1x1_za64_s16_s16(uint64_t, svint16_t, svint16_t);
	3285	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za64_u16_u16)))
	3286	void svmop4a_1x1_za64_u16_u16(uint64_t, svuint16_t, svuint16_t);
	3287	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za64_s16_u16)))
	3288	void svmop4a_1x2_za64_s16_u16(uint64_t, svint16_t, svuint16x2_t);
	3289	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za64_u16_s16)))
	3290	void svmop4a_1x2_za64_u16_s16(uint64_t, svuint16_t, svint16x2_t);
	3291	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za64_s16_s16)))
	3292	void svmop4a_1x2_za64_s16_s16(uint64_t, svint16_t, svint16x2_t);
	3293	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za64_u16_u16)))
	3294	void svmop4a_1x2_za64_u16_u16(uint64_t, svuint16_t, svuint16x2_t);
	3295	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za64_s16_u16)))
	3296	void svmop4a_2x1_za64_s16_u16(uint64_t, svint16x2_t, svuint16_t);
	3297	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za64_u16_s16)))
	3298	void svmop4a_2x1_za64_u16_s16(uint64_t, svuint16x2_t, svint16_t);
	3299	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za64_s16_s16)))
	3300	void svmop4a_2x1_za64_s16_s16(uint64_t, svint16x2_t, svint16_t);
	3301	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za64_u16_u16)))
	3302	void svmop4a_2x1_za64_u16_u16(uint64_t, svuint16x2_t, svuint16_t);
	3303	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za64_s16_u16)))
	3304	void svmop4a_2x2_za64_s16_u16(uint64_t, svint16x2_t, svuint16x2_t);
	3305	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za64_u16_s16)))
	3306	void svmop4a_2x2_za64_u16_s16(uint64_t, svuint16x2_t, svint16x2_t);
	3307	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za64_s16_s16)))
	3308	void svmop4a_2x2_za64_s16_s16(uint64_t, svint16x2_t, svint16x2_t);
	3309	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za64_u16_u16)))
	3310	void svmop4a_2x2_za64_u16_u16(uint64_t, svuint16x2_t, svuint16x2_t);
	3311	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za64_s16_u16)))
	3312	void svmop4s_1x1_za64_s16_u16(uint64_t, svint16_t, svuint16_t);
	3313	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za64_u16_s16)))
	3314	void svmop4s_1x1_za64_u16_s16(uint64_t, svuint16_t, svint16_t);
	3315	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za64_s16_s16)))
	3316	void svmop4s_1x1_za64_s16_s16(uint64_t, svint16_t, svint16_t);
	3317	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za64_u16_u16)))
	3318	void svmop4s_1x1_za64_u16_u16(uint64_t, svuint16_t, svuint16_t);
	3319	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za64_s16_u16)))
	3320	void svmop4s_1x2_za64_s16_u16(uint64_t, svint16_t, svuint16x2_t);
	3321	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za64_u16_s16)))
	3322	void svmop4s_1x2_za64_u16_s16(uint64_t, svuint16_t, svint16x2_t);
	3323	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za64_s16_s16)))
	3324	void svmop4s_1x2_za64_s16_s16(uint64_t, svint16_t, svint16x2_t);
	3325	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za64_u16_u16)))
	3326	void svmop4s_1x2_za64_u16_u16(uint64_t, svuint16_t, svuint16x2_t);
	3327	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za64_s16_u16)))
	3328	void svmop4s_2x1_za64_s16_u16(uint64_t, svint16x2_t, svuint16_t);
	3329	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za64_u16_s16)))
	3330	void svmop4s_2x1_za64_u16_s16(uint64_t, svuint16x2_t, svint16_t);
	3331	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za64_s16_s16)))
	3332	void svmop4s_2x1_za64_s16_s16(uint64_t, svint16x2_t, svint16_t);
	3333	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za64_u16_u16)))
	3334	void svmop4s_2x1_za64_u16_u16(uint64_t, svuint16x2_t, svuint16_t);
	3335	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za64_s16_u16)))
	3336	void svmop4s_2x2_za64_s16_u16(uint64_t, svint16x2_t, svuint16x2_t);
	3337	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za64_u16_s16)))
	3338	void svmop4s_2x2_za64_u16_s16(uint64_t, svuint16x2_t, svint16x2_t);
	3339	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za64_s16_s16)))
	3340	void svmop4s_2x2_za64_s16_s16(uint64_t, svint16x2_t, svint16x2_t);
	3341	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za64_u16_u16)))
	3342	void svmop4s_2x2_za64_u16_u16(uint64_t, svuint16x2_t, svuint16x2_t);
	3343	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za64_s16_u16)))
	3344	void svmop4a_za64(uint64_t, svint16_t, svuint16_t);
	3345	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za64_u16_s16)))
	3346	void svmop4a_za64(uint64_t, svuint16_t, svint16_t);
	3347	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za64_s16_s16)))
	3348	void svmop4a_za64(uint64_t, svint16_t, svint16_t);
	3349	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x1_za64_u16_u16)))
	3350	void svmop4a_za64(uint64_t, svuint16_t, svuint16_t);
	3351	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za64_s16_u16)))
	3352	void svmop4a_za64(uint64_t, svint16_t, svuint16x2_t);
	3353	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za64_u16_s16)))
	3354	void svmop4a_za64(uint64_t, svuint16_t, svint16x2_t);
	3355	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za64_s16_s16)))
	3356	void svmop4a_za64(uint64_t, svint16_t, svint16x2_t);
	3357	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_1x2_za64_u16_u16)))
	3358	void svmop4a_za64(uint64_t, svuint16_t, svuint16x2_t);
	3359	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za64_s16_u16)))
	3360	void svmop4a_za64(uint64_t, svint16x2_t, svuint16_t);
	3361	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za64_u16_s16)))
	3362	void svmop4a_za64(uint64_t, svuint16x2_t, svint16_t);
	3363	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za64_s16_s16)))
	3364	void svmop4a_za64(uint64_t, svint16x2_t, svint16_t);
	3365	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x1_za64_u16_u16)))
	3366	void svmop4a_za64(uint64_t, svuint16x2_t, svuint16_t);
	3367	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za64_s16_u16)))
	3368	void svmop4a_za64(uint64_t, svint16x2_t, svuint16x2_t);
	3369	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za64_u16_s16)))
	3370	void svmop4a_za64(uint64_t, svuint16x2_t, svint16x2_t);
	3371	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za64_s16_s16)))
	3372	void svmop4a_za64(uint64_t, svint16x2_t, svint16x2_t);
	3373	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4a_2x2_za64_u16_u16)))
	3374	void svmop4a_za64(uint64_t, svuint16x2_t, svuint16x2_t);
	3375	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za64_s16_u16)))
	3376	void svmop4s_za64(uint64_t, svint16_t, svuint16_t);
	3377	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za64_u16_s16)))
	3378	void svmop4s_za64(uint64_t, svuint16_t, svint16_t);
	3379	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za64_s16_s16)))
	3380	void svmop4s_za64(uint64_t, svint16_t, svint16_t);
	3381	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x1_za64_u16_u16)))
	3382	void svmop4s_za64(uint64_t, svuint16_t, svuint16_t);
	3383	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za64_s16_u16)))
	3384	void svmop4s_za64(uint64_t, svint16_t, svuint16x2_t);
	3385	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za64_u16_s16)))
	3386	void svmop4s_za64(uint64_t, svuint16_t, svint16x2_t);
	3387	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za64_s16_s16)))
	3388	void svmop4s_za64(uint64_t, svint16_t, svint16x2_t);
	3389	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_1x2_za64_u16_u16)))
	3390	void svmop4s_za64(uint64_t, svuint16_t, svuint16x2_t);
	3391	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za64_s16_u16)))
	3392	void svmop4s_za64(uint64_t, svint16x2_t, svuint16_t);
	3393	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za64_u16_s16)))
	3394	void svmop4s_za64(uint64_t, svuint16x2_t, svint16_t);
	3395	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za64_s16_s16)))
	3396	void svmop4s_za64(uint64_t, svint16x2_t, svint16_t);
	3397	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x1_za64_u16_u16)))
	3398	void svmop4s_za64(uint64_t, svuint16x2_t, svuint16_t);
	3399	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za64_s16_u16)))
	3400	void svmop4s_za64(uint64_t, svint16x2_t, svuint16x2_t);
	3401	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za64_u16_s16)))
	3402	void svmop4s_za64(uint64_t, svuint16x2_t, svint16x2_t);
	3403	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za64_s16_s16)))
	3404	void svmop4s_za64(uint64_t, svint16x2_t, svint16x2_t);
	3405	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svmop4s_2x2_za64_u16_u16)))
	3406	void svmop4s_za64(uint64_t, svuint16x2_t, svuint16x2_t);
	3407	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svtmopa_lane_za32_s8_u8)))
	3408	void svtmopa_lane_za32_s8_u8(uint64_t, svint8x2_t, svuint8_t, svuint8_t, uint64_t);
	3409	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svtmopa_lane_za32_u8_s8)))
	3410	void svtmopa_lane_za32_u8_s8(uint64_t, svuint8x2_t, svint8_t, svuint8_t, uint64_t);
	3411	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svtmopa_lane_za32_s8_s8)))
	3412	void svtmopa_lane_za32_s8_s8(uint64_t, svint8x2_t, svint8_t, svuint8_t, uint64_t);
	3413	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svtmopa_lane_za32_s16_s16)))
	3414	void svtmopa_lane_za32_s16_s16(uint64_t, svint16x2_t, svint16_t, svuint8_t, uint64_t);
	3415	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svtmopa_lane_za32_bf16_bf16)))
	3416	void svtmopa_lane_za32_bf16_bf16(uint64_t, svbfloat16x2_t, svbfloat16_t, svuint8_t, uint64_t);
	3417	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svtmopa_lane_za32_f32_f32)))
	3418	void svtmopa_lane_za32_f32_f32(uint64_t, svfloat32x2_t, svfloat32_t, svuint8_t, uint64_t);
	3419	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svtmopa_lane_za32_f16_f16)))
	3420	void svtmopa_lane_za32_f16_f16(uint64_t, svfloat16x2_t, svfloat16_t, svuint8_t, uint64_t);
	3421	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svtmopa_lane_za32_u8_u8)))
	3422	void svtmopa_lane_za32_u8_u8(uint64_t, svuint8x2_t, svuint8_t, svuint8_t, uint64_t);
	3423	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svtmopa_lane_za32_u16_u16)))
	3424	void svtmopa_lane_za32_u16_u16(uint64_t, svuint16x2_t, svuint16_t, svuint8_t, uint64_t);
	3425	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svtmopa_lane_za32_s8_u8)))
	3426	void svtmopa_lane_za32(uint64_t, svint8x2_t, svuint8_t, svuint8_t, uint64_t);
	3427	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svtmopa_lane_za32_u8_s8)))
	3428	void svtmopa_lane_za32(uint64_t, svuint8x2_t, svint8_t, svuint8_t, uint64_t);
	3429	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svtmopa_lane_za32_s8_s8)))
	3430	void svtmopa_lane_za32(uint64_t, svint8x2_t, svint8_t, svuint8_t, uint64_t);
	3431	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svtmopa_lane_za32_s16_s16)))
	3432	void svtmopa_lane_za32(uint64_t, svint16x2_t, svint16_t, svuint8_t, uint64_t);
	3433	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svtmopa_lane_za32_bf16_bf16)))
	3434	void svtmopa_lane_za32(uint64_t, svbfloat16x2_t, svbfloat16_t, svuint8_t, uint64_t);
	3435	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svtmopa_lane_za32_f32_f32)))
	3436	void svtmopa_lane_za32(uint64_t, svfloat32x2_t, svfloat32_t, svuint8_t, uint64_t);
	3437	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svtmopa_lane_za32_f16_f16)))
	3438	void svtmopa_lane_za32(uint64_t, svfloat16x2_t, svfloat16_t, svuint8_t, uint64_t);
	3439	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svtmopa_lane_za32_u8_u8)))
	3440	void svtmopa_lane_za32(uint64_t, svuint8x2_t, svuint8_t, svuint8_t, uint64_t);
	3441	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svtmopa_lane_za32_u16_u16)))
	3442	void svtmopa_lane_za32(uint64_t, svuint16x2_t, svuint16_t, svuint8_t, uint64_t);
	3443	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svtmopa_lane_za16_bf16_bf16)))
	3444	void svtmopa_lane_za16_bf16_bf16(uint64_t, svbfloat16x2_t, svbfloat16_t, svuint8_t, uint64_t);
	3445	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svtmopa_lane_za16_bf16_bf16)))
	3446	void svtmopa_lane_za16(uint64_t, svbfloat16x2_t, svbfloat16_t, svuint8_t, uint64_t);
	3447	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svtmopa_lane_za16_f16_f16)))
	3448	void svtmopa_lane_za16_f16_f16(uint64_t, svfloat16x2_t, svfloat16_t, svuint8_t, uint64_t);
	3449	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svtmopa_lane_za16_f16_f16)))
	3450	void svtmopa_lane_za16(uint64_t, svfloat16x2_t, svfloat16_t, svuint8_t, uint64_t);
	3451	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svtmopa_lane_za16_mf8_mf8_fpm)))
	3452	void svtmopa_lane_za16_mf8_mf8_fpm(uint64_t, svmfloat8x2_t, svmfloat8_t, svuint8_t, uint64_t, fpm_t);
	3453	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svtmopa_lane_za16_mf8_mf8_fpm)))
	3454	void svtmopa_lane_za16_fpm(uint64_t, svmfloat8x2_t, svmfloat8_t, svuint8_t, uint64_t, fpm_t);
	3455	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svtmopa_lane_za32_mf8_mf8_fpm)))
	3456	void svtmopa_lane_za32_mf8_mf8_fpm(uint64_t, svmfloat8x2_t, svmfloat8_t, svuint8_t, uint64_t, fpm_t);
	3457	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sme_svtmopa_lane_za32_mf8_mf8_fpm)))
	3458	void svtmopa_lane_za32_fpm(uint64_t, svmfloat8x2_t, svmfloat8_t, svuint8_t, uint64_t, fpm_t);
2785	3459	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_hor_za128_u8)))
2786	3460	svuint8_t svreadz_hor_za128_u8(uint64_t, uint32_t);
2787	3461	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_hor_za128_u32)))
...	...	@@ -2804,6 +3478,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_hor_za128_s3
2804	3478	svint32_t svreadz_hor_za128_s32(uint64_t, uint32_t);
2805	3479	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_hor_za128_s64)))
2806	3480	svint64_t svreadz_hor_za128_s64(uint64_t, uint32_t);
	3481	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_hor_za128_mf8)))
	3482	svmfloat8_t svreadz_hor_za128_mf8(uint64_t, uint32_t);
2807	3483	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_hor_za128_s16)))
2808	3484	svint16_t svreadz_hor_za128_s16(uint64_t, uint32_t);
2809	3485	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_hor_za16_u16)))
...	...	@@ -2870,14 +3546,20 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_hor_za8_u8))
2870	3546	svuint8_t svreadz_hor_za8_u8(uint64_t, uint32_t);
2871	3547	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_hor_za8_s8)))
2872	3548	svint8_t svreadz_hor_za8_s8(uint64_t, uint32_t);
	3549	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_hor_za8_mf8)))
	3550	svmfloat8_t svreadz_hor_za8_mf8(uint64_t, uint32_t);
2873	3551	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_hor_za8_u8_vg2)))
2874	3552	svuint8x2_t svreadz_hor_za8_u8_vg2(uint64_t, uint32_t);
2875	3553	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_hor_za8_s8_vg2)))
2876	3554	svint8x2_t svreadz_hor_za8_s8_vg2(uint64_t, uint32_t);
	3555	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_hor_za8_mf8_vg2)))
	3556	svmfloat8x2_t svreadz_hor_za8_mf8_vg2(uint64_t, uint32_t);
2877	3557	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_hor_za8_u8_vg4)))
2878	3558	svuint8x4_t svreadz_hor_za8_u8_vg4(uint64_t, uint32_t);
2879	3559	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_hor_za8_s8_vg4)))
2880	3560	svint8x4_t svreadz_hor_za8_s8_vg4(uint64_t, uint32_t);
	3561	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_hor_za8_mf8_vg4)))
	3562	svmfloat8x4_t svreadz_hor_za8_mf8_vg4(uint64_t, uint32_t);
2881	3563	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_ver_za128_u8)))
2882	3564	svuint8_t svreadz_ver_za128_u8(uint64_t, uint32_t);
2883	3565	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_ver_za128_u32)))
...	...	@@ -2900,6 +3582,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_ver_za128_s3
2900	3582	svint32_t svreadz_ver_za128_s32(uint64_t, uint32_t);
2901	3583	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_ver_za128_s64)))
2902	3584	svint64_t svreadz_ver_za128_s64(uint64_t, uint32_t);
	3585	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_ver_za128_mf8)))
	3586	svmfloat8_t svreadz_ver_za128_mf8(uint64_t, uint32_t);
2903	3587	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_ver_za128_s16)))
2904	3588	svint16_t svreadz_ver_za128_s16(uint64_t, uint32_t);
2905	3589	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_ver_za16_u16)))
...	...	@@ -2966,14 +3650,20 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_ver_za8_u8))
2966	3650	svuint8_t svreadz_ver_za8_u8(uint64_t, uint32_t);
2967	3651	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_ver_za8_s8)))
2968	3652	svint8_t svreadz_ver_za8_s8(uint64_t, uint32_t);
	3653	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_ver_za8_mf8)))
	3654	svmfloat8_t svreadz_ver_za8_mf8(uint64_t, uint32_t);
2969	3655	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_ver_za8_u8_vg2)))
2970	3656	svuint8x2_t svreadz_ver_za8_u8_vg2(uint64_t, uint32_t);
2971	3657	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_ver_za8_s8_vg2)))
2972	3658	svint8x2_t svreadz_ver_za8_s8_vg2(uint64_t, uint32_t);
	3659	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_ver_za8_mf8_vg2)))
	3660	svmfloat8x2_t svreadz_ver_za8_mf8_vg2(uint64_t, uint32_t);
2973	3661	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_ver_za8_u8_vg4)))
2974	3662	svuint8x4_t svreadz_ver_za8_u8_vg4(uint64_t, uint32_t);
2975	3663	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_ver_za8_s8_vg4)))
2976	3664	svint8x4_t svreadz_ver_za8_s8_vg4(uint64_t, uint32_t);
	3665	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_ver_za8_mf8_vg4)))
	3666	svmfloat8x4_t svreadz_ver_za8_mf8_vg4(uint64_t, uint32_t);
2977	3667	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_za16_u16_vg1x2)))
2978	3668	svuint16x2_t svreadz_za16_u16_vg1x2(uint32_t);
2979	3669	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_za16_bf16_vg1x2)))
...	...	@@ -3018,10 +3708,14 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_za8_u8_vg1x2
3018	3708	svuint8x2_t svreadz_za8_u8_vg1x2(uint32_t);
3019	3709	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_za8_s8_vg1x2)))
3020	3710	svint8x2_t svreadz_za8_s8_vg1x2(uint32_t);
	3711	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_za8_mf8_vg1x2)))
	3712	svmfloat8x2_t svreadz_za8_mf8_vg1x2(uint32_t);
3021	3713	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_za8_u8_vg1x4)))
3022	3714	svuint8x4_t svreadz_za8_u8_vg1x4(uint32_t);
3023	3715	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_za8_s8_vg1x4)))
3024	3716	svint8x4_t svreadz_za8_s8_vg1x4(uint32_t);
	3717	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svreadz_za8_mf8_vg1x4)))
	3718	svmfloat8x4_t svreadz_za8_mf8_vg1x4(uint32_t);
3025	3719	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svzero_za64_vg1x2)))
3026	3720	void svzero_za64_vg1x2(uint32_t);
3027	3721	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sme_svzero_za64_vg1x4)))

lib/include/arm_sve.h+5110-4770

...	...	@@ -3359,6 +3359,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsel_s32_x2)))
3359	3359	svint32x2_t svsel_s32_x2(svcount_t, svint32x2_t, svint32x2_t);
3360	3360	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsel_s64_x2)))
3361	3361	svint64x2_t svsel_s64_x2(svcount_t, svint64x2_t, svint64x2_t);
	3362	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsel_mf8_x2)))
	3363	svmfloat8x2_t svsel_mf8_x2(svcount_t, svmfloat8x2_t, svmfloat8x2_t);
3362	3364	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsel_s16_x2)))
3363	3365	svint16x2_t svsel_s16_x2(svcount_t, svint16x2_t, svint16x2_t);
3364	3366	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsel_u8_x4)))
...	...	@@ -3383,6 +3385,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsel_s32_x4)))
3383	3385	svint32x4_t svsel_s32_x4(svcount_t, svint32x4_t, svint32x4_t);
3384	3386	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsel_s64_x4)))
3385	3387	svint64x4_t svsel_s64_x4(svcount_t, svint64x4_t, svint64x4_t);
	3388	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsel_mf8_x4)))
	3389	svmfloat8x4_t svsel_mf8_x4(svcount_t, svmfloat8x4_t, svmfloat8x4_t);
3386	3390	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsel_s16_x4)))
3387	3391	svint16x4_t svsel_s16_x4(svcount_t, svint16x4_t, svint16x4_t);
3388	3392	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svunpk_s32_s16_x2)))
...	...	@@ -3431,6 +3435,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp_s32_x2)))
3431	3435	svint32x2_t svuzp_s32_x2(svint32x2_t);
3432	3436	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp_s64_x2)))
3433	3437	svint64x2_t svuzp_s64_x2(svint64x2_t);
	3438	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp_mf8_x2)))
	3439	svmfloat8x2_t svuzp_mf8_x2(svmfloat8x2_t);
3434	3440	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp_s16_x2)))
3435	3441	svint16x2_t svuzp_s16_x2(svint16x2_t);
3436	3442	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp_u8_x4)))
...	...	@@ -3455,6 +3461,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp_s32_x4)))
3455	3461	svint32x4_t svuzp_s32_x4(svint32x4_t);
3456	3462	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp_s64_x4)))
3457	3463	svint64x4_t svuzp_s64_x4(svint64x4_t);
	3464	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp_mf8_x4)))
	3465	svmfloat8x4_t svuzp_mf8_x4(svmfloat8x4_t);
3458	3466	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp_s16_x4)))
3459	3467	svint16x4_t svuzp_s16_x4(svint16x4_t);
3460	3468	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq_u8_x2)))
...	...	@@ -3479,6 +3487,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq_s32_x2)))
3479	3487	svint32x2_t svuzpq_s32_x2(svint32x2_t);
3480	3488	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq_s64_x2)))
3481	3489	svint64x2_t svuzpq_s64_x2(svint64x2_t);
	3490	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq_mf8_x2)))
	3491	svmfloat8x2_t svuzpq_mf8_x2(svmfloat8x2_t);
3482	3492	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq_s16_x2)))
3483	3493	svint16x2_t svuzpq_s16_x2(svint16x2_t);
3484	3494	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq_u8_x4)))
...	...	@@ -3503,6 +3513,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq_s32_x4)))
3503	3513	svint32x4_t svuzpq_s32_x4(svint32x4_t);
3504	3514	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq_s64_x4)))
3505	3515	svint64x4_t svuzpq_s64_x4(svint64x4_t);
	3516	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq_mf8_x4)))
	3517	svmfloat8x4_t svuzpq_mf8_x4(svmfloat8x4_t);
3506	3518	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq_s16_x4)))
3507	3519	svint16x4_t svuzpq_s16_x4(svint16x4_t);
3508	3520	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip_u8_x2)))
...	...	@@ -3527,6 +3539,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip_s32_x2)))
3527	3539	svint32x2_t svzip_s32_x2(svint32x2_t);
3528	3540	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip_s64_x2)))
3529	3541	svint64x2_t svzip_s64_x2(svint64x2_t);
	3542	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip_mf8_x2)))
	3543	svmfloat8x2_t svzip_mf8_x2(svmfloat8x2_t);
3530	3544	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip_s16_x2)))
3531	3545	svint16x2_t svzip_s16_x2(svint16x2_t);
3532	3546	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip_u8_x4)))
...	...	@@ -3551,6 +3565,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip_s32_x4)))
3551	3565	svint32x4_t svzip_s32_x4(svint32x4_t);
3552	3566	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip_s64_x4)))
3553	3567	svint64x4_t svzip_s64_x4(svint64x4_t);
	3568	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip_mf8_x4)))
	3569	svmfloat8x4_t svzip_mf8_x4(svmfloat8x4_t);
3554	3570	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip_s16_x4)))
3555	3571	svint16x4_t svzip_s16_x4(svint16x4_t);
3556	3572	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq_u8_x2)))
...	...	@@ -3575,6 +3591,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq_s32_x2)))
3575	3591	svint32x2_t svzipq_s32_x2(svint32x2_t);
3576	3592	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq_s64_x2)))
3577	3593	svint64x2_t svzipq_s64_x2(svint64x2_t);
	3594	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq_mf8_x2)))
	3595	svmfloat8x2_t svzipq_mf8_x2(svmfloat8x2_t);
3578	3596	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq_s16_x2)))
3579	3597	svint16x2_t svzipq_s16_x2(svint16x2_t);
3580	3598	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq_u8_x4)))
...	...	@@ -3599,6 +3617,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq_s32_x4)))
3599	3617	svint32x4_t svzipq_s32_x4(svint32x4_t);
3600	3618	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq_s64_x4)))
3601	3619	svint64x4_t svzipq_s64_x4(svint64x4_t);
	3620	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq_mf8_x4)))
	3621	svmfloat8x4_t svzipq_mf8_x4(svmfloat8x4_t);
3602	3622	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq_s16_x4)))
3603	3623	svint16x4_t svzipq_s16_x4(svint16x4_t);
3604	3624	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svadd_single_u8_x2)))
...	...	@@ -4119,6 +4139,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsel_s32_x2)))
4119	4139	svint32x2_t svsel(svcount_t, svint32x2_t, svint32x2_t);
4120	4140	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsel_s64_x2)))
4121	4141	svint64x2_t svsel(svcount_t, svint64x2_t, svint64x2_t);
	4142	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsel_mf8_x2)))
	4143	svmfloat8x2_t svsel(svcount_t, svmfloat8x2_t, svmfloat8x2_t);
4122	4144	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsel_s16_x2)))
4123	4145	svint16x2_t svsel(svcount_t, svint16x2_t, svint16x2_t);
4124	4146	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsel_u8_x4)))
...	...	@@ -4143,6 +4165,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsel_s32_x4)))
4143	4165	svint32x4_t svsel(svcount_t, svint32x4_t, svint32x4_t);
4144	4166	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsel_s64_x4)))
4145	4167	svint64x4_t svsel(svcount_t, svint64x4_t, svint64x4_t);
	4168	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsel_mf8_x4)))
	4169	svmfloat8x4_t svsel(svcount_t, svmfloat8x4_t, svmfloat8x4_t);
4146	4170	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsel_s16_x4)))
4147	4171	svint16x4_t svsel(svcount_t, svint16x4_t, svint16x4_t);
4148	4172	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svunpk_s32_s16_x2)))
...	...	@@ -4191,6 +4215,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp_s32_x2)))
4191	4215	svint32x2_t svuzp(svint32x2_t);
4192	4216	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp_s64_x2)))
4193	4217	svint64x2_t svuzp(svint64x2_t);
	4218	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp_mf8_x2)))
	4219	svmfloat8x2_t svuzp(svmfloat8x2_t);
4194	4220	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp_s16_x2)))
4195	4221	svint16x2_t svuzp(svint16x2_t);
4196	4222	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp_u8_x4)))
...	...	@@ -4215,6 +4241,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp_s32_x4)))
4215	4241	svint32x4_t svuzp(svint32x4_t);
4216	4242	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp_s64_x4)))
4217	4243	svint64x4_t svuzp(svint64x4_t);
	4244	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp_mf8_x4)))
	4245	svmfloat8x4_t svuzp(svmfloat8x4_t);
4218	4246	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp_s16_x4)))
4219	4247	svint16x4_t svuzp(svint16x4_t);
4220	4248	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq_u8_x2)))
...	...	@@ -4239,6 +4267,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq_s32_x2)))
4239	4267	svint32x2_t svuzpq(svint32x2_t);
4240	4268	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq_s64_x2)))
4241	4269	svint64x2_t svuzpq(svint64x2_t);
	4270	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq_mf8_x2)))
	4271	svmfloat8x2_t svuzpq(svmfloat8x2_t);
4242	4272	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq_s16_x2)))
4243	4273	svint16x2_t svuzpq(svint16x2_t);
4244	4274	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq_u8_x4)))
...	...	@@ -4263,6 +4293,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq_s32_x4)))
4263	4293	svint32x4_t svuzpq(svint32x4_t);
4264	4294	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq_s64_x4)))
4265	4295	svint64x4_t svuzpq(svint64x4_t);
	4296	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq_mf8_x4)))
	4297	svmfloat8x4_t svuzpq(svmfloat8x4_t);
4266	4298	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq_s16_x4)))
4267	4299	svint16x4_t svuzpq(svint16x4_t);
4268	4300	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip_u8_x2)))
...	...	@@ -4287,6 +4319,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip_s32_x2)))
4287	4319	svint32x2_t svzip(svint32x2_t);
4288	4320	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip_s64_x2)))
4289	4321	svint64x2_t svzip(svint64x2_t);
	4322	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip_mf8_x2)))
	4323	svmfloat8x2_t svzip(svmfloat8x2_t);
4290	4324	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip_s16_x2)))
4291	4325	svint16x2_t svzip(svint16x2_t);
4292	4326	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip_u8_x4)))
...	...	@@ -4311,6 +4345,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip_s32_x4)))
4311	4345	svint32x4_t svzip(svint32x4_t);
4312	4346	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip_s64_x4)))
4313	4347	svint64x4_t svzip(svint64x4_t);
	4348	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip_mf8_x4)))
	4349	svmfloat8x4_t svzip(svmfloat8x4_t);
4314	4350	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip_s16_x4)))
4315	4351	svint16x4_t svzip(svint16x4_t);
4316	4352	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq_u8_x2)))
...	...	@@ -4335,6 +4371,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq_s32_x2)))
4335	4371	svint32x2_t svzipq(svint32x2_t);
4336	4372	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq_s64_x2)))
4337	4373	svint64x2_t svzipq(svint64x2_t);
	4374	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq_mf8_x2)))
	4375	svmfloat8x2_t svzipq(svmfloat8x2_t);
4338	4376	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq_s16_x2)))
4339	4377	svint16x2_t svzipq(svint16x2_t);
4340	4378	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq_u8_x4)))
...	...	@@ -4359,6 +4397,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq_s32_x4)))
4359	4397	svint32x4_t svzipq(svint32x4_t);
4360	4398	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq_s64_x4)))
4361	4399	svint64x4_t svzipq(svint64x4_t);
	4400	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq_mf8_x4)))
	4401	svmfloat8x4_t svzipq(svmfloat8x4_t);
4362	4402	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq_s16_x4)))
4363	4403	svint16x4_t svzipq(svint16x4_t);
4364	4404	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f64_x2)))
...	...	@@ -4627,12 +4667,6 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcompact_s32)))
4627	4667	svint32_t svcompact_s32(svbool_t, svint32_t);
4628	4668	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcompact_s64)))
4629	4669	svint64_t svcompact_s64(svbool_t, svint64_t);
4630		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svexpa_f64)))
4631		svfloat64_t svexpa_f64(svuint64_t);
4632		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svexpa_f32)))
4633		svfloat32_t svexpa_f32(svuint32_t);
4634		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svexpa_f16)))
4635		svfloat16_t svexpa_f16(svuint16_t);
4636	4670	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_gather_u32base_index_u32)))
4637	4671	svuint32_t svld1_gather_u32base_index_u32(svbool_t, svuint32_t, int64_t);
4638	4672	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_gather_u64base_index_u64)))
...	...	@@ -4957,6 +4991,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_u64)))
4957	4991	svuint64_t svldff1_u64(svbool_t, uint64_t const *);
4958	4992	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_u16)))
4959	4993	svuint16_t svldff1_u16(svbool_t, uint16_t const *);
	4994	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_bf16)))
	4995	svbfloat16_t svldff1_bf16(svbool_t, bfloat16_t const *);
4960	4996	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_s8)))
4961	4997	svint8_t svldff1_s8(svbool_t, int8_t const *);
4962	4998	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_f64)))
...	...	@@ -4969,6 +5005,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_s32)))
4969	5005	svint32_t svldff1_s32(svbool_t, int32_t const *);
4970	5006	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_s64)))
4971	5007	svint64_t svldff1_s64(svbool_t, int64_t const *);
	5008	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_mf8)))
	5009	svmfloat8_t svldff1_mf8(svbool_t, mfloat8_t const *);
4972	5010	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_s16)))
4973	5011	svint16_t svldff1_s16(svbool_t, int16_t const *);
4974	5012	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_gather_u32base_index_u32)))
...	...	@@ -5063,6 +5101,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_vnum_u64)))
5063	5101	svuint64_t svldff1_vnum_u64(svbool_t, uint64_t const *, int64_t);
5064	5102	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_vnum_u16)))
5065	5103	svuint16_t svldff1_vnum_u16(svbool_t, uint16_t const *, int64_t);
	5104	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_vnum_bf16)))
	5105	svbfloat16_t svldff1_vnum_bf16(svbool_t, bfloat16_t const *, int64_t);
5066	5106	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_vnum_s8)))
5067	5107	svint8_t svldff1_vnum_s8(svbool_t, int8_t const *, int64_t);
5068	5108	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_vnum_f64)))
...	...	@@ -5075,6 +5115,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_vnum_s32)))
5075	5115	svint32_t svldff1_vnum_s32(svbool_t, int32_t const *, int64_t);
5076	5116	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_vnum_s64)))
5077	5117	svint64_t svldff1_vnum_s64(svbool_t, int64_t const *, int64_t);
	5118	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_vnum_mf8)))
	5119	svmfloat8_t svldff1_vnum_mf8(svbool_t, mfloat8_t const *, int64_t);
5078	5120	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_vnum_s16)))
5079	5121	svint16_t svldff1_vnum_s16(svbool_t, int16_t const *, int64_t);
5080	5122	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1sb_gather_u32base_offset_u32)))
...	...	@@ -5413,6 +5455,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_u64)))
5413	5455	svuint64_t svldnf1_u64(svbool_t, uint64_t const *);
5414	5456	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_u16)))
5415	5457	svuint16_t svldnf1_u16(svbool_t, uint16_t const *);
	5458	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_bf16)))
	5459	svbfloat16_t svldnf1_bf16(svbool_t, bfloat16_t const *);
5416	5460	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_s8)))
5417	5461	svint8_t svldnf1_s8(svbool_t, int8_t const *);
5418	5462	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_f64)))
...	...	@@ -5425,6 +5469,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_s32)))
5425	5469	svint32_t svldnf1_s32(svbool_t, int32_t const *);
5426	5470	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_s64)))
5427	5471	svint64_t svldnf1_s64(svbool_t, int64_t const *);
	5472	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_mf8)))
	5473	svmfloat8_t svldnf1_mf8(svbool_t, mfloat8_t const *);
5428	5474	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_s16)))
5429	5475	svint16_t svldnf1_s16(svbool_t, int16_t const *);
5430	5476	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_vnum_u8)))
...	...	@@ -5435,6 +5481,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_vnum_u64)))
5435	5481	svuint64_t svldnf1_vnum_u64(svbool_t, uint64_t const *, int64_t);
5436	5482	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_vnum_u16)))
5437	5483	svuint16_t svldnf1_vnum_u16(svbool_t, uint16_t const *, int64_t);
	5484	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_vnum_bf16)))
	5485	svbfloat16_t svldnf1_vnum_bf16(svbool_t, bfloat16_t const *, int64_t);
5438	5486	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_vnum_s8)))
5439	5487	svint8_t svldnf1_vnum_s8(svbool_t, int8_t const *, int64_t);
5440	5488	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_vnum_f64)))
...	...	@@ -5447,6 +5495,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_vnum_s32)))
5447	5495	svint32_t svldnf1_vnum_s32(svbool_t, int32_t const *, int64_t);
5448	5496	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_vnum_s64)))
5449	5497	svint64_t svldnf1_vnum_s64(svbool_t, int64_t const *, int64_t);
	5498	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_vnum_mf8)))
	5499	svmfloat8_t svldnf1_vnum_mf8(svbool_t, mfloat8_t const *, int64_t);
5450	5500	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_vnum_s16)))
5451	5501	svint16_t svldnf1_vnum_s16(svbool_t, int16_t const *, int64_t);
5452	5502	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1sb_vnum_u32)))
...	...	@@ -5885,12 +5935,6 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcompact_s32)))
5885	5935	svint32_t svcompact(svbool_t, svint32_t);
5886	5936	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcompact_s64)))
5887	5937	svint64_t svcompact(svbool_t, svint64_t);
5888		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svexpa_f64)))
5889		svfloat64_t svexpa(svuint64_t);
5890		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svexpa_f32)))
5891		svfloat32_t svexpa(svuint32_t);
5892		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svexpa_f16)))
5893		svfloat16_t svexpa(svuint16_t);
5894	5938	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_gather_u32base_index_u32)))
5895	5939	svuint32_t svld1_gather_index_u32(svbool_t, svuint32_t, int64_t);
5896	5940	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_gather_u64base_index_u64)))
...	...	@@ -6215,6 +6259,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_u64)))
6215	6259	svuint64_t svldff1(svbool_t, uint64_t const *);
6216	6260	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_u16)))
6217	6261	svuint16_t svldff1(svbool_t, uint16_t const *);
	6262	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_bf16)))
	6263	svbfloat16_t svldff1(svbool_t, bfloat16_t const *);
6218	6264	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_s8)))
6219	6265	svint8_t svldff1(svbool_t, int8_t const *);
6220	6266	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_f64)))
...	...	@@ -6227,6 +6273,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_s32)))
6227	6273	svint32_t svldff1(svbool_t, int32_t const *);
6228	6274	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_s64)))
6229	6275	svint64_t svldff1(svbool_t, int64_t const *);
	6276	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_mf8)))
	6277	svmfloat8_t svldff1(svbool_t, mfloat8_t const *);
6230	6278	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_s16)))
6231	6279	svint16_t svldff1(svbool_t, int16_t const *);
6232	6280	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_gather_u32base_index_u32)))
...	...	@@ -6321,6 +6369,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_vnum_u64)))
6321	6369	svuint64_t svldff1_vnum(svbool_t, uint64_t const *, int64_t);
6322	6370	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_vnum_u16)))
6323	6371	svuint16_t svldff1_vnum(svbool_t, uint16_t const *, int64_t);
	6372	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_vnum_bf16)))
	6373	svbfloat16_t svldff1_vnum(svbool_t, bfloat16_t const *, int64_t);
6324	6374	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_vnum_s8)))
6325	6375	svint8_t svldff1_vnum(svbool_t, int8_t const *, int64_t);
6326	6376	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_vnum_f64)))
...	...	@@ -6333,6 +6383,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_vnum_s32)))
6333	6383	svint32_t svldff1_vnum(svbool_t, int32_t const *, int64_t);
6334	6384	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_vnum_s64)))
6335	6385	svint64_t svldff1_vnum(svbool_t, int64_t const *, int64_t);
	6386	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_vnum_mf8)))
	6387	svmfloat8_t svldff1_vnum(svbool_t, mfloat8_t const *, int64_t);
6336	6388	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_vnum_s16)))
6337	6389	svint16_t svldff1_vnum(svbool_t, int16_t const *, int64_t);
6338	6390	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1sb_gather_u32base_offset_u32)))
...	...	@@ -6575,6 +6627,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_u64)))
6575	6627	svuint64_t svldnf1(svbool_t, uint64_t const *);
6576	6628	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_u16)))
6577	6629	svuint16_t svldnf1(svbool_t, uint16_t const *);
	6630	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_bf16)))
	6631	svbfloat16_t svldnf1(svbool_t, bfloat16_t const *);
6578	6632	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_s8)))
6579	6633	svint8_t svldnf1(svbool_t, int8_t const *);
6580	6634	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_f64)))
...	...	@@ -6587,6 +6641,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_s32)))
6587	6641	svint32_t svldnf1(svbool_t, int32_t const *);
6588	6642	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_s64)))
6589	6643	svint64_t svldnf1(svbool_t, int64_t const *);
	6644	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_mf8)))
	6645	svmfloat8_t svldnf1(svbool_t, mfloat8_t const *);
6590	6646	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_s16)))
6591	6647	svint16_t svldnf1(svbool_t, int16_t const *);
6592	6648	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_vnum_u8)))
...	...	@@ -6597,6 +6653,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_vnum_u64)))
6597	6653	svuint64_t svldnf1_vnum(svbool_t, uint64_t const *, int64_t);
6598	6654	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_vnum_u16)))
6599	6655	svuint16_t svldnf1_vnum(svbool_t, uint16_t const *, int64_t);
	6656	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_vnum_bf16)))
	6657	svbfloat16_t svldnf1_vnum(svbool_t, bfloat16_t const *, int64_t);
6600	6658	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_vnum_s8)))
6601	6659	svint8_t svldnf1_vnum(svbool_t, int8_t const *, int64_t);
6602	6660	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_vnum_f64)))
...	...	@@ -6609,6 +6667,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_vnum_s32)))
6609	6667	svint32_t svldnf1_vnum(svbool_t, int32_t const *, int64_t);
6610	6668	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_vnum_s64)))
6611	6669	svint64_t svldnf1_vnum(svbool_t, int64_t const *, int64_t);
	6670	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_vnum_mf8)))
	6671	svmfloat8_t svldnf1_vnum(svbool_t, mfloat8_t const *, int64_t);
6612	6672	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_vnum_s16)))
6613	6673	svint16_t svldnf1_vnum(svbool_t, int16_t const *, int64_t);
6614	6674	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svprfb_gather_u32base)))
...	...	@@ -6893,4496 +6953,4578 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtssel_f32)))
6893	6953	svfloat32_t svtssel(svfloat32_t, svuint32_t);
6894	6954	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtssel_f16)))
6895	6955	svfloat16_t svtssel(svfloat16_t, svuint16_t);
6896		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmmla_f32)))
6897		svfloat32_t svbfmmla_f32(svfloat32_t, svbfloat16_t, svbfloat16_t);
6898		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_bf16)))
6899		svbfloat16_t svldff1_bf16(svbool_t, bfloat16_t const *);
6900		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_vnum_bf16)))
6901		svbfloat16_t svldff1_vnum_bf16(svbool_t, bfloat16_t const *, int64_t);
6902		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_bf16)))
6903		svbfloat16_t svldnf1_bf16(svbool_t, bfloat16_t const *);
6904		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_vnum_bf16)))
6905		svbfloat16_t svldnf1_vnum_bf16(svbool_t, bfloat16_t const *, int64_t);
6906		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmmla_f32)))
6907		svfloat32_t svbfmmla(svfloat32_t, svbfloat16_t, svbfloat16_t);
6908		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_bf16)))
6909		svbfloat16_t svldff1(svbool_t, bfloat16_t const *);
6910		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldff1_vnum_bf16)))
6911		svbfloat16_t svldff1_vnum(svbool_t, bfloat16_t const *, int64_t);
6912		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_bf16)))
6913		svbfloat16_t svldnf1(svbool_t, bfloat16_t const *);
6914		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnf1_vnum_bf16)))
6915		svbfloat16_t svldnf1_vnum(svbool_t, bfloat16_t const *, int64_t);
6916		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_bf16)))
6917		svbfloat16_t svtrn1q_bf16(svbfloat16_t, svbfloat16_t);
6918		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_bf16)))
6919		svbfloat16_t svtrn2q_bf16(svbfloat16_t, svbfloat16_t);
6920		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_bf16)))
6921		svbfloat16_t svuzp1q_bf16(svbfloat16_t, svbfloat16_t);
6922		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_bf16)))
6923		svbfloat16_t svuzp2q_bf16(svbfloat16_t, svbfloat16_t);
6924		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_bf16)))
6925		svbfloat16_t svzip1q_bf16(svbfloat16_t, svbfloat16_t);
6926		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_bf16)))
6927		svbfloat16_t svzip2q_bf16(svbfloat16_t, svbfloat16_t);
6928		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_bf16)))
6929		svbfloat16_t svtrn1q(svbfloat16_t, svbfloat16_t);
6930		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_bf16)))
6931		svbfloat16_t svtrn2q(svbfloat16_t, svbfloat16_t);
6932		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_bf16)))
6933		svbfloat16_t svuzp1q(svbfloat16_t, svbfloat16_t);
6934		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_bf16)))
6935		svbfloat16_t svuzp2q(svbfloat16_t, svbfloat16_t);
6936		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_bf16)))
6937		svbfloat16_t svzip1q(svbfloat16_t, svbfloat16_t);
6938		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_bf16)))
6939		svbfloat16_t svzip2q(svbfloat16_t, svbfloat16_t);
6940		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfdot_n_f32)))
6941		svfloat32_t svbfdot_n_f32(svfloat32_t, svbfloat16_t, bfloat16_t);
6942		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfdot_f32)))
6943		svfloat32_t svbfdot_f32(svfloat32_t, svbfloat16_t, svbfloat16_t);
6944		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfdot_lane_f32)))
6945		svfloat32_t svbfdot_lane_f32(svfloat32_t, svbfloat16_t, svbfloat16_t, uint64_t);
6946		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlalb_n_f32)))
6947		svfloat32_t svbfmlalb_n_f32(svfloat32_t, svbfloat16_t, bfloat16_t);
6948		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlalb_f32)))
6949		svfloat32_t svbfmlalb_f32(svfloat32_t, svbfloat16_t, svbfloat16_t);
6950		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlalb_lane_f32)))
6951		svfloat32_t svbfmlalb_lane_f32(svfloat32_t, svbfloat16_t, svbfloat16_t, uint64_t);
6952		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlalt_n_f32)))
6953		svfloat32_t svbfmlalt_n_f32(svfloat32_t, svbfloat16_t, bfloat16_t);
6954		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlalt_f32)))
6955		svfloat32_t svbfmlalt_f32(svfloat32_t, svbfloat16_t, svbfloat16_t);
6956		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlalt_lane_f32)))
6957		svfloat32_t svbfmlalt_lane_f32(svfloat32_t, svbfloat16_t, svbfloat16_t, uint64_t);
6958		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclasta_n_bf16)))
6959		bfloat16_t svclasta_n_bf16(svbool_t, bfloat16_t, svbfloat16_t);
6960		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclasta_bf16)))
6961		svbfloat16_t svclasta_bf16(svbool_t, svbfloat16_t, svbfloat16_t);
6962		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclastb_n_bf16)))
6963		bfloat16_t svclastb_n_bf16(svbool_t, bfloat16_t, svbfloat16_t);
6964		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclastb_bf16)))
6965		svbfloat16_t svclastb_bf16(svbool_t, svbfloat16_t, svbfloat16_t);
6966		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_bf16_m)))
6967		svuint16_t svcnt_bf16_m(svuint16_t, svbool_t, svbfloat16_t);
6968		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_bf16_x)))
6969		svuint16_t svcnt_bf16_x(svbool_t, svbfloat16_t);
6970		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_bf16_z)))
6971		svuint16_t svcnt_bf16_z(svbool_t, svbfloat16_t);
6972		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate2_bf16)))
6973		svbfloat16x2_t svcreate2_bf16(svbfloat16_t, svbfloat16_t);
6974		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate3_bf16)))
6975		svbfloat16x3_t svcreate3_bf16(svbfloat16_t, svbfloat16_t, svbfloat16_t);
6976		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate4_bf16)))
6977		svbfloat16x4_t svcreate4_bf16(svbfloat16_t, svbfloat16_t, svbfloat16_t, svbfloat16_t);
6978		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvt_bf16_f32_m)))
6979		svbfloat16_t svcvt_bf16_f32_m(svbfloat16_t, svbool_t, svfloat32_t);
6980		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvt_bf16_f32_x)))
6981		svbfloat16_t svcvt_bf16_f32_x(svbool_t, svfloat32_t);
6982		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvt_bf16_f32_z)))
6983		svbfloat16_t svcvt_bf16_f32_z(svbool_t, svfloat32_t);
6984		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtnt_bf16_f32_m)))
6985		svbfloat16_t svcvtnt_bf16_f32_m(svbfloat16_t, svbool_t, svfloat32_t);
6986		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_bf16)))
6987		svbfloat16_t svdup_n_bf16(bfloat16_t);
6988		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_bf16_m)))
6989		svbfloat16_t svdup_n_bf16_m(svbfloat16_t, svbool_t, bfloat16_t);
6990		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_bf16_x)))
6991		svbfloat16_t svdup_n_bf16_x(svbool_t, bfloat16_t);
6992		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_bf16_z)))
6993		svbfloat16_t svdup_n_bf16_z(svbool_t, bfloat16_t);
6994		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_lane_bf16)))
6995		svbfloat16_t svdup_lane_bf16(svbfloat16_t, uint16_t);
6996		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdupq_n_bf16)))
6997		svbfloat16_t svdupq_n_bf16(bfloat16_t, bfloat16_t, bfloat16_t, bfloat16_t, bfloat16_t, bfloat16_t, bfloat16_t, bfloat16_t);
6998		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdupq_lane_bf16)))
6999		svbfloat16_t svdupq_lane_bf16(svbfloat16_t, uint64_t);
7000		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svext_bf16)))
7001		svbfloat16_t svext_bf16(svbfloat16_t, svbfloat16_t, uint64_t);
7002		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget2_bf16)))
7003		svbfloat16_t svget2_bf16(svbfloat16x2_t, uint64_t);
7004		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget3_bf16)))
7005		svbfloat16_t svget3_bf16(svbfloat16x3_t, uint64_t);
7006		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget4_bf16)))
7007		svbfloat16_t svget4_bf16(svbfloat16x4_t, uint64_t);
7008		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svinsr_n_bf16)))
7009		svbfloat16_t svinsr_n_bf16(svbfloat16_t, bfloat16_t);
7010		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlasta_bf16)))
7011		bfloat16_t svlasta_bf16(svbool_t, svbfloat16_t);
7012		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlastb_bf16)))
7013		bfloat16_t svlastb_bf16(svbool_t, svbfloat16_t);
7014		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_bf16)))
7015		svbfloat16_t svld1_bf16(svbool_t, bfloat16_t const *);
7016		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_bf16)))
7017		svbfloat16_t svld1_vnum_bf16(svbool_t, bfloat16_t const *, int64_t);
7018		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1rq_bf16)))
7019		svbfloat16_t svld1rq_bf16(svbool_t, bfloat16_t const *);
7020		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_bf16)))
7021		svbfloat16x2_t svld2_bf16(svbool_t, bfloat16_t const *);
7022		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_vnum_bf16)))
7023		svbfloat16x2_t svld2_vnum_bf16(svbool_t, bfloat16_t const *, int64_t);
7024		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_bf16)))
7025		svbfloat16x3_t svld3_bf16(svbool_t, bfloat16_t const *);
7026		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_vnum_bf16)))
7027		svbfloat16x3_t svld3_vnum_bf16(svbool_t, bfloat16_t const *, int64_t);
7028		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_bf16)))
7029		svbfloat16x4_t svld4_bf16(svbool_t, bfloat16_t const *);
7030		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_vnum_bf16)))
7031		svbfloat16x4_t svld4_vnum_bf16(svbool_t, bfloat16_t const *, int64_t);
7032		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_bf16)))
7033		svbfloat16_t svldnt1_bf16(svbool_t, bfloat16_t const *);
7034		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_bf16)))
7035		svbfloat16_t svldnt1_vnum_bf16(svbool_t, bfloat16_t const *, int64_t);
7036		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlen_bf16)))
7037		uint64_t svlen_bf16(svbfloat16_t);
7038		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrev_bf16)))
7039		svbfloat16_t svrev_bf16(svbfloat16_t);
7040		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsel_bf16)))
7041		svbfloat16_t svsel_bf16(svbool_t, svbfloat16_t, svbfloat16_t);
7042		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset2_bf16)))
7043		svbfloat16x2_t svset2_bf16(svbfloat16x2_t, uint64_t, svbfloat16_t);
7044		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset3_bf16)))
7045		svbfloat16x3_t svset3_bf16(svbfloat16x3_t, uint64_t, svbfloat16_t);
7046		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset4_bf16)))
7047		svbfloat16x4_t svset4_bf16(svbfloat16x4_t, uint64_t, svbfloat16_t);
7048		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsplice_bf16)))
7049		svbfloat16_t svsplice_bf16(svbool_t, svbfloat16_t, svbfloat16_t);
7050		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_bf16)))
7051		void svst1_bf16(svbool_t, bfloat16_t *, svbfloat16_t);
7052		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_bf16)))
7053		void svst1_vnum_bf16(svbool_t, bfloat16_t *, int64_t, svbfloat16_t);
7054		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_bf16)))
7055		void svst2_bf16(svbool_t, bfloat16_t *, svbfloat16x2_t);
7056		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_vnum_bf16)))
7057		void svst2_vnum_bf16(svbool_t, bfloat16_t *, int64_t, svbfloat16x2_t);
7058		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_bf16)))
7059		void svst3_bf16(svbool_t, bfloat16_t *, svbfloat16x3_t);
7060		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_vnum_bf16)))
7061		void svst3_vnum_bf16(svbool_t, bfloat16_t *, int64_t, svbfloat16x3_t);
7062		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_bf16)))
7063		void svst4_bf16(svbool_t, bfloat16_t *, svbfloat16x4_t);
7064		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_vnum_bf16)))
7065		void svst4_vnum_bf16(svbool_t, bfloat16_t *, int64_t, svbfloat16x4_t);
7066		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_bf16)))
7067		void svstnt1_bf16(svbool_t, bfloat16_t *, svbfloat16_t);
7068		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_bf16)))
7069		void svstnt1_vnum_bf16(svbool_t, bfloat16_t *, int64_t, svbfloat16_t);
7070		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbl_bf16)))
7071		svbfloat16_t svtbl_bf16(svbfloat16_t, svuint16_t);
7072		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1_bf16)))
7073		svbfloat16_t svtrn1_bf16(svbfloat16_t, svbfloat16_t);
7074		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2_bf16)))
7075		svbfloat16_t svtrn2_bf16(svbfloat16_t, svbfloat16_t);
7076		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef2_bf16)))
7077		svbfloat16x2_t svundef2_bf16(void);
7078		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef3_bf16)))
7079		svbfloat16x3_t svundef3_bf16(void);
7080		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef4_bf16)))
7081		svbfloat16x4_t svundef4_bf16(void);
7082		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef_bf16)))
7083		svbfloat16_t svundef_bf16(void);
7084		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1_bf16)))
7085		svbfloat16_t svuzp1_bf16(svbfloat16_t, svbfloat16_t);
7086		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2_bf16)))
7087		svbfloat16_t svuzp2_bf16(svbfloat16_t, svbfloat16_t);
7088		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1_bf16)))
7089		svbfloat16_t svzip1_bf16(svbfloat16_t, svbfloat16_t);
7090		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2_bf16)))
7091		svbfloat16_t svzip2_bf16(svbfloat16_t, svbfloat16_t);
7092		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfdot_n_f32)))
7093		svfloat32_t svbfdot(svfloat32_t, svbfloat16_t, bfloat16_t);
7094		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfdot_f32)))
7095		svfloat32_t svbfdot(svfloat32_t, svbfloat16_t, svbfloat16_t);
7096		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfdot_lane_f32)))
7097		svfloat32_t svbfdot_lane(svfloat32_t, svbfloat16_t, svbfloat16_t, uint64_t);
7098		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlalb_n_f32)))
7099		svfloat32_t svbfmlalb(svfloat32_t, svbfloat16_t, bfloat16_t);
7100		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlalb_f32)))
7101		svfloat32_t svbfmlalb(svfloat32_t, svbfloat16_t, svbfloat16_t);
7102		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlalb_lane_f32)))
7103		svfloat32_t svbfmlalb_lane(svfloat32_t, svbfloat16_t, svbfloat16_t, uint64_t);
7104		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlalt_n_f32)))
7105		svfloat32_t svbfmlalt(svfloat32_t, svbfloat16_t, bfloat16_t);
7106		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlalt_f32)))
7107		svfloat32_t svbfmlalt(svfloat32_t, svbfloat16_t, svbfloat16_t);
7108		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlalt_lane_f32)))
7109		svfloat32_t svbfmlalt_lane(svfloat32_t, svbfloat16_t, svbfloat16_t, uint64_t);
7110		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclasta_n_bf16)))
7111		bfloat16_t svclasta(svbool_t, bfloat16_t, svbfloat16_t);
7112		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclasta_bf16)))
7113		svbfloat16_t svclasta(svbool_t, svbfloat16_t, svbfloat16_t);
7114		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclastb_n_bf16)))
7115		bfloat16_t svclastb(svbool_t, bfloat16_t, svbfloat16_t);
7116		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclastb_bf16)))
7117		svbfloat16_t svclastb(svbool_t, svbfloat16_t, svbfloat16_t);
7118		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_bf16_m)))
7119		svuint16_t svcnt_m(svuint16_t, svbool_t, svbfloat16_t);
7120		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_bf16_x)))
7121		svuint16_t svcnt_x(svbool_t, svbfloat16_t);
7122		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_bf16_z)))
7123		svuint16_t svcnt_z(svbool_t, svbfloat16_t);
7124		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate2_bf16)))
7125		svbfloat16x2_t svcreate2(svbfloat16_t, svbfloat16_t);
7126		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate3_bf16)))
7127		svbfloat16x3_t svcreate3(svbfloat16_t, svbfloat16_t, svbfloat16_t);
7128		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate4_bf16)))
7129		svbfloat16x4_t svcreate4(svbfloat16_t, svbfloat16_t, svbfloat16_t, svbfloat16_t);
7130		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvt_bf16_f32_m)))
7131		svbfloat16_t svcvt_bf16_m(svbfloat16_t, svbool_t, svfloat32_t);
7132		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvt_bf16_f32_x)))
7133		svbfloat16_t svcvt_bf16_x(svbool_t, svfloat32_t);
7134		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvt_bf16_f32_z)))
7135		svbfloat16_t svcvt_bf16_z(svbool_t, svfloat32_t);
7136		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtnt_bf16_f32_m)))
7137		svbfloat16_t svcvtnt_bf16_m(svbfloat16_t, svbool_t, svfloat32_t);
7138		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_bf16)))
7139		svbfloat16_t svdup_bf16(bfloat16_t);
7140		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_bf16_m)))
7141		svbfloat16_t svdup_bf16_m(svbfloat16_t, svbool_t, bfloat16_t);
7142		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_bf16_x)))
7143		svbfloat16_t svdup_bf16_x(svbool_t, bfloat16_t);
7144		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_bf16_z)))
7145		svbfloat16_t svdup_bf16_z(svbool_t, bfloat16_t);
7146		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_lane_bf16)))
7147		svbfloat16_t svdup_lane(svbfloat16_t, uint16_t);
7148		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdupq_n_bf16)))
7149		svbfloat16_t svdupq_bf16(bfloat16_t, bfloat16_t, bfloat16_t, bfloat16_t, bfloat16_t, bfloat16_t, bfloat16_t, bfloat16_t);
7150		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdupq_lane_bf16)))
7151		svbfloat16_t svdupq_lane(svbfloat16_t, uint64_t);
7152		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svext_bf16)))
7153		svbfloat16_t svext(svbfloat16_t, svbfloat16_t, uint64_t);
7154		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget2_bf16)))
7155		svbfloat16_t svget2(svbfloat16x2_t, uint64_t);
7156		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget3_bf16)))
7157		svbfloat16_t svget3(svbfloat16x3_t, uint64_t);
7158		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget4_bf16)))
7159		svbfloat16_t svget4(svbfloat16x4_t, uint64_t);
7160		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svinsr_n_bf16)))
7161		svbfloat16_t svinsr(svbfloat16_t, bfloat16_t);
7162		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlasta_bf16)))
7163		bfloat16_t svlasta(svbool_t, svbfloat16_t);
7164		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlastb_bf16)))
7165		bfloat16_t svlastb(svbool_t, svbfloat16_t);
7166		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_bf16)))
7167		svbfloat16_t svld1(svbool_t, bfloat16_t const *);
7168		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_bf16)))
7169		svbfloat16_t svld1_vnum(svbool_t, bfloat16_t const *, int64_t);
7170		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1rq_bf16)))
7171		svbfloat16_t svld1rq(svbool_t, bfloat16_t const *);
7172		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_bf16)))
7173		svbfloat16x2_t svld2(svbool_t, bfloat16_t const *);
7174		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_vnum_bf16)))
7175		svbfloat16x2_t svld2_vnum(svbool_t, bfloat16_t const *, int64_t);
7176		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_bf16)))
7177		svbfloat16x3_t svld3(svbool_t, bfloat16_t const *);
7178		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_vnum_bf16)))
7179		svbfloat16x3_t svld3_vnum(svbool_t, bfloat16_t const *, int64_t);
7180		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_bf16)))
7181		svbfloat16x4_t svld4(svbool_t, bfloat16_t const *);
7182		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_vnum_bf16)))
7183		svbfloat16x4_t svld4_vnum(svbool_t, bfloat16_t const *, int64_t);
7184		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_bf16)))
7185		svbfloat16_t svldnt1(svbool_t, bfloat16_t const *);
7186		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_bf16)))
7187		svbfloat16_t svldnt1_vnum(svbool_t, bfloat16_t const *, int64_t);
7188		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlen_bf16)))
7189		uint64_t svlen(svbfloat16_t);
7190		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrev_bf16)))
7191		svbfloat16_t svrev(svbfloat16_t);
7192		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsel_bf16)))
7193		svbfloat16_t svsel(svbool_t, svbfloat16_t, svbfloat16_t);
7194		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset2_bf16)))
7195		svbfloat16x2_t svset2(svbfloat16x2_t, uint64_t, svbfloat16_t);
7196		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset3_bf16)))
7197		svbfloat16x3_t svset3(svbfloat16x3_t, uint64_t, svbfloat16_t);
7198		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset4_bf16)))
7199		svbfloat16x4_t svset4(svbfloat16x4_t, uint64_t, svbfloat16_t);
7200		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsplice_bf16)))
7201		svbfloat16_t svsplice(svbool_t, svbfloat16_t, svbfloat16_t);
7202		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_bf16)))
7203		void svst1(svbool_t, bfloat16_t *, svbfloat16_t);
7204		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_bf16)))
7205		void svst1_vnum(svbool_t, bfloat16_t *, int64_t, svbfloat16_t);
7206		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_bf16)))
7207		void svst2(svbool_t, bfloat16_t *, svbfloat16x2_t);
7208		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_vnum_bf16)))
7209		void svst2_vnum(svbool_t, bfloat16_t *, int64_t, svbfloat16x2_t);
7210		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_bf16)))
7211		void svst3(svbool_t, bfloat16_t *, svbfloat16x3_t);
7212		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_vnum_bf16)))
7213		void svst3_vnum(svbool_t, bfloat16_t *, int64_t, svbfloat16x3_t);
7214		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_bf16)))
7215		void svst4(svbool_t, bfloat16_t *, svbfloat16x4_t);
7216		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_vnum_bf16)))
7217		void svst4_vnum(svbool_t, bfloat16_t *, int64_t, svbfloat16x4_t);
7218		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_bf16)))
7219		void svstnt1(svbool_t, bfloat16_t *, svbfloat16_t);
7220		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_bf16)))
7221		void svstnt1_vnum(svbool_t, bfloat16_t *, int64_t, svbfloat16_t);
7222		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbl_bf16)))
7223		svbfloat16_t svtbl(svbfloat16_t, svuint16_t);
7224		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1_bf16)))
7225		svbfloat16_t svtrn1(svbfloat16_t, svbfloat16_t);
7226		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2_bf16)))
7227		svbfloat16_t svtrn2(svbfloat16_t, svbfloat16_t);
7228		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1_bf16)))
7229		svbfloat16_t svuzp1(svbfloat16_t, svbfloat16_t);
7230		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2_bf16)))
7231		svbfloat16_t svuzp2(svbfloat16_t, svbfloat16_t);
7232		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1_bf16)))
7233		svbfloat16_t svzip1(svbfloat16_t, svbfloat16_t);
7234		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2_bf16)))
7235		svbfloat16_t svzip2(svbfloat16_t, svbfloat16_t);
7236		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmmla_f32)))
7237		svfloat32_t svmmla_f32(svfloat32_t, svfloat32_t, svfloat32_t);
7238		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmmla_f32)))
7239		svfloat32_t svmmla(svfloat32_t, svfloat32_t, svfloat32_t);
7240		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_u8)))
7241		svuint8_t svld1ro_u8(svbool_t, uint8_t const *);
7242		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_u32)))
7243		svuint32_t svld1ro_u32(svbool_t, uint32_t const *);
7244		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_u64)))
7245		svuint64_t svld1ro_u64(svbool_t, uint64_t const *);
7246		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_u16)))
7247		svuint16_t svld1ro_u16(svbool_t, uint16_t const *);
7248		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_s8)))
7249		svint8_t svld1ro_s8(svbool_t, int8_t const *);
7250		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_f64)))
7251		svfloat64_t svld1ro_f64(svbool_t, float64_t const *);
7252		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_f32)))
7253		svfloat32_t svld1ro_f32(svbool_t, float32_t const *);
7254		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_f16)))
7255		svfloat16_t svld1ro_f16(svbool_t, float16_t const *);
7256		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_s32)))
7257		svint32_t svld1ro_s32(svbool_t, int32_t const *);
7258		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_s64)))
7259		svint64_t svld1ro_s64(svbool_t, int64_t const *);
7260		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_s16)))
7261		svint16_t svld1ro_s16(svbool_t, int16_t const *);
7262		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmmla_f64)))
7263		svfloat64_t svmmla_f64(svfloat64_t, svfloat64_t, svfloat64_t);
7264		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_u8)))
7265		svuint8_t svtrn1q_u8(svuint8_t, svuint8_t);
7266		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_u32)))
7267		svuint32_t svtrn1q_u32(svuint32_t, svuint32_t);
7268		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_u64)))
7269		svuint64_t svtrn1q_u64(svuint64_t, svuint64_t);
7270		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_u16)))
7271		svuint16_t svtrn1q_u16(svuint16_t, svuint16_t);
7272		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_s8)))
7273		svint8_t svtrn1q_s8(svint8_t, svint8_t);
7274		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_f64)))
7275		svfloat64_t svtrn1q_f64(svfloat64_t, svfloat64_t);
7276		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_f32)))
7277		svfloat32_t svtrn1q_f32(svfloat32_t, svfloat32_t);
7278		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_f16)))
7279		svfloat16_t svtrn1q_f16(svfloat16_t, svfloat16_t);
7280		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_s32)))
7281		svint32_t svtrn1q_s32(svint32_t, svint32_t);
7282		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_s64)))
7283		svint64_t svtrn1q_s64(svint64_t, svint64_t);
7284		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_s16)))
7285		svint16_t svtrn1q_s16(svint16_t, svint16_t);
7286		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_u8)))
7287		svuint8_t svtrn2q_u8(svuint8_t, svuint8_t);
7288		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_u32)))
7289		svuint32_t svtrn2q_u32(svuint32_t, svuint32_t);
7290		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_u64)))
7291		svuint64_t svtrn2q_u64(svuint64_t, svuint64_t);
7292		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_u16)))
7293		svuint16_t svtrn2q_u16(svuint16_t, svuint16_t);
7294		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_s8)))
7295		svint8_t svtrn2q_s8(svint8_t, svint8_t);
7296		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_f64)))
7297		svfloat64_t svtrn2q_f64(svfloat64_t, svfloat64_t);
7298		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_f32)))
7299		svfloat32_t svtrn2q_f32(svfloat32_t, svfloat32_t);
7300		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_f16)))
7301		svfloat16_t svtrn2q_f16(svfloat16_t, svfloat16_t);
7302		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_s32)))
7303		svint32_t svtrn2q_s32(svint32_t, svint32_t);
7304		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_s64)))
7305		svint64_t svtrn2q_s64(svint64_t, svint64_t);
7306		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_s16)))
7307		svint16_t svtrn2q_s16(svint16_t, svint16_t);
7308		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_u8)))
7309		svuint8_t svuzp1q_u8(svuint8_t, svuint8_t);
7310		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_u32)))
7311		svuint32_t svuzp1q_u32(svuint32_t, svuint32_t);
7312		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_u64)))
7313		svuint64_t svuzp1q_u64(svuint64_t, svuint64_t);
7314		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_u16)))
7315		svuint16_t svuzp1q_u16(svuint16_t, svuint16_t);
7316		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_s8)))
7317		svint8_t svuzp1q_s8(svint8_t, svint8_t);
7318		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_f64)))
7319		svfloat64_t svuzp1q_f64(svfloat64_t, svfloat64_t);
7320		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_f32)))
7321		svfloat32_t svuzp1q_f32(svfloat32_t, svfloat32_t);
7322		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_f16)))
7323		svfloat16_t svuzp1q_f16(svfloat16_t, svfloat16_t);
7324		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_s32)))
7325		svint32_t svuzp1q_s32(svint32_t, svint32_t);
7326		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_s64)))
7327		svint64_t svuzp1q_s64(svint64_t, svint64_t);
7328		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_s16)))
7329		svint16_t svuzp1q_s16(svint16_t, svint16_t);
7330		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_u8)))
7331		svuint8_t svuzp2q_u8(svuint8_t, svuint8_t);
7332		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_u32)))
7333		svuint32_t svuzp2q_u32(svuint32_t, svuint32_t);
7334		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_u64)))
7335		svuint64_t svuzp2q_u64(svuint64_t, svuint64_t);
7336		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_u16)))
7337		svuint16_t svuzp2q_u16(svuint16_t, svuint16_t);
7338		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_s8)))
7339		svint8_t svuzp2q_s8(svint8_t, svint8_t);
7340		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_f64)))
7341		svfloat64_t svuzp2q_f64(svfloat64_t, svfloat64_t);
7342		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_f32)))
7343		svfloat32_t svuzp2q_f32(svfloat32_t, svfloat32_t);
7344		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_f16)))
7345		svfloat16_t svuzp2q_f16(svfloat16_t, svfloat16_t);
7346		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_s32)))
7347		svint32_t svuzp2q_s32(svint32_t, svint32_t);
7348		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_s64)))
7349		svint64_t svuzp2q_s64(svint64_t, svint64_t);
7350		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_s16)))
7351		svint16_t svuzp2q_s16(svint16_t, svint16_t);
7352		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_u8)))
7353		svuint8_t svzip1q_u8(svuint8_t, svuint8_t);
7354		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_u32)))
7355		svuint32_t svzip1q_u32(svuint32_t, svuint32_t);
7356		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_u64)))
7357		svuint64_t svzip1q_u64(svuint64_t, svuint64_t);
7358		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_u16)))
7359		svuint16_t svzip1q_u16(svuint16_t, svuint16_t);
7360		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_s8)))
7361		svint8_t svzip1q_s8(svint8_t, svint8_t);
7362		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_f64)))
7363		svfloat64_t svzip1q_f64(svfloat64_t, svfloat64_t);
7364		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_f32)))
7365		svfloat32_t svzip1q_f32(svfloat32_t, svfloat32_t);
7366		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_f16)))
7367		svfloat16_t svzip1q_f16(svfloat16_t, svfloat16_t);
7368		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_s32)))
7369		svint32_t svzip1q_s32(svint32_t, svint32_t);
7370		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_s64)))
7371		svint64_t svzip1q_s64(svint64_t, svint64_t);
7372		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_s16)))
7373		svint16_t svzip1q_s16(svint16_t, svint16_t);
7374		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_u8)))
7375		svuint8_t svzip2q_u8(svuint8_t, svuint8_t);
7376		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_u32)))
7377		svuint32_t svzip2q_u32(svuint32_t, svuint32_t);
7378		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_u64)))
7379		svuint64_t svzip2q_u64(svuint64_t, svuint64_t);
7380		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_u16)))
7381		svuint16_t svzip2q_u16(svuint16_t, svuint16_t);
7382		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_s8)))
7383		svint8_t svzip2q_s8(svint8_t, svint8_t);
7384		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_f64)))
7385		svfloat64_t svzip2q_f64(svfloat64_t, svfloat64_t);
7386		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_f32)))
7387		svfloat32_t svzip2q_f32(svfloat32_t, svfloat32_t);
7388		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_f16)))
7389		svfloat16_t svzip2q_f16(svfloat16_t, svfloat16_t);
7390		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_s32)))
7391		svint32_t svzip2q_s32(svint32_t, svint32_t);
7392		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_s64)))
7393		svint64_t svzip2q_s64(svint64_t, svint64_t);
7394		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_s16)))
7395		svint16_t svzip2q_s16(svint16_t, svint16_t);
7396		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_u8)))
7397		svuint8_t svld1ro(svbool_t, uint8_t const *);
7398		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_u32)))
7399		svuint32_t svld1ro(svbool_t, uint32_t const *);
7400		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_u64)))
7401		svuint64_t svld1ro(svbool_t, uint64_t const *);
7402		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_u16)))
7403		svuint16_t svld1ro(svbool_t, uint16_t const *);
7404		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_s8)))
7405		svint8_t svld1ro(svbool_t, int8_t const *);
7406		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_f64)))
7407		svfloat64_t svld1ro(svbool_t, float64_t const *);
7408		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_f32)))
7409		svfloat32_t svld1ro(svbool_t, float32_t const *);
7410		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_f16)))
7411		svfloat16_t svld1ro(svbool_t, float16_t const *);
7412		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_s32)))
7413		svint32_t svld1ro(svbool_t, int32_t const *);
7414		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_s64)))
7415		svint64_t svld1ro(svbool_t, int64_t const *);
7416		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_s16)))
7417		svint16_t svld1ro(svbool_t, int16_t const *);
7418		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmmla_f64)))
7419		svfloat64_t svmmla(svfloat64_t, svfloat64_t, svfloat64_t);
7420		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_u8)))
7421		svuint8_t svtrn1q(svuint8_t, svuint8_t);
7422		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_u32)))
7423		svuint32_t svtrn1q(svuint32_t, svuint32_t);
7424		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_u64)))
7425		svuint64_t svtrn1q(svuint64_t, svuint64_t);
7426		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_u16)))
7427		svuint16_t svtrn1q(svuint16_t, svuint16_t);
7428		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_s8)))
7429		svint8_t svtrn1q(svint8_t, svint8_t);
7430		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_f64)))
7431		svfloat64_t svtrn1q(svfloat64_t, svfloat64_t);
7432		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_f32)))
7433		svfloat32_t svtrn1q(svfloat32_t, svfloat32_t);
7434		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_f16)))
7435		svfloat16_t svtrn1q(svfloat16_t, svfloat16_t);
7436		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_s32)))
7437		svint32_t svtrn1q(svint32_t, svint32_t);
7438		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_s64)))
7439		svint64_t svtrn1q(svint64_t, svint64_t);
7440		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_s16)))
7441		svint16_t svtrn1q(svint16_t, svint16_t);
7442		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_u8)))
7443		svuint8_t svtrn2q(svuint8_t, svuint8_t);
7444		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_u32)))
7445		svuint32_t svtrn2q(svuint32_t, svuint32_t);
7446		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_u64)))
7447		svuint64_t svtrn2q(svuint64_t, svuint64_t);
7448		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_u16)))
7449		svuint16_t svtrn2q(svuint16_t, svuint16_t);
7450		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_s8)))
7451		svint8_t svtrn2q(svint8_t, svint8_t);
7452		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_f64)))
7453		svfloat64_t svtrn2q(svfloat64_t, svfloat64_t);
7454		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_f32)))
7455		svfloat32_t svtrn2q(svfloat32_t, svfloat32_t);
7456		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_f16)))
7457		svfloat16_t svtrn2q(svfloat16_t, svfloat16_t);
7458		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_s32)))
7459		svint32_t svtrn2q(svint32_t, svint32_t);
7460		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_s64)))
7461		svint64_t svtrn2q(svint64_t, svint64_t);
7462		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_s16)))
7463		svint16_t svtrn2q(svint16_t, svint16_t);
7464		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_u8)))
7465		svuint8_t svuzp1q(svuint8_t, svuint8_t);
7466		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_u32)))
7467		svuint32_t svuzp1q(svuint32_t, svuint32_t);
7468		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_u64)))
7469		svuint64_t svuzp1q(svuint64_t, svuint64_t);
7470		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_u16)))
7471		svuint16_t svuzp1q(svuint16_t, svuint16_t);
7472		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_s8)))
7473		svint8_t svuzp1q(svint8_t, svint8_t);
7474		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_f64)))
7475		svfloat64_t svuzp1q(svfloat64_t, svfloat64_t);
7476		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_f32)))
7477		svfloat32_t svuzp1q(svfloat32_t, svfloat32_t);
7478		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_f16)))
7479		svfloat16_t svuzp1q(svfloat16_t, svfloat16_t);
7480		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_s32)))
7481		svint32_t svuzp1q(svint32_t, svint32_t);
7482		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_s64)))
7483		svint64_t svuzp1q(svint64_t, svint64_t);
7484		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_s16)))
7485		svint16_t svuzp1q(svint16_t, svint16_t);
7486		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_u8)))
7487		svuint8_t svuzp2q(svuint8_t, svuint8_t);
7488		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_u32)))
7489		svuint32_t svuzp2q(svuint32_t, svuint32_t);
7490		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_u64)))
7491		svuint64_t svuzp2q(svuint64_t, svuint64_t);
7492		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_u16)))
7493		svuint16_t svuzp2q(svuint16_t, svuint16_t);
7494		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_s8)))
7495		svint8_t svuzp2q(svint8_t, svint8_t);
7496		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_f64)))
7497		svfloat64_t svuzp2q(svfloat64_t, svfloat64_t);
7498		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_f32)))
7499		svfloat32_t svuzp2q(svfloat32_t, svfloat32_t);
7500		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_f16)))
7501		svfloat16_t svuzp2q(svfloat16_t, svfloat16_t);
7502		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_s32)))
7503		svint32_t svuzp2q(svint32_t, svint32_t);
7504		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_s64)))
7505		svint64_t svuzp2q(svint64_t, svint64_t);
7506		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_s16)))
7507		svint16_t svuzp2q(svint16_t, svint16_t);
7508		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_u8)))
7509		svuint8_t svzip1q(svuint8_t, svuint8_t);
7510		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_u32)))
7511		svuint32_t svzip1q(svuint32_t, svuint32_t);
7512		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_u64)))
7513		svuint64_t svzip1q(svuint64_t, svuint64_t);
7514		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_u16)))
7515		svuint16_t svzip1q(svuint16_t, svuint16_t);
7516		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_s8)))
7517		svint8_t svzip1q(svint8_t, svint8_t);
7518		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_f64)))
7519		svfloat64_t svzip1q(svfloat64_t, svfloat64_t);
7520		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_f32)))
7521		svfloat32_t svzip1q(svfloat32_t, svfloat32_t);
7522		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_f16)))
7523		svfloat16_t svzip1q(svfloat16_t, svfloat16_t);
7524		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_s32)))
7525		svint32_t svzip1q(svint32_t, svint32_t);
7526		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_s64)))
7527		svint64_t svzip1q(svint64_t, svint64_t);
7528		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_s16)))
7529		svint16_t svzip1q(svint16_t, svint16_t);
7530		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_u8)))
7531		svuint8_t svzip2q(svuint8_t, svuint8_t);
7532		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_u32)))
7533		svuint32_t svzip2q(svuint32_t, svuint32_t);
7534		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_u64)))
7535		svuint64_t svzip2q(svuint64_t, svuint64_t);
7536		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_u16)))
7537		svuint16_t svzip2q(svuint16_t, svuint16_t);
7538		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_s8)))
7539		svint8_t svzip2q(svint8_t, svint8_t);
7540		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_f64)))
7541		svfloat64_t svzip2q(svfloat64_t, svfloat64_t);
7542		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_f32)))
7543		svfloat32_t svzip2q(svfloat32_t, svfloat32_t);
7544		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_f16)))
7545		svfloat16_t svzip2q(svfloat16_t, svfloat16_t);
7546		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_s32)))
7547		svint32_t svzip2q(svint32_t, svint32_t);
7548		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_s64)))
7549		svint64_t svzip2q(svint64_t, svint64_t);
7550		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_s16)))
7551		svint16_t svzip2q(svint16_t, svint16_t);
7552		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_bf16)))
7553		svbfloat16_t svld1ro_bf16(svbool_t, bfloat16_t const *);
7554		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_bf16)))
7555		svbfloat16_t svld1ro(svbool_t, bfloat16_t const *);
7556		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmmla_s32)))
7557		svint32_t svmmla_s32(svint32_t, svint8_t, svint8_t);
7558		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmmla_u32)))
7559		svuint32_t svmmla_u32(svuint32_t, svuint8_t, svuint8_t);
7560		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svusmmla_s32)))
7561		svint32_t svusmmla_s32(svint32_t, svuint8_t, svint8_t);
7562		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmmla_s32)))
7563		svint32_t svmmla(svint32_t, svint8_t, svint8_t);
7564		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmmla_u32)))
7565		svuint32_t svmmla(svuint32_t, svuint8_t, svuint8_t);
7566		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svusmmla_s32)))
7567		svint32_t svusmmla(svint32_t, svuint8_t, svint8_t);
7568		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsudot_n_s32)))
7569		svint32_t svsudot_n_s32(svint32_t, svint8_t, uint8_t);
7570		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsudot_s32)))
7571		svint32_t svsudot_s32(svint32_t, svint8_t, svuint8_t);
7572		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsudot_lane_s32)))
7573		svint32_t svsudot_lane_s32(svint32_t, svint8_t, svuint8_t, uint64_t);
7574		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svusdot_n_s32)))
7575		svint32_t svusdot_n_s32(svint32_t, svuint8_t, int8_t);
7576		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svusdot_s32)))
7577		svint32_t svusdot_s32(svint32_t, svuint8_t, svint8_t);
7578		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svusdot_lane_s32)))
7579		svint32_t svusdot_lane_s32(svint32_t, svuint8_t, svint8_t, uint64_t);
7580		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsudot_n_s32)))
7581		svint32_t svsudot(svint32_t, svint8_t, uint8_t);
7582		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsudot_s32)))
7583		svint32_t svsudot(svint32_t, svint8_t, svuint8_t);
7584		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsudot_lane_s32)))
7585		svint32_t svsudot_lane(svint32_t, svint8_t, svuint8_t, uint64_t);
7586		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svusdot_n_s32)))
7587		svint32_t svusdot(svint32_t, svuint8_t, int8_t);
7588		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svusdot_s32)))
7589		svint32_t svusdot(svint32_t, svuint8_t, svint8_t);
7590		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svusdot_lane_s32)))
7591		svint32_t svusdot_lane(svint32_t, svuint8_t, svint8_t, uint64_t);
7592		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svhistcnt_u32_z)))
7593		svuint32_t svhistcnt_u32_z(svbool_t, svuint32_t, svuint32_t);
7594		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svhistcnt_u64_z)))
7595		svuint64_t svhistcnt_u64_z(svbool_t, svuint64_t, svuint64_t);
7596		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svhistcnt_s32_z)))
7597		svuint32_t svhistcnt_s32_z(svbool_t, svint32_t, svint32_t);
7598		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svhistcnt_s64_z)))
7599		svuint64_t svhistcnt_s64_z(svbool_t, svint64_t, svint64_t);
7600		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svhistseg_u8)))
7601		svuint8_t svhistseg_u8(svuint8_t, svuint8_t);
7602		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svhistseg_s8)))
7603		svuint8_t svhistseg_s8(svint8_t, svint8_t);
7604		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_index_u32)))
7605		svuint32_t svldnt1_gather_u32base_index_u32(svbool_t, svuint32_t, int64_t);
7606		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_index_u64)))
7607		svuint64_t svldnt1_gather_u64base_index_u64(svbool_t, svuint64_t, int64_t);
7608		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_index_f64)))
7609		svfloat64_t svldnt1_gather_u64base_index_f64(svbool_t, svuint64_t, int64_t);
7610		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_index_f32)))
7611		svfloat32_t svldnt1_gather_u32base_index_f32(svbool_t, svuint32_t, int64_t);
7612		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_index_s32)))
7613		svint32_t svldnt1_gather_u32base_index_s32(svbool_t, svuint32_t, int64_t);
7614		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_index_s64)))
7615		svint64_t svldnt1_gather_u64base_index_s64(svbool_t, svuint64_t, int64_t);
7616		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_offset_u32)))
7617		svuint32_t svldnt1_gather_u32base_offset_u32(svbool_t, svuint32_t, int64_t);
7618		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_offset_u64)))
7619		svuint64_t svldnt1_gather_u64base_offset_u64(svbool_t, svuint64_t, int64_t);
7620		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_offset_f64)))
7621		svfloat64_t svldnt1_gather_u64base_offset_f64(svbool_t, svuint64_t, int64_t);
7622		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_offset_f32)))
7623		svfloat32_t svldnt1_gather_u32base_offset_f32(svbool_t, svuint32_t, int64_t);
7624		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_offset_s32)))
7625		svint32_t svldnt1_gather_u32base_offset_s32(svbool_t, svuint32_t, int64_t);
7626		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_offset_s64)))
7627		svint64_t svldnt1_gather_u64base_offset_s64(svbool_t, svuint64_t, int64_t);
7628		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_u32)))
7629		svuint32_t svldnt1_gather_u32base_u32(svbool_t, svuint32_t);
7630		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_u64)))
7631		svuint64_t svldnt1_gather_u64base_u64(svbool_t, svuint64_t);
7632		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_f64)))
7633		svfloat64_t svldnt1_gather_u64base_f64(svbool_t, svuint64_t);
7634		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_f32)))
7635		svfloat32_t svldnt1_gather_u32base_f32(svbool_t, svuint32_t);
7636		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_s32)))
7637		svint32_t svldnt1_gather_u32base_s32(svbool_t, svuint32_t);
7638		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_s64)))
7639		svint64_t svldnt1_gather_u64base_s64(svbool_t, svuint64_t);
7640		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_s64index_u64)))
7641		svuint64_t svldnt1_gather_s64index_u64(svbool_t, uint64_t const *, svint64_t);
7642		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_s64index_f64)))
7643		svfloat64_t svldnt1_gather_s64index_f64(svbool_t, float64_t const *, svint64_t);
7644		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_s64index_s64)))
7645		svint64_t svldnt1_gather_s64index_s64(svbool_t, int64_t const *, svint64_t);
7646		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64index_u64)))
7647		svuint64_t svldnt1_gather_u64index_u64(svbool_t, uint64_t const *, svuint64_t);
7648		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64index_f64)))
7649		svfloat64_t svldnt1_gather_u64index_f64(svbool_t, float64_t const *, svuint64_t);
7650		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64index_s64)))
7651		svint64_t svldnt1_gather_u64index_s64(svbool_t, int64_t const *, svuint64_t);
7652		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32offset_u32)))
7653		svuint32_t svldnt1_gather_u32offset_u32(svbool_t, uint32_t const *, svuint32_t);
7654		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32offset_f32)))
7655		svfloat32_t svldnt1_gather_u32offset_f32(svbool_t, float32_t const *, svuint32_t);
7656		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32offset_s32)))
7657		svint32_t svldnt1_gather_u32offset_s32(svbool_t, int32_t const *, svuint32_t);
7658		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_s64offset_u64)))
7659		svuint64_t svldnt1_gather_s64offset_u64(svbool_t, uint64_t const *, svint64_t);
7660		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_s64offset_f64)))
7661		svfloat64_t svldnt1_gather_s64offset_f64(svbool_t, float64_t const *, svint64_t);
7662		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_s64offset_s64)))
7663		svint64_t svldnt1_gather_s64offset_s64(svbool_t, int64_t const *, svint64_t);
7664		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64offset_u64)))
7665		svuint64_t svldnt1_gather_u64offset_u64(svbool_t, uint64_t const *, svuint64_t);
7666		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64offset_f64)))
7667		svfloat64_t svldnt1_gather_u64offset_f64(svbool_t, float64_t const *, svuint64_t);
7668		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64offset_s64)))
7669		svint64_t svldnt1_gather_u64offset_s64(svbool_t, int64_t const *, svuint64_t);
7670		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u32base_offset_u32)))
7671		svuint32_t svldnt1sb_gather_u32base_offset_u32(svbool_t, svuint32_t, int64_t);
7672		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u64base_offset_u64)))
7673		svuint64_t svldnt1sb_gather_u64base_offset_u64(svbool_t, svuint64_t, int64_t);
7674		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u32base_offset_s32)))
7675		svint32_t svldnt1sb_gather_u32base_offset_s32(svbool_t, svuint32_t, int64_t);
7676		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u64base_offset_s64)))
7677		svint64_t svldnt1sb_gather_u64base_offset_s64(svbool_t, svuint64_t, int64_t);
7678		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u32base_u32)))
7679		svuint32_t svldnt1sb_gather_u32base_u32(svbool_t, svuint32_t);
7680		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u64base_u64)))
7681		svuint64_t svldnt1sb_gather_u64base_u64(svbool_t, svuint64_t);
7682		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u32base_s32)))
7683		svint32_t svldnt1sb_gather_u32base_s32(svbool_t, svuint32_t);
7684		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u64base_s64)))
7685		svint64_t svldnt1sb_gather_u64base_s64(svbool_t, svuint64_t);
7686		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u32offset_u32)))
7687		svuint32_t svldnt1sb_gather_u32offset_u32(svbool_t, int8_t const *, svuint32_t);
7688		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u32offset_s32)))
7689		svint32_t svldnt1sb_gather_u32offset_s32(svbool_t, int8_t const *, svuint32_t);
7690		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_s64offset_u64)))
7691		svuint64_t svldnt1sb_gather_s64offset_u64(svbool_t, int8_t const *, svint64_t);
7692		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_s64offset_s64)))
7693		svint64_t svldnt1sb_gather_s64offset_s64(svbool_t, int8_t const *, svint64_t);
7694		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u64offset_u64)))
7695		svuint64_t svldnt1sb_gather_u64offset_u64(svbool_t, int8_t const *, svuint64_t);
7696		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u64offset_s64)))
7697		svint64_t svldnt1sb_gather_u64offset_s64(svbool_t, int8_t const *, svuint64_t);
7698		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32base_index_u32)))
7699		svuint32_t svldnt1sh_gather_u32base_index_u32(svbool_t, svuint32_t, int64_t);
7700		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64base_index_u64)))
7701		svuint64_t svldnt1sh_gather_u64base_index_u64(svbool_t, svuint64_t, int64_t);
7702		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32base_index_s32)))
7703		svint32_t svldnt1sh_gather_u32base_index_s32(svbool_t, svuint32_t, int64_t);
7704		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64base_index_s64)))
7705		svint64_t svldnt1sh_gather_u64base_index_s64(svbool_t, svuint64_t, int64_t);
7706		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32base_offset_u32)))
7707		svuint32_t svldnt1sh_gather_u32base_offset_u32(svbool_t, svuint32_t, int64_t);
7708		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64base_offset_u64)))
7709		svuint64_t svldnt1sh_gather_u64base_offset_u64(svbool_t, svuint64_t, int64_t);
7710		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32base_offset_s32)))
7711		svint32_t svldnt1sh_gather_u32base_offset_s32(svbool_t, svuint32_t, int64_t);
7712		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64base_offset_s64)))
7713		svint64_t svldnt1sh_gather_u64base_offset_s64(svbool_t, svuint64_t, int64_t);
7714		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32base_u32)))
7715		svuint32_t svldnt1sh_gather_u32base_u32(svbool_t, svuint32_t);
7716		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64base_u64)))
7717		svuint64_t svldnt1sh_gather_u64base_u64(svbool_t, svuint64_t);
7718		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32base_s32)))
7719		svint32_t svldnt1sh_gather_u32base_s32(svbool_t, svuint32_t);
7720		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64base_s64)))
7721		svint64_t svldnt1sh_gather_u64base_s64(svbool_t, svuint64_t);
7722		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_s64index_u64)))
7723		svuint64_t svldnt1sh_gather_s64index_u64(svbool_t, int16_t const *, svint64_t);
7724		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_s64index_s64)))
7725		svint64_t svldnt1sh_gather_s64index_s64(svbool_t, int16_t const *, svint64_t);
7726		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64index_u64)))
7727		svuint64_t svldnt1sh_gather_u64index_u64(svbool_t, int16_t const *, svuint64_t);
7728		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64index_s64)))
7729		svint64_t svldnt1sh_gather_u64index_s64(svbool_t, int16_t const *, svuint64_t);
7730		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32offset_u32)))
7731		svuint32_t svldnt1sh_gather_u32offset_u32(svbool_t, int16_t const *, svuint32_t);
7732		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32offset_s32)))
7733		svint32_t svldnt1sh_gather_u32offset_s32(svbool_t, int16_t const *, svuint32_t);
7734		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_s64offset_u64)))
7735		svuint64_t svldnt1sh_gather_s64offset_u64(svbool_t, int16_t const *, svint64_t);
7736		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_s64offset_s64)))
7737		svint64_t svldnt1sh_gather_s64offset_s64(svbool_t, int16_t const *, svint64_t);
7738		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64offset_u64)))
7739		svuint64_t svldnt1sh_gather_u64offset_u64(svbool_t, int16_t const *, svuint64_t);
7740		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64offset_s64)))
7741		svint64_t svldnt1sh_gather_u64offset_s64(svbool_t, int16_t const *, svuint64_t);
7742		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64base_index_u64)))
7743		svuint64_t svldnt1sw_gather_u64base_index_u64(svbool_t, svuint64_t, int64_t);
7744		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64base_index_s64)))
7745		svint64_t svldnt1sw_gather_u64base_index_s64(svbool_t, svuint64_t, int64_t);
7746		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64base_offset_u64)))
7747		svuint64_t svldnt1sw_gather_u64base_offset_u64(svbool_t, svuint64_t, int64_t);
7748		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64base_offset_s64)))
7749		svint64_t svldnt1sw_gather_u64base_offset_s64(svbool_t, svuint64_t, int64_t);
7750		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64base_u64)))
7751		svuint64_t svldnt1sw_gather_u64base_u64(svbool_t, svuint64_t);
7752		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64base_s64)))
7753		svint64_t svldnt1sw_gather_u64base_s64(svbool_t, svuint64_t);
7754		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_s64index_u64)))
7755		svuint64_t svldnt1sw_gather_s64index_u64(svbool_t, int32_t const *, svint64_t);
7756		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_s64index_s64)))
7757		svint64_t svldnt1sw_gather_s64index_s64(svbool_t, int32_t const *, svint64_t);
7758		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64index_u64)))
7759		svuint64_t svldnt1sw_gather_u64index_u64(svbool_t, int32_t const *, svuint64_t);
7760		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64index_s64)))
7761		svint64_t svldnt1sw_gather_u64index_s64(svbool_t, int32_t const *, svuint64_t);
7762		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_s64offset_u64)))
7763		svuint64_t svldnt1sw_gather_s64offset_u64(svbool_t, int32_t const *, svint64_t);
7764		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_s64offset_s64)))
7765		svint64_t svldnt1sw_gather_s64offset_s64(svbool_t, int32_t const *, svint64_t);
7766		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64offset_u64)))
7767		svuint64_t svldnt1sw_gather_u64offset_u64(svbool_t, int32_t const *, svuint64_t);
7768		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64offset_s64)))
7769		svint64_t svldnt1sw_gather_u64offset_s64(svbool_t, int32_t const *, svuint64_t);
7770		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u32base_offset_u32)))
7771		svuint32_t svldnt1ub_gather_u32base_offset_u32(svbool_t, svuint32_t, int64_t);
7772		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u64base_offset_u64)))
7773		svuint64_t svldnt1ub_gather_u64base_offset_u64(svbool_t, svuint64_t, int64_t);
7774		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u32base_offset_s32)))
7775		svint32_t svldnt1ub_gather_u32base_offset_s32(svbool_t, svuint32_t, int64_t);
7776		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u64base_offset_s64)))
7777		svint64_t svldnt1ub_gather_u64base_offset_s64(svbool_t, svuint64_t, int64_t);
7778		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u32base_u32)))
7779		svuint32_t svldnt1ub_gather_u32base_u32(svbool_t, svuint32_t);
7780		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u64base_u64)))
7781		svuint64_t svldnt1ub_gather_u64base_u64(svbool_t, svuint64_t);
7782		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u32base_s32)))
7783		svint32_t svldnt1ub_gather_u32base_s32(svbool_t, svuint32_t);
7784		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u64base_s64)))
7785		svint64_t svldnt1ub_gather_u64base_s64(svbool_t, svuint64_t);
7786		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u32offset_u32)))
7787		svuint32_t svldnt1ub_gather_u32offset_u32(svbool_t, uint8_t const *, svuint32_t);
7788		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u32offset_s32)))
7789		svint32_t svldnt1ub_gather_u32offset_s32(svbool_t, uint8_t const *, svuint32_t);
7790		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_s64offset_u64)))
7791		svuint64_t svldnt1ub_gather_s64offset_u64(svbool_t, uint8_t const *, svint64_t);
7792		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_s64offset_s64)))
7793		svint64_t svldnt1ub_gather_s64offset_s64(svbool_t, uint8_t const *, svint64_t);
7794		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u64offset_u64)))
7795		svuint64_t svldnt1ub_gather_u64offset_u64(svbool_t, uint8_t const *, svuint64_t);
7796		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u64offset_s64)))
7797		svint64_t svldnt1ub_gather_u64offset_s64(svbool_t, uint8_t const *, svuint64_t);
7798		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32base_index_u32)))
7799		svuint32_t svldnt1uh_gather_u32base_index_u32(svbool_t, svuint32_t, int64_t);
7800		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64base_index_u64)))
7801		svuint64_t svldnt1uh_gather_u64base_index_u64(svbool_t, svuint64_t, int64_t);
7802		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32base_index_s32)))
7803		svint32_t svldnt1uh_gather_u32base_index_s32(svbool_t, svuint32_t, int64_t);
7804		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64base_index_s64)))
7805		svint64_t svldnt1uh_gather_u64base_index_s64(svbool_t, svuint64_t, int64_t);
7806		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32base_offset_u32)))
7807		svuint32_t svldnt1uh_gather_u32base_offset_u32(svbool_t, svuint32_t, int64_t);
7808		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64base_offset_u64)))
7809		svuint64_t svldnt1uh_gather_u64base_offset_u64(svbool_t, svuint64_t, int64_t);
7810		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32base_offset_s32)))
7811		svint32_t svldnt1uh_gather_u32base_offset_s32(svbool_t, svuint32_t, int64_t);
7812		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64base_offset_s64)))
7813		svint64_t svldnt1uh_gather_u64base_offset_s64(svbool_t, svuint64_t, int64_t);
7814		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32base_u32)))
7815		svuint32_t svldnt1uh_gather_u32base_u32(svbool_t, svuint32_t);
7816		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64base_u64)))
7817		svuint64_t svldnt1uh_gather_u64base_u64(svbool_t, svuint64_t);
7818		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32base_s32)))
7819		svint32_t svldnt1uh_gather_u32base_s32(svbool_t, svuint32_t);
7820		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64base_s64)))
7821		svint64_t svldnt1uh_gather_u64base_s64(svbool_t, svuint64_t);
7822		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_s64index_u64)))
7823		svuint64_t svldnt1uh_gather_s64index_u64(svbool_t, uint16_t const *, svint64_t);
7824		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_s64index_s64)))
7825		svint64_t svldnt1uh_gather_s64index_s64(svbool_t, uint16_t const *, svint64_t);
7826		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64index_u64)))
7827		svuint64_t svldnt1uh_gather_u64index_u64(svbool_t, uint16_t const *, svuint64_t);
7828		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64index_s64)))
7829		svint64_t svldnt1uh_gather_u64index_s64(svbool_t, uint16_t const *, svuint64_t);
7830		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32offset_u32)))
7831		svuint32_t svldnt1uh_gather_u32offset_u32(svbool_t, uint16_t const *, svuint32_t);
7832		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32offset_s32)))
7833		svint32_t svldnt1uh_gather_u32offset_s32(svbool_t, uint16_t const *, svuint32_t);
7834		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_s64offset_u64)))
7835		svuint64_t svldnt1uh_gather_s64offset_u64(svbool_t, uint16_t const *, svint64_t);
7836		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_s64offset_s64)))
7837		svint64_t svldnt1uh_gather_s64offset_s64(svbool_t, uint16_t const *, svint64_t);
7838		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64offset_u64)))
7839		svuint64_t svldnt1uh_gather_u64offset_u64(svbool_t, uint16_t const *, svuint64_t);
7840		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64offset_s64)))
7841		svint64_t svldnt1uh_gather_u64offset_s64(svbool_t, uint16_t const *, svuint64_t);
7842		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64base_index_u64)))
7843		svuint64_t svldnt1uw_gather_u64base_index_u64(svbool_t, svuint64_t, int64_t);
7844		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64base_index_s64)))
7845		svint64_t svldnt1uw_gather_u64base_index_s64(svbool_t, svuint64_t, int64_t);
7846		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64base_offset_u64)))
7847		svuint64_t svldnt1uw_gather_u64base_offset_u64(svbool_t, svuint64_t, int64_t);
7848		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64base_offset_s64)))
7849		svint64_t svldnt1uw_gather_u64base_offset_s64(svbool_t, svuint64_t, int64_t);
7850		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64base_u64)))
7851		svuint64_t svldnt1uw_gather_u64base_u64(svbool_t, svuint64_t);
7852		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64base_s64)))
7853		svint64_t svldnt1uw_gather_u64base_s64(svbool_t, svuint64_t);
7854		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_s64index_u64)))
7855		svuint64_t svldnt1uw_gather_s64index_u64(svbool_t, uint32_t const *, svint64_t);
7856		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_s64index_s64)))
7857		svint64_t svldnt1uw_gather_s64index_s64(svbool_t, uint32_t const *, svint64_t);
7858		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64index_u64)))
7859		svuint64_t svldnt1uw_gather_u64index_u64(svbool_t, uint32_t const *, svuint64_t);
7860		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64index_s64)))
7861		svint64_t svldnt1uw_gather_u64index_s64(svbool_t, uint32_t const *, svuint64_t);
7862		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_s64offset_u64)))
7863		svuint64_t svldnt1uw_gather_s64offset_u64(svbool_t, uint32_t const *, svint64_t);
7864		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_s64offset_s64)))
7865		svint64_t svldnt1uw_gather_s64offset_s64(svbool_t, uint32_t const *, svint64_t);
7866		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64offset_u64)))
7867		svuint64_t svldnt1uw_gather_u64offset_u64(svbool_t, uint32_t const *, svuint64_t);
7868		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64offset_s64)))
7869		svint64_t svldnt1uw_gather_u64offset_s64(svbool_t, uint32_t const *, svuint64_t);
7870		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmatch_u8)))
7871		svbool_t svmatch_u8(svbool_t, svuint8_t, svuint8_t);
7872		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmatch_u16)))
7873		svbool_t svmatch_u16(svbool_t, svuint16_t, svuint16_t);
7874		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmatch_s8)))
7875		svbool_t svmatch_s8(svbool_t, svint8_t, svint8_t);
7876		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmatch_s16)))
7877		svbool_t svmatch_s16(svbool_t, svint16_t, svint16_t);
7878		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svnmatch_u8)))
7879		svbool_t svnmatch_u8(svbool_t, svuint8_t, svuint8_t);
7880		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svnmatch_u16)))
7881		svbool_t svnmatch_u16(svbool_t, svuint16_t, svuint16_t);
7882		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svnmatch_s8)))
7883		svbool_t svnmatch_s8(svbool_t, svint8_t, svint8_t);
7884		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svnmatch_s16)))
7885		svbool_t svnmatch_s16(svbool_t, svint16_t, svint16_t);
7886		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_index_u32)))
7887		void svstnt1_scatter_u32base_index_u32(svbool_t, svuint32_t, int64_t, svuint32_t);
7888		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_index_u64)))
7889		void svstnt1_scatter_u64base_index_u64(svbool_t, svuint64_t, int64_t, svuint64_t);
7890		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_index_f64)))
7891		void svstnt1_scatter_u64base_index_f64(svbool_t, svuint64_t, int64_t, svfloat64_t);
7892		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_index_f32)))
7893		void svstnt1_scatter_u32base_index_f32(svbool_t, svuint32_t, int64_t, svfloat32_t);
7894		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_index_s32)))
7895		void svstnt1_scatter_u32base_index_s32(svbool_t, svuint32_t, int64_t, svint32_t);
7896		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_index_s64)))
7897		void svstnt1_scatter_u64base_index_s64(svbool_t, svuint64_t, int64_t, svint64_t);
7898		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_offset_u32)))
7899		void svstnt1_scatter_u32base_offset_u32(svbool_t, svuint32_t, int64_t, svuint32_t);
7900		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_offset_u64)))
7901		void svstnt1_scatter_u64base_offset_u64(svbool_t, svuint64_t, int64_t, svuint64_t);
7902		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_offset_f64)))
7903		void svstnt1_scatter_u64base_offset_f64(svbool_t, svuint64_t, int64_t, svfloat64_t);
7904		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_offset_f32)))
7905		void svstnt1_scatter_u32base_offset_f32(svbool_t, svuint32_t, int64_t, svfloat32_t);
7906		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_offset_s32)))
7907		void svstnt1_scatter_u32base_offset_s32(svbool_t, svuint32_t, int64_t, svint32_t);
7908		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_offset_s64)))
7909		void svstnt1_scatter_u64base_offset_s64(svbool_t, svuint64_t, int64_t, svint64_t);
7910		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_u32)))
7911		void svstnt1_scatter_u32base_u32(svbool_t, svuint32_t, svuint32_t);
7912		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_u64)))
7913		void svstnt1_scatter_u64base_u64(svbool_t, svuint64_t, svuint64_t);
7914		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_f64)))
7915		void svstnt1_scatter_u64base_f64(svbool_t, svuint64_t, svfloat64_t);
7916		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_f32)))
7917		void svstnt1_scatter_u32base_f32(svbool_t, svuint32_t, svfloat32_t);
7918		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_s32)))
7919		void svstnt1_scatter_u32base_s32(svbool_t, svuint32_t, svint32_t);
7920		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_s64)))
7921		void svstnt1_scatter_u64base_s64(svbool_t, svuint64_t, svint64_t);
7922		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_s64index_u64)))
7923		void svstnt1_scatter_s64index_u64(svbool_t, uint64_t *, svint64_t, svuint64_t);
7924		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_s64index_f64)))
7925		void svstnt1_scatter_s64index_f64(svbool_t, float64_t *, svint64_t, svfloat64_t);
7926		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_s64index_s64)))
7927		void svstnt1_scatter_s64index_s64(svbool_t, int64_t *, svint64_t, svint64_t);
7928		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64index_u64)))
7929		void svstnt1_scatter_u64index_u64(svbool_t, uint64_t *, svuint64_t, svuint64_t);
7930		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64index_f64)))
7931		void svstnt1_scatter_u64index_f64(svbool_t, float64_t *, svuint64_t, svfloat64_t);
7932		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64index_s64)))
7933		void svstnt1_scatter_u64index_s64(svbool_t, int64_t *, svuint64_t, svint64_t);
7934		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32offset_u32)))
7935		void svstnt1_scatter_u32offset_u32(svbool_t, uint32_t *, svuint32_t, svuint32_t);
7936		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32offset_f32)))
7937		void svstnt1_scatter_u32offset_f32(svbool_t, float32_t *, svuint32_t, svfloat32_t);
7938		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32offset_s32)))
7939		void svstnt1_scatter_u32offset_s32(svbool_t, int32_t *, svuint32_t, svint32_t);
7940		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_s64offset_u64)))
7941		void svstnt1_scatter_s64offset_u64(svbool_t, uint64_t *, svint64_t, svuint64_t);
7942		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_s64offset_f64)))
7943		void svstnt1_scatter_s64offset_f64(svbool_t, float64_t *, svint64_t, svfloat64_t);
7944		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_s64offset_s64)))
7945		void svstnt1_scatter_s64offset_s64(svbool_t, int64_t *, svint64_t, svint64_t);
7946		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64offset_u64)))
7947		void svstnt1_scatter_u64offset_u64(svbool_t, uint64_t *, svuint64_t, svuint64_t);
7948		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64offset_f64)))
7949		void svstnt1_scatter_u64offset_f64(svbool_t, float64_t *, svuint64_t, svfloat64_t);
7950		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64offset_s64)))
7951		void svstnt1_scatter_u64offset_s64(svbool_t, int64_t *, svuint64_t, svint64_t);
7952		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u32base_offset_u32)))
7953		void svstnt1b_scatter_u32base_offset_u32(svbool_t, svuint32_t, int64_t, svuint32_t);
7954		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u64base_offset_u64)))
7955		void svstnt1b_scatter_u64base_offset_u64(svbool_t, svuint64_t, int64_t, svuint64_t);
7956		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u32base_offset_s32)))
7957		void svstnt1b_scatter_u32base_offset_s32(svbool_t, svuint32_t, int64_t, svint32_t);
7958		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u64base_offset_s64)))
7959		void svstnt1b_scatter_u64base_offset_s64(svbool_t, svuint64_t, int64_t, svint64_t);
7960		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u32base_u32)))
7961		void svstnt1b_scatter_u32base_u32(svbool_t, svuint32_t, svuint32_t);
7962		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u64base_u64)))
7963		void svstnt1b_scatter_u64base_u64(svbool_t, svuint64_t, svuint64_t);
7964		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u32base_s32)))
7965		void svstnt1b_scatter_u32base_s32(svbool_t, svuint32_t, svint32_t);
7966		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u64base_s64)))
7967		void svstnt1b_scatter_u64base_s64(svbool_t, svuint64_t, svint64_t);
7968		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u32offset_s32)))
7969		void svstnt1b_scatter_u32offset_s32(svbool_t, int8_t *, svuint32_t, svint32_t);
7970		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u32offset_u32)))
7971		void svstnt1b_scatter_u32offset_u32(svbool_t, uint8_t *, svuint32_t, svuint32_t);
7972		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_s64offset_s64)))
7973		void svstnt1b_scatter_s64offset_s64(svbool_t, int8_t *, svint64_t, svint64_t);
7974		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_s64offset_u64)))
7975		void svstnt1b_scatter_s64offset_u64(svbool_t, uint8_t *, svint64_t, svuint64_t);
7976		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u64offset_s64)))
7977		void svstnt1b_scatter_u64offset_s64(svbool_t, int8_t *, svuint64_t, svint64_t);
7978		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u64offset_u64)))
7979		void svstnt1b_scatter_u64offset_u64(svbool_t, uint8_t *, svuint64_t, svuint64_t);
7980		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32base_index_u32)))
7981		void svstnt1h_scatter_u32base_index_u32(svbool_t, svuint32_t, int64_t, svuint32_t);
7982		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64base_index_u64)))
7983		void svstnt1h_scatter_u64base_index_u64(svbool_t, svuint64_t, int64_t, svuint64_t);
7984		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32base_index_s32)))
7985		void svstnt1h_scatter_u32base_index_s32(svbool_t, svuint32_t, int64_t, svint32_t);
7986		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64base_index_s64)))
7987		void svstnt1h_scatter_u64base_index_s64(svbool_t, svuint64_t, int64_t, svint64_t);
7988		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32base_offset_u32)))
7989		void svstnt1h_scatter_u32base_offset_u32(svbool_t, svuint32_t, int64_t, svuint32_t);
7990		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64base_offset_u64)))
7991		void svstnt1h_scatter_u64base_offset_u64(svbool_t, svuint64_t, int64_t, svuint64_t);
7992		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32base_offset_s32)))
7993		void svstnt1h_scatter_u32base_offset_s32(svbool_t, svuint32_t, int64_t, svint32_t);
7994		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64base_offset_s64)))
7995		void svstnt1h_scatter_u64base_offset_s64(svbool_t, svuint64_t, int64_t, svint64_t);
7996		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32base_u32)))
7997		void svstnt1h_scatter_u32base_u32(svbool_t, svuint32_t, svuint32_t);
7998		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64base_u64)))
7999		void svstnt1h_scatter_u64base_u64(svbool_t, svuint64_t, svuint64_t);
8000		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32base_s32)))
8001		void svstnt1h_scatter_u32base_s32(svbool_t, svuint32_t, svint32_t);
8002		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64base_s64)))
8003		void svstnt1h_scatter_u64base_s64(svbool_t, svuint64_t, svint64_t);
8004		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_s64index_s64)))
8005		void svstnt1h_scatter_s64index_s64(svbool_t, int16_t *, svint64_t, svint64_t);
8006		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_s64index_u64)))
8007		void svstnt1h_scatter_s64index_u64(svbool_t, uint16_t *, svint64_t, svuint64_t);
8008		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64index_s64)))
8009		void svstnt1h_scatter_u64index_s64(svbool_t, int16_t *, svuint64_t, svint64_t);
8010		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64index_u64)))
8011		void svstnt1h_scatter_u64index_u64(svbool_t, uint16_t *, svuint64_t, svuint64_t);
8012		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32offset_s32)))
8013		void svstnt1h_scatter_u32offset_s32(svbool_t, int16_t *, svuint32_t, svint32_t);
8014		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32offset_u32)))
8015		void svstnt1h_scatter_u32offset_u32(svbool_t, uint16_t *, svuint32_t, svuint32_t);
8016		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_s64offset_s64)))
8017		void svstnt1h_scatter_s64offset_s64(svbool_t, int16_t *, svint64_t, svint64_t);
8018		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_s64offset_u64)))
8019		void svstnt1h_scatter_s64offset_u64(svbool_t, uint16_t *, svint64_t, svuint64_t);
8020		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64offset_s64)))
8021		void svstnt1h_scatter_u64offset_s64(svbool_t, int16_t *, svuint64_t, svint64_t);
8022		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64offset_u64)))
8023		void svstnt1h_scatter_u64offset_u64(svbool_t, uint16_t *, svuint64_t, svuint64_t);
8024		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64base_index_u64)))
8025		void svstnt1w_scatter_u64base_index_u64(svbool_t, svuint64_t, int64_t, svuint64_t);
8026		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64base_index_s64)))
8027		void svstnt1w_scatter_u64base_index_s64(svbool_t, svuint64_t, int64_t, svint64_t);
8028		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64base_offset_u64)))
8029		void svstnt1w_scatter_u64base_offset_u64(svbool_t, svuint64_t, int64_t, svuint64_t);
8030		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64base_offset_s64)))
8031		void svstnt1w_scatter_u64base_offset_s64(svbool_t, svuint64_t, int64_t, svint64_t);
8032		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64base_u64)))
8033		void svstnt1w_scatter_u64base_u64(svbool_t, svuint64_t, svuint64_t);
8034		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64base_s64)))
8035		void svstnt1w_scatter_u64base_s64(svbool_t, svuint64_t, svint64_t);
8036		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_s64index_s64)))
8037		void svstnt1w_scatter_s64index_s64(svbool_t, int32_t *, svint64_t, svint64_t);
8038		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_s64index_u64)))
8039		void svstnt1w_scatter_s64index_u64(svbool_t, uint32_t *, svint64_t, svuint64_t);
8040		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64index_s64)))
8041		void svstnt1w_scatter_u64index_s64(svbool_t, int32_t *, svuint64_t, svint64_t);
8042		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64index_u64)))
8043		void svstnt1w_scatter_u64index_u64(svbool_t, uint32_t *, svuint64_t, svuint64_t);
8044		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_s64offset_s64)))
8045		void svstnt1w_scatter_s64offset_s64(svbool_t, int32_t *, svint64_t, svint64_t);
8046		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_s64offset_u64)))
8047		void svstnt1w_scatter_s64offset_u64(svbool_t, uint32_t *, svint64_t, svuint64_t);
8048		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64offset_s64)))
8049		void svstnt1w_scatter_u64offset_s64(svbool_t, int32_t *, svuint64_t, svint64_t);
8050		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64offset_u64)))
8051		void svstnt1w_scatter_u64offset_u64(svbool_t, uint32_t *, svuint64_t, svuint64_t);
8052		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svhistcnt_u32_z)))
8053		svuint32_t svhistcnt_z(svbool_t, svuint32_t, svuint32_t);
8054		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svhistcnt_u64_z)))
8055		svuint64_t svhistcnt_z(svbool_t, svuint64_t, svuint64_t);
8056		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svhistcnt_s32_z)))
8057		svuint32_t svhistcnt_z(svbool_t, svint32_t, svint32_t);
8058		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svhistcnt_s64_z)))
8059		svuint64_t svhistcnt_z(svbool_t, svint64_t, svint64_t);
8060		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svhistseg_u8)))
8061		svuint8_t svhistseg(svuint8_t, svuint8_t);
8062		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svhistseg_s8)))
8063		svuint8_t svhistseg(svint8_t, svint8_t);
8064		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_index_u32)))
8065		svuint32_t svldnt1_gather_index_u32(svbool_t, svuint32_t, int64_t);
8066		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_index_u64)))
8067		svuint64_t svldnt1_gather_index_u64(svbool_t, svuint64_t, int64_t);
8068		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_index_f64)))
8069		svfloat64_t svldnt1_gather_index_f64(svbool_t, svuint64_t, int64_t);
8070		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_index_f32)))
8071		svfloat32_t svldnt1_gather_index_f32(svbool_t, svuint32_t, int64_t);
8072		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_index_s32)))
8073		svint32_t svldnt1_gather_index_s32(svbool_t, svuint32_t, int64_t);
8074		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_index_s64)))
8075		svint64_t svldnt1_gather_index_s64(svbool_t, svuint64_t, int64_t);
8076		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_offset_u32)))
8077		svuint32_t svldnt1_gather_offset_u32(svbool_t, svuint32_t, int64_t);
8078		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_offset_u64)))
8079		svuint64_t svldnt1_gather_offset_u64(svbool_t, svuint64_t, int64_t);
8080		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_offset_f64)))
8081		svfloat64_t svldnt1_gather_offset_f64(svbool_t, svuint64_t, int64_t);
8082		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_offset_f32)))
8083		svfloat32_t svldnt1_gather_offset_f32(svbool_t, svuint32_t, int64_t);
8084		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_offset_s32)))
8085		svint32_t svldnt1_gather_offset_s32(svbool_t, svuint32_t, int64_t);
8086		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_offset_s64)))
8087		svint64_t svldnt1_gather_offset_s64(svbool_t, svuint64_t, int64_t);
8088		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_u32)))
8089		svuint32_t svldnt1_gather_u32(svbool_t, svuint32_t);
8090		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_u64)))
8091		svuint64_t svldnt1_gather_u64(svbool_t, svuint64_t);
8092		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_f64)))
8093		svfloat64_t svldnt1_gather_f64(svbool_t, svuint64_t);
8094		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_f32)))
8095		svfloat32_t svldnt1_gather_f32(svbool_t, svuint32_t);
8096		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_s32)))
8097		svint32_t svldnt1_gather_s32(svbool_t, svuint32_t);
8098		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_s64)))
8099		svint64_t svldnt1_gather_s64(svbool_t, svuint64_t);
8100		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_s64index_u64)))
8101		svuint64_t svldnt1_gather_index(svbool_t, uint64_t const *, svint64_t);
8102		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_s64index_f64)))
8103		svfloat64_t svldnt1_gather_index(svbool_t, float64_t const *, svint64_t);
8104		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_s64index_s64)))
8105		svint64_t svldnt1_gather_index(svbool_t, int64_t const *, svint64_t);
8106		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64index_u64)))
8107		svuint64_t svldnt1_gather_index(svbool_t, uint64_t const *, svuint64_t);
8108		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64index_f64)))
8109		svfloat64_t svldnt1_gather_index(svbool_t, float64_t const *, svuint64_t);
8110		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64index_s64)))
8111		svint64_t svldnt1_gather_index(svbool_t, int64_t const *, svuint64_t);
8112		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32offset_u32)))
8113		svuint32_t svldnt1_gather_offset(svbool_t, uint32_t const *, svuint32_t);
8114		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32offset_f32)))
8115		svfloat32_t svldnt1_gather_offset(svbool_t, float32_t const *, svuint32_t);
8116		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32offset_s32)))
8117		svint32_t svldnt1_gather_offset(svbool_t, int32_t const *, svuint32_t);
8118		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_s64offset_u64)))
8119		svuint64_t svldnt1_gather_offset(svbool_t, uint64_t const *, svint64_t);
8120		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_s64offset_f64)))
8121		svfloat64_t svldnt1_gather_offset(svbool_t, float64_t const *, svint64_t);
8122		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_s64offset_s64)))
8123		svint64_t svldnt1_gather_offset(svbool_t, int64_t const *, svint64_t);
8124		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64offset_u64)))
8125		svuint64_t svldnt1_gather_offset(svbool_t, uint64_t const *, svuint64_t);
8126		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64offset_f64)))
8127		svfloat64_t svldnt1_gather_offset(svbool_t, float64_t const *, svuint64_t);
8128		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64offset_s64)))
8129		svint64_t svldnt1_gather_offset(svbool_t, int64_t const *, svuint64_t);
8130		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u32base_offset_u32)))
8131		svuint32_t svldnt1sb_gather_offset_u32(svbool_t, svuint32_t, int64_t);
8132		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u64base_offset_u64)))
8133		svuint64_t svldnt1sb_gather_offset_u64(svbool_t, svuint64_t, int64_t);
8134		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u32base_offset_s32)))
8135		svint32_t svldnt1sb_gather_offset_s32(svbool_t, svuint32_t, int64_t);
8136		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u64base_offset_s64)))
8137		svint64_t svldnt1sb_gather_offset_s64(svbool_t, svuint64_t, int64_t);
8138		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u32base_u32)))
8139		svuint32_t svldnt1sb_gather_u32(svbool_t, svuint32_t);
8140		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u64base_u64)))
8141		svuint64_t svldnt1sb_gather_u64(svbool_t, svuint64_t);
8142		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u32base_s32)))
8143		svint32_t svldnt1sb_gather_s32(svbool_t, svuint32_t);
8144		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u64base_s64)))
8145		svint64_t svldnt1sb_gather_s64(svbool_t, svuint64_t);
8146		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u32offset_u32)))
8147		svuint32_t svldnt1sb_gather_offset_u32(svbool_t, int8_t const *, svuint32_t);
8148		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u32offset_s32)))
8149		svint32_t svldnt1sb_gather_offset_s32(svbool_t, int8_t const *, svuint32_t);
8150		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_s64offset_u64)))
8151		svuint64_t svldnt1sb_gather_offset_u64(svbool_t, int8_t const *, svint64_t);
8152		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_s64offset_s64)))
8153		svint64_t svldnt1sb_gather_offset_s64(svbool_t, int8_t const *, svint64_t);
8154		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u64offset_u64)))
8155		svuint64_t svldnt1sb_gather_offset_u64(svbool_t, int8_t const *, svuint64_t);
8156		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u64offset_s64)))
8157		svint64_t svldnt1sb_gather_offset_s64(svbool_t, int8_t const *, svuint64_t);
8158		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32base_index_u32)))
8159		svuint32_t svldnt1sh_gather_index_u32(svbool_t, svuint32_t, int64_t);
8160		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64base_index_u64)))
8161		svuint64_t svldnt1sh_gather_index_u64(svbool_t, svuint64_t, int64_t);
8162		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32base_index_s32)))
8163		svint32_t svldnt1sh_gather_index_s32(svbool_t, svuint32_t, int64_t);
8164		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64base_index_s64)))
8165		svint64_t svldnt1sh_gather_index_s64(svbool_t, svuint64_t, int64_t);
8166		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32base_offset_u32)))
8167		svuint32_t svldnt1sh_gather_offset_u32(svbool_t, svuint32_t, int64_t);
8168		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64base_offset_u64)))
8169		svuint64_t svldnt1sh_gather_offset_u64(svbool_t, svuint64_t, int64_t);
8170		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32base_offset_s32)))
8171		svint32_t svldnt1sh_gather_offset_s32(svbool_t, svuint32_t, int64_t);
8172		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64base_offset_s64)))
8173		svint64_t svldnt1sh_gather_offset_s64(svbool_t, svuint64_t, int64_t);
8174		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32base_u32)))
8175		svuint32_t svldnt1sh_gather_u32(svbool_t, svuint32_t);
8176		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64base_u64)))
8177		svuint64_t svldnt1sh_gather_u64(svbool_t, svuint64_t);
8178		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32base_s32)))
8179		svint32_t svldnt1sh_gather_s32(svbool_t, svuint32_t);
8180		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64base_s64)))
8181		svint64_t svldnt1sh_gather_s64(svbool_t, svuint64_t);
8182		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_s64index_u64)))
8183		svuint64_t svldnt1sh_gather_index_u64(svbool_t, int16_t const *, svint64_t);
8184		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_s64index_s64)))
8185		svint64_t svldnt1sh_gather_index_s64(svbool_t, int16_t const *, svint64_t);
8186		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64index_u64)))
8187		svuint64_t svldnt1sh_gather_index_u64(svbool_t, int16_t const *, svuint64_t);
8188		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64index_s64)))
8189		svint64_t svldnt1sh_gather_index_s64(svbool_t, int16_t const *, svuint64_t);
8190		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32offset_u32)))
8191		svuint32_t svldnt1sh_gather_offset_u32(svbool_t, int16_t const *, svuint32_t);
8192		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32offset_s32)))
8193		svint32_t svldnt1sh_gather_offset_s32(svbool_t, int16_t const *, svuint32_t);
8194		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_s64offset_u64)))
8195		svuint64_t svldnt1sh_gather_offset_u64(svbool_t, int16_t const *, svint64_t);
8196		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_s64offset_s64)))
8197		svint64_t svldnt1sh_gather_offset_s64(svbool_t, int16_t const *, svint64_t);
8198		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64offset_u64)))
8199		svuint64_t svldnt1sh_gather_offset_u64(svbool_t, int16_t const *, svuint64_t);
8200		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64offset_s64)))
8201		svint64_t svldnt1sh_gather_offset_s64(svbool_t, int16_t const *, svuint64_t);
8202		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64base_index_u64)))
8203		svuint64_t svldnt1sw_gather_index_u64(svbool_t, svuint64_t, int64_t);
8204		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64base_index_s64)))
8205		svint64_t svldnt1sw_gather_index_s64(svbool_t, svuint64_t, int64_t);
8206		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64base_offset_u64)))
8207		svuint64_t svldnt1sw_gather_offset_u64(svbool_t, svuint64_t, int64_t);
8208		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64base_offset_s64)))
8209		svint64_t svldnt1sw_gather_offset_s64(svbool_t, svuint64_t, int64_t);
8210		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64base_u64)))
8211		svuint64_t svldnt1sw_gather_u64(svbool_t, svuint64_t);
8212		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64base_s64)))
8213		svint64_t svldnt1sw_gather_s64(svbool_t, svuint64_t);
8214		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_s64index_u64)))
8215		svuint64_t svldnt1sw_gather_index_u64(svbool_t, int32_t const *, svint64_t);
8216		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_s64index_s64)))
8217		svint64_t svldnt1sw_gather_index_s64(svbool_t, int32_t const *, svint64_t);
8218		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64index_u64)))
8219		svuint64_t svldnt1sw_gather_index_u64(svbool_t, int32_t const *, svuint64_t);
8220		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64index_s64)))
8221		svint64_t svldnt1sw_gather_index_s64(svbool_t, int32_t const *, svuint64_t);
8222		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_s64offset_u64)))
8223		svuint64_t svldnt1sw_gather_offset_u64(svbool_t, int32_t const *, svint64_t);
8224		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_s64offset_s64)))
8225		svint64_t svldnt1sw_gather_offset_s64(svbool_t, int32_t const *, svint64_t);
8226		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64offset_u64)))
8227		svuint64_t svldnt1sw_gather_offset_u64(svbool_t, int32_t const *, svuint64_t);
8228		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64offset_s64)))
8229		svint64_t svldnt1sw_gather_offset_s64(svbool_t, int32_t const *, svuint64_t);
8230		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u32base_offset_u32)))
8231		svuint32_t svldnt1ub_gather_offset_u32(svbool_t, svuint32_t, int64_t);
8232		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u64base_offset_u64)))
8233		svuint64_t svldnt1ub_gather_offset_u64(svbool_t, svuint64_t, int64_t);
8234		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u32base_offset_s32)))
8235		svint32_t svldnt1ub_gather_offset_s32(svbool_t, svuint32_t, int64_t);
8236		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u64base_offset_s64)))
8237		svint64_t svldnt1ub_gather_offset_s64(svbool_t, svuint64_t, int64_t);
8238		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u32base_u32)))
8239		svuint32_t svldnt1ub_gather_u32(svbool_t, svuint32_t);
8240		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u64base_u64)))
8241		svuint64_t svldnt1ub_gather_u64(svbool_t, svuint64_t);
8242		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u32base_s32)))
8243		svint32_t svldnt1ub_gather_s32(svbool_t, svuint32_t);
8244		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u64base_s64)))
8245		svint64_t svldnt1ub_gather_s64(svbool_t, svuint64_t);
8246		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u32offset_u32)))
8247		svuint32_t svldnt1ub_gather_offset_u32(svbool_t, uint8_t const *, svuint32_t);
8248		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u32offset_s32)))
8249		svint32_t svldnt1ub_gather_offset_s32(svbool_t, uint8_t const *, svuint32_t);
8250		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_s64offset_u64)))
8251		svuint64_t svldnt1ub_gather_offset_u64(svbool_t, uint8_t const *, svint64_t);
8252		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_s64offset_s64)))
8253		svint64_t svldnt1ub_gather_offset_s64(svbool_t, uint8_t const *, svint64_t);
8254		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u64offset_u64)))
8255		svuint64_t svldnt1ub_gather_offset_u64(svbool_t, uint8_t const *, svuint64_t);
8256		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u64offset_s64)))
8257		svint64_t svldnt1ub_gather_offset_s64(svbool_t, uint8_t const *, svuint64_t);
8258		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32base_index_u32)))
8259		svuint32_t svldnt1uh_gather_index_u32(svbool_t, svuint32_t, int64_t);
8260		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64base_index_u64)))
8261		svuint64_t svldnt1uh_gather_index_u64(svbool_t, svuint64_t, int64_t);
8262		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32base_index_s32)))
8263		svint32_t svldnt1uh_gather_index_s32(svbool_t, svuint32_t, int64_t);
8264		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64base_index_s64)))
8265		svint64_t svldnt1uh_gather_index_s64(svbool_t, svuint64_t, int64_t);
8266		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32base_offset_u32)))
8267		svuint32_t svldnt1uh_gather_offset_u32(svbool_t, svuint32_t, int64_t);
8268		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64base_offset_u64)))
8269		svuint64_t svldnt1uh_gather_offset_u64(svbool_t, svuint64_t, int64_t);
8270		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32base_offset_s32)))
8271		svint32_t svldnt1uh_gather_offset_s32(svbool_t, svuint32_t, int64_t);
8272		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64base_offset_s64)))
8273		svint64_t svldnt1uh_gather_offset_s64(svbool_t, svuint64_t, int64_t);
8274		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32base_u32)))
8275		svuint32_t svldnt1uh_gather_u32(svbool_t, svuint32_t);
8276		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64base_u64)))
8277		svuint64_t svldnt1uh_gather_u64(svbool_t, svuint64_t);
8278		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32base_s32)))
8279		svint32_t svldnt1uh_gather_s32(svbool_t, svuint32_t);
8280		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64base_s64)))
8281		svint64_t svldnt1uh_gather_s64(svbool_t, svuint64_t);
8282		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_s64index_u64)))
8283		svuint64_t svldnt1uh_gather_index_u64(svbool_t, uint16_t const *, svint64_t);
8284		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_s64index_s64)))
8285		svint64_t svldnt1uh_gather_index_s64(svbool_t, uint16_t const *, svint64_t);
8286		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64index_u64)))
8287		svuint64_t svldnt1uh_gather_index_u64(svbool_t, uint16_t const *, svuint64_t);
8288		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64index_s64)))
8289		svint64_t svldnt1uh_gather_index_s64(svbool_t, uint16_t const *, svuint64_t);
8290		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32offset_u32)))
8291		svuint32_t svldnt1uh_gather_offset_u32(svbool_t, uint16_t const *, svuint32_t);
8292		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32offset_s32)))
8293		svint32_t svldnt1uh_gather_offset_s32(svbool_t, uint16_t const *, svuint32_t);
8294		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_s64offset_u64)))
8295		svuint64_t svldnt1uh_gather_offset_u64(svbool_t, uint16_t const *, svint64_t);
8296		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_s64offset_s64)))
8297		svint64_t svldnt1uh_gather_offset_s64(svbool_t, uint16_t const *, svint64_t);
8298		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64offset_u64)))
8299		svuint64_t svldnt1uh_gather_offset_u64(svbool_t, uint16_t const *, svuint64_t);
8300		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64offset_s64)))
8301		svint64_t svldnt1uh_gather_offset_s64(svbool_t, uint16_t const *, svuint64_t);
8302		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64base_index_u64)))
8303		svuint64_t svldnt1uw_gather_index_u64(svbool_t, svuint64_t, int64_t);
8304		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64base_index_s64)))
8305		svint64_t svldnt1uw_gather_index_s64(svbool_t, svuint64_t, int64_t);
8306		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64base_offset_u64)))
8307		svuint64_t svldnt1uw_gather_offset_u64(svbool_t, svuint64_t, int64_t);
8308		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64base_offset_s64)))
8309		svint64_t svldnt1uw_gather_offset_s64(svbool_t, svuint64_t, int64_t);
8310		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64base_u64)))
8311		svuint64_t svldnt1uw_gather_u64(svbool_t, svuint64_t);
8312		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64base_s64)))
8313		svint64_t svldnt1uw_gather_s64(svbool_t, svuint64_t);
8314		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_s64index_u64)))
8315		svuint64_t svldnt1uw_gather_index_u64(svbool_t, uint32_t const *, svint64_t);
8316		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_s64index_s64)))
8317		svint64_t svldnt1uw_gather_index_s64(svbool_t, uint32_t const *, svint64_t);
8318		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64index_u64)))
8319		svuint64_t svldnt1uw_gather_index_u64(svbool_t, uint32_t const *, svuint64_t);
8320		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64index_s64)))
8321		svint64_t svldnt1uw_gather_index_s64(svbool_t, uint32_t const *, svuint64_t);
8322		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_s64offset_u64)))
8323		svuint64_t svldnt1uw_gather_offset_u64(svbool_t, uint32_t const *, svint64_t);
8324		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_s64offset_s64)))
8325		svint64_t svldnt1uw_gather_offset_s64(svbool_t, uint32_t const *, svint64_t);
8326		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64offset_u64)))
8327		svuint64_t svldnt1uw_gather_offset_u64(svbool_t, uint32_t const *, svuint64_t);
8328		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64offset_s64)))
8329		svint64_t svldnt1uw_gather_offset_s64(svbool_t, uint32_t const *, svuint64_t);
8330		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmatch_u8)))
8331		svbool_t svmatch(svbool_t, svuint8_t, svuint8_t);
8332		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmatch_u16)))
8333		svbool_t svmatch(svbool_t, svuint16_t, svuint16_t);
8334		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmatch_s8)))
8335		svbool_t svmatch(svbool_t, svint8_t, svint8_t);
8336		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmatch_s16)))
8337		svbool_t svmatch(svbool_t, svint16_t, svint16_t);
8338		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svnmatch_u8)))
8339		svbool_t svnmatch(svbool_t, svuint8_t, svuint8_t);
8340		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svnmatch_u16)))
8341		svbool_t svnmatch(svbool_t, svuint16_t, svuint16_t);
8342		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svnmatch_s8)))
8343		svbool_t svnmatch(svbool_t, svint8_t, svint8_t);
8344		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svnmatch_s16)))
8345		svbool_t svnmatch(svbool_t, svint16_t, svint16_t);
8346		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_index_u32)))
8347		void svstnt1_scatter_index(svbool_t, svuint32_t, int64_t, svuint32_t);
8348		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_index_u64)))
8349		void svstnt1_scatter_index(svbool_t, svuint64_t, int64_t, svuint64_t);
8350		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_index_f64)))
8351		void svstnt1_scatter_index(svbool_t, svuint64_t, int64_t, svfloat64_t);
8352		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_index_f32)))
8353		void svstnt1_scatter_index(svbool_t, svuint32_t, int64_t, svfloat32_t);
8354		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_index_s32)))
8355		void svstnt1_scatter_index(svbool_t, svuint32_t, int64_t, svint32_t);
8356		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_index_s64)))
8357		void svstnt1_scatter_index(svbool_t, svuint64_t, int64_t, svint64_t);
8358		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_offset_u32)))
8359		void svstnt1_scatter_offset(svbool_t, svuint32_t, int64_t, svuint32_t);
8360		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_offset_u64)))
8361		void svstnt1_scatter_offset(svbool_t, svuint64_t, int64_t, svuint64_t);
8362		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_offset_f64)))
8363		void svstnt1_scatter_offset(svbool_t, svuint64_t, int64_t, svfloat64_t);
8364		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_offset_f32)))
8365		void svstnt1_scatter_offset(svbool_t, svuint32_t, int64_t, svfloat32_t);
8366		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_offset_s32)))
8367		void svstnt1_scatter_offset(svbool_t, svuint32_t, int64_t, svint32_t);
8368		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_offset_s64)))
8369		void svstnt1_scatter_offset(svbool_t, svuint64_t, int64_t, svint64_t);
8370		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_u32)))
8371		void svstnt1_scatter(svbool_t, svuint32_t, svuint32_t);
8372		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_u64)))
8373		void svstnt1_scatter(svbool_t, svuint64_t, svuint64_t);
8374		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_f64)))
8375		void svstnt1_scatter(svbool_t, svuint64_t, svfloat64_t);
8376		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_f32)))
8377		void svstnt1_scatter(svbool_t, svuint32_t, svfloat32_t);
8378		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_s32)))
8379		void svstnt1_scatter(svbool_t, svuint32_t, svint32_t);
8380		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_s64)))
8381		void svstnt1_scatter(svbool_t, svuint64_t, svint64_t);
8382		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_s64index_u64)))
8383		void svstnt1_scatter_index(svbool_t, uint64_t *, svint64_t, svuint64_t);
8384		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_s64index_f64)))
8385		void svstnt1_scatter_index(svbool_t, float64_t *, svint64_t, svfloat64_t);
8386		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_s64index_s64)))
8387		void svstnt1_scatter_index(svbool_t, int64_t *, svint64_t, svint64_t);
8388		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64index_u64)))
8389		void svstnt1_scatter_index(svbool_t, uint64_t *, svuint64_t, svuint64_t);
8390		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64index_f64)))
8391		void svstnt1_scatter_index(svbool_t, float64_t *, svuint64_t, svfloat64_t);
8392		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64index_s64)))
8393		void svstnt1_scatter_index(svbool_t, int64_t *, svuint64_t, svint64_t);
8394		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32offset_u32)))
8395		void svstnt1_scatter_offset(svbool_t, uint32_t *, svuint32_t, svuint32_t);
8396		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32offset_f32)))
8397		void svstnt1_scatter_offset(svbool_t, float32_t *, svuint32_t, svfloat32_t);
8398		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32offset_s32)))
8399		void svstnt1_scatter_offset(svbool_t, int32_t *, svuint32_t, svint32_t);
8400		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_s64offset_u64)))
8401		void svstnt1_scatter_offset(svbool_t, uint64_t *, svint64_t, svuint64_t);
8402		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_s64offset_f64)))
8403		void svstnt1_scatter_offset(svbool_t, float64_t *, svint64_t, svfloat64_t);
8404		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_s64offset_s64)))
8405		void svstnt1_scatter_offset(svbool_t, int64_t *, svint64_t, svint64_t);
8406		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64offset_u64)))
8407		void svstnt1_scatter_offset(svbool_t, uint64_t *, svuint64_t, svuint64_t);
8408		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64offset_f64)))
8409		void svstnt1_scatter_offset(svbool_t, float64_t *, svuint64_t, svfloat64_t);
8410		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64offset_s64)))
8411		void svstnt1_scatter_offset(svbool_t, int64_t *, svuint64_t, svint64_t);
8412		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u32base_offset_u32)))
8413		void svstnt1b_scatter_offset(svbool_t, svuint32_t, int64_t, svuint32_t);
8414		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u64base_offset_u64)))
8415		void svstnt1b_scatter_offset(svbool_t, svuint64_t, int64_t, svuint64_t);
8416		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u32base_offset_s32)))
8417		void svstnt1b_scatter_offset(svbool_t, svuint32_t, int64_t, svint32_t);
8418		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u64base_offset_s64)))
8419		void svstnt1b_scatter_offset(svbool_t, svuint64_t, int64_t, svint64_t);
8420		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u32base_u32)))
8421		void svstnt1b_scatter(svbool_t, svuint32_t, svuint32_t);
8422		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u64base_u64)))
8423		void svstnt1b_scatter(svbool_t, svuint64_t, svuint64_t);
8424		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u32base_s32)))
8425		void svstnt1b_scatter(svbool_t, svuint32_t, svint32_t);
8426		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u64base_s64)))
8427		void svstnt1b_scatter(svbool_t, svuint64_t, svint64_t);
8428		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u32offset_s32)))
8429		void svstnt1b_scatter_offset(svbool_t, int8_t *, svuint32_t, svint32_t);
8430		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u32offset_u32)))
8431		void svstnt1b_scatter_offset(svbool_t, uint8_t *, svuint32_t, svuint32_t);
8432		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_s64offset_s64)))
8433		void svstnt1b_scatter_offset(svbool_t, int8_t *, svint64_t, svint64_t);
8434		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_s64offset_u64)))
8435		void svstnt1b_scatter_offset(svbool_t, uint8_t *, svint64_t, svuint64_t);
8436		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u64offset_s64)))
8437		void svstnt1b_scatter_offset(svbool_t, int8_t *, svuint64_t, svint64_t);
8438		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u64offset_u64)))
8439		void svstnt1b_scatter_offset(svbool_t, uint8_t *, svuint64_t, svuint64_t);
8440		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32base_index_u32)))
8441		void svstnt1h_scatter_index(svbool_t, svuint32_t, int64_t, svuint32_t);
8442		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64base_index_u64)))
8443		void svstnt1h_scatter_index(svbool_t, svuint64_t, int64_t, svuint64_t);
8444		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32base_index_s32)))
8445		void svstnt1h_scatter_index(svbool_t, svuint32_t, int64_t, svint32_t);
8446		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64base_index_s64)))
8447		void svstnt1h_scatter_index(svbool_t, svuint64_t, int64_t, svint64_t);
8448		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32base_offset_u32)))
8449		void svstnt1h_scatter_offset(svbool_t, svuint32_t, int64_t, svuint32_t);
8450		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64base_offset_u64)))
8451		void svstnt1h_scatter_offset(svbool_t, svuint64_t, int64_t, svuint64_t);
8452		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32base_offset_s32)))
8453		void svstnt1h_scatter_offset(svbool_t, svuint32_t, int64_t, svint32_t);
8454		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64base_offset_s64)))
8455		void svstnt1h_scatter_offset(svbool_t, svuint64_t, int64_t, svint64_t);
8456		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32base_u32)))
8457		void svstnt1h_scatter(svbool_t, svuint32_t, svuint32_t);
8458		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64base_u64)))
8459		void svstnt1h_scatter(svbool_t, svuint64_t, svuint64_t);
8460		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32base_s32)))
8461		void svstnt1h_scatter(svbool_t, svuint32_t, svint32_t);
8462		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64base_s64)))
8463		void svstnt1h_scatter(svbool_t, svuint64_t, svint64_t);
8464		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_s64index_s64)))
8465		void svstnt1h_scatter_index(svbool_t, int16_t *, svint64_t, svint64_t);
8466		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_s64index_u64)))
8467		void svstnt1h_scatter_index(svbool_t, uint16_t *, svint64_t, svuint64_t);
8468		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64index_s64)))
8469		void svstnt1h_scatter_index(svbool_t, int16_t *, svuint64_t, svint64_t);
8470		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64index_u64)))
8471		void svstnt1h_scatter_index(svbool_t, uint16_t *, svuint64_t, svuint64_t);
8472		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32offset_s32)))
8473		void svstnt1h_scatter_offset(svbool_t, int16_t *, svuint32_t, svint32_t);
8474		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32offset_u32)))
8475		void svstnt1h_scatter_offset(svbool_t, uint16_t *, svuint32_t, svuint32_t);
8476		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_s64offset_s64)))
8477		void svstnt1h_scatter_offset(svbool_t, int16_t *, svint64_t, svint64_t);
8478		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_s64offset_u64)))
8479		void svstnt1h_scatter_offset(svbool_t, uint16_t *, svint64_t, svuint64_t);
8480		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64offset_s64)))
8481		void svstnt1h_scatter_offset(svbool_t, int16_t *, svuint64_t, svint64_t);
8482		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64offset_u64)))
8483		void svstnt1h_scatter_offset(svbool_t, uint16_t *, svuint64_t, svuint64_t);
8484		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64base_index_u64)))
8485		void svstnt1w_scatter_index(svbool_t, svuint64_t, int64_t, svuint64_t);
8486		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64base_index_s64)))
8487		void svstnt1w_scatter_index(svbool_t, svuint64_t, int64_t, svint64_t);
8488		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64base_offset_u64)))
8489		void svstnt1w_scatter_offset(svbool_t, svuint64_t, int64_t, svuint64_t);
8490		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64base_offset_s64)))
8491		void svstnt1w_scatter_offset(svbool_t, svuint64_t, int64_t, svint64_t);
8492		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64base_u64)))
8493		void svstnt1w_scatter(svbool_t, svuint64_t, svuint64_t);
8494		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64base_s64)))
8495		void svstnt1w_scatter(svbool_t, svuint64_t, svint64_t);
8496		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_s64index_s64)))
8497		void svstnt1w_scatter_index(svbool_t, int32_t *, svint64_t, svint64_t);
8498		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_s64index_u64)))
8499		void svstnt1w_scatter_index(svbool_t, uint32_t *, svint64_t, svuint64_t);
8500		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64index_s64)))
8501		void svstnt1w_scatter_index(svbool_t, int32_t *, svuint64_t, svint64_t);
8502		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64index_u64)))
8503		void svstnt1w_scatter_index(svbool_t, uint32_t *, svuint64_t, svuint64_t);
8504		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_s64offset_s64)))
8505		void svstnt1w_scatter_offset(svbool_t, int32_t *, svint64_t, svint64_t);
8506		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_s64offset_u64)))
8507		void svstnt1w_scatter_offset(svbool_t, uint32_t *, svint64_t, svuint64_t);
8508		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64offset_s64)))
8509		void svstnt1w_scatter_offset(svbool_t, int32_t *, svuint64_t, svint64_t);
8510		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64offset_u64)))
8511		void svstnt1w_scatter_offset(svbool_t, uint32_t *, svuint64_t, svuint64_t);
8512		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbl2_bf16)))
8513		svbfloat16_t svtbl2_bf16(svbfloat16x2_t, svuint16_t);
8514		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbx_bf16)))
8515		svbfloat16_t svtbx_bf16(svbfloat16_t, svbfloat16_t, svuint16_t);
8516		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilerw_bf16)))
8517		svbool_t svwhilerw_bf16(bfloat16_t const , bfloat16_t const );
8518		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilewr_bf16)))
8519		svbool_t svwhilewr_bf16(bfloat16_t const , bfloat16_t const );
8520		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbl2_bf16)))
8521		svbfloat16_t svtbl2(svbfloat16x2_t, svuint16_t);
8522		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbx_bf16)))
8523		svbfloat16_t svtbx(svbfloat16_t, svbfloat16_t, svuint16_t);
8524		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilerw_bf16)))
8525		svbool_t svwhilerw(bfloat16_t const , bfloat16_t const );
8526		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilewr_bf16)))
8527		svbool_t svwhilewr(bfloat16_t const , bfloat16_t const );
8528		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f64_m)))
8529		svfloat64_t svamax_n_f64_m(svbool_t, svfloat64_t, float64_t);
8530		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f32_m)))
8531		svfloat32_t svamax_n_f32_m(svbool_t, svfloat32_t, float32_t);
8532		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f16_m)))
8533		svfloat16_t svamax_n_f16_m(svbool_t, svfloat16_t, float16_t);
8534		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f64_x)))
8535		svfloat64_t svamax_n_f64_x(svbool_t, svfloat64_t, float64_t);
8536		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f32_x)))
8537		svfloat32_t svamax_n_f32_x(svbool_t, svfloat32_t, float32_t);
8538		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f16_x)))
8539		svfloat16_t svamax_n_f16_x(svbool_t, svfloat16_t, float16_t);
8540		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f64_z)))
8541		svfloat64_t svamax_n_f64_z(svbool_t, svfloat64_t, float64_t);
8542		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f32_z)))
8543		svfloat32_t svamax_n_f32_z(svbool_t, svfloat32_t, float32_t);
8544		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f16_z)))
8545		svfloat16_t svamax_n_f16_z(svbool_t, svfloat16_t, float16_t);
8546		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f64_m)))
8547		svfloat64_t svamax_f64_m(svbool_t, svfloat64_t, svfloat64_t);
8548		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f32_m)))
8549		svfloat32_t svamax_f32_m(svbool_t, svfloat32_t, svfloat32_t);
8550		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f16_m)))
8551		svfloat16_t svamax_f16_m(svbool_t, svfloat16_t, svfloat16_t);
8552		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f64_x)))
8553		svfloat64_t svamax_f64_x(svbool_t, svfloat64_t, svfloat64_t);
8554		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f32_x)))
8555		svfloat32_t svamax_f32_x(svbool_t, svfloat32_t, svfloat32_t);
8556		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f16_x)))
8557		svfloat16_t svamax_f16_x(svbool_t, svfloat16_t, svfloat16_t);
8558		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f64_z)))
8559		svfloat64_t svamax_f64_z(svbool_t, svfloat64_t, svfloat64_t);
8560		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f32_z)))
8561		svfloat32_t svamax_f32_z(svbool_t, svfloat32_t, svfloat32_t);
8562		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f16_z)))
8563		svfloat16_t svamax_f16_z(svbool_t, svfloat16_t, svfloat16_t);
8564		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f64_m)))
8565		svfloat64_t svamin_n_f64_m(svbool_t, svfloat64_t, float64_t);
8566		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f32_m)))
8567		svfloat32_t svamin_n_f32_m(svbool_t, svfloat32_t, float32_t);
8568		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f16_m)))
8569		svfloat16_t svamin_n_f16_m(svbool_t, svfloat16_t, float16_t);
8570		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f64_x)))
8571		svfloat64_t svamin_n_f64_x(svbool_t, svfloat64_t, float64_t);
8572		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f32_x)))
8573		svfloat32_t svamin_n_f32_x(svbool_t, svfloat32_t, float32_t);
8574		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f16_x)))
8575		svfloat16_t svamin_n_f16_x(svbool_t, svfloat16_t, float16_t);
8576		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f64_z)))
8577		svfloat64_t svamin_n_f64_z(svbool_t, svfloat64_t, float64_t);
8578		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f32_z)))
8579		svfloat32_t svamin_n_f32_z(svbool_t, svfloat32_t, float32_t);
8580		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f16_z)))
8581		svfloat16_t svamin_n_f16_z(svbool_t, svfloat16_t, float16_t);
8582		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f64_m)))
8583		svfloat64_t svamin_f64_m(svbool_t, svfloat64_t, svfloat64_t);
8584		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f32_m)))
8585		svfloat32_t svamin_f32_m(svbool_t, svfloat32_t, svfloat32_t);
8586		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f16_m)))
8587		svfloat16_t svamin_f16_m(svbool_t, svfloat16_t, svfloat16_t);
8588		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f64_x)))
8589		svfloat64_t svamin_f64_x(svbool_t, svfloat64_t, svfloat64_t);
8590		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f32_x)))
8591		svfloat32_t svamin_f32_x(svbool_t, svfloat32_t, svfloat32_t);
8592		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f16_x)))
8593		svfloat16_t svamin_f16_x(svbool_t, svfloat16_t, svfloat16_t);
8594		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f64_z)))
8595		svfloat64_t svamin_f64_z(svbool_t, svfloat64_t, svfloat64_t);
8596		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f32_z)))
8597		svfloat32_t svamin_f32_z(svbool_t, svfloat32_t, svfloat32_t);
8598		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f16_z)))
8599		svfloat16_t svamin_f16_z(svbool_t, svfloat16_t, svfloat16_t);
8600		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f64_m)))
8601		svfloat64_t svamax_m(svbool_t, svfloat64_t, float64_t);
8602		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f32_m)))
8603		svfloat32_t svamax_m(svbool_t, svfloat32_t, float32_t);
8604		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f16_m)))
8605		svfloat16_t svamax_m(svbool_t, svfloat16_t, float16_t);
8606		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f64_x)))
8607		svfloat64_t svamax_x(svbool_t, svfloat64_t, float64_t);
8608		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f32_x)))
8609		svfloat32_t svamax_x(svbool_t, svfloat32_t, float32_t);
8610		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f16_x)))
8611		svfloat16_t svamax_x(svbool_t, svfloat16_t, float16_t);
8612		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f64_z)))
8613		svfloat64_t svamax_z(svbool_t, svfloat64_t, float64_t);
8614		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f32_z)))
8615		svfloat32_t svamax_z(svbool_t, svfloat32_t, float32_t);
8616		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f16_z)))
8617		svfloat16_t svamax_z(svbool_t, svfloat16_t, float16_t);
8618		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f64_m)))
8619		svfloat64_t svamax_m(svbool_t, svfloat64_t, svfloat64_t);
8620		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f32_m)))
8621		svfloat32_t svamax_m(svbool_t, svfloat32_t, svfloat32_t);
8622		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f16_m)))
8623		svfloat16_t svamax_m(svbool_t, svfloat16_t, svfloat16_t);
8624		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f64_x)))
8625		svfloat64_t svamax_x(svbool_t, svfloat64_t, svfloat64_t);
8626		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f32_x)))
8627		svfloat32_t svamax_x(svbool_t, svfloat32_t, svfloat32_t);
8628		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f16_x)))
8629		svfloat16_t svamax_x(svbool_t, svfloat16_t, svfloat16_t);
8630		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f64_z)))
8631		svfloat64_t svamax_z(svbool_t, svfloat64_t, svfloat64_t);
8632		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f32_z)))
8633		svfloat32_t svamax_z(svbool_t, svfloat32_t, svfloat32_t);
8634		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f16_z)))
8635		svfloat16_t svamax_z(svbool_t, svfloat16_t, svfloat16_t);
8636		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f64_m)))
8637		svfloat64_t svamin_m(svbool_t, svfloat64_t, float64_t);
8638		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f32_m)))
8639		svfloat32_t svamin_m(svbool_t, svfloat32_t, float32_t);
8640		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f16_m)))
8641		svfloat16_t svamin_m(svbool_t, svfloat16_t, float16_t);
8642		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f64_x)))
8643		svfloat64_t svamin_x(svbool_t, svfloat64_t, float64_t);
8644		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f32_x)))
8645		svfloat32_t svamin_x(svbool_t, svfloat32_t, float32_t);
8646		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f16_x)))
8647		svfloat16_t svamin_x(svbool_t, svfloat16_t, float16_t);
8648		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f64_z)))
8649		svfloat64_t svamin_z(svbool_t, svfloat64_t, float64_t);
8650		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f32_z)))
8651		svfloat32_t svamin_z(svbool_t, svfloat32_t, float32_t);
8652		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f16_z)))
8653		svfloat16_t svamin_z(svbool_t, svfloat16_t, float16_t);
8654		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f64_m)))
8655		svfloat64_t svamin_m(svbool_t, svfloat64_t, svfloat64_t);
8656		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f32_m)))
8657		svfloat32_t svamin_m(svbool_t, svfloat32_t, svfloat32_t);
8658		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f16_m)))
8659		svfloat16_t svamin_m(svbool_t, svfloat16_t, svfloat16_t);
8660		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f64_x)))
8661		svfloat64_t svamin_x(svbool_t, svfloat64_t, svfloat64_t);
8662		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f32_x)))
8663		svfloat32_t svamin_x(svbool_t, svfloat32_t, svfloat32_t);
8664		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f16_x)))
8665		svfloat16_t svamin_x(svbool_t, svfloat16_t, svfloat16_t);
8666		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f64_z)))
8667		svfloat64_t svamin_z(svbool_t, svfloat64_t, svfloat64_t);
8668		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f32_z)))
8669		svfloat32_t svamin_z(svbool_t, svfloat32_t, svfloat32_t);
8670		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f16_z)))
8671		svfloat16_t svamin_z(svbool_t, svfloat16_t, svfloat16_t);
8672		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_f16_mf8_fpm)))
8673		svfloat16_t svdot_f16_mf8_fpm(svfloat16_t, svmfloat8_t, svmfloat8_t, fpm_t);
8674		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_n_f16_mf8_fpm)))
8675		svfloat16_t svdot_n_f16_mf8_fpm(svfloat16_t, svmfloat8_t, mfloat8_t, fpm_t);
8676		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_lane_f16_mf8_fpm)))
8677		svfloat16_t svdot_lane_f16_mf8_fpm(svfloat16_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
8678		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_f16_mf8_fpm)))
8679		svfloat16_t svdot_fpm(svfloat16_t, svmfloat8_t, svmfloat8_t, fpm_t);
8680		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_n_f16_mf8_fpm)))
8681		svfloat16_t svdot_fpm(svfloat16_t, svmfloat8_t, mfloat8_t, fpm_t);
8682		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_lane_f16_mf8_fpm)))
8683		svfloat16_t svdot_lane_fpm(svfloat16_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
8684		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_f32_mf8_fpm)))
8685		svfloat32_t svdot_f32_mf8_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, fpm_t);
8686		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_n_f32_mf8_fpm)))
8687		svfloat32_t svdot_n_f32_mf8_fpm(svfloat32_t, svmfloat8_t, mfloat8_t, fpm_t);
8688		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_lane_f32_mf8_fpm)))
8689		svfloat32_t svdot_lane_f32_mf8_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
8690		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_f32_mf8_fpm)))
8691		svfloat32_t svdot_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, fpm_t);
8692		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_n_f32_mf8_fpm)))
8693		svfloat32_t svdot_fpm(svfloat32_t, svmfloat8_t, mfloat8_t, fpm_t);
8694		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_lane_f32_mf8_fpm)))
8695		svfloat32_t svdot_lane_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
8696		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalb_f16_mf8_fpm)))
8697		svfloat16_t svmlalb_f16_mf8_fpm(svfloat16_t, svmfloat8_t, svmfloat8_t, fpm_t);
8698		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalb_n_f16_mf8_fpm)))
8699		svfloat16_t svmlalb_n_f16_mf8_fpm(svfloat16_t, svmfloat8_t, mfloat8_t, fpm_t);
8700		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalb_lane_f16_mf8_fpm)))
8701		svfloat16_t svmlalb_lane_f16_mf8_fpm(svfloat16_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
8702		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlallbb_f32_mf8_fpm)))
8703		svfloat32_t svmlallbb_f32_mf8_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, fpm_t);
8704		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlallbb_n_f32_mf8_fpm)))
8705		svfloat32_t svmlallbb_n_f32_mf8_fpm(svfloat32_t, svmfloat8_t, mfloat8_t, fpm_t);
8706		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlallbb_lane_f32_mf8_fpm)))
8707		svfloat32_t svmlallbb_lane_f32_mf8_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
8708		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlallbt_f32_mf8_fpm)))
8709		svfloat32_t svmlallbt_f32_mf8_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, fpm_t);
8710		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlallbt_n_f32_mf8_fpm)))
8711		svfloat32_t svmlallbt_n_f32_mf8_fpm(svfloat32_t, svmfloat8_t, mfloat8_t, fpm_t);
8712		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlallbt_lane_f32_mf8_fpm)))
8713		svfloat32_t svmlallbt_lane_f32_mf8_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
8714		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalltb_f32_mf8_fpm)))
8715		svfloat32_t svmlalltb_f32_mf8_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, fpm_t);
8716		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalltb_n_f32_mf8_fpm)))
8717		svfloat32_t svmlalltb_n_f32_mf8_fpm(svfloat32_t, svmfloat8_t, mfloat8_t, fpm_t);
8718		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalltb_lane_f32_mf8_fpm)))
8719		svfloat32_t svmlalltb_lane_f32_mf8_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
8720		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalltt_f32_mf8_fpm)))
8721		svfloat32_t svmlalltt_f32_mf8_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, fpm_t);
8722		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalltt_n_f32_mf8_fpm)))
8723		svfloat32_t svmlalltt_n_f32_mf8_fpm(svfloat32_t, svmfloat8_t, mfloat8_t, fpm_t);
8724		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalltt_lane_f32_mf8_fpm)))
8725		svfloat32_t svmlalltt_lane_f32_mf8_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
8726		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalt_f16_mf8_fpm)))
8727		svfloat16_t svmlalt_f16_mf8_fpm(svfloat16_t, svmfloat8_t, svmfloat8_t, fpm_t);
8728		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalt_n_f16_mf8_fpm)))
8729		svfloat16_t svmlalt_n_f16_mf8_fpm(svfloat16_t, svmfloat8_t, mfloat8_t, fpm_t);
8730		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalt_lane_f16_mf8_fpm)))
8731		svfloat16_t svmlalt_lane_f16_mf8_fpm(svfloat16_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
8732		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalb_f16_mf8_fpm)))
8733		svfloat16_t svmlalb_fpm(svfloat16_t, svmfloat8_t, svmfloat8_t, fpm_t);
8734		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalb_n_f16_mf8_fpm)))
8735		svfloat16_t svmlalb_fpm(svfloat16_t, svmfloat8_t, mfloat8_t, fpm_t);
8736		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalb_lane_f16_mf8_fpm)))
8737		svfloat16_t svmlalb_lane_fpm(svfloat16_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
8738		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlallbb_f32_mf8_fpm)))
8739		svfloat32_t svmlallbb_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, fpm_t);
8740		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlallbb_n_f32_mf8_fpm)))
8741		svfloat32_t svmlallbb_fpm(svfloat32_t, svmfloat8_t, mfloat8_t, fpm_t);
8742		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlallbb_lane_f32_mf8_fpm)))
8743		svfloat32_t svmlallbb_lane_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
8744		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlallbt_f32_mf8_fpm)))
8745		svfloat32_t svmlallbt_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, fpm_t);
8746		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlallbt_n_f32_mf8_fpm)))
8747		svfloat32_t svmlallbt_fpm(svfloat32_t, svmfloat8_t, mfloat8_t, fpm_t);
8748		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlallbt_lane_f32_mf8_fpm)))
8749		svfloat32_t svmlallbt_lane_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
8750		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalltb_f32_mf8_fpm)))
8751		svfloat32_t svmlalltb_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, fpm_t);
8752		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalltb_n_f32_mf8_fpm)))
8753		svfloat32_t svmlalltb_fpm(svfloat32_t, svmfloat8_t, mfloat8_t, fpm_t);
8754		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalltb_lane_f32_mf8_fpm)))
8755		svfloat32_t svmlalltb_lane_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
8756		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalltt_f32_mf8_fpm)))
8757		svfloat32_t svmlalltt_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, fpm_t);
8758		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalltt_n_f32_mf8_fpm)))
8759		svfloat32_t svmlalltt_fpm(svfloat32_t, svmfloat8_t, mfloat8_t, fpm_t);
8760		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalltt_lane_f32_mf8_fpm)))
8761		svfloat32_t svmlalltt_lane_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
8762		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalt_f16_mf8_fpm)))
8763		svfloat16_t svmlalt_fpm(svfloat16_t, svmfloat8_t, svmfloat8_t, fpm_t);
8764		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalt_n_f16_mf8_fpm)))
8765		svfloat16_t svmlalt_fpm(svfloat16_t, svmfloat8_t, mfloat8_t, fpm_t);
8766		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalt_lane_f16_mf8_fpm)))
8767		svfloat16_t svmlalt_lane_fpm(svfloat16_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
8768		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvt1_bf16_mf8_fpm)))
8769		svbfloat16_t svcvt1_bf16_mf8_fpm(svmfloat8_t, fpm_t);
8770		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvt1_f16_mf8_fpm)))
8771		svfloat16_t svcvt1_f16_mf8_fpm(svmfloat8_t, fpm_t);
8772		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvt2_bf16_mf8_fpm)))
8773		svbfloat16_t svcvt2_bf16_mf8_fpm(svmfloat8_t, fpm_t);
8774		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvt2_f16_mf8_fpm)))
8775		svfloat16_t svcvt2_f16_mf8_fpm(svmfloat8_t, fpm_t);
8776		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtlt1_bf16_mf8_fpm)))
8777		svbfloat16_t svcvtlt1_bf16_mf8_fpm(svmfloat8_t, fpm_t);
8778		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtlt1_f16_mf8_fpm)))
8779		svfloat16_t svcvtlt1_f16_mf8_fpm(svmfloat8_t, fpm_t);
8780		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtlt2_bf16_mf8_fpm)))
8781		svbfloat16_t svcvtlt2_bf16_mf8_fpm(svmfloat8_t, fpm_t);
8782		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtlt2_f16_mf8_fpm)))
8783		svfloat16_t svcvtlt2_f16_mf8_fpm(svmfloat8_t, fpm_t);
8784		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtn_mf8_bf16_x2_fpm)))
8785		svmfloat8_t svcvtn_mf8_bf16_x2_fpm(svbfloat16x2_t, fpm_t);
8786		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtn_mf8_f16_x2_fpm)))
8787		svmfloat8_t svcvtn_mf8_f16_x2_fpm(svfloat16x2_t, fpm_t);
8788		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtnb_mf8_f32_x2_fpm)))
8789		svmfloat8_t svcvtnb_mf8_f32_x2_fpm(svfloat32x2_t, fpm_t);
8790		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtnt_mf8_f32_x2_fpm)))
8791		svmfloat8_t svcvtnt_mf8_f32_x2_fpm(svmfloat8_t, svfloat32x2_t, fpm_t);
8792		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvt1_bf16_mf8_fpm)))
8793		svbfloat16_t svcvt1_bf16_fpm(svmfloat8_t, fpm_t);
8794		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvt1_f16_mf8_fpm)))
8795		svfloat16_t svcvt1_f16_fpm(svmfloat8_t, fpm_t);
8796		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvt2_bf16_mf8_fpm)))
8797		svbfloat16_t svcvt2_bf16_fpm(svmfloat8_t, fpm_t);
8798		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvt2_f16_mf8_fpm)))
8799		svfloat16_t svcvt2_f16_fpm(svmfloat8_t, fpm_t);
8800		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtlt1_bf16_mf8_fpm)))
8801		svbfloat16_t svcvtlt1_bf16_fpm(svmfloat8_t, fpm_t);
8802		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtlt1_f16_mf8_fpm)))
8803		svfloat16_t svcvtlt1_f16_fpm(svmfloat8_t, fpm_t);
8804		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtlt2_bf16_mf8_fpm)))
8805		svbfloat16_t svcvtlt2_bf16_fpm(svmfloat8_t, fpm_t);
8806		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtlt2_f16_mf8_fpm)))
8807		svfloat16_t svcvtlt2_f16_fpm(svmfloat8_t, fpm_t);
8808		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtn_mf8_bf16_x2_fpm)))
8809		svmfloat8_t svcvtn_mf8_fpm(svbfloat16x2_t, fpm_t);
8810		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtn_mf8_f16_x2_fpm)))
8811		svmfloat8_t svcvtn_mf8_fpm(svfloat16x2_t, fpm_t);
8812		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtnb_mf8_f32_x2_fpm)))
8813		svmfloat8_t svcvtnb_mf8_fpm(svfloat32x2_t, fpm_t);
8814		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtnt_mf8_f32_x2_fpm)))
8815		svmfloat8_t svcvtnt_mf8_fpm(svmfloat8_t, svfloat32x2_t, fpm_t);
8816		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti2_lane_bf16)))
8817		svbfloat16_t svluti2_lane_bf16(svbfloat16_t, svuint8_t, uint64_t);
8818		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_bf16)))
8819		svbfloat16_t svluti4_lane_bf16(svbfloat16_t, svuint8_t, uint64_t);
8820		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_bf16_x2)))
8821		svbfloat16_t svluti4_lane_bf16_x2(svbfloat16x2_t, svuint8_t, uint64_t);
8822		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti2_lane_bf16)))
8823		svbfloat16_t svluti2_lane(svbfloat16_t, svuint8_t, uint64_t);
8824		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_bf16)))
8825		svbfloat16_t svluti4_lane(svbfloat16_t, svuint8_t, uint64_t);
8826		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_bf16_x2)))
8827		svbfloat16_t svluti4_lane(svbfloat16x2_t, svuint8_t, uint64_t);
8828		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti2_lane_u8)))
8829		svuint8_t svluti2_lane_u8(svuint8_t, svuint8_t, uint64_t);
8830		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti2_lane_s8)))
8831		svint8_t svluti2_lane_s8(svint8_t, svuint8_t, uint64_t);
8832		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti2_lane_u16)))
8833		svuint16_t svluti2_lane_u16(svuint16_t, svuint8_t, uint64_t);
8834		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti2_lane_f16)))
8835		svfloat16_t svluti2_lane_f16(svfloat16_t, svuint8_t, uint64_t);
8836		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti2_lane_s16)))
8837		svint16_t svluti2_lane_s16(svint16_t, svuint8_t, uint64_t);
8838		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_u8)))
8839		svuint8_t svluti4_lane_u8(svuint8_t, svuint8_t, uint64_t);
8840		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_s8)))
8841		svint8_t svluti4_lane_s8(svint8_t, svuint8_t, uint64_t);
8842		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_u16)))
8843		svuint16_t svluti4_lane_u16(svuint16_t, svuint8_t, uint64_t);
8844		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_f16)))
8845		svfloat16_t svluti4_lane_f16(svfloat16_t, svuint8_t, uint64_t);
8846		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_s16)))
8847		svint16_t svluti4_lane_s16(svint16_t, svuint8_t, uint64_t);
8848		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_u16_x2)))
8849		svuint16_t svluti4_lane_u16_x2(svuint16x2_t, svuint8_t, uint64_t);
8850		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_f16_x2)))
8851		svfloat16_t svluti4_lane_f16_x2(svfloat16x2_t, svuint8_t, uint64_t);
8852		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_s16_x2)))
8853		svint16_t svluti4_lane_s16_x2(svint16x2_t, svuint8_t, uint64_t);
8854		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti2_lane_u8)))
8855		svuint8_t svluti2_lane(svuint8_t, svuint8_t, uint64_t);
8856		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti2_lane_s8)))
8857		svint8_t svluti2_lane(svint8_t, svuint8_t, uint64_t);
8858		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti2_lane_u16)))
8859		svuint16_t svluti2_lane(svuint16_t, svuint8_t, uint64_t);
8860		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti2_lane_f16)))
8861		svfloat16_t svluti2_lane(svfloat16_t, svuint8_t, uint64_t);
8862		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti2_lane_s16)))
8863		svint16_t svluti2_lane(svint16_t, svuint8_t, uint64_t);
8864		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_u8)))
8865		svuint8_t svluti4_lane(svuint8_t, svuint8_t, uint64_t);
8866		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_s8)))
8867		svint8_t svluti4_lane(svint8_t, svuint8_t, uint64_t);
8868		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_u16)))
8869		svuint16_t svluti4_lane(svuint16_t, svuint8_t, uint64_t);
8870		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_f16)))
8871		svfloat16_t svluti4_lane(svfloat16_t, svuint8_t, uint64_t);
8872		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_s16)))
8873		svint16_t svluti4_lane(svint16_t, svuint8_t, uint64_t);
8874		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_u16_x2)))
8875		svuint16_t svluti4_lane(svuint16x2_t, svuint8_t, uint64_t);
8876		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_f16_x2)))
8877		svfloat16_t svluti4_lane(svfloat16x2_t, svuint8_t, uint64_t);
8878		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_s16_x2)))
8879		svint16_t svluti4_lane(svint16x2_t, svuint8_t, uint64_t);
8880		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaesd_u8)))
8881		svuint8_t svaesd_u8(svuint8_t, svuint8_t);
8882		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaese_u8)))
8883		svuint8_t svaese_u8(svuint8_t, svuint8_t);
8884		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaesimc_u8)))
8885		svuint8_t svaesimc_u8(svuint8_t);
8886		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaesmc_u8)))
8887		svuint8_t svaesmc_u8(svuint8_t);
8888		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmullb_pair_n_u64)))
8889		svuint64_t svpmullb_pair_n_u64(svuint64_t, uint64_t);
8890		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmullb_pair_u64)))
8891		svuint64_t svpmullb_pair_u64(svuint64_t, svuint64_t);
8892		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmullt_pair_n_u64)))
8893		svuint64_t svpmullt_pair_n_u64(svuint64_t, uint64_t);
8894		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmullt_pair_u64)))
8895		svuint64_t svpmullt_pair_u64(svuint64_t, svuint64_t);
8896		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaesd_u8)))
8897		svuint8_t svaesd(svuint8_t, svuint8_t);
8898		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaese_u8)))
8899		svuint8_t svaese(svuint8_t, svuint8_t);
8900		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaesimc_u8)))
8901		svuint8_t svaesimc(svuint8_t);
8902		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaesmc_u8)))
8903		svuint8_t svaesmc(svuint8_t);
8904		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmullb_pair_n_u64)))
8905		svuint64_t svpmullb_pair(svuint64_t, uint64_t);
8906		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmullb_pair_u64)))
8907		svuint64_t svpmullb_pair(svuint64_t, svuint64_t);
8908		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmullt_pair_n_u64)))
8909		svuint64_t svpmullt_pair(svuint64_t, uint64_t);
8910		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmullt_pair_u64)))
8911		svuint64_t svpmullt_pair(svuint64_t, svuint64_t);
8912		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svadd_n_bf16_m)))
8913		svbfloat16_t svadd_n_bf16_m(svbool_t, svbfloat16_t, bfloat16_t);
8914		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svadd_n_bf16_x)))
8915		svbfloat16_t svadd_n_bf16_x(svbool_t, svbfloat16_t, bfloat16_t);
8916		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svadd_n_bf16_z)))
8917		svbfloat16_t svadd_n_bf16_z(svbool_t, svbfloat16_t, bfloat16_t);
8918		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svadd_bf16_m)))
8919		svbfloat16_t svadd_bf16_m(svbool_t, svbfloat16_t, svbfloat16_t);
8920		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svadd_bf16_x)))
8921		svbfloat16_t svadd_bf16_x(svbool_t, svbfloat16_t, svbfloat16_t);
8922		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svadd_bf16_z)))
8923		svbfloat16_t svadd_bf16_z(svbool_t, svbfloat16_t, svbfloat16_t);
8924		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_bf16)))
8925		svbfloat16_t svclamp_bf16(svbfloat16_t, svbfloat16_t, svbfloat16_t);
8926		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmax_n_bf16_m)))
8927		svbfloat16_t svmax_n_bf16_m(svbool_t, svbfloat16_t, bfloat16_t);
8928		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmax_n_bf16_x)))
8929		svbfloat16_t svmax_n_bf16_x(svbool_t, svbfloat16_t, bfloat16_t);
8930		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmax_n_bf16_z)))
8931		svbfloat16_t svmax_n_bf16_z(svbool_t, svbfloat16_t, bfloat16_t);
8932		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmax_bf16_m)))
8933		svbfloat16_t svmax_bf16_m(svbool_t, svbfloat16_t, svbfloat16_t);
8934		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmax_bf16_x)))
8935		svbfloat16_t svmax_bf16_x(svbool_t, svbfloat16_t, svbfloat16_t);
8936		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmax_bf16_z)))
8937		svbfloat16_t svmax_bf16_z(svbool_t, svbfloat16_t, svbfloat16_t);
8938		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnm_n_bf16_m)))
8939		svbfloat16_t svmaxnm_n_bf16_m(svbool_t, svbfloat16_t, bfloat16_t);
8940		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnm_n_bf16_x)))
8941		svbfloat16_t svmaxnm_n_bf16_x(svbool_t, svbfloat16_t, bfloat16_t);
8942		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnm_n_bf16_z)))
8943		svbfloat16_t svmaxnm_n_bf16_z(svbool_t, svbfloat16_t, bfloat16_t);
8944		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnm_bf16_m)))
8945		svbfloat16_t svmaxnm_bf16_m(svbool_t, svbfloat16_t, svbfloat16_t);
8946		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnm_bf16_x)))
8947		svbfloat16_t svmaxnm_bf16_x(svbool_t, svbfloat16_t, svbfloat16_t);
8948		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnm_bf16_z)))
8949		svbfloat16_t svmaxnm_bf16_z(svbool_t, svbfloat16_t, svbfloat16_t);
8950		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmin_n_bf16_m)))
8951		svbfloat16_t svmin_n_bf16_m(svbool_t, svbfloat16_t, bfloat16_t);
8952		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmin_n_bf16_x)))
8953		svbfloat16_t svmin_n_bf16_x(svbool_t, svbfloat16_t, bfloat16_t);
8954		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmin_n_bf16_z)))
8955		svbfloat16_t svmin_n_bf16_z(svbool_t, svbfloat16_t, bfloat16_t);
8956		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmin_bf16_m)))
8957		svbfloat16_t svmin_bf16_m(svbool_t, svbfloat16_t, svbfloat16_t);
8958		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmin_bf16_x)))
8959		svbfloat16_t svmin_bf16_x(svbool_t, svbfloat16_t, svbfloat16_t);
8960		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmin_bf16_z)))
8961		svbfloat16_t svmin_bf16_z(svbool_t, svbfloat16_t, svbfloat16_t);
8962		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnm_n_bf16_m)))
8963		svbfloat16_t svminnm_n_bf16_m(svbool_t, svbfloat16_t, bfloat16_t);
8964		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnm_n_bf16_x)))
8965		svbfloat16_t svminnm_n_bf16_x(svbool_t, svbfloat16_t, bfloat16_t);
8966		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnm_n_bf16_z)))
8967		svbfloat16_t svminnm_n_bf16_z(svbool_t, svbfloat16_t, bfloat16_t);
8968		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnm_bf16_m)))
8969		svbfloat16_t svminnm_bf16_m(svbool_t, svbfloat16_t, svbfloat16_t);
8970		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnm_bf16_x)))
8971		svbfloat16_t svminnm_bf16_x(svbool_t, svbfloat16_t, svbfloat16_t);
8972		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnm_bf16_z)))
8973		svbfloat16_t svminnm_bf16_z(svbool_t, svbfloat16_t, svbfloat16_t);
8974		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmla_n_bf16_m)))
8975		svbfloat16_t svmla_n_bf16_m(svbool_t, svbfloat16_t, svbfloat16_t, bfloat16_t);
8976		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmla_n_bf16_x)))
8977		svbfloat16_t svmla_n_bf16_x(svbool_t, svbfloat16_t, svbfloat16_t, bfloat16_t);
8978		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmla_n_bf16_z)))
8979		svbfloat16_t svmla_n_bf16_z(svbool_t, svbfloat16_t, svbfloat16_t, bfloat16_t);
8980		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmla_bf16_m)))
8981		svbfloat16_t svmla_bf16_m(svbool_t, svbfloat16_t, svbfloat16_t, svbfloat16_t);
8982		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmla_bf16_x)))
8983		svbfloat16_t svmla_bf16_x(svbool_t, svbfloat16_t, svbfloat16_t, svbfloat16_t);
8984		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmla_bf16_z)))
8985		svbfloat16_t svmla_bf16_z(svbool_t, svbfloat16_t, svbfloat16_t, svbfloat16_t);
8986		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmla_lane_bf16)))
8987		svbfloat16_t svmla_lane_bf16(svbfloat16_t, svbfloat16_t, svbfloat16_t, uint64_t);
8988		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmls_n_bf16_m)))
8989		svbfloat16_t svmls_n_bf16_m(svbool_t, svbfloat16_t, svbfloat16_t, bfloat16_t);
8990		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmls_n_bf16_x)))
8991		svbfloat16_t svmls_n_bf16_x(svbool_t, svbfloat16_t, svbfloat16_t, bfloat16_t);
8992		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmls_n_bf16_z)))
8993		svbfloat16_t svmls_n_bf16_z(svbool_t, svbfloat16_t, svbfloat16_t, bfloat16_t);
8994		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmls_bf16_m)))
8995		svbfloat16_t svmls_bf16_m(svbool_t, svbfloat16_t, svbfloat16_t, svbfloat16_t);
8996		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmls_bf16_x)))
8997		svbfloat16_t svmls_bf16_x(svbool_t, svbfloat16_t, svbfloat16_t, svbfloat16_t);
8998		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmls_bf16_z)))
8999		svbfloat16_t svmls_bf16_z(svbool_t, svbfloat16_t, svbfloat16_t, svbfloat16_t);
9000		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmls_lane_bf16)))
9001		svbfloat16_t svmls_lane_bf16(svbfloat16_t, svbfloat16_t, svbfloat16_t, uint64_t);
9002		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmul_n_bf16_m)))
9003		svbfloat16_t svmul_n_bf16_m(svbool_t, svbfloat16_t, bfloat16_t);
9004		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmul_n_bf16_x)))
9005		svbfloat16_t svmul_n_bf16_x(svbool_t, svbfloat16_t, bfloat16_t);
9006		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmul_n_bf16_z)))
9007		svbfloat16_t svmul_n_bf16_z(svbool_t, svbfloat16_t, bfloat16_t);
9008		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmul_bf16_m)))
9009		svbfloat16_t svmul_bf16_m(svbool_t, svbfloat16_t, svbfloat16_t);
9010		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmul_bf16_x)))
9011		svbfloat16_t svmul_bf16_x(svbool_t, svbfloat16_t, svbfloat16_t);
9012		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmul_bf16_z)))
9013		svbfloat16_t svmul_bf16_z(svbool_t, svbfloat16_t, svbfloat16_t);
9014		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmul_lane_bf16)))
9015		svbfloat16_t svmul_lane_bf16(svbfloat16_t, svbfloat16_t, uint64_t);
9016		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsub_n_bf16_m)))
9017		svbfloat16_t svsub_n_bf16_m(svbool_t, svbfloat16_t, bfloat16_t);
9018		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsub_n_bf16_x)))
9019		svbfloat16_t svsub_n_bf16_x(svbool_t, svbfloat16_t, bfloat16_t);
9020		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsub_n_bf16_z)))
9021		svbfloat16_t svsub_n_bf16_z(svbool_t, svbfloat16_t, bfloat16_t);
9022		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsub_bf16_m)))
9023		svbfloat16_t svsub_bf16_m(svbool_t, svbfloat16_t, svbfloat16_t);
9024		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsub_bf16_x)))
9025		svbfloat16_t svsub_bf16_x(svbool_t, svbfloat16_t, svbfloat16_t);
9026		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsub_bf16_z)))
9027		svbfloat16_t svsub_bf16_z(svbool_t, svbfloat16_t, svbfloat16_t);
9028		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svadd_n_bf16_m)))
9029		svbfloat16_t svadd_m(svbool_t, svbfloat16_t, bfloat16_t);
9030		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svadd_n_bf16_x)))
9031		svbfloat16_t svadd_x(svbool_t, svbfloat16_t, bfloat16_t);
9032		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svadd_n_bf16_z)))
9033		svbfloat16_t svadd_z(svbool_t, svbfloat16_t, bfloat16_t);
9034		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svadd_bf16_m)))
9035		svbfloat16_t svadd_m(svbool_t, svbfloat16_t, svbfloat16_t);
9036		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svadd_bf16_x)))
9037		svbfloat16_t svadd_x(svbool_t, svbfloat16_t, svbfloat16_t);
9038		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svadd_bf16_z)))
9039		svbfloat16_t svadd_z(svbool_t, svbfloat16_t, svbfloat16_t);
9040		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_bf16)))
9041		svbfloat16_t svclamp(svbfloat16_t, svbfloat16_t, svbfloat16_t);
9042		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmax_n_bf16_m)))
9043		svbfloat16_t svmax_m(svbool_t, svbfloat16_t, bfloat16_t);
9044		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmax_n_bf16_x)))
9045		svbfloat16_t svmax_x(svbool_t, svbfloat16_t, bfloat16_t);
9046		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmax_n_bf16_z)))
9047		svbfloat16_t svmax_z(svbool_t, svbfloat16_t, bfloat16_t);
9048		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmax_bf16_m)))
9049		svbfloat16_t svmax_m(svbool_t, svbfloat16_t, svbfloat16_t);
9050		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmax_bf16_x)))
9051		svbfloat16_t svmax_x(svbool_t, svbfloat16_t, svbfloat16_t);
9052		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmax_bf16_z)))
9053		svbfloat16_t svmax_z(svbool_t, svbfloat16_t, svbfloat16_t);
9054		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnm_n_bf16_m)))
9055		svbfloat16_t svmaxnm_m(svbool_t, svbfloat16_t, bfloat16_t);
9056		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnm_n_bf16_x)))
9057		svbfloat16_t svmaxnm_x(svbool_t, svbfloat16_t, bfloat16_t);
9058		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnm_n_bf16_z)))
9059		svbfloat16_t svmaxnm_z(svbool_t, svbfloat16_t, bfloat16_t);
9060		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnm_bf16_m)))
9061		svbfloat16_t svmaxnm_m(svbool_t, svbfloat16_t, svbfloat16_t);
9062		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnm_bf16_x)))
9063		svbfloat16_t svmaxnm_x(svbool_t, svbfloat16_t, svbfloat16_t);
9064		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnm_bf16_z)))
9065		svbfloat16_t svmaxnm_z(svbool_t, svbfloat16_t, svbfloat16_t);
9066		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmin_n_bf16_m)))
9067		svbfloat16_t svmin_m(svbool_t, svbfloat16_t, bfloat16_t);
9068		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmin_n_bf16_x)))
9069		svbfloat16_t svmin_x(svbool_t, svbfloat16_t, bfloat16_t);
9070		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmin_n_bf16_z)))
9071		svbfloat16_t svmin_z(svbool_t, svbfloat16_t, bfloat16_t);
9072		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmin_bf16_m)))
9073		svbfloat16_t svmin_m(svbool_t, svbfloat16_t, svbfloat16_t);
9074		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmin_bf16_x)))
9075		svbfloat16_t svmin_x(svbool_t, svbfloat16_t, svbfloat16_t);
9076		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmin_bf16_z)))
9077		svbfloat16_t svmin_z(svbool_t, svbfloat16_t, svbfloat16_t);
9078		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnm_n_bf16_m)))
9079		svbfloat16_t svminnm_m(svbool_t, svbfloat16_t, bfloat16_t);
9080		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnm_n_bf16_x)))
9081		svbfloat16_t svminnm_x(svbool_t, svbfloat16_t, bfloat16_t);
9082		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnm_n_bf16_z)))
9083		svbfloat16_t svminnm_z(svbool_t, svbfloat16_t, bfloat16_t);
9084		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnm_bf16_m)))
9085		svbfloat16_t svminnm_m(svbool_t, svbfloat16_t, svbfloat16_t);
9086		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnm_bf16_x)))
9087		svbfloat16_t svminnm_x(svbool_t, svbfloat16_t, svbfloat16_t);
9088		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnm_bf16_z)))
9089		svbfloat16_t svminnm_z(svbool_t, svbfloat16_t, svbfloat16_t);
9090		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmla_n_bf16_m)))
9091		svbfloat16_t svmla_m(svbool_t, svbfloat16_t, svbfloat16_t, bfloat16_t);
9092		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmla_n_bf16_x)))
9093		svbfloat16_t svmla_x(svbool_t, svbfloat16_t, svbfloat16_t, bfloat16_t);
9094		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmla_n_bf16_z)))
9095		svbfloat16_t svmla_z(svbool_t, svbfloat16_t, svbfloat16_t, bfloat16_t);
9096		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmla_bf16_m)))
9097		svbfloat16_t svmla_m(svbool_t, svbfloat16_t, svbfloat16_t, svbfloat16_t);
9098		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmla_bf16_x)))
9099		svbfloat16_t svmla_x(svbool_t, svbfloat16_t, svbfloat16_t, svbfloat16_t);
9100		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmla_bf16_z)))
9101		svbfloat16_t svmla_z(svbool_t, svbfloat16_t, svbfloat16_t, svbfloat16_t);
9102		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmla_lane_bf16)))
9103		svbfloat16_t svmla_lane(svbfloat16_t, svbfloat16_t, svbfloat16_t, uint64_t);
9104		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmls_n_bf16_m)))
9105		svbfloat16_t svmls_m(svbool_t, svbfloat16_t, svbfloat16_t, bfloat16_t);
9106		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmls_n_bf16_x)))
9107		svbfloat16_t svmls_x(svbool_t, svbfloat16_t, svbfloat16_t, bfloat16_t);
9108		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmls_n_bf16_z)))
9109		svbfloat16_t svmls_z(svbool_t, svbfloat16_t, svbfloat16_t, bfloat16_t);
9110		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmls_bf16_m)))
9111		svbfloat16_t svmls_m(svbool_t, svbfloat16_t, svbfloat16_t, svbfloat16_t);
9112		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmls_bf16_x)))
9113		svbfloat16_t svmls_x(svbool_t, svbfloat16_t, svbfloat16_t, svbfloat16_t);
9114		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmls_bf16_z)))
9115		svbfloat16_t svmls_z(svbool_t, svbfloat16_t, svbfloat16_t, svbfloat16_t);
9116		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmls_lane_bf16)))
9117		svbfloat16_t svmls_lane(svbfloat16_t, svbfloat16_t, svbfloat16_t, uint64_t);
9118		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmul_n_bf16_m)))
9119		svbfloat16_t svmul_m(svbool_t, svbfloat16_t, bfloat16_t);
9120		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmul_n_bf16_x)))
9121		svbfloat16_t svmul_x(svbool_t, svbfloat16_t, bfloat16_t);
9122		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmul_n_bf16_z)))
9123		svbfloat16_t svmul_z(svbool_t, svbfloat16_t, bfloat16_t);
9124		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmul_bf16_m)))
9125		svbfloat16_t svmul_m(svbool_t, svbfloat16_t, svbfloat16_t);
9126		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmul_bf16_x)))
9127		svbfloat16_t svmul_x(svbool_t, svbfloat16_t, svbfloat16_t);
9128		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmul_bf16_z)))
9129		svbfloat16_t svmul_z(svbool_t, svbfloat16_t, svbfloat16_t);
9130		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmul_lane_bf16)))
9131		svbfloat16_t svmul_lane(svbfloat16_t, svbfloat16_t, uint64_t);
9132		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsub_n_bf16_m)))
9133		svbfloat16_t svsub_m(svbool_t, svbfloat16_t, bfloat16_t);
9134		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsub_n_bf16_x)))
9135		svbfloat16_t svsub_x(svbool_t, svbfloat16_t, bfloat16_t);
9136		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsub_n_bf16_z)))
9137		svbfloat16_t svsub_z(svbool_t, svbfloat16_t, bfloat16_t);
9138		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsub_bf16_m)))
9139		svbfloat16_t svsub_m(svbool_t, svbfloat16_t, svbfloat16_t);
9140		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsub_bf16_x)))
9141		svbfloat16_t svsub_x(svbool_t, svbfloat16_t, svbfloat16_t);
9142		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsub_bf16_z)))
9143		svbfloat16_t svsub_z(svbool_t, svbfloat16_t, svbfloat16_t);
9144		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_n_u8)))
9145		svuint8_t svbdep_n_u8(svuint8_t, uint8_t);
9146		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_n_u32)))
9147		svuint32_t svbdep_n_u32(svuint32_t, uint32_t);
9148		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_n_u64)))
9149		svuint64_t svbdep_n_u64(svuint64_t, uint64_t);
9150		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_n_u16)))
9151		svuint16_t svbdep_n_u16(svuint16_t, uint16_t);
9152		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_u8)))
9153		svuint8_t svbdep_u8(svuint8_t, svuint8_t);
9154		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_u32)))
9155		svuint32_t svbdep_u32(svuint32_t, svuint32_t);
9156		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_u64)))
9157		svuint64_t svbdep_u64(svuint64_t, svuint64_t);
9158		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_u16)))
9159		svuint16_t svbdep_u16(svuint16_t, svuint16_t);
9160		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_n_u8)))
9161		svuint8_t svbext_n_u8(svuint8_t, uint8_t);
9162		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_n_u32)))
9163		svuint32_t svbext_n_u32(svuint32_t, uint32_t);
9164		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_n_u64)))
9165		svuint64_t svbext_n_u64(svuint64_t, uint64_t);
9166		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_n_u16)))
9167		svuint16_t svbext_n_u16(svuint16_t, uint16_t);
9168		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_u8)))
9169		svuint8_t svbext_u8(svuint8_t, svuint8_t);
9170		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_u32)))
9171		svuint32_t svbext_u32(svuint32_t, svuint32_t);
9172		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_u64)))
9173		svuint64_t svbext_u64(svuint64_t, svuint64_t);
9174		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_u16)))
9175		svuint16_t svbext_u16(svuint16_t, svuint16_t);
9176		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_n_u8)))
9177		svuint8_t svbgrp_n_u8(svuint8_t, uint8_t);
9178		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_n_u32)))
9179		svuint32_t svbgrp_n_u32(svuint32_t, uint32_t);
9180		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_n_u64)))
9181		svuint64_t svbgrp_n_u64(svuint64_t, uint64_t);
9182		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_n_u16)))
9183		svuint16_t svbgrp_n_u16(svuint16_t, uint16_t);
9184		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_u8)))
9185		svuint8_t svbgrp_u8(svuint8_t, svuint8_t);
9186		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_u32)))
9187		svuint32_t svbgrp_u32(svuint32_t, svuint32_t);
9188		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_u64)))
9189		svuint64_t svbgrp_u64(svuint64_t, svuint64_t);
9190		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_u16)))
9191		svuint16_t svbgrp_u16(svuint16_t, svuint16_t);
9192		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_n_u8)))
9193		svuint8_t svbdep(svuint8_t, uint8_t);
9194		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_n_u32)))
9195		svuint32_t svbdep(svuint32_t, uint32_t);
9196		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_n_u64)))
9197		svuint64_t svbdep(svuint64_t, uint64_t);
9198		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_n_u16)))
9199		svuint16_t svbdep(svuint16_t, uint16_t);
9200		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_u8)))
9201		svuint8_t svbdep(svuint8_t, svuint8_t);
9202		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_u32)))
9203		svuint32_t svbdep(svuint32_t, svuint32_t);
9204		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_u64)))
9205		svuint64_t svbdep(svuint64_t, svuint64_t);
9206		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_u16)))
9207		svuint16_t svbdep(svuint16_t, svuint16_t);
9208		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_n_u8)))
9209		svuint8_t svbext(svuint8_t, uint8_t);
9210		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_n_u32)))
9211		svuint32_t svbext(svuint32_t, uint32_t);
9212		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_n_u64)))
9213		svuint64_t svbext(svuint64_t, uint64_t);
9214		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_n_u16)))
9215		svuint16_t svbext(svuint16_t, uint16_t);
9216		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_u8)))
9217		svuint8_t svbext(svuint8_t, svuint8_t);
9218		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_u32)))
9219		svuint32_t svbext(svuint32_t, svuint32_t);
9220		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_u64)))
9221		svuint64_t svbext(svuint64_t, svuint64_t);
9222		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_u16)))
9223		svuint16_t svbext(svuint16_t, svuint16_t);
9224		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_n_u8)))
9225		svuint8_t svbgrp(svuint8_t, uint8_t);
9226		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_n_u32)))
9227		svuint32_t svbgrp(svuint32_t, uint32_t);
9228		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_n_u64)))
9229		svuint64_t svbgrp(svuint64_t, uint64_t);
9230		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_n_u16)))
9231		svuint16_t svbgrp(svuint16_t, uint16_t);
9232		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_u8)))
9233		svuint8_t svbgrp(svuint8_t, svuint8_t);
9234		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_u32)))
9235		svuint32_t svbgrp(svuint32_t, svuint32_t);
9236		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_u64)))
9237		svuint64_t svbgrp(svuint64_t, svuint64_t);
9238		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_u16)))
9239		svuint16_t svbgrp(svuint16_t, svuint16_t);
9240		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrax1_u64)))
9241		svuint64_t svrax1_u64(svuint64_t, svuint64_t);
9242		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrax1_s64)))
9243		svint64_t svrax1_s64(svint64_t, svint64_t);
9244		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrax1_u64)))
9245		svuint64_t svrax1(svuint64_t, svuint64_t);
9246		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrax1_s64)))
9247		svint64_t svrax1(svint64_t, svint64_t);
9248		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsm4e_u32)))
9249		svuint32_t svsm4e_u32(svuint32_t, svuint32_t);
9250		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsm4ekey_u32)))
9251		svuint32_t svsm4ekey_u32(svuint32_t, svuint32_t);
9252		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsm4e_u32)))
9253		svuint32_t svsm4e(svuint32_t, svuint32_t);
9254		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsm4ekey_u32)))
9255		svuint32_t svsm4ekey(svuint32_t, svuint32_t);
9256		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_u8)))
9257		uint8x16_t svaddqv_u8(svbool_t, svuint8_t);
9258		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_u32)))
9259		uint32x4_t svaddqv_u32(svbool_t, svuint32_t);
9260		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_u64)))
9261		uint64x2_t svaddqv_u64(svbool_t, svuint64_t);
9262		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_u16)))
9263		uint16x8_t svaddqv_u16(svbool_t, svuint16_t);
9264		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_s8)))
9265		int8x16_t svaddqv_s8(svbool_t, svint8_t);
9266		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_s32)))
9267		int32x4_t svaddqv_s32(svbool_t, svint32_t);
9268		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_s64)))
9269		int64x2_t svaddqv_s64(svbool_t, svint64_t);
9270		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_s16)))
9271		int16x8_t svaddqv_s16(svbool_t, svint16_t);
9272		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_f64)))
9273		float64x2_t svaddqv_f64(svbool_t, svfloat64_t);
9274		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_f32)))
9275		float32x4_t svaddqv_f32(svbool_t, svfloat32_t);
9276		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_f16)))
9277		float16x8_t svaddqv_f16(svbool_t, svfloat16_t);
9278		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_u8)))
9279		uint8x16_t svandqv_u8(svbool_t, svuint8_t);
9280		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_u32)))
9281		uint32x4_t svandqv_u32(svbool_t, svuint32_t);
9282		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_u64)))
9283		uint64x2_t svandqv_u64(svbool_t, svuint64_t);
9284		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_u16)))
9285		uint16x8_t svandqv_u16(svbool_t, svuint16_t);
9286		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_s8)))
9287		int8x16_t svandqv_s8(svbool_t, svint8_t);
9288		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_s32)))
9289		int32x4_t svandqv_s32(svbool_t, svint32_t);
9290		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_s64)))
9291		int64x2_t svandqv_s64(svbool_t, svint64_t);
9292		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_s16)))
9293		int16x8_t svandqv_s16(svbool_t, svint16_t);
9294		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_u8)))
9295		uint8x16_t sveorqv_u8(svbool_t, svuint8_t);
9296		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_u32)))
9297		uint32x4_t sveorqv_u32(svbool_t, svuint32_t);
9298		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_u64)))
9299		uint64x2_t sveorqv_u64(svbool_t, svuint64_t);
9300		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_u16)))
9301		uint16x8_t sveorqv_u16(svbool_t, svuint16_t);
9302		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_s8)))
9303		int8x16_t sveorqv_s8(svbool_t, svint8_t);
9304		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_s32)))
9305		int32x4_t sveorqv_s32(svbool_t, svint32_t);
9306		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_s64)))
9307		int64x2_t sveorqv_s64(svbool_t, svint64_t);
9308		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_s16)))
9309		int16x8_t sveorqv_s16(svbool_t, svint16_t);
9310		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_u8)))
9311		svuint8_t svextq_u8(svuint8_t, svuint8_t, int32_t);
9312		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_u32)))
9313		svuint32_t svextq_u32(svuint32_t, svuint32_t, int32_t);
9314		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_u64)))
9315		svuint64_t svextq_u64(svuint64_t, svuint64_t, int32_t);
9316		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_u16)))
9317		svuint16_t svextq_u16(svuint16_t, svuint16_t, int32_t);
9318		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_bf16)))
9319		svbfloat16_t svextq_bf16(svbfloat16_t, svbfloat16_t, int32_t);
9320		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_s8)))
9321		svint8_t svextq_s8(svint8_t, svint8_t, int32_t);
9322		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_f64)))
9323		svfloat64_t svextq_f64(svfloat64_t, svfloat64_t, int32_t);
9324		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_f32)))
9325		svfloat32_t svextq_f32(svfloat32_t, svfloat32_t, int32_t);
9326		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_f16)))
9327		svfloat16_t svextq_f16(svfloat16_t, svfloat16_t, int32_t);
9328		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_s32)))
9329		svint32_t svextq_s32(svint32_t, svint32_t, int32_t);
9330		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_s64)))
9331		svint64_t svextq_s64(svint64_t, svint64_t, int32_t);
9332		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_mf8)))
9333		svmfloat8_t svextq_mf8(svmfloat8_t, svmfloat8_t, int32_t);
9334		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_s16)))
9335		svint16_t svextq_s16(svint16_t, svint16_t, int32_t);
9336		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_u32)))
9337		svuint32_t svld1q_gather_u64base_index_u32(svbool_t, svuint64_t, int64_t);
9338		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_u64)))
9339		svuint64_t svld1q_gather_u64base_index_u64(svbool_t, svuint64_t, int64_t);
9340		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_u16)))
9341		svuint16_t svld1q_gather_u64base_index_u16(svbool_t, svuint64_t, int64_t);
9342		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_bf16)))
9343		svbfloat16_t svld1q_gather_u64base_index_bf16(svbool_t, svuint64_t, int64_t);
9344		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_f64)))
9345		svfloat64_t svld1q_gather_u64base_index_f64(svbool_t, svuint64_t, int64_t);
9346		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_f32)))
9347		svfloat32_t svld1q_gather_u64base_index_f32(svbool_t, svuint64_t, int64_t);
9348		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_f16)))
9349		svfloat16_t svld1q_gather_u64base_index_f16(svbool_t, svuint64_t, int64_t);
9350		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_s32)))
9351		svint32_t svld1q_gather_u64base_index_s32(svbool_t, svuint64_t, int64_t);
9352		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_s64)))
9353		svint64_t svld1q_gather_u64base_index_s64(svbool_t, svuint64_t, int64_t);
9354		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_s16)))
9355		svint16_t svld1q_gather_u64base_index_s16(svbool_t, svuint64_t, int64_t);
9356		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_u8)))
9357		svuint8_t svld1q_gather_u64base_offset_u8(svbool_t, svuint64_t, int64_t);
9358		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_u32)))
9359		svuint32_t svld1q_gather_u64base_offset_u32(svbool_t, svuint64_t, int64_t);
9360		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_u64)))
9361		svuint64_t svld1q_gather_u64base_offset_u64(svbool_t, svuint64_t, int64_t);
9362		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_u16)))
9363		svuint16_t svld1q_gather_u64base_offset_u16(svbool_t, svuint64_t, int64_t);
9364		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_bf16)))
9365		svbfloat16_t svld1q_gather_u64base_offset_bf16(svbool_t, svuint64_t, int64_t);
9366		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_s8)))
9367		svint8_t svld1q_gather_u64base_offset_s8(svbool_t, svuint64_t, int64_t);
9368		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_f64)))
9369		svfloat64_t svld1q_gather_u64base_offset_f64(svbool_t, svuint64_t, int64_t);
9370		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_f32)))
9371		svfloat32_t svld1q_gather_u64base_offset_f32(svbool_t, svuint64_t, int64_t);
9372		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_f16)))
9373		svfloat16_t svld1q_gather_u64base_offset_f16(svbool_t, svuint64_t, int64_t);
9374		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_s32)))
9375		svint32_t svld1q_gather_u64base_offset_s32(svbool_t, svuint64_t, int64_t);
9376		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_s64)))
9377		svint64_t svld1q_gather_u64base_offset_s64(svbool_t, svuint64_t, int64_t);
9378		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_s16)))
9379		svint16_t svld1q_gather_u64base_offset_s16(svbool_t, svuint64_t, int64_t);
9380		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_u8)))
9381		svuint8_t svld1q_gather_u64base_u8(svbool_t, svuint64_t);
9382		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_u32)))
9383		svuint32_t svld1q_gather_u64base_u32(svbool_t, svuint64_t);
9384		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_u64)))
9385		svuint64_t svld1q_gather_u64base_u64(svbool_t, svuint64_t);
9386		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_u16)))
9387		svuint16_t svld1q_gather_u64base_u16(svbool_t, svuint64_t);
9388		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_bf16)))
9389		svbfloat16_t svld1q_gather_u64base_bf16(svbool_t, svuint64_t);
9390		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_s8)))
9391		svint8_t svld1q_gather_u64base_s8(svbool_t, svuint64_t);
9392		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_f64)))
9393		svfloat64_t svld1q_gather_u64base_f64(svbool_t, svuint64_t);
9394		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_f32)))
9395		svfloat32_t svld1q_gather_u64base_f32(svbool_t, svuint64_t);
9396		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_f16)))
9397		svfloat16_t svld1q_gather_u64base_f16(svbool_t, svuint64_t);
9398		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_s32)))
9399		svint32_t svld1q_gather_u64base_s32(svbool_t, svuint64_t);
9400		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_s64)))
9401		svint64_t svld1q_gather_u64base_s64(svbool_t, svuint64_t);
9402		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_s16)))
9403		svint16_t svld1q_gather_u64base_s16(svbool_t, svuint64_t);
9404		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_u32)))
9405		svuint32_t svld1q_gather_u64index_u32(svbool_t, uint32_t const *, svuint64_t);
9406		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_u64)))
9407		svuint64_t svld1q_gather_u64index_u64(svbool_t, uint64_t const *, svuint64_t);
9408		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_u16)))
9409		svuint16_t svld1q_gather_u64index_u16(svbool_t, uint16_t const *, svuint64_t);
9410		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_bf16)))
9411		svbfloat16_t svld1q_gather_u64index_bf16(svbool_t, bfloat16_t const *, svuint64_t);
9412		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_f64)))
9413		svfloat64_t svld1q_gather_u64index_f64(svbool_t, float64_t const *, svuint64_t);
9414		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_f32)))
9415		svfloat32_t svld1q_gather_u64index_f32(svbool_t, float32_t const *, svuint64_t);
9416		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_f16)))
9417		svfloat16_t svld1q_gather_u64index_f16(svbool_t, float16_t const *, svuint64_t);
9418		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_s32)))
9419		svint32_t svld1q_gather_u64index_s32(svbool_t, int32_t const *, svuint64_t);
9420		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_s64)))
9421		svint64_t svld1q_gather_u64index_s64(svbool_t, int64_t const *, svuint64_t);
9422		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_s16)))
9423		svint16_t svld1q_gather_u64index_s16(svbool_t, int16_t const *, svuint64_t);
9424		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_u8)))
9425		svuint8_t svld1q_gather_u64offset_u8(svbool_t, uint8_t const *, svuint64_t);
9426		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_u32)))
9427		svuint32_t svld1q_gather_u64offset_u32(svbool_t, uint32_t const *, svuint64_t);
9428		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_u64)))
9429		svuint64_t svld1q_gather_u64offset_u64(svbool_t, uint64_t const *, svuint64_t);
9430		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_u16)))
9431		svuint16_t svld1q_gather_u64offset_u16(svbool_t, uint16_t const *, svuint64_t);
9432		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_bf16)))
9433		svbfloat16_t svld1q_gather_u64offset_bf16(svbool_t, bfloat16_t const *, svuint64_t);
9434		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_s8)))
9435		svint8_t svld1q_gather_u64offset_s8(svbool_t, int8_t const *, svuint64_t);
9436		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_f64)))
9437		svfloat64_t svld1q_gather_u64offset_f64(svbool_t, float64_t const *, svuint64_t);
9438		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_f32)))
9439		svfloat32_t svld1q_gather_u64offset_f32(svbool_t, float32_t const *, svuint64_t);
9440		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_f16)))
9441		svfloat16_t svld1q_gather_u64offset_f16(svbool_t, float16_t const *, svuint64_t);
9442		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_s32)))
9443		svint32_t svld1q_gather_u64offset_s32(svbool_t, int32_t const *, svuint64_t);
9444		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_s64)))
9445		svint64_t svld1q_gather_u64offset_s64(svbool_t, int64_t const *, svuint64_t);
9446		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_s16)))
9447		svint16_t svld1q_gather_u64offset_s16(svbool_t, int16_t const *, svuint64_t);
9448		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1udq_u64)))
9449		svuint64_t svld1udq_u64(svbool_t, uint64_t const *);
9450		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1udq_f64)))
9451		svfloat64_t svld1udq_f64(svbool_t, float64_t const *);
9452		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1udq_s64)))
9453		svint64_t svld1udq_s64(svbool_t, int64_t const *);
9454		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1udq_vnum_u64)))
9455		svuint64_t svld1udq_vnum_u64(svbool_t, uint64_t const *, int64_t);
9456		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1udq_vnum_f64)))
9457		svfloat64_t svld1udq_vnum_f64(svbool_t, float64_t const *, int64_t);
9458		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1udq_vnum_s64)))
9459		svint64_t svld1udq_vnum_s64(svbool_t, int64_t const *, int64_t);
9460		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1uwq_u32)))
9461		svuint32_t svld1uwq_u32(svbool_t, uint32_t const *);
9462		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1uwq_f32)))
9463		svfloat32_t svld1uwq_f32(svbool_t, float32_t const *);
9464		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1uwq_s32)))
9465		svint32_t svld1uwq_s32(svbool_t, int32_t const *);
9466		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1uwq_vnum_u32)))
9467		svuint32_t svld1uwq_vnum_u32(svbool_t, uint32_t const *, int64_t);
9468		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1uwq_vnum_f32)))
9469		svfloat32_t svld1uwq_vnum_f32(svbool_t, float32_t const *, int64_t);
9470		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1uwq_vnum_s32)))
9471		svint32_t svld1uwq_vnum_s32(svbool_t, int32_t const *, int64_t);
9472		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_u8)))
9473		svuint8x2_t svld2q_u8(svbool_t, uint8_t const *);
9474		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_u32)))
9475		svuint32x2_t svld2q_u32(svbool_t, uint32_t const *);
9476		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_u64)))
9477		svuint64x2_t svld2q_u64(svbool_t, uint64_t const *);
9478		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_u16)))
9479		svuint16x2_t svld2q_u16(svbool_t, uint16_t const *);
9480		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_s8)))
9481		svint8x2_t svld2q_s8(svbool_t, int8_t const *);
9482		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_f64)))
9483		svfloat64x2_t svld2q_f64(svbool_t, float64_t const *);
9484		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_f32)))
9485		svfloat32x2_t svld2q_f32(svbool_t, float32_t const *);
9486		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_f16)))
9487		svfloat16x2_t svld2q_f16(svbool_t, float16_t const *);
9488		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_s32)))
9489		svint32x2_t svld2q_s32(svbool_t, int32_t const *);
9490		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_s64)))
9491		svint64x2_t svld2q_s64(svbool_t, int64_t const *);
9492		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_s16)))
9493		svint16x2_t svld2q_s16(svbool_t, int16_t const *);
9494		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_bf16)))
9495		svbfloat16x2_t svld2q_bf16(svbool_t, bfloat16_t const *);
9496		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_u8)))
9497		svuint8x2_t svld2q_vnum_u8(svbool_t, uint8_t const *, int64_t);
9498		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_u32)))
9499		svuint32x2_t svld2q_vnum_u32(svbool_t, uint32_t const *, int64_t);
9500		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_u64)))
9501		svuint64x2_t svld2q_vnum_u64(svbool_t, uint64_t const *, int64_t);
9502		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_u16)))
9503		svuint16x2_t svld2q_vnum_u16(svbool_t, uint16_t const *, int64_t);
9504		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_s8)))
9505		svint8x2_t svld2q_vnum_s8(svbool_t, int8_t const *, int64_t);
9506		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_f64)))
9507		svfloat64x2_t svld2q_vnum_f64(svbool_t, float64_t const *, int64_t);
9508		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_f32)))
9509		svfloat32x2_t svld2q_vnum_f32(svbool_t, float32_t const *, int64_t);
9510		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_f16)))
9511		svfloat16x2_t svld2q_vnum_f16(svbool_t, float16_t const *, int64_t);
9512		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_s32)))
9513		svint32x2_t svld2q_vnum_s32(svbool_t, int32_t const *, int64_t);
9514		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_s64)))
9515		svint64x2_t svld2q_vnum_s64(svbool_t, int64_t const *, int64_t);
9516		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_s16)))
9517		svint16x2_t svld2q_vnum_s16(svbool_t, int16_t const *, int64_t);
9518		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_bf16)))
9519		svbfloat16x2_t svld2q_vnum_bf16(svbool_t, bfloat16_t const *, int64_t);
9520		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_u8)))
9521		svuint8x3_t svld3q_u8(svbool_t, uint8_t const *);
9522		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_u32)))
9523		svuint32x3_t svld3q_u32(svbool_t, uint32_t const *);
9524		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_u64)))
9525		svuint64x3_t svld3q_u64(svbool_t, uint64_t const *);
9526		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_u16)))
9527		svuint16x3_t svld3q_u16(svbool_t, uint16_t const *);
9528		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_s8)))
9529		svint8x3_t svld3q_s8(svbool_t, int8_t const *);
9530		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_f64)))
9531		svfloat64x3_t svld3q_f64(svbool_t, float64_t const *);
9532		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_f32)))
9533		svfloat32x3_t svld3q_f32(svbool_t, float32_t const *);
9534		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_f16)))
9535		svfloat16x3_t svld3q_f16(svbool_t, float16_t const *);
9536		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_s32)))
9537		svint32x3_t svld3q_s32(svbool_t, int32_t const *);
9538		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_s64)))
9539		svint64x3_t svld3q_s64(svbool_t, int64_t const *);
9540		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_s16)))
9541		svint16x3_t svld3q_s16(svbool_t, int16_t const *);
9542		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_bf16)))
9543		svbfloat16x3_t svld3q_bf16(svbool_t, bfloat16_t const *);
9544		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_u8)))
9545		svuint8x3_t svld3q_vnum_u8(svbool_t, uint8_t const *, int64_t);
9546		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_u32)))
9547		svuint32x3_t svld3q_vnum_u32(svbool_t, uint32_t const *, int64_t);
9548		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_u64)))
9549		svuint64x3_t svld3q_vnum_u64(svbool_t, uint64_t const *, int64_t);
9550		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_u16)))
9551		svuint16x3_t svld3q_vnum_u16(svbool_t, uint16_t const *, int64_t);
9552		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_s8)))
9553		svint8x3_t svld3q_vnum_s8(svbool_t, int8_t const *, int64_t);
9554		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_f64)))
9555		svfloat64x3_t svld3q_vnum_f64(svbool_t, float64_t const *, int64_t);
9556		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_f32)))
9557		svfloat32x3_t svld3q_vnum_f32(svbool_t, float32_t const *, int64_t);
9558		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_f16)))
9559		svfloat16x3_t svld3q_vnum_f16(svbool_t, float16_t const *, int64_t);
9560		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_s32)))
9561		svint32x3_t svld3q_vnum_s32(svbool_t, int32_t const *, int64_t);
9562		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_s64)))
9563		svint64x3_t svld3q_vnum_s64(svbool_t, int64_t const *, int64_t);
9564		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_s16)))
9565		svint16x3_t svld3q_vnum_s16(svbool_t, int16_t const *, int64_t);
9566		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_bf16)))
9567		svbfloat16x3_t svld3q_vnum_bf16(svbool_t, bfloat16_t const *, int64_t);
9568		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_u8)))
9569		svuint8x4_t svld4q_u8(svbool_t, uint8_t const *);
9570		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_u32)))
9571		svuint32x4_t svld4q_u32(svbool_t, uint32_t const *);
9572		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_u64)))
9573		svuint64x4_t svld4q_u64(svbool_t, uint64_t const *);
9574		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_u16)))
9575		svuint16x4_t svld4q_u16(svbool_t, uint16_t const *);
9576		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_s8)))
9577		svint8x4_t svld4q_s8(svbool_t, int8_t const *);
9578		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_f64)))
9579		svfloat64x4_t svld4q_f64(svbool_t, float64_t const *);
9580		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_f32)))
9581		svfloat32x4_t svld4q_f32(svbool_t, float32_t const *);
9582		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_f16)))
9583		svfloat16x4_t svld4q_f16(svbool_t, float16_t const *);
9584		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_s32)))
9585		svint32x4_t svld4q_s32(svbool_t, int32_t const *);
9586		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_s64)))
9587		svint64x4_t svld4q_s64(svbool_t, int64_t const *);
9588		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_s16)))
9589		svint16x4_t svld4q_s16(svbool_t, int16_t const *);
9590		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_bf16)))
9591		svbfloat16x4_t svld4q_bf16(svbool_t, bfloat16_t const *);
9592		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_u8)))
9593		svuint8x4_t svld4q_vnum_u8(svbool_t, uint8_t const *, int64_t);
9594		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_u32)))
9595		svuint32x4_t svld4q_vnum_u32(svbool_t, uint32_t const *, int64_t);
9596		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_u64)))
9597		svuint64x4_t svld4q_vnum_u64(svbool_t, uint64_t const *, int64_t);
9598		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_u16)))
9599		svuint16x4_t svld4q_vnum_u16(svbool_t, uint16_t const *, int64_t);
9600		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_s8)))
9601		svint8x4_t svld4q_vnum_s8(svbool_t, int8_t const *, int64_t);
9602		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_f64)))
9603		svfloat64x4_t svld4q_vnum_f64(svbool_t, float64_t const *, int64_t);
9604		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_f32)))
9605		svfloat32x4_t svld4q_vnum_f32(svbool_t, float32_t const *, int64_t);
9606		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_f16)))
9607		svfloat16x4_t svld4q_vnum_f16(svbool_t, float16_t const *, int64_t);
9608		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_s32)))
9609		svint32x4_t svld4q_vnum_s32(svbool_t, int32_t const *, int64_t);
9610		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_s64)))
9611		svint64x4_t svld4q_vnum_s64(svbool_t, int64_t const *, int64_t);
9612		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_s16)))
9613		svint16x4_t svld4q_vnum_s16(svbool_t, int16_t const *, int64_t);
9614		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_bf16)))
9615		svbfloat16x4_t svld4q_vnum_bf16(svbool_t, bfloat16_t const *, int64_t);
9616		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnmqv_f64)))
9617		float64x2_t svmaxnmqv_f64(svbool_t, svfloat64_t);
9618		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnmqv_f32)))
9619		float32x4_t svmaxnmqv_f32(svbool_t, svfloat32_t);
9620		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnmqv_f16)))
9621		float16x8_t svmaxnmqv_f16(svbool_t, svfloat16_t);
9622		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_f64)))
9623		float64x2_t svmaxqv_f64(svbool_t, svfloat64_t);
9624		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_f32)))
9625		float32x4_t svmaxqv_f32(svbool_t, svfloat32_t);
9626		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_f16)))
9627		float16x8_t svmaxqv_f16(svbool_t, svfloat16_t);
9628		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_s8)))
9629		int8x16_t svmaxqv_s8(svbool_t, svint8_t);
9630		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_s32)))
9631		int32x4_t svmaxqv_s32(svbool_t, svint32_t);
9632		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_s64)))
9633		int64x2_t svmaxqv_s64(svbool_t, svint64_t);
9634		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_s16)))
9635		int16x8_t svmaxqv_s16(svbool_t, svint16_t);
9636		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_u8)))
9637		uint8x16_t svmaxqv_u8(svbool_t, svuint8_t);
9638		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_u32)))
9639		uint32x4_t svmaxqv_u32(svbool_t, svuint32_t);
9640		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_u64)))
9641		uint64x2_t svmaxqv_u64(svbool_t, svuint64_t);
9642		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_u16)))
9643		uint16x8_t svmaxqv_u16(svbool_t, svuint16_t);
9644		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnmqv_f64)))
9645		float64x2_t svminnmqv_f64(svbool_t, svfloat64_t);
9646		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnmqv_f32)))
9647		float32x4_t svminnmqv_f32(svbool_t, svfloat32_t);
9648		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnmqv_f16)))
9649		float16x8_t svminnmqv_f16(svbool_t, svfloat16_t);
9650		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_f64)))
9651		float64x2_t svminqv_f64(svbool_t, svfloat64_t);
9652		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_f32)))
9653		float32x4_t svminqv_f32(svbool_t, svfloat32_t);
9654		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_f16)))
9655		float16x8_t svminqv_f16(svbool_t, svfloat16_t);
9656		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_s8)))
9657		int8x16_t svminqv_s8(svbool_t, svint8_t);
9658		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_s32)))
9659		int32x4_t svminqv_s32(svbool_t, svint32_t);
9660		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_s64)))
9661		int64x2_t svminqv_s64(svbool_t, svint64_t);
9662		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_s16)))
9663		int16x8_t svminqv_s16(svbool_t, svint16_t);
9664		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_u8)))
9665		uint8x16_t svminqv_u8(svbool_t, svuint8_t);
9666		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_u32)))
9667		uint32x4_t svminqv_u32(svbool_t, svuint32_t);
9668		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_u64)))
9669		uint64x2_t svminqv_u64(svbool_t, svuint64_t);
9670		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_u16)))
9671		uint16x8_t svminqv_u16(svbool_t, svuint16_t);
9672		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_u8)))
9673		uint8x16_t svorqv_u8(svbool_t, svuint8_t);
9674		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_u32)))
9675		uint32x4_t svorqv_u32(svbool_t, svuint32_t);
9676		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_u64)))
9677		uint64x2_t svorqv_u64(svbool_t, svuint64_t);
9678		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_u16)))
9679		uint16x8_t svorqv_u16(svbool_t, svuint16_t);
9680		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_s8)))
9681		int8x16_t svorqv_s8(svbool_t, svint8_t);
9682		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_s32)))
9683		int32x4_t svorqv_s32(svbool_t, svint32_t);
9684		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_s64)))
9685		int64x2_t svorqv_s64(svbool_t, svint64_t);
9686		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_s16)))
9687		int16x8_t svorqv_s16(svbool_t, svint16_t);
9688		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_u8)))
9689		svbool_t svpmov_u8(svuint8_t);
9690		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_s8)))
9691		svbool_t svpmov_s8(svint8_t);
9692		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_u64)))
9693		svbool_t svpmov_u64(svuint64_t);
9694		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_s64)))
9695		svbool_t svpmov_s64(svint64_t);
9696		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_u16)))
9697		svbool_t svpmov_u16(svuint16_t);
9698		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_s16)))
9699		svbool_t svpmov_s16(svint16_t);
9700		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_u32)))
9701		svbool_t svpmov_u32(svuint32_t);
9702		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_s32)))
9703		svbool_t svpmov_s32(svint32_t);
9704		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_u8)))
9705		svbool_t svpmov_lane_u8(svuint8_t, uint64_t);
9706		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_s8)))
9707		svbool_t svpmov_lane_s8(svint8_t, uint64_t);
9708		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_u64)))
9709		svbool_t svpmov_lane_u64(svuint64_t, uint64_t);
9710		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_s64)))
9711		svbool_t svpmov_lane_s64(svint64_t, uint64_t);
9712		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_u16)))
9713		svbool_t svpmov_lane_u16(svuint16_t, uint64_t);
9714		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_s16)))
9715		svbool_t svpmov_lane_s16(svint16_t, uint64_t);
9716		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_u32)))
9717		svbool_t svpmov_lane_u32(svuint32_t, uint64_t);
9718		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_s32)))
9719		svbool_t svpmov_lane_s32(svint32_t, uint64_t);
9720		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_u64_m)))
9721		svuint64_t svpmov_lane_u64_m(svuint64_t, svbool_t, uint64_t);
9722		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_s64_m)))
9723		svint64_t svpmov_lane_s64_m(svint64_t, svbool_t, uint64_t);
9724		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_u16_m)))
9725		svuint16_t svpmov_lane_u16_m(svuint16_t, svbool_t, uint64_t);
9726		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_s16_m)))
9727		svint16_t svpmov_lane_s16_m(svint16_t, svbool_t, uint64_t);
9728		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_u32_m)))
9729		svuint32_t svpmov_lane_u32_m(svuint32_t, svbool_t, uint64_t);
9730		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_s32_m)))
9731		svint32_t svpmov_lane_s32_m(svint32_t, svbool_t, uint64_t);
9732		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_u8_z)))
9733		svuint8_t svpmov_u8_z(svbool_t);
9734		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_s8_z)))
9735		svint8_t svpmov_s8_z(svbool_t);
9736		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_u64_z)))
9737		svuint64_t svpmov_u64_z(svbool_t);
9738		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_s64_z)))
9739		svint64_t svpmov_s64_z(svbool_t);
9740		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_u16_z)))
9741		svuint16_t svpmov_u16_z(svbool_t);
9742		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_s16_z)))
9743		svint16_t svpmov_s16_z(svbool_t);
9744		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_u32_z)))
9745		svuint32_t svpmov_u32_z(svbool_t);
9746		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_s32_z)))
9747		svint32_t svpmov_s32_z(svbool_t);
9748		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1dq_u64)))
9749		void svst1dq_u64(svbool_t, uint64_t *, svuint64_t);
9750		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1dq_f64)))
9751		void svst1dq_f64(svbool_t, float64_t *, svfloat64_t);
9752		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1dq_s64)))
9753		void svst1dq_s64(svbool_t, int64_t *, svint64_t);
9754		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1dq_vnum_u64)))
9755		void svst1dq_vnum_u64(svbool_t, uint64_t *, int64_t, svuint64_t);
9756		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1dq_vnum_f64)))
9757		void svst1dq_vnum_f64(svbool_t, float64_t *, int64_t, svfloat64_t);
9758		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1dq_vnum_s64)))
9759		void svst1dq_vnum_s64(svbool_t, int64_t *, int64_t, svint64_t);
9760		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_u8)))
9761		void svst1q_scatter_u64base_u8(svbool_t, svuint64_t, svuint8_t);
9762		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_u32)))
9763		void svst1q_scatter_u64base_u32(svbool_t, svuint64_t, svuint32_t);
9764		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_u64)))
9765		void svst1q_scatter_u64base_u64(svbool_t, svuint64_t, svuint64_t);
9766		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_u16)))
9767		void svst1q_scatter_u64base_u16(svbool_t, svuint64_t, svuint16_t);
9768		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_bf16)))
9769		void svst1q_scatter_u64base_bf16(svbool_t, svuint64_t, svbfloat16_t);
9770		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_s8)))
9771		void svst1q_scatter_u64base_s8(svbool_t, svuint64_t, svint8_t);
9772		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_f64)))
9773		void svst1q_scatter_u64base_f64(svbool_t, svuint64_t, svfloat64_t);
9774		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_f32)))
9775		void svst1q_scatter_u64base_f32(svbool_t, svuint64_t, svfloat32_t);
9776		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_f16)))
9777		void svst1q_scatter_u64base_f16(svbool_t, svuint64_t, svfloat16_t);
9778		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_s32)))
9779		void svst1q_scatter_u64base_s32(svbool_t, svuint64_t, svint32_t);
9780		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_s64)))
9781		void svst1q_scatter_u64base_s64(svbool_t, svuint64_t, svint64_t);
9782		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_s16)))
9783		void svst1q_scatter_u64base_s16(svbool_t, svuint64_t, svint16_t);
9784		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_u32)))
9785		void svst1q_scatter_u64base_index_u32(svbool_t, svuint64_t, int64_t, svuint32_t);
9786		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_u64)))
9787		void svst1q_scatter_u64base_index_u64(svbool_t, svuint64_t, int64_t, svuint64_t);
9788		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_u16)))
9789		void svst1q_scatter_u64base_index_u16(svbool_t, svuint64_t, int64_t, svuint16_t);
9790		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_bf16)))
9791		void svst1q_scatter_u64base_index_bf16(svbool_t, svuint64_t, int64_t, svbfloat16_t);
9792		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_f64)))
9793		void svst1q_scatter_u64base_index_f64(svbool_t, svuint64_t, int64_t, svfloat64_t);
9794		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_f32)))
9795		void svst1q_scatter_u64base_index_f32(svbool_t, svuint64_t, int64_t, svfloat32_t);
9796		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_f16)))
9797		void svst1q_scatter_u64base_index_f16(svbool_t, svuint64_t, int64_t, svfloat16_t);
9798		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_s32)))
9799		void svst1q_scatter_u64base_index_s32(svbool_t, svuint64_t, int64_t, svint32_t);
9800		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_s64)))
9801		void svst1q_scatter_u64base_index_s64(svbool_t, svuint64_t, int64_t, svint64_t);
9802		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_s16)))
9803		void svst1q_scatter_u64base_index_s16(svbool_t, svuint64_t, int64_t, svint16_t);
9804		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_u8)))
9805		void svst1q_scatter_u64base_offset_u8(svbool_t, svuint64_t, int64_t, svuint8_t);
9806		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_u32)))
9807		void svst1q_scatter_u64base_offset_u32(svbool_t, svuint64_t, int64_t, svuint32_t);
9808		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_u64)))
9809		void svst1q_scatter_u64base_offset_u64(svbool_t, svuint64_t, int64_t, svuint64_t);
9810		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_u16)))
9811		void svst1q_scatter_u64base_offset_u16(svbool_t, svuint64_t, int64_t, svuint16_t);
9812		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_bf16)))
9813		void svst1q_scatter_u64base_offset_bf16(svbool_t, svuint64_t, int64_t, svbfloat16_t);
9814		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_s8)))
9815		void svst1q_scatter_u64base_offset_s8(svbool_t, svuint64_t, int64_t, svint8_t);
9816		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_f64)))
9817		void svst1q_scatter_u64base_offset_f64(svbool_t, svuint64_t, int64_t, svfloat64_t);
9818		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_f32)))
9819		void svst1q_scatter_u64base_offset_f32(svbool_t, svuint64_t, int64_t, svfloat32_t);
9820		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_f16)))
9821		void svst1q_scatter_u64base_offset_f16(svbool_t, svuint64_t, int64_t, svfloat16_t);
9822		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_s32)))
9823		void svst1q_scatter_u64base_offset_s32(svbool_t, svuint64_t, int64_t, svint32_t);
9824		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_s64)))
9825		void svst1q_scatter_u64base_offset_s64(svbool_t, svuint64_t, int64_t, svint64_t);
9826		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_s16)))
9827		void svst1q_scatter_u64base_offset_s16(svbool_t, svuint64_t, int64_t, svint16_t);
9828		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_u32)))
9829		void svst1q_scatter_s64index_u32(svbool_t, uint32_t *, svint64_t, svuint32_t);
9830		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_u64)))
9831		void svst1q_scatter_s64index_u64(svbool_t, uint64_t *, svint64_t, svuint64_t);
9832		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_u16)))
9833		void svst1q_scatter_s64index_u16(svbool_t, uint16_t *, svint64_t, svuint16_t);
9834		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_bf16)))
9835		void svst1q_scatter_s64index_bf16(svbool_t, bfloat16_t *, svint64_t, svbfloat16_t);
9836		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_f64)))
9837		void svst1q_scatter_s64index_f64(svbool_t, float64_t *, svint64_t, svfloat64_t);
9838		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_f32)))
9839		void svst1q_scatter_s64index_f32(svbool_t, float32_t *, svint64_t, svfloat32_t);
9840		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_f16)))
9841		void svst1q_scatter_s64index_f16(svbool_t, float16_t *, svint64_t, svfloat16_t);
9842		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_s32)))
9843		void svst1q_scatter_s64index_s32(svbool_t, int32_t *, svint64_t, svint32_t);
9844		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_s64)))
9845		void svst1q_scatter_s64index_s64(svbool_t, int64_t *, svint64_t, svint64_t);
9846		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_s16)))
9847		void svst1q_scatter_s64index_s16(svbool_t, int16_t *, svint64_t, svint16_t);
9848		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_u32)))
9849		void svst1q_scatter_u64index_u32(svbool_t, uint32_t *, svuint64_t, svuint32_t);
9850		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_u64)))
9851		void svst1q_scatter_u64index_u64(svbool_t, uint64_t *, svuint64_t, svuint64_t);
9852		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_u16)))
9853		void svst1q_scatter_u64index_u16(svbool_t, uint16_t *, svuint64_t, svuint16_t);
9854		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_bf16)))
9855		void svst1q_scatter_u64index_bf16(svbool_t, bfloat16_t *, svuint64_t, svbfloat16_t);
9856		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_f64)))
9857		void svst1q_scatter_u64index_f64(svbool_t, float64_t *, svuint64_t, svfloat64_t);
9858		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_f32)))
9859		void svst1q_scatter_u64index_f32(svbool_t, float32_t *, svuint64_t, svfloat32_t);
9860		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_f16)))
9861		void svst1q_scatter_u64index_f16(svbool_t, float16_t *, svuint64_t, svfloat16_t);
9862		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_s32)))
9863		void svst1q_scatter_u64index_s32(svbool_t, int32_t *, svuint64_t, svint32_t);
9864		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_s64)))
9865		void svst1q_scatter_u64index_s64(svbool_t, int64_t *, svuint64_t, svint64_t);
9866		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_s16)))
9867		void svst1q_scatter_u64index_s16(svbool_t, int16_t *, svuint64_t, svint16_t);
9868		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_u8)))
9869		void svst1q_scatter_s64offset_u8(svbool_t, uint8_t *, svint64_t, svuint8_t);
9870		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_u32)))
9871		void svst1q_scatter_s64offset_u32(svbool_t, uint32_t *, svint64_t, svuint32_t);
9872		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_u64)))
9873		void svst1q_scatter_s64offset_u64(svbool_t, uint64_t *, svint64_t, svuint64_t);
9874		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_u16)))
9875		void svst1q_scatter_s64offset_u16(svbool_t, uint16_t *, svint64_t, svuint16_t);
9876		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_bf16)))
9877		void svst1q_scatter_s64offset_bf16(svbool_t, bfloat16_t *, svint64_t, svbfloat16_t);
9878		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_s8)))
9879		void svst1q_scatter_s64offset_s8(svbool_t, int8_t *, svint64_t, svint8_t);
9880		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_f64)))
9881		void svst1q_scatter_s64offset_f64(svbool_t, float64_t *, svint64_t, svfloat64_t);
9882		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_f32)))
9883		void svst1q_scatter_s64offset_f32(svbool_t, float32_t *, svint64_t, svfloat32_t);
9884		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_f16)))
9885		void svst1q_scatter_s64offset_f16(svbool_t, float16_t *, svint64_t, svfloat16_t);
9886		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_s32)))
9887		void svst1q_scatter_s64offset_s32(svbool_t, int32_t *, svint64_t, svint32_t);
9888		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_s64)))
9889		void svst1q_scatter_s64offset_s64(svbool_t, int64_t *, svint64_t, svint64_t);
9890		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_s16)))
9891		void svst1q_scatter_s64offset_s16(svbool_t, int16_t *, svint64_t, svint16_t);
9892		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_u8)))
9893		void svst1q_scatter_u64offset_u8(svbool_t, uint8_t *, svuint64_t, svuint8_t);
9894		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_u32)))
9895		void svst1q_scatter_u64offset_u32(svbool_t, uint32_t *, svuint64_t, svuint32_t);
9896		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_u64)))
9897		void svst1q_scatter_u64offset_u64(svbool_t, uint64_t *, svuint64_t, svuint64_t);
9898		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_u16)))
9899		void svst1q_scatter_u64offset_u16(svbool_t, uint16_t *, svuint64_t, svuint16_t);
9900		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_bf16)))
9901		void svst1q_scatter_u64offset_bf16(svbool_t, bfloat16_t *, svuint64_t, svbfloat16_t);
9902		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_s8)))
9903		void svst1q_scatter_u64offset_s8(svbool_t, int8_t *, svuint64_t, svint8_t);
9904		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_f64)))
9905		void svst1q_scatter_u64offset_f64(svbool_t, float64_t *, svuint64_t, svfloat64_t);
9906		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_f32)))
9907		void svst1q_scatter_u64offset_f32(svbool_t, float32_t *, svuint64_t, svfloat32_t);
9908		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_f16)))
9909		void svst1q_scatter_u64offset_f16(svbool_t, float16_t *, svuint64_t, svfloat16_t);
9910		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_s32)))
9911		void svst1q_scatter_u64offset_s32(svbool_t, int32_t *, svuint64_t, svint32_t);
9912		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_s64)))
9913		void svst1q_scatter_u64offset_s64(svbool_t, int64_t *, svuint64_t, svint64_t);
9914		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_s16)))
9915		void svst1q_scatter_u64offset_s16(svbool_t, int16_t *, svuint64_t, svint16_t);
9916		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1wq_u32)))
9917		void svst1wq_u32(svbool_t, uint32_t *, svuint32_t);
9918		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1wq_f32)))
9919		void svst1wq_f32(svbool_t, float32_t *, svfloat32_t);
9920		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1wq_s32)))
9921		void svst1wq_s32(svbool_t, int32_t *, svint32_t);
9922		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1wq_vnum_u32)))
9923		void svst1wq_vnum_u32(svbool_t, uint32_t *, int64_t, svuint32_t);
9924		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1wq_vnum_f32)))
9925		void svst1wq_vnum_f32(svbool_t, float32_t *, int64_t, svfloat32_t);
9926		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1wq_vnum_s32)))
9927		void svst1wq_vnum_s32(svbool_t, int32_t *, int64_t, svint32_t);
9928		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_u8)))
9929		void svst2q_u8(svbool_t, uint8_t const *, svuint8x2_t);
9930		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_u32)))
9931		void svst2q_u32(svbool_t, uint32_t const *, svuint32x2_t);
9932		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_u64)))
9933		void svst2q_u64(svbool_t, uint64_t const *, svuint64x2_t);
9934		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_u16)))
9935		void svst2q_u16(svbool_t, uint16_t const *, svuint16x2_t);
9936		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_s8)))
9937		void svst2q_s8(svbool_t, int8_t const *, svint8x2_t);
9938		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_f64)))
9939		void svst2q_f64(svbool_t, float64_t const *, svfloat64x2_t);
9940		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_f32)))
9941		void svst2q_f32(svbool_t, float32_t const *, svfloat32x2_t);
9942		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_f16)))
9943		void svst2q_f16(svbool_t, float16_t const *, svfloat16x2_t);
9944		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_s32)))
9945		void svst2q_s32(svbool_t, int32_t const *, svint32x2_t);
9946		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_s64)))
9947		void svst2q_s64(svbool_t, int64_t const *, svint64x2_t);
9948		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_s16)))
9949		void svst2q_s16(svbool_t, int16_t const *, svint16x2_t);
9950		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_bf16)))
9951		void svst2q_bf16(svbool_t, bfloat16_t const *, svbfloat16x2_t);
9952		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_u8)))
9953		void svst2q_vnum_u8(svbool_t, uint8_t const *, int64_t, svuint8x2_t);
9954		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_u32)))
9955		void svst2q_vnum_u32(svbool_t, uint32_t const *, int64_t, svuint32x2_t);
9956		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_u64)))
9957		void svst2q_vnum_u64(svbool_t, uint64_t const *, int64_t, svuint64x2_t);
9958		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_u16)))
9959		void svst2q_vnum_u16(svbool_t, uint16_t const *, int64_t, svuint16x2_t);
9960		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_s8)))
9961		void svst2q_vnum_s8(svbool_t, int8_t const *, int64_t, svint8x2_t);
9962		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_f64)))
9963		void svst2q_vnum_f64(svbool_t, float64_t const *, int64_t, svfloat64x2_t);
9964		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_f32)))
9965		void svst2q_vnum_f32(svbool_t, float32_t const *, int64_t, svfloat32x2_t);
9966		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_f16)))
9967		void svst2q_vnum_f16(svbool_t, float16_t const *, int64_t, svfloat16x2_t);
9968		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_s32)))
9969		void svst2q_vnum_s32(svbool_t, int32_t const *, int64_t, svint32x2_t);
9970		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_s64)))
9971		void svst2q_vnum_s64(svbool_t, int64_t const *, int64_t, svint64x2_t);
9972		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_s16)))
9973		void svst2q_vnum_s16(svbool_t, int16_t const *, int64_t, svint16x2_t);
9974		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_bf16)))
9975		void svst2q_vnum_bf16(svbool_t, bfloat16_t const *, int64_t, svbfloat16x2_t);
9976		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_u8)))
9977		void svst3q_u8(svbool_t, uint8_t const *, svuint8x3_t);
9978		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_u32)))
9979		void svst3q_u32(svbool_t, uint32_t const *, svuint32x3_t);
9980		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_u64)))
9981		void svst3q_u64(svbool_t, uint64_t const *, svuint64x3_t);
9982		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_u16)))
9983		void svst3q_u16(svbool_t, uint16_t const *, svuint16x3_t);
9984		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_s8)))
9985		void svst3q_s8(svbool_t, int8_t const *, svint8x3_t);
9986		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_f64)))
9987		void svst3q_f64(svbool_t, float64_t const *, svfloat64x3_t);
9988		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_f32)))
9989		void svst3q_f32(svbool_t, float32_t const *, svfloat32x3_t);
9990		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_f16)))
9991		void svst3q_f16(svbool_t, float16_t const *, svfloat16x3_t);
9992		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_s32)))
9993		void svst3q_s32(svbool_t, int32_t const *, svint32x3_t);
9994		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_s64)))
9995		void svst3q_s64(svbool_t, int64_t const *, svint64x3_t);
9996		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_s16)))
9997		void svst3q_s16(svbool_t, int16_t const *, svint16x3_t);
9998		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_bf16)))
9999		void svst3q_bf16(svbool_t, bfloat16_t const *, svbfloat16x3_t);
10000		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_u8)))
10001		void svst3q_vnum_u8(svbool_t, uint8_t const *, int64_t, svuint8x3_t);
10002		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_u32)))
10003		void svst3q_vnum_u32(svbool_t, uint32_t const *, int64_t, svuint32x3_t);
10004		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_u64)))
10005		void svst3q_vnum_u64(svbool_t, uint64_t const *, int64_t, svuint64x3_t);
10006		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_u16)))
10007		void svst3q_vnum_u16(svbool_t, uint16_t const *, int64_t, svuint16x3_t);
10008		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_s8)))
10009		void svst3q_vnum_s8(svbool_t, int8_t const *, int64_t, svint8x3_t);
10010		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_f64)))
10011		void svst3q_vnum_f64(svbool_t, float64_t const *, int64_t, svfloat64x3_t);
10012		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_f32)))
10013		void svst3q_vnum_f32(svbool_t, float32_t const *, int64_t, svfloat32x3_t);
10014		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_f16)))
10015		void svst3q_vnum_f16(svbool_t, float16_t const *, int64_t, svfloat16x3_t);
10016		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_s32)))
10017		void svst3q_vnum_s32(svbool_t, int32_t const *, int64_t, svint32x3_t);
10018		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_s64)))
10019		void svst3q_vnum_s64(svbool_t, int64_t const *, int64_t, svint64x3_t);
10020		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_s16)))
10021		void svst3q_vnum_s16(svbool_t, int16_t const *, int64_t, svint16x3_t);
10022		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_bf16)))
10023		void svst3q_vnum_bf16(svbool_t, bfloat16_t const *, int64_t, svbfloat16x3_t);
10024		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_u8)))
10025		void svst4q_u8(svbool_t, uint8_t const *, svuint8x4_t);
10026		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_u32)))
10027		void svst4q_u32(svbool_t, uint32_t const *, svuint32x4_t);
10028		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_u64)))
10029		void svst4q_u64(svbool_t, uint64_t const *, svuint64x4_t);
10030		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_u16)))
10031		void svst4q_u16(svbool_t, uint16_t const *, svuint16x4_t);
10032		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_s8)))
10033		void svst4q_s8(svbool_t, int8_t const *, svint8x4_t);
10034		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_f64)))
10035		void svst4q_f64(svbool_t, float64_t const *, svfloat64x4_t);
10036		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_f32)))
10037		void svst4q_f32(svbool_t, float32_t const *, svfloat32x4_t);
10038		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_f16)))
10039		void svst4q_f16(svbool_t, float16_t const *, svfloat16x4_t);
10040		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_s32)))
10041		void svst4q_s32(svbool_t, int32_t const *, svint32x4_t);
10042		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_s64)))
10043		void svst4q_s64(svbool_t, int64_t const *, svint64x4_t);
10044		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_s16)))
10045		void svst4q_s16(svbool_t, int16_t const *, svint16x4_t);
10046		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_bf16)))
10047		void svst4q_bf16(svbool_t, bfloat16_t const *, svbfloat16x4_t);
10048		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_u8)))
10049		void svst4q_vnum_u8(svbool_t, uint8_t const *, int64_t, svuint8x4_t);
10050		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_u32)))
10051		void svst4q_vnum_u32(svbool_t, uint32_t const *, int64_t, svuint32x4_t);
10052		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_u64)))
10053		void svst4q_vnum_u64(svbool_t, uint64_t const *, int64_t, svuint64x4_t);
10054		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_u16)))
10055		void svst4q_vnum_u16(svbool_t, uint16_t const *, int64_t, svuint16x4_t);
10056		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_s8)))
10057		void svst4q_vnum_s8(svbool_t, int8_t const *, int64_t, svint8x4_t);
10058		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_f64)))
10059		void svst4q_vnum_f64(svbool_t, float64_t const *, int64_t, svfloat64x4_t);
10060		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_f32)))
10061		void svst4q_vnum_f32(svbool_t, float32_t const *, int64_t, svfloat32x4_t);
10062		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_f16)))
10063		void svst4q_vnum_f16(svbool_t, float16_t const *, int64_t, svfloat16x4_t);
10064		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_s32)))
10065		void svst4q_vnum_s32(svbool_t, int32_t const *, int64_t, svint32x4_t);
10066		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_s64)))
10067		void svst4q_vnum_s64(svbool_t, int64_t const *, int64_t, svint64x4_t);
10068		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_s16)))
10069		void svst4q_vnum_s16(svbool_t, int16_t const *, int64_t, svint16x4_t);
10070		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_bf16)))
10071		void svst4q_vnum_bf16(svbool_t, bfloat16_t const *, int64_t, svbfloat16x4_t);
10072		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_u8)))
10073		svuint8_t svtblq_u8(svuint8_t, svuint8_t);
10074		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_u32)))
10075		svuint32_t svtblq_u32(svuint32_t, svuint32_t);
10076		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_u64)))
10077		svuint64_t svtblq_u64(svuint64_t, svuint64_t);
10078		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_u16)))
10079		svuint16_t svtblq_u16(svuint16_t, svuint16_t);
10080		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_bf16)))
10081		svbfloat16_t svtblq_bf16(svbfloat16_t, svuint16_t);
10082		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_s8)))
10083		svint8_t svtblq_s8(svint8_t, svuint8_t);
10084		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_f64)))
10085		svfloat64_t svtblq_f64(svfloat64_t, svuint64_t);
10086		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_f32)))
10087		svfloat32_t svtblq_f32(svfloat32_t, svuint32_t);
10088		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_f16)))
10089		svfloat16_t svtblq_f16(svfloat16_t, svuint16_t);
10090		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_s32)))
10091		svint32_t svtblq_s32(svint32_t, svuint32_t);
10092		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_s64)))
10093		svint64_t svtblq_s64(svint64_t, svuint64_t);
10094		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_mf8)))
10095		svmfloat8_t svtblq_mf8(svmfloat8_t, svuint8_t);
10096		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_s16)))
10097		svint16_t svtblq_s16(svint16_t, svuint16_t);
10098		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_u8)))
10099		svuint8_t svtbxq_u8(svuint8_t, svuint8_t, svuint8_t);
10100		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_u32)))
10101		svuint32_t svtbxq_u32(svuint32_t, svuint32_t, svuint32_t);
10102		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_u64)))
10103		svuint64_t svtbxq_u64(svuint64_t, svuint64_t, svuint64_t);
10104		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_u16)))
10105		svuint16_t svtbxq_u16(svuint16_t, svuint16_t, svuint16_t);
10106		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_bf16)))
10107		svbfloat16_t svtbxq_bf16(svbfloat16_t, svbfloat16_t, svuint16_t);
10108		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_s8)))
10109		svint8_t svtbxq_s8(svint8_t, svint8_t, svuint8_t);
10110		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_f64)))
10111		svfloat64_t svtbxq_f64(svfloat64_t, svfloat64_t, svuint64_t);
10112		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_f32)))
10113		svfloat32_t svtbxq_f32(svfloat32_t, svfloat32_t, svuint32_t);
10114		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_f16)))
10115		svfloat16_t svtbxq_f16(svfloat16_t, svfloat16_t, svuint16_t);
10116		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_s32)))
10117		svint32_t svtbxq_s32(svint32_t, svint32_t, svuint32_t);
10118		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_s64)))
10119		svint64_t svtbxq_s64(svint64_t, svint64_t, svuint64_t);
10120		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_mf8)))
10121		svmfloat8_t svtbxq_mf8(svmfloat8_t, svmfloat8_t, svuint8_t);
10122		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_s16)))
10123		svint16_t svtbxq_s16(svint16_t, svint16_t, svuint16_t);
10124		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_u8)))
10125		svuint8_t svuzpq1_u8(svuint8_t, svuint8_t);
10126		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_u32)))
10127		svuint32_t svuzpq1_u32(svuint32_t, svuint32_t);
10128		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_u64)))
10129		svuint64_t svuzpq1_u64(svuint64_t, svuint64_t);
10130		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_u16)))
10131		svuint16_t svuzpq1_u16(svuint16_t, svuint16_t);
10132		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_bf16)))
10133		svbfloat16_t svuzpq1_bf16(svbfloat16_t, svbfloat16_t);
10134		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_s8)))
10135		svint8_t svuzpq1_s8(svint8_t, svint8_t);
10136		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_f64)))
10137		svfloat64_t svuzpq1_f64(svfloat64_t, svfloat64_t);
10138		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_f32)))
10139		svfloat32_t svuzpq1_f32(svfloat32_t, svfloat32_t);
10140		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_f16)))
10141		svfloat16_t svuzpq1_f16(svfloat16_t, svfloat16_t);
10142		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_s32)))
10143		svint32_t svuzpq1_s32(svint32_t, svint32_t);
10144		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_s64)))
10145		svint64_t svuzpq1_s64(svint64_t, svint64_t);
10146		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_mf8)))
10147		svmfloat8_t svuzpq1_mf8(svmfloat8_t, svmfloat8_t);
10148		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_s16)))
10149		svint16_t svuzpq1_s16(svint16_t, svint16_t);
10150		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_u8)))
10151		svuint8_t svuzpq2_u8(svuint8_t, svuint8_t);
10152		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_u32)))
10153		svuint32_t svuzpq2_u32(svuint32_t, svuint32_t);
10154		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_u64)))
10155		svuint64_t svuzpq2_u64(svuint64_t, svuint64_t);
10156		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_u16)))
10157		svuint16_t svuzpq2_u16(svuint16_t, svuint16_t);
10158		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_bf16)))
10159		svbfloat16_t svuzpq2_bf16(svbfloat16_t, svbfloat16_t);
10160		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_s8)))
10161		svint8_t svuzpq2_s8(svint8_t, svint8_t);
10162		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_f64)))
10163		svfloat64_t svuzpq2_f64(svfloat64_t, svfloat64_t);
10164		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_f32)))
10165		svfloat32_t svuzpq2_f32(svfloat32_t, svfloat32_t);
10166		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_f16)))
10167		svfloat16_t svuzpq2_f16(svfloat16_t, svfloat16_t);
10168		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_s32)))
10169		svint32_t svuzpq2_s32(svint32_t, svint32_t);
10170		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_s64)))
10171		svint64_t svuzpq2_s64(svint64_t, svint64_t);
10172		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_mf8)))
10173		svmfloat8_t svuzpq2_mf8(svmfloat8_t, svmfloat8_t);
10174		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_s16)))
10175		svint16_t svuzpq2_s16(svint16_t, svint16_t);
10176		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_u8)))
10177		svuint8_t svzipq1_u8(svuint8_t, svuint8_t);
10178		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_u32)))
10179		svuint32_t svzipq1_u32(svuint32_t, svuint32_t);
10180		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_u64)))
10181		svuint64_t svzipq1_u64(svuint64_t, svuint64_t);
10182		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_u16)))
10183		svuint16_t svzipq1_u16(svuint16_t, svuint16_t);
10184		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_bf16)))
10185		svbfloat16_t svzipq1_bf16(svbfloat16_t, svbfloat16_t);
10186		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_s8)))
10187		svint8_t svzipq1_s8(svint8_t, svint8_t);
10188		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_f64)))
10189		svfloat64_t svzipq1_f64(svfloat64_t, svfloat64_t);
10190		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_f32)))
10191		svfloat32_t svzipq1_f32(svfloat32_t, svfloat32_t);
10192		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_f16)))
10193		svfloat16_t svzipq1_f16(svfloat16_t, svfloat16_t);
10194		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_s32)))
10195		svint32_t svzipq1_s32(svint32_t, svint32_t);
10196		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_s64)))
10197		svint64_t svzipq1_s64(svint64_t, svint64_t);
10198		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_mf8)))
10199		svmfloat8_t svzipq1_mf8(svmfloat8_t, svmfloat8_t);
10200		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_s16)))
10201		svint16_t svzipq1_s16(svint16_t, svint16_t);
10202		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_u8)))
10203		svuint8_t svzipq2_u8(svuint8_t, svuint8_t);
10204		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_u32)))
10205		svuint32_t svzipq2_u32(svuint32_t, svuint32_t);
10206		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_u64)))
10207		svuint64_t svzipq2_u64(svuint64_t, svuint64_t);
10208		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_u16)))
10209		svuint16_t svzipq2_u16(svuint16_t, svuint16_t);
10210		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_bf16)))
10211		svbfloat16_t svzipq2_bf16(svbfloat16_t, svbfloat16_t);
10212		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_s8)))
10213		svint8_t svzipq2_s8(svint8_t, svint8_t);
10214		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_f64)))
10215		svfloat64_t svzipq2_f64(svfloat64_t, svfloat64_t);
10216		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_f32)))
10217		svfloat32_t svzipq2_f32(svfloat32_t, svfloat32_t);
10218		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_f16)))
10219		svfloat16_t svzipq2_f16(svfloat16_t, svfloat16_t);
10220		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_s32)))
10221		svint32_t svzipq2_s32(svint32_t, svint32_t);
10222		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_s64)))
10223		svint64_t svzipq2_s64(svint64_t, svint64_t);
10224		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_mf8)))
10225		svmfloat8_t svzipq2_mf8(svmfloat8_t, svmfloat8_t);
10226		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_s16)))
10227		svint16_t svzipq2_s16(svint16_t, svint16_t);
10228		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_u8)))
10229		uint8x16_t svaddqv(svbool_t, svuint8_t);
10230		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_u32)))
10231		uint32x4_t svaddqv(svbool_t, svuint32_t);
10232		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_u64)))
10233		uint64x2_t svaddqv(svbool_t, svuint64_t);
10234		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_u16)))
10235		uint16x8_t svaddqv(svbool_t, svuint16_t);
10236		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_s8)))
10237		int8x16_t svaddqv(svbool_t, svint8_t);
10238		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_s32)))
10239		int32x4_t svaddqv(svbool_t, svint32_t);
10240		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_s64)))
10241		int64x2_t svaddqv(svbool_t, svint64_t);
10242		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_s16)))
10243		int16x8_t svaddqv(svbool_t, svint16_t);
10244		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_f64)))
10245		float64x2_t svaddqv(svbool_t, svfloat64_t);
10246		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_f32)))
10247		float32x4_t svaddqv(svbool_t, svfloat32_t);
10248		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_f16)))
10249		float16x8_t svaddqv(svbool_t, svfloat16_t);
10250		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_u8)))
10251		uint8x16_t svandqv(svbool_t, svuint8_t);
10252		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_u32)))
10253		uint32x4_t svandqv(svbool_t, svuint32_t);
10254		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_u64)))
10255		uint64x2_t svandqv(svbool_t, svuint64_t);
10256		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_u16)))
10257		uint16x8_t svandqv(svbool_t, svuint16_t);
10258		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_s8)))
10259		int8x16_t svandqv(svbool_t, svint8_t);
10260		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_s32)))
10261		int32x4_t svandqv(svbool_t, svint32_t);
10262		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_s64)))
10263		int64x2_t svandqv(svbool_t, svint64_t);
10264		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_s16)))
10265		int16x8_t svandqv(svbool_t, svint16_t);
10266		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_u8)))
10267		uint8x16_t sveorqv(svbool_t, svuint8_t);
10268		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_u32)))
10269		uint32x4_t sveorqv(svbool_t, svuint32_t);
10270		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_u64)))
10271		uint64x2_t sveorqv(svbool_t, svuint64_t);
10272		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_u16)))
10273		uint16x8_t sveorqv(svbool_t, svuint16_t);
10274		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_s8)))
10275		int8x16_t sveorqv(svbool_t, svint8_t);
10276		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_s32)))
10277		int32x4_t sveorqv(svbool_t, svint32_t);
10278		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_s64)))
10279		int64x2_t sveorqv(svbool_t, svint64_t);
10280		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_s16)))
10281		int16x8_t sveorqv(svbool_t, svint16_t);
10282		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_u8)))
10283		svuint8_t svextq(svuint8_t, svuint8_t, int32_t);
10284		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_u32)))
10285		svuint32_t svextq(svuint32_t, svuint32_t, int32_t);
10286		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_u64)))
10287		svuint64_t svextq(svuint64_t, svuint64_t, int32_t);
10288		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_u16)))
10289		svuint16_t svextq(svuint16_t, svuint16_t, int32_t);
10290		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_bf16)))
10291		svbfloat16_t svextq(svbfloat16_t, svbfloat16_t, int32_t);
10292		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_s8)))
10293		svint8_t svextq(svint8_t, svint8_t, int32_t);
10294		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_f64)))
10295		svfloat64_t svextq(svfloat64_t, svfloat64_t, int32_t);
10296		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_f32)))
10297		svfloat32_t svextq(svfloat32_t, svfloat32_t, int32_t);
10298		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_f16)))
10299		svfloat16_t svextq(svfloat16_t, svfloat16_t, int32_t);
10300		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_s32)))
10301		svint32_t svextq(svint32_t, svint32_t, int32_t);
10302		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_s64)))
10303		svint64_t svextq(svint64_t, svint64_t, int32_t);
10304		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_mf8)))
10305		svmfloat8_t svextq(svmfloat8_t, svmfloat8_t, int32_t);
10306		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_s16)))
10307		svint16_t svextq(svint16_t, svint16_t, int32_t);
10308		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_u32)))
10309		svuint32_t svld1q_gather_index_u32(svbool_t, svuint64_t, int64_t);
10310		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_u64)))
10311		svuint64_t svld1q_gather_index_u64(svbool_t, svuint64_t, int64_t);
10312		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_u16)))
10313		svuint16_t svld1q_gather_index_u16(svbool_t, svuint64_t, int64_t);
10314		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_bf16)))
10315		svbfloat16_t svld1q_gather_index_bf16(svbool_t, svuint64_t, int64_t);
10316		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_f64)))
10317		svfloat64_t svld1q_gather_index_f64(svbool_t, svuint64_t, int64_t);
10318		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_f32)))
10319		svfloat32_t svld1q_gather_index_f32(svbool_t, svuint64_t, int64_t);
10320		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_f16)))
10321		svfloat16_t svld1q_gather_index_f16(svbool_t, svuint64_t, int64_t);
10322		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_s32)))
10323		svint32_t svld1q_gather_index_s32(svbool_t, svuint64_t, int64_t);
10324		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_s64)))
10325		svint64_t svld1q_gather_index_s64(svbool_t, svuint64_t, int64_t);
10326		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_s16)))
10327		svint16_t svld1q_gather_index_s16(svbool_t, svuint64_t, int64_t);
10328		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_u8)))
10329		svuint8_t svld1q_gather_offset_u8(svbool_t, svuint64_t, int64_t);
10330		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_u32)))
10331		svuint32_t svld1q_gather_offset_u32(svbool_t, svuint64_t, int64_t);
10332		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_u64)))
10333		svuint64_t svld1q_gather_offset_u64(svbool_t, svuint64_t, int64_t);
10334		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_u16)))
10335		svuint16_t svld1q_gather_offset_u16(svbool_t, svuint64_t, int64_t);
10336		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_bf16)))
10337		svbfloat16_t svld1q_gather_offset_bf16(svbool_t, svuint64_t, int64_t);
10338		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_s8)))
10339		svint8_t svld1q_gather_offset_s8(svbool_t, svuint64_t, int64_t);
10340		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_f64)))
10341		svfloat64_t svld1q_gather_offset_f64(svbool_t, svuint64_t, int64_t);
10342		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_f32)))
10343		svfloat32_t svld1q_gather_offset_f32(svbool_t, svuint64_t, int64_t);
10344		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_f16)))
10345		svfloat16_t svld1q_gather_offset_f16(svbool_t, svuint64_t, int64_t);
10346		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_s32)))
10347		svint32_t svld1q_gather_offset_s32(svbool_t, svuint64_t, int64_t);
10348		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_s64)))
10349		svint64_t svld1q_gather_offset_s64(svbool_t, svuint64_t, int64_t);
10350		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_s16)))
10351		svint16_t svld1q_gather_offset_s16(svbool_t, svuint64_t, int64_t);
10352		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_u8)))
10353		svuint8_t svld1q_gather_u8(svbool_t, svuint64_t);
10354		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_u32)))
10355		svuint32_t svld1q_gather_u32(svbool_t, svuint64_t);
10356		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_u64)))
10357		svuint64_t svld1q_gather_u64(svbool_t, svuint64_t);
10358		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_u16)))
10359		svuint16_t svld1q_gather_u16(svbool_t, svuint64_t);
10360		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_bf16)))
10361		svbfloat16_t svld1q_gather_bf16(svbool_t, svuint64_t);
10362		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_s8)))
10363		svint8_t svld1q_gather_s8(svbool_t, svuint64_t);
10364		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_f64)))
10365		svfloat64_t svld1q_gather_f64(svbool_t, svuint64_t);
10366		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_f32)))
10367		svfloat32_t svld1q_gather_f32(svbool_t, svuint64_t);
10368		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_f16)))
10369		svfloat16_t svld1q_gather_f16(svbool_t, svuint64_t);
10370		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_s32)))
10371		svint32_t svld1q_gather_s32(svbool_t, svuint64_t);
10372		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_s64)))
10373		svint64_t svld1q_gather_s64(svbool_t, svuint64_t);
10374		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_s16)))
10375		svint16_t svld1q_gather_s16(svbool_t, svuint64_t);
10376		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_u32)))
10377		svuint32_t svld1q_gather_index(svbool_t, uint32_t const *, svuint64_t);
10378		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_u64)))
10379		svuint64_t svld1q_gather_index(svbool_t, uint64_t const *, svuint64_t);
10380		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_u16)))
10381		svuint16_t svld1q_gather_index(svbool_t, uint16_t const *, svuint64_t);
10382		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_bf16)))
10383		svbfloat16_t svld1q_gather_index(svbool_t, bfloat16_t const *, svuint64_t);
10384		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_f64)))
10385		svfloat64_t svld1q_gather_index(svbool_t, float64_t const *, svuint64_t);
10386		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_f32)))
10387		svfloat32_t svld1q_gather_index(svbool_t, float32_t const *, svuint64_t);
10388		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_f16)))
10389		svfloat16_t svld1q_gather_index(svbool_t, float16_t const *, svuint64_t);
10390		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_s32)))
10391		svint32_t svld1q_gather_index(svbool_t, int32_t const *, svuint64_t);
10392		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_s64)))
10393		svint64_t svld1q_gather_index(svbool_t, int64_t const *, svuint64_t);
10394		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_s16)))
10395		svint16_t svld1q_gather_index(svbool_t, int16_t const *, svuint64_t);
10396		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_u8)))
10397		svuint8_t svld1q_gather_offset(svbool_t, uint8_t const *, svuint64_t);
10398		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_u32)))
10399		svuint32_t svld1q_gather_offset(svbool_t, uint32_t const *, svuint64_t);
10400		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_u64)))
10401		svuint64_t svld1q_gather_offset(svbool_t, uint64_t const *, svuint64_t);
10402		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_u16)))
10403		svuint16_t svld1q_gather_offset(svbool_t, uint16_t const *, svuint64_t);
10404		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_bf16)))
10405		svbfloat16_t svld1q_gather_offset(svbool_t, bfloat16_t const *, svuint64_t);
10406		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_s8)))
10407		svint8_t svld1q_gather_offset(svbool_t, int8_t const *, svuint64_t);
10408		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_f64)))
10409		svfloat64_t svld1q_gather_offset(svbool_t, float64_t const *, svuint64_t);
10410		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_f32)))
10411		svfloat32_t svld1q_gather_offset(svbool_t, float32_t const *, svuint64_t);
10412		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_f16)))
10413		svfloat16_t svld1q_gather_offset(svbool_t, float16_t const *, svuint64_t);
10414		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_s32)))
10415		svint32_t svld1q_gather_offset(svbool_t, int32_t const *, svuint64_t);
10416		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_s64)))
10417		svint64_t svld1q_gather_offset(svbool_t, int64_t const *, svuint64_t);
10418		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_s16)))
10419		svint16_t svld1q_gather_offset(svbool_t, int16_t const *, svuint64_t);
10420		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1udq_u64)))
10421		svuint64_t svld1udq(svbool_t, uint64_t const *);
10422		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1udq_f64)))
10423		svfloat64_t svld1udq(svbool_t, float64_t const *);
10424		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1udq_s64)))
10425		svint64_t svld1udq(svbool_t, int64_t const *);
10426		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1udq_vnum_u64)))
10427		svuint64_t svld1udq_vnum(svbool_t, uint64_t const *, int64_t);
10428		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1udq_vnum_f64)))
10429		svfloat64_t svld1udq_vnum(svbool_t, float64_t const *, int64_t);
10430		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1udq_vnum_s64)))
10431		svint64_t svld1udq_vnum(svbool_t, int64_t const *, int64_t);
10432		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1uwq_u32)))
10433		svuint32_t svld1uwq(svbool_t, uint32_t const *);
10434		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1uwq_f32)))
10435		svfloat32_t svld1uwq(svbool_t, float32_t const *);
10436		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1uwq_s32)))
10437		svint32_t svld1uwq(svbool_t, int32_t const *);
10438		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1uwq_vnum_u32)))
10439		svuint32_t svld1uwq_vnum(svbool_t, uint32_t const *, int64_t);
10440		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1uwq_vnum_f32)))
10441		svfloat32_t svld1uwq_vnum(svbool_t, float32_t const *, int64_t);
10442		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1uwq_vnum_s32)))
10443		svint32_t svld1uwq_vnum(svbool_t, int32_t const *, int64_t);
10444		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_u8)))
10445		svuint8x2_t svld2q(svbool_t, uint8_t const *);
10446		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_u32)))
10447		svuint32x2_t svld2q(svbool_t, uint32_t const *);
10448		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_u64)))
10449		svuint64x2_t svld2q(svbool_t, uint64_t const *);
10450		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_u16)))
10451		svuint16x2_t svld2q(svbool_t, uint16_t const *);
10452		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_s8)))
10453		svint8x2_t svld2q(svbool_t, int8_t const *);
10454		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_f64)))
10455		svfloat64x2_t svld2q(svbool_t, float64_t const *);
10456		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_f32)))
10457		svfloat32x2_t svld2q(svbool_t, float32_t const *);
10458		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_f16)))
10459		svfloat16x2_t svld2q(svbool_t, float16_t const *);
10460		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_s32)))
10461		svint32x2_t svld2q(svbool_t, int32_t const *);
10462		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_s64)))
10463		svint64x2_t svld2q(svbool_t, int64_t const *);
10464		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_s16)))
10465		svint16x2_t svld2q(svbool_t, int16_t const *);
10466		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_bf16)))
10467		svbfloat16x2_t svld2q(svbool_t, bfloat16_t const *);
10468		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_u8)))
10469		svuint8x2_t svld2q_vnum(svbool_t, uint8_t const *, int64_t);
10470		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_u32)))
10471		svuint32x2_t svld2q_vnum(svbool_t, uint32_t const *, int64_t);
10472		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_u64)))
10473		svuint64x2_t svld2q_vnum(svbool_t, uint64_t const *, int64_t);
10474		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_u16)))
10475		svuint16x2_t svld2q_vnum(svbool_t, uint16_t const *, int64_t);
10476		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_s8)))
10477		svint8x2_t svld2q_vnum(svbool_t, int8_t const *, int64_t);
10478		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_f64)))
10479		svfloat64x2_t svld2q_vnum(svbool_t, float64_t const *, int64_t);
10480		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_f32)))
10481		svfloat32x2_t svld2q_vnum(svbool_t, float32_t const *, int64_t);
10482		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_f16)))
10483		svfloat16x2_t svld2q_vnum(svbool_t, float16_t const *, int64_t);
10484		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_s32)))
10485		svint32x2_t svld2q_vnum(svbool_t, int32_t const *, int64_t);
10486		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_s64)))
10487		svint64x2_t svld2q_vnum(svbool_t, int64_t const *, int64_t);
10488		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_s16)))
10489		svint16x2_t svld2q_vnum(svbool_t, int16_t const *, int64_t);
10490		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_bf16)))
10491		svbfloat16x2_t svld2q_vnum(svbool_t, bfloat16_t const *, int64_t);
10492		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_u8)))
10493		svuint8x3_t svld3q(svbool_t, uint8_t const *);
10494		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_u32)))
10495		svuint32x3_t svld3q(svbool_t, uint32_t const *);
10496		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_u64)))
10497		svuint64x3_t svld3q(svbool_t, uint64_t const *);
10498		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_u16)))
10499		svuint16x3_t svld3q(svbool_t, uint16_t const *);
10500		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_s8)))
10501		svint8x3_t svld3q(svbool_t, int8_t const *);
10502		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_f64)))
10503		svfloat64x3_t svld3q(svbool_t, float64_t const *);
10504		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_f32)))
10505		svfloat32x3_t svld3q(svbool_t, float32_t const *);
10506		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_f16)))
10507		svfloat16x3_t svld3q(svbool_t, float16_t const *);
10508		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_s32)))
10509		svint32x3_t svld3q(svbool_t, int32_t const *);
10510		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_s64)))
10511		svint64x3_t svld3q(svbool_t, int64_t const *);
10512		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_s16)))
10513		svint16x3_t svld3q(svbool_t, int16_t const *);
10514		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_bf16)))
10515		svbfloat16x3_t svld3q(svbool_t, bfloat16_t const *);
10516		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_u8)))
10517		svuint8x3_t svld3q_vnum(svbool_t, uint8_t const *, int64_t);
10518		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_u32)))
10519		svuint32x3_t svld3q_vnum(svbool_t, uint32_t const *, int64_t);
10520		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_u64)))
10521		svuint64x3_t svld3q_vnum(svbool_t, uint64_t const *, int64_t);
10522		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_u16)))
10523		svuint16x3_t svld3q_vnum(svbool_t, uint16_t const *, int64_t);
10524		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_s8)))
10525		svint8x3_t svld3q_vnum(svbool_t, int8_t const *, int64_t);
10526		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_f64)))
10527		svfloat64x3_t svld3q_vnum(svbool_t, float64_t const *, int64_t);
10528		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_f32)))
10529		svfloat32x3_t svld3q_vnum(svbool_t, float32_t const *, int64_t);
10530		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_f16)))
10531		svfloat16x3_t svld3q_vnum(svbool_t, float16_t const *, int64_t);
10532		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_s32)))
10533		svint32x3_t svld3q_vnum(svbool_t, int32_t const *, int64_t);
10534		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_s64)))
10535		svint64x3_t svld3q_vnum(svbool_t, int64_t const *, int64_t);
10536		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_s16)))
10537		svint16x3_t svld3q_vnum(svbool_t, int16_t const *, int64_t);
10538		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_bf16)))
10539		svbfloat16x3_t svld3q_vnum(svbool_t, bfloat16_t const *, int64_t);
10540		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_u8)))
10541		svuint8x4_t svld4q(svbool_t, uint8_t const *);
10542		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_u32)))
10543		svuint32x4_t svld4q(svbool_t, uint32_t const *);
10544		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_u64)))
10545		svuint64x4_t svld4q(svbool_t, uint64_t const *);
10546		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_u16)))
10547		svuint16x4_t svld4q(svbool_t, uint16_t const *);
10548		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_s8)))
10549		svint8x4_t svld4q(svbool_t, int8_t const *);
10550		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_f64)))
10551		svfloat64x4_t svld4q(svbool_t, float64_t const *);
10552		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_f32)))
10553		svfloat32x4_t svld4q(svbool_t, float32_t const *);
10554		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_f16)))
10555		svfloat16x4_t svld4q(svbool_t, float16_t const *);
10556		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_s32)))
10557		svint32x4_t svld4q(svbool_t, int32_t const *);
10558		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_s64)))
10559		svint64x4_t svld4q(svbool_t, int64_t const *);
10560		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_s16)))
10561		svint16x4_t svld4q(svbool_t, int16_t const *);
10562		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_bf16)))
10563		svbfloat16x4_t svld4q(svbool_t, bfloat16_t const *);
10564		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_u8)))
10565		svuint8x4_t svld4q_vnum(svbool_t, uint8_t const *, int64_t);
10566		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_u32)))
10567		svuint32x4_t svld4q_vnum(svbool_t, uint32_t const *, int64_t);
10568		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_u64)))
10569		svuint64x4_t svld4q_vnum(svbool_t, uint64_t const *, int64_t);
10570		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_u16)))
10571		svuint16x4_t svld4q_vnum(svbool_t, uint16_t const *, int64_t);
10572		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_s8)))
10573		svint8x4_t svld4q_vnum(svbool_t, int8_t const *, int64_t);
10574		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_f64)))
10575		svfloat64x4_t svld4q_vnum(svbool_t, float64_t const *, int64_t);
10576		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_f32)))
10577		svfloat32x4_t svld4q_vnum(svbool_t, float32_t const *, int64_t);
10578		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_f16)))
10579		svfloat16x4_t svld4q_vnum(svbool_t, float16_t const *, int64_t);
10580		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_s32)))
10581		svint32x4_t svld4q_vnum(svbool_t, int32_t const *, int64_t);
10582		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_s64)))
10583		svint64x4_t svld4q_vnum(svbool_t, int64_t const *, int64_t);
10584		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_s16)))
10585		svint16x4_t svld4q_vnum(svbool_t, int16_t const *, int64_t);
10586		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_bf16)))
10587		svbfloat16x4_t svld4q_vnum(svbool_t, bfloat16_t const *, int64_t);
10588		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnmqv_f64)))
10589		float64x2_t svmaxnmqv(svbool_t, svfloat64_t);
10590		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnmqv_f32)))
10591		float32x4_t svmaxnmqv(svbool_t, svfloat32_t);
10592		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnmqv_f16)))
10593		float16x8_t svmaxnmqv(svbool_t, svfloat16_t);
10594		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_f64)))
10595		float64x2_t svmaxqv(svbool_t, svfloat64_t);
10596		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_f32)))
10597		float32x4_t svmaxqv(svbool_t, svfloat32_t);
10598		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_f16)))
10599		float16x8_t svmaxqv(svbool_t, svfloat16_t);
10600		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_s8)))
10601		int8x16_t svmaxqv(svbool_t, svint8_t);
10602		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_s32)))
10603		int32x4_t svmaxqv(svbool_t, svint32_t);
10604		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_s64)))
10605		int64x2_t svmaxqv(svbool_t, svint64_t);
10606		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_s16)))
10607		int16x8_t svmaxqv(svbool_t, svint16_t);
10608		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_u8)))
10609		uint8x16_t svmaxqv(svbool_t, svuint8_t);
10610		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_u32)))
10611		uint32x4_t svmaxqv(svbool_t, svuint32_t);
10612		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_u64)))
10613		uint64x2_t svmaxqv(svbool_t, svuint64_t);
10614		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_u16)))
10615		uint16x8_t svmaxqv(svbool_t, svuint16_t);
10616		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnmqv_f64)))
10617		float64x2_t svminnmqv(svbool_t, svfloat64_t);
10618		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnmqv_f32)))
10619		float32x4_t svminnmqv(svbool_t, svfloat32_t);
10620		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnmqv_f16)))
10621		float16x8_t svminnmqv(svbool_t, svfloat16_t);
10622		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_f64)))
10623		float64x2_t svminqv(svbool_t, svfloat64_t);
10624		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_f32)))
10625		float32x4_t svminqv(svbool_t, svfloat32_t);
10626		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_f16)))
10627		float16x8_t svminqv(svbool_t, svfloat16_t);
10628		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_s8)))
10629		int8x16_t svminqv(svbool_t, svint8_t);
10630		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_s32)))
10631		int32x4_t svminqv(svbool_t, svint32_t);
10632		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_s64)))
10633		int64x2_t svminqv(svbool_t, svint64_t);
10634		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_s16)))
10635		int16x8_t svminqv(svbool_t, svint16_t);
10636		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_u8)))
10637		uint8x16_t svminqv(svbool_t, svuint8_t);
10638		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_u32)))
10639		uint32x4_t svminqv(svbool_t, svuint32_t);
10640		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_u64)))
10641		uint64x2_t svminqv(svbool_t, svuint64_t);
10642		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_u16)))
10643		uint16x8_t svminqv(svbool_t, svuint16_t);
10644		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_u8)))
10645		uint8x16_t svorqv(svbool_t, svuint8_t);
10646		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_u32)))
10647		uint32x4_t svorqv(svbool_t, svuint32_t);
10648		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_u64)))
10649		uint64x2_t svorqv(svbool_t, svuint64_t);
10650		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_u16)))
10651		uint16x8_t svorqv(svbool_t, svuint16_t);
10652		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_s8)))
10653		int8x16_t svorqv(svbool_t, svint8_t);
10654		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_s32)))
10655		int32x4_t svorqv(svbool_t, svint32_t);
10656		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_s64)))
10657		int64x2_t svorqv(svbool_t, svint64_t);
10658		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_s16)))
10659		int16x8_t svorqv(svbool_t, svint16_t);
10660		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_u8)))
10661		svbool_t svpmov(svuint8_t);
10662		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_s8)))
10663		svbool_t svpmov(svint8_t);
10664		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_u64)))
10665		svbool_t svpmov(svuint64_t);
10666		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_s64)))
10667		svbool_t svpmov(svint64_t);
10668		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_u16)))
10669		svbool_t svpmov(svuint16_t);
10670		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_s16)))
10671		svbool_t svpmov(svint16_t);
10672		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_u32)))
10673		svbool_t svpmov(svuint32_t);
10674		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_s32)))
10675		svbool_t svpmov(svint32_t);
10676		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_u8)))
10677		svbool_t svpmov_lane(svuint8_t, uint64_t);
10678		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_s8)))
10679		svbool_t svpmov_lane(svint8_t, uint64_t);
10680		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_u64)))
10681		svbool_t svpmov_lane(svuint64_t, uint64_t);
10682		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_s64)))
10683		svbool_t svpmov_lane(svint64_t, uint64_t);
10684		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_u16)))
10685		svbool_t svpmov_lane(svuint16_t, uint64_t);
10686		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_s16)))
10687		svbool_t svpmov_lane(svint16_t, uint64_t);
10688		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_u32)))
10689		svbool_t svpmov_lane(svuint32_t, uint64_t);
10690		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_s32)))
10691		svbool_t svpmov_lane(svint32_t, uint64_t);
10692		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_u64_m)))
10693		svuint64_t svpmov_lane_m(svuint64_t, svbool_t, uint64_t);
10694		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_s64_m)))
10695		svint64_t svpmov_lane_m(svint64_t, svbool_t, uint64_t);
10696		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_u16_m)))
10697		svuint16_t svpmov_lane_m(svuint16_t, svbool_t, uint64_t);
10698		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_s16_m)))
10699		svint16_t svpmov_lane_m(svint16_t, svbool_t, uint64_t);
10700		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_u32_m)))
10701		svuint32_t svpmov_lane_m(svuint32_t, svbool_t, uint64_t);
10702		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_s32_m)))
10703		svint32_t svpmov_lane_m(svint32_t, svbool_t, uint64_t);
10704		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1dq_u64)))
10705		void svst1dq(svbool_t, uint64_t *, svuint64_t);
10706		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1dq_f64)))
10707		void svst1dq(svbool_t, float64_t *, svfloat64_t);
10708		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1dq_s64)))
10709		void svst1dq(svbool_t, int64_t *, svint64_t);
10710		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1dq_vnum_u64)))
10711		void svst1dq_vnum(svbool_t, uint64_t *, int64_t, svuint64_t);
10712		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1dq_vnum_f64)))
10713		void svst1dq_vnum(svbool_t, float64_t *, int64_t, svfloat64_t);
10714		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1dq_vnum_s64)))
10715		void svst1dq_vnum(svbool_t, int64_t *, int64_t, svint64_t);
10716		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_u8)))
10717		void svst1q_scatter(svbool_t, svuint64_t, svuint8_t);
10718		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_u32)))
10719		void svst1q_scatter(svbool_t, svuint64_t, svuint32_t);
10720		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_u64)))
10721		void svst1q_scatter(svbool_t, svuint64_t, svuint64_t);
10722		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_u16)))
10723		void svst1q_scatter(svbool_t, svuint64_t, svuint16_t);
10724		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_bf16)))
10725		void svst1q_scatter(svbool_t, svuint64_t, svbfloat16_t);
10726		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_s8)))
10727		void svst1q_scatter(svbool_t, svuint64_t, svint8_t);
10728		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_f64)))
10729		void svst1q_scatter(svbool_t, svuint64_t, svfloat64_t);
10730		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_f32)))
10731		void svst1q_scatter(svbool_t, svuint64_t, svfloat32_t);
10732		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_f16)))
10733		void svst1q_scatter(svbool_t, svuint64_t, svfloat16_t);
10734		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_s32)))
10735		void svst1q_scatter(svbool_t, svuint64_t, svint32_t);
10736		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_s64)))
10737		void svst1q_scatter(svbool_t, svuint64_t, svint64_t);
10738		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_s16)))
10739		void svst1q_scatter(svbool_t, svuint64_t, svint16_t);
10740		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_u32)))
10741		void svst1q_scatter_index(svbool_t, svuint64_t, int64_t, svuint32_t);
10742		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_u64)))
10743		void svst1q_scatter_index(svbool_t, svuint64_t, int64_t, svuint64_t);
10744		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_u16)))
10745		void svst1q_scatter_index(svbool_t, svuint64_t, int64_t, svuint16_t);
10746		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_bf16)))
10747		void svst1q_scatter_index(svbool_t, svuint64_t, int64_t, svbfloat16_t);
10748		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_f64)))
10749		void svst1q_scatter_index(svbool_t, svuint64_t, int64_t, svfloat64_t);
10750		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_f32)))
10751		void svst1q_scatter_index(svbool_t, svuint64_t, int64_t, svfloat32_t);
10752		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_f16)))
10753		void svst1q_scatter_index(svbool_t, svuint64_t, int64_t, svfloat16_t);
10754		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_s32)))
10755		void svst1q_scatter_index(svbool_t, svuint64_t, int64_t, svint32_t);
10756		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_s64)))
10757		void svst1q_scatter_index(svbool_t, svuint64_t, int64_t, svint64_t);
10758		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_s16)))
10759		void svst1q_scatter_index(svbool_t, svuint64_t, int64_t, svint16_t);
10760		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_u8)))
10761		void svst1q_scatter_offset(svbool_t, svuint64_t, int64_t, svuint8_t);
10762		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_u32)))
10763		void svst1q_scatter_offset(svbool_t, svuint64_t, int64_t, svuint32_t);
10764		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_u64)))
10765		void svst1q_scatter_offset(svbool_t, svuint64_t, int64_t, svuint64_t);
10766		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_u16)))
10767		void svst1q_scatter_offset(svbool_t, svuint64_t, int64_t, svuint16_t);
10768		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_bf16)))
10769		void svst1q_scatter_offset(svbool_t, svuint64_t, int64_t, svbfloat16_t);
10770		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_s8)))
10771		void svst1q_scatter_offset(svbool_t, svuint64_t, int64_t, svint8_t);
10772		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_f64)))
10773		void svst1q_scatter_offset(svbool_t, svuint64_t, int64_t, svfloat64_t);
10774		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_f32)))
10775		void svst1q_scatter_offset(svbool_t, svuint64_t, int64_t, svfloat32_t);
10776		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_f16)))
10777		void svst1q_scatter_offset(svbool_t, svuint64_t, int64_t, svfloat16_t);
10778		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_s32)))
10779		void svst1q_scatter_offset(svbool_t, svuint64_t, int64_t, svint32_t);
10780		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_s64)))
10781		void svst1q_scatter_offset(svbool_t, svuint64_t, int64_t, svint64_t);
10782		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_s16)))
10783		void svst1q_scatter_offset(svbool_t, svuint64_t, int64_t, svint16_t);
10784		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_u32)))
10785		void svst1q_scatter_index(svbool_t, uint32_t *, svint64_t, svuint32_t);
10786		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_u64)))
10787		void svst1q_scatter_index(svbool_t, uint64_t *, svint64_t, svuint64_t);
10788		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_u16)))
10789		void svst1q_scatter_index(svbool_t, uint16_t *, svint64_t, svuint16_t);
10790		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_bf16)))
10791		void svst1q_scatter_index(svbool_t, bfloat16_t *, svint64_t, svbfloat16_t);
10792		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_f64)))
10793		void svst1q_scatter_index(svbool_t, float64_t *, svint64_t, svfloat64_t);
10794		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_f32)))
10795		void svst1q_scatter_index(svbool_t, float32_t *, svint64_t, svfloat32_t);
10796		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_f16)))
10797		void svst1q_scatter_index(svbool_t, float16_t *, svint64_t, svfloat16_t);
10798		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_s32)))
10799		void svst1q_scatter_index(svbool_t, int32_t *, svint64_t, svint32_t);
10800		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_s64)))
10801		void svst1q_scatter_index(svbool_t, int64_t *, svint64_t, svint64_t);
10802		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_s16)))
10803		void svst1q_scatter_index(svbool_t, int16_t *, svint64_t, svint16_t);
10804		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_u32)))
10805		void svst1q_scatter_index(svbool_t, uint32_t *, svuint64_t, svuint32_t);
10806		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_u64)))
10807		void svst1q_scatter_index(svbool_t, uint64_t *, svuint64_t, svuint64_t);
10808		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_u16)))
10809		void svst1q_scatter_index(svbool_t, uint16_t *, svuint64_t, svuint16_t);
10810		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_bf16)))
10811		void svst1q_scatter_index(svbool_t, bfloat16_t *, svuint64_t, svbfloat16_t);
10812		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_f64)))
10813		void svst1q_scatter_index(svbool_t, float64_t *, svuint64_t, svfloat64_t);
10814		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_f32)))
10815		void svst1q_scatter_index(svbool_t, float32_t *, svuint64_t, svfloat32_t);
10816		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_f16)))
10817		void svst1q_scatter_index(svbool_t, float16_t *, svuint64_t, svfloat16_t);
10818		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_s32)))
10819		void svst1q_scatter_index(svbool_t, int32_t *, svuint64_t, svint32_t);
10820		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_s64)))
10821		void svst1q_scatter_index(svbool_t, int64_t *, svuint64_t, svint64_t);
10822		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_s16)))
10823		void svst1q_scatter_index(svbool_t, int16_t *, svuint64_t, svint16_t);
10824		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_u8)))
10825		void svst1q_scatter_offset(svbool_t, uint8_t *, svint64_t, svuint8_t);
10826		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_u32)))
10827		void svst1q_scatter_offset(svbool_t, uint32_t *, svint64_t, svuint32_t);
10828		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_u64)))
10829		void svst1q_scatter_offset(svbool_t, uint64_t *, svint64_t, svuint64_t);
10830		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_u16)))
10831		void svst1q_scatter_offset(svbool_t, uint16_t *, svint64_t, svuint16_t);
10832		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_bf16)))
10833		void svst1q_scatter_offset(svbool_t, bfloat16_t *, svint64_t, svbfloat16_t);
10834		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_s8)))
10835		void svst1q_scatter_offset(svbool_t, int8_t *, svint64_t, svint8_t);
10836		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_f64)))
10837		void svst1q_scatter_offset(svbool_t, float64_t *, svint64_t, svfloat64_t);
10838		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_f32)))
10839		void svst1q_scatter_offset(svbool_t, float32_t *, svint64_t, svfloat32_t);
10840		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_f16)))
10841		void svst1q_scatter_offset(svbool_t, float16_t *, svint64_t, svfloat16_t);
10842		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_s32)))
10843		void svst1q_scatter_offset(svbool_t, int32_t *, svint64_t, svint32_t);
10844		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_s64)))
10845		void svst1q_scatter_offset(svbool_t, int64_t *, svint64_t, svint64_t);
10846		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_s16)))
10847		void svst1q_scatter_offset(svbool_t, int16_t *, svint64_t, svint16_t);
10848		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_u8)))
10849		void svst1q_scatter_offset(svbool_t, uint8_t *, svuint64_t, svuint8_t);
10850		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_u32)))
10851		void svst1q_scatter_offset(svbool_t, uint32_t *, svuint64_t, svuint32_t);
10852		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_u64)))
10853		void svst1q_scatter_offset(svbool_t, uint64_t *, svuint64_t, svuint64_t);
10854		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_u16)))
10855		void svst1q_scatter_offset(svbool_t, uint16_t *, svuint64_t, svuint16_t);
10856		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_bf16)))
10857		void svst1q_scatter_offset(svbool_t, bfloat16_t *, svuint64_t, svbfloat16_t);
10858		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_s8)))
10859		void svst1q_scatter_offset(svbool_t, int8_t *, svuint64_t, svint8_t);
10860		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_f64)))
10861		void svst1q_scatter_offset(svbool_t, float64_t *, svuint64_t, svfloat64_t);
10862		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_f32)))
10863		void svst1q_scatter_offset(svbool_t, float32_t *, svuint64_t, svfloat32_t);
10864		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_f16)))
10865		void svst1q_scatter_offset(svbool_t, float16_t *, svuint64_t, svfloat16_t);
10866		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_s32)))
10867		void svst1q_scatter_offset(svbool_t, int32_t *, svuint64_t, svint32_t);
10868		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_s64)))
10869		void svst1q_scatter_offset(svbool_t, int64_t *, svuint64_t, svint64_t);
10870		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_s16)))
10871		void svst1q_scatter_offset(svbool_t, int16_t *, svuint64_t, svint16_t);
10872		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1wq_u32)))
10873		void svst1wq(svbool_t, uint32_t *, svuint32_t);
10874		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1wq_f32)))
10875		void svst1wq(svbool_t, float32_t *, svfloat32_t);
10876		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1wq_s32)))
10877		void svst1wq(svbool_t, int32_t *, svint32_t);
10878		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1wq_vnum_u32)))
10879		void svst1wq_vnum(svbool_t, uint32_t *, int64_t, svuint32_t);
10880		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1wq_vnum_f32)))
10881		void svst1wq_vnum(svbool_t, float32_t *, int64_t, svfloat32_t);
10882		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1wq_vnum_s32)))
10883		void svst1wq_vnum(svbool_t, int32_t *, int64_t, svint32_t);
10884		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_u8)))
10885		void svst2q(svbool_t, uint8_t const *, svuint8x2_t);
10886		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_u32)))
10887		void svst2q(svbool_t, uint32_t const *, svuint32x2_t);
10888		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_u64)))
10889		void svst2q(svbool_t, uint64_t const *, svuint64x2_t);
10890		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_u16)))
10891		void svst2q(svbool_t, uint16_t const *, svuint16x2_t);
10892		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_s8)))
10893		void svst2q(svbool_t, int8_t const *, svint8x2_t);
10894		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_f64)))
10895		void svst2q(svbool_t, float64_t const *, svfloat64x2_t);
10896		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_f32)))
10897		void svst2q(svbool_t, float32_t const *, svfloat32x2_t);
10898		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_f16)))
10899		void svst2q(svbool_t, float16_t const *, svfloat16x2_t);
10900		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_s32)))
10901		void svst2q(svbool_t, int32_t const *, svint32x2_t);
10902		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_s64)))
10903		void svst2q(svbool_t, int64_t const *, svint64x2_t);
10904		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_s16)))
10905		void svst2q(svbool_t, int16_t const *, svint16x2_t);
10906		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_bf16)))
10907		void svst2q(svbool_t, bfloat16_t const *, svbfloat16x2_t);
10908		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_u8)))
10909		void svst2q_vnum(svbool_t, uint8_t const *, int64_t, svuint8x2_t);
10910		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_u32)))
10911		void svst2q_vnum(svbool_t, uint32_t const *, int64_t, svuint32x2_t);
10912		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_u64)))
10913		void svst2q_vnum(svbool_t, uint64_t const *, int64_t, svuint64x2_t);
10914		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_u16)))
10915		void svst2q_vnum(svbool_t, uint16_t const *, int64_t, svuint16x2_t);
10916		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_s8)))
10917		void svst2q_vnum(svbool_t, int8_t const *, int64_t, svint8x2_t);
10918		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_f64)))
10919		void svst2q_vnum(svbool_t, float64_t const *, int64_t, svfloat64x2_t);
10920		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_f32)))
10921		void svst2q_vnum(svbool_t, float32_t const *, int64_t, svfloat32x2_t);
10922		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_f16)))
10923		void svst2q_vnum(svbool_t, float16_t const *, int64_t, svfloat16x2_t);
10924		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_s32)))
10925		void svst2q_vnum(svbool_t, int32_t const *, int64_t, svint32x2_t);
10926		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_s64)))
10927		void svst2q_vnum(svbool_t, int64_t const *, int64_t, svint64x2_t);
10928		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_s16)))
10929		void svst2q_vnum(svbool_t, int16_t const *, int64_t, svint16x2_t);
10930		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_bf16)))
10931		void svst2q_vnum(svbool_t, bfloat16_t const *, int64_t, svbfloat16x2_t);
10932		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_u8)))
10933		void svst3q(svbool_t, uint8_t const *, svuint8x3_t);
10934		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_u32)))
10935		void svst3q(svbool_t, uint32_t const *, svuint32x3_t);
10936		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_u64)))
10937		void svst3q(svbool_t, uint64_t const *, svuint64x3_t);
10938		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_u16)))
10939		void svst3q(svbool_t, uint16_t const *, svuint16x3_t);
10940		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_s8)))
10941		void svst3q(svbool_t, int8_t const *, svint8x3_t);
10942		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_f64)))
10943		void svst3q(svbool_t, float64_t const *, svfloat64x3_t);
10944		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_f32)))
10945		void svst3q(svbool_t, float32_t const *, svfloat32x3_t);
10946		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_f16)))
10947		void svst3q(svbool_t, float16_t const *, svfloat16x3_t);
10948		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_s32)))
10949		void svst3q(svbool_t, int32_t const *, svint32x3_t);
10950		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_s64)))
10951		void svst3q(svbool_t, int64_t const *, svint64x3_t);
10952		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_s16)))
10953		void svst3q(svbool_t, int16_t const *, svint16x3_t);
10954		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_bf16)))
10955		void svst3q(svbool_t, bfloat16_t const *, svbfloat16x3_t);
10956		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_u8)))
10957		void svst3q_vnum(svbool_t, uint8_t const *, int64_t, svuint8x3_t);
10958		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_u32)))
10959		void svst3q_vnum(svbool_t, uint32_t const *, int64_t, svuint32x3_t);
10960		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_u64)))
10961		void svst3q_vnum(svbool_t, uint64_t const *, int64_t, svuint64x3_t);
10962		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_u16)))
10963		void svst3q_vnum(svbool_t, uint16_t const *, int64_t, svuint16x3_t);
10964		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_s8)))
10965		void svst3q_vnum(svbool_t, int8_t const *, int64_t, svint8x3_t);
10966		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_f64)))
10967		void svst3q_vnum(svbool_t, float64_t const *, int64_t, svfloat64x3_t);
10968		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_f32)))
10969		void svst3q_vnum(svbool_t, float32_t const *, int64_t, svfloat32x3_t);
10970		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_f16)))
10971		void svst3q_vnum(svbool_t, float16_t const *, int64_t, svfloat16x3_t);
10972		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_s32)))
10973		void svst3q_vnum(svbool_t, int32_t const *, int64_t, svint32x3_t);
10974		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_s64)))
10975		void svst3q_vnum(svbool_t, int64_t const *, int64_t, svint64x3_t);
10976		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_s16)))
10977		void svst3q_vnum(svbool_t, int16_t const *, int64_t, svint16x3_t);
10978		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_bf16)))
10979		void svst3q_vnum(svbool_t, bfloat16_t const *, int64_t, svbfloat16x3_t);
10980		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_u8)))
10981		void svst4q(svbool_t, uint8_t const *, svuint8x4_t);
10982		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_u32)))
10983		void svst4q(svbool_t, uint32_t const *, svuint32x4_t);
10984		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_u64)))
10985		void svst4q(svbool_t, uint64_t const *, svuint64x4_t);
10986		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_u16)))
10987		void svst4q(svbool_t, uint16_t const *, svuint16x4_t);
10988		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_s8)))
10989		void svst4q(svbool_t, int8_t const *, svint8x4_t);
10990		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_f64)))
10991		void svst4q(svbool_t, float64_t const *, svfloat64x4_t);
10992		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_f32)))
10993		void svst4q(svbool_t, float32_t const *, svfloat32x4_t);
10994		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_f16)))
10995		void svst4q(svbool_t, float16_t const *, svfloat16x4_t);
10996		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_s32)))
10997		void svst4q(svbool_t, int32_t const *, svint32x4_t);
10998		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_s64)))
10999		void svst4q(svbool_t, int64_t const *, svint64x4_t);
11000		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_s16)))
11001		void svst4q(svbool_t, int16_t const *, svint16x4_t);
11002		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_bf16)))
11003		void svst4q(svbool_t, bfloat16_t const *, svbfloat16x4_t);
11004		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_u8)))
11005		void svst4q_vnum(svbool_t, uint8_t const *, int64_t, svuint8x4_t);
11006		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_u32)))
11007		void svst4q_vnum(svbool_t, uint32_t const *, int64_t, svuint32x4_t);
11008		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_u64)))
11009		void svst4q_vnum(svbool_t, uint64_t const *, int64_t, svuint64x4_t);
11010		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_u16)))
11011		void svst4q_vnum(svbool_t, uint16_t const *, int64_t, svuint16x4_t);
11012		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_s8)))
11013		void svst4q_vnum(svbool_t, int8_t const *, int64_t, svint8x4_t);
11014		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_f64)))
11015		void svst4q_vnum(svbool_t, float64_t const *, int64_t, svfloat64x4_t);
11016		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_f32)))
11017		void svst4q_vnum(svbool_t, float32_t const *, int64_t, svfloat32x4_t);
11018		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_f16)))
11019		void svst4q_vnum(svbool_t, float16_t const *, int64_t, svfloat16x4_t);
11020		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_s32)))
11021		void svst4q_vnum(svbool_t, int32_t const *, int64_t, svint32x4_t);
11022		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_s64)))
11023		void svst4q_vnum(svbool_t, int64_t const *, int64_t, svint64x4_t);
11024		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_s16)))
11025		void svst4q_vnum(svbool_t, int16_t const *, int64_t, svint16x4_t);
11026		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_bf16)))
11027		void svst4q_vnum(svbool_t, bfloat16_t const *, int64_t, svbfloat16x4_t);
11028		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_u8)))
11029		svuint8_t svtblq(svuint8_t, svuint8_t);
11030		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_u32)))
11031		svuint32_t svtblq(svuint32_t, svuint32_t);
11032		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_u64)))
11033		svuint64_t svtblq(svuint64_t, svuint64_t);
11034		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_u16)))
11035		svuint16_t svtblq(svuint16_t, svuint16_t);
11036		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_bf16)))
11037		svbfloat16_t svtblq(svbfloat16_t, svuint16_t);
11038		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_s8)))
11039		svint8_t svtblq(svint8_t, svuint8_t);
11040		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_f64)))
11041		svfloat64_t svtblq(svfloat64_t, svuint64_t);
11042		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_f32)))
11043		svfloat32_t svtblq(svfloat32_t, svuint32_t);
11044		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_f16)))
11045		svfloat16_t svtblq(svfloat16_t, svuint16_t);
11046		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_s32)))
11047		svint32_t svtblq(svint32_t, svuint32_t);
11048		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_s64)))
11049		svint64_t svtblq(svint64_t, svuint64_t);
11050		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_mf8)))
11051		svmfloat8_t svtblq(svmfloat8_t, svuint8_t);
11052		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_s16)))
11053		svint16_t svtblq(svint16_t, svuint16_t);
11054		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_u8)))
11055		svuint8_t svtbxq(svuint8_t, svuint8_t, svuint8_t);
11056		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_u32)))
11057		svuint32_t svtbxq(svuint32_t, svuint32_t, svuint32_t);
11058		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_u64)))
11059		svuint64_t svtbxq(svuint64_t, svuint64_t, svuint64_t);
11060		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_u16)))
11061		svuint16_t svtbxq(svuint16_t, svuint16_t, svuint16_t);
11062		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_bf16)))
11063		svbfloat16_t svtbxq(svbfloat16_t, svbfloat16_t, svuint16_t);
11064		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_s8)))
11065		svint8_t svtbxq(svint8_t, svint8_t, svuint8_t);
11066		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_f64)))
11067		svfloat64_t svtbxq(svfloat64_t, svfloat64_t, svuint64_t);
11068		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_f32)))
11069		svfloat32_t svtbxq(svfloat32_t, svfloat32_t, svuint32_t);
11070		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_f16)))
11071		svfloat16_t svtbxq(svfloat16_t, svfloat16_t, svuint16_t);
11072		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_s32)))
11073		svint32_t svtbxq(svint32_t, svint32_t, svuint32_t);
11074		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_s64)))
11075		svint64_t svtbxq(svint64_t, svint64_t, svuint64_t);
11076		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_mf8)))
11077		svmfloat8_t svtbxq(svmfloat8_t, svmfloat8_t, svuint8_t);
11078		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_s16)))
11079		svint16_t svtbxq(svint16_t, svint16_t, svuint16_t);
11080		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_u8)))
11081		svuint8_t svuzpq1(svuint8_t, svuint8_t);
11082		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_u32)))
11083		svuint32_t svuzpq1(svuint32_t, svuint32_t);
11084		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_u64)))
11085		svuint64_t svuzpq1(svuint64_t, svuint64_t);
11086		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_u16)))
11087		svuint16_t svuzpq1(svuint16_t, svuint16_t);
11088		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_bf16)))
11089		svbfloat16_t svuzpq1(svbfloat16_t, svbfloat16_t);
11090		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_s8)))
11091		svint8_t svuzpq1(svint8_t, svint8_t);
11092		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_f64)))
11093		svfloat64_t svuzpq1(svfloat64_t, svfloat64_t);
11094		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_f32)))
11095		svfloat32_t svuzpq1(svfloat32_t, svfloat32_t);
11096		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_f16)))
11097		svfloat16_t svuzpq1(svfloat16_t, svfloat16_t);
11098		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_s32)))
11099		svint32_t svuzpq1(svint32_t, svint32_t);
11100		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_s64)))
11101		svint64_t svuzpq1(svint64_t, svint64_t);
11102		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_mf8)))
11103		svmfloat8_t svuzpq1(svmfloat8_t, svmfloat8_t);
11104		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_s16)))
11105		svint16_t svuzpq1(svint16_t, svint16_t);
11106		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_u8)))
11107		svuint8_t svuzpq2(svuint8_t, svuint8_t);
11108		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_u32)))
11109		svuint32_t svuzpq2(svuint32_t, svuint32_t);
11110		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_u64)))
11111		svuint64_t svuzpq2(svuint64_t, svuint64_t);
11112		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_u16)))
11113		svuint16_t svuzpq2(svuint16_t, svuint16_t);
11114		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_bf16)))
11115		svbfloat16_t svuzpq2(svbfloat16_t, svbfloat16_t);
11116		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_s8)))
11117		svint8_t svuzpq2(svint8_t, svint8_t);
11118		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_f64)))
11119		svfloat64_t svuzpq2(svfloat64_t, svfloat64_t);
11120		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_f32)))
11121		svfloat32_t svuzpq2(svfloat32_t, svfloat32_t);
11122		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_f16)))
11123		svfloat16_t svuzpq2(svfloat16_t, svfloat16_t);
11124		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_s32)))
11125		svint32_t svuzpq2(svint32_t, svint32_t);
11126		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_s64)))
11127		svint64_t svuzpq2(svint64_t, svint64_t);
11128		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_mf8)))
11129		svmfloat8_t svuzpq2(svmfloat8_t, svmfloat8_t);
11130		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_s16)))
11131		svint16_t svuzpq2(svint16_t, svint16_t);
11132		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_u8)))
11133		svuint8_t svzipq1(svuint8_t, svuint8_t);
11134		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_u32)))
11135		svuint32_t svzipq1(svuint32_t, svuint32_t);
11136		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_u64)))
11137		svuint64_t svzipq1(svuint64_t, svuint64_t);
11138		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_u16)))
11139		svuint16_t svzipq1(svuint16_t, svuint16_t);
11140		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_bf16)))
11141		svbfloat16_t svzipq1(svbfloat16_t, svbfloat16_t);
11142		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_s8)))
11143		svint8_t svzipq1(svint8_t, svint8_t);
11144		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_f64)))
11145		svfloat64_t svzipq1(svfloat64_t, svfloat64_t);
11146		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_f32)))
11147		svfloat32_t svzipq1(svfloat32_t, svfloat32_t);
11148		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_f16)))
11149		svfloat16_t svzipq1(svfloat16_t, svfloat16_t);
11150		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_s32)))
11151		svint32_t svzipq1(svint32_t, svint32_t);
11152		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_s64)))
11153		svint64_t svzipq1(svint64_t, svint64_t);
11154		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_mf8)))
11155		svmfloat8_t svzipq1(svmfloat8_t, svmfloat8_t);
11156		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_s16)))
11157		svint16_t svzipq1(svint16_t, svint16_t);
11158		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_u8)))
11159		svuint8_t svzipq2(svuint8_t, svuint8_t);
11160		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_u32)))
11161		svuint32_t svzipq2(svuint32_t, svuint32_t);
11162		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_u64)))
11163		svuint64_t svzipq2(svuint64_t, svuint64_t);
11164		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_u16)))
11165		svuint16_t svzipq2(svuint16_t, svuint16_t);
11166		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_bf16)))
11167		svbfloat16_t svzipq2(svbfloat16_t, svbfloat16_t);
11168		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_s8)))
11169		svint8_t svzipq2(svint8_t, svint8_t);
11170		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_f64)))
11171		svfloat64_t svzipq2(svfloat64_t, svfloat64_t);
11172		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_f32)))
11173		svfloat32_t svzipq2(svfloat32_t, svfloat32_t);
11174		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_f16)))
11175		svfloat16_t svzipq2(svfloat16_t, svfloat16_t);
11176		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_s32)))
11177		svint32_t svzipq2(svint32_t, svint32_t);
11178		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_s64)))
11179		svint64_t svzipq2(svint64_t, svint64_t);
11180		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_mf8)))
11181		svmfloat8_t svzipq2(svmfloat8_t, svmfloat8_t);
11182		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_s16)))
11183		svint16_t svzipq2(svint16_t, svint16_t);
	6956	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_s8)))
	6957	svint8_t svclamp_s8(svint8_t, svint8_t, svint8_t);
	6958	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_s32)))
	6959	svint32_t svclamp_s32(svint32_t, svint32_t, svint32_t);
	6960	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_s64)))
	6961	svint64_t svclamp_s64(svint64_t, svint64_t, svint64_t);
	6962	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_s16)))
	6963	svint16_t svclamp_s16(svint16_t, svint16_t, svint16_t);
	6964	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_u8)))
	6965	svuint8_t svclamp_u8(svuint8_t, svuint8_t, svuint8_t);
	6966	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_u32)))
	6967	svuint32_t svclamp_u32(svuint32_t, svuint32_t, svuint32_t);
	6968	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_u64)))
	6969	svuint64_t svclamp_u64(svuint64_t, svuint64_t, svuint64_t);
	6970	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_u16)))
	6971	svuint16_t svclamp_u16(svuint16_t, svuint16_t, svuint16_t);
	6972	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpsel_lane_b16)))
	6973	svbool_t svpsel_lane_b16(svbool_t, svbool_t, uint32_t);
	6974	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpsel_lane_b32)))
	6975	svbool_t svpsel_lane_b32(svbool_t, svbool_t, uint32_t);
	6976	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpsel_lane_b64)))
	6977	svbool_t svpsel_lane_b64(svbool_t, svbool_t, uint32_t);
	6978	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpsel_lane_b8)))
	6979	svbool_t svpsel_lane_b8(svbool_t, svbool_t, uint32_t);
	6980	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u8_m)))
	6981	svuint8_t svrevd_u8_m(svuint8_t, svbool_t, svuint8_t);
	6982	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u32_m)))
	6983	svuint32_t svrevd_u32_m(svuint32_t, svbool_t, svuint32_t);
	6984	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u64_m)))
	6985	svuint64_t svrevd_u64_m(svuint64_t, svbool_t, svuint64_t);
	6986	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u16_m)))
	6987	svuint16_t svrevd_u16_m(svuint16_t, svbool_t, svuint16_t);
	6988	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_bf16_m)))
	6989	svbfloat16_t svrevd_bf16_m(svbfloat16_t, svbool_t, svbfloat16_t);
	6990	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s8_m)))
	6991	svint8_t svrevd_s8_m(svint8_t, svbool_t, svint8_t);
	6992	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f64_m)))
	6993	svfloat64_t svrevd_f64_m(svfloat64_t, svbool_t, svfloat64_t);
	6994	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f32_m)))
	6995	svfloat32_t svrevd_f32_m(svfloat32_t, svbool_t, svfloat32_t);
	6996	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f16_m)))
	6997	svfloat16_t svrevd_f16_m(svfloat16_t, svbool_t, svfloat16_t);
	6998	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s32_m)))
	6999	svint32_t svrevd_s32_m(svint32_t, svbool_t, svint32_t);
	7000	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s64_m)))
	7001	svint64_t svrevd_s64_m(svint64_t, svbool_t, svint64_t);
	7002	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_mf8_m)))
	7003	svmfloat8_t svrevd_mf8_m(svmfloat8_t, svbool_t, svmfloat8_t);
	7004	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s16_m)))
	7005	svint16_t svrevd_s16_m(svint16_t, svbool_t, svint16_t);
	7006	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u8_x)))
	7007	svuint8_t svrevd_u8_x(svbool_t, svuint8_t);
	7008	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u32_x)))
	7009	svuint32_t svrevd_u32_x(svbool_t, svuint32_t);
	7010	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u64_x)))
	7011	svuint64_t svrevd_u64_x(svbool_t, svuint64_t);
	7012	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u16_x)))
	7013	svuint16_t svrevd_u16_x(svbool_t, svuint16_t);
	7014	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_bf16_x)))
	7015	svbfloat16_t svrevd_bf16_x(svbool_t, svbfloat16_t);
	7016	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s8_x)))
	7017	svint8_t svrevd_s8_x(svbool_t, svint8_t);
	7018	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f64_x)))
	7019	svfloat64_t svrevd_f64_x(svbool_t, svfloat64_t);
	7020	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f32_x)))
	7021	svfloat32_t svrevd_f32_x(svbool_t, svfloat32_t);
	7022	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f16_x)))
	7023	svfloat16_t svrevd_f16_x(svbool_t, svfloat16_t);
	7024	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s32_x)))
	7025	svint32_t svrevd_s32_x(svbool_t, svint32_t);
	7026	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s64_x)))
	7027	svint64_t svrevd_s64_x(svbool_t, svint64_t);
	7028	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_mf8_x)))
	7029	svmfloat8_t svrevd_mf8_x(svbool_t, svmfloat8_t);
	7030	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s16_x)))
	7031	svint16_t svrevd_s16_x(svbool_t, svint16_t);
	7032	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u8_z)))
	7033	svuint8_t svrevd_u8_z(svbool_t, svuint8_t);
	7034	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u32_z)))
	7035	svuint32_t svrevd_u32_z(svbool_t, svuint32_t);
	7036	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u64_z)))
	7037	svuint64_t svrevd_u64_z(svbool_t, svuint64_t);
	7038	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u16_z)))
	7039	svuint16_t svrevd_u16_z(svbool_t, svuint16_t);
	7040	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_bf16_z)))
	7041	svbfloat16_t svrevd_bf16_z(svbool_t, svbfloat16_t);
	7042	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s8_z)))
	7043	svint8_t svrevd_s8_z(svbool_t, svint8_t);
	7044	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f64_z)))
	7045	svfloat64_t svrevd_f64_z(svbool_t, svfloat64_t);
	7046	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f32_z)))
	7047	svfloat32_t svrevd_f32_z(svbool_t, svfloat32_t);
	7048	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f16_z)))
	7049	svfloat16_t svrevd_f16_z(svbool_t, svfloat16_t);
	7050	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s32_z)))
	7051	svint32_t svrevd_s32_z(svbool_t, svint32_t);
	7052	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s64_z)))
	7053	svint64_t svrevd_s64_z(svbool_t, svint64_t);
	7054	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_mf8_z)))
	7055	svmfloat8_t svrevd_mf8_z(svbool_t, svmfloat8_t);
	7056	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s16_z)))
	7057	svint16_t svrevd_s16_z(svbool_t, svint16_t);
	7058	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_s8)))
	7059	svint8_t svclamp(svint8_t, svint8_t, svint8_t);
	7060	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_s32)))
	7061	svint32_t svclamp(svint32_t, svint32_t, svint32_t);
	7062	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_s64)))
	7063	svint64_t svclamp(svint64_t, svint64_t, svint64_t);
	7064	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_s16)))
	7065	svint16_t svclamp(svint16_t, svint16_t, svint16_t);
	7066	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_u8)))
	7067	svuint8_t svclamp(svuint8_t, svuint8_t, svuint8_t);
	7068	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_u32)))
	7069	svuint32_t svclamp(svuint32_t, svuint32_t, svuint32_t);
	7070	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_u64)))
	7071	svuint64_t svclamp(svuint64_t, svuint64_t, svuint64_t);
	7072	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_u16)))
	7073	svuint16_t svclamp(svuint16_t, svuint16_t, svuint16_t);
	7074	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u8_m)))
	7075	svuint8_t svrevd_m(svuint8_t, svbool_t, svuint8_t);
	7076	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u32_m)))
	7077	svuint32_t svrevd_m(svuint32_t, svbool_t, svuint32_t);
	7078	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u64_m)))
	7079	svuint64_t svrevd_m(svuint64_t, svbool_t, svuint64_t);
	7080	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u16_m)))
	7081	svuint16_t svrevd_m(svuint16_t, svbool_t, svuint16_t);
	7082	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_bf16_m)))
	7083	svbfloat16_t svrevd_m(svbfloat16_t, svbool_t, svbfloat16_t);
	7084	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s8_m)))
	7085	svint8_t svrevd_m(svint8_t, svbool_t, svint8_t);
	7086	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f64_m)))
	7087	svfloat64_t svrevd_m(svfloat64_t, svbool_t, svfloat64_t);
	7088	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f32_m)))
	7089	svfloat32_t svrevd_m(svfloat32_t, svbool_t, svfloat32_t);
	7090	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f16_m)))
	7091	svfloat16_t svrevd_m(svfloat16_t, svbool_t, svfloat16_t);
	7092	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s32_m)))
	7093	svint32_t svrevd_m(svint32_t, svbool_t, svint32_t);
	7094	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s64_m)))
	7095	svint64_t svrevd_m(svint64_t, svbool_t, svint64_t);
	7096	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_mf8_m)))
	7097	svmfloat8_t svrevd_m(svmfloat8_t, svbool_t, svmfloat8_t);
	7098	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s16_m)))
	7099	svint16_t svrevd_m(svint16_t, svbool_t, svint16_t);
	7100	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u8_x)))
	7101	svuint8_t svrevd_x(svbool_t, svuint8_t);
	7102	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u32_x)))
	7103	svuint32_t svrevd_x(svbool_t, svuint32_t);
	7104	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u64_x)))
	7105	svuint64_t svrevd_x(svbool_t, svuint64_t);
	7106	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u16_x)))
	7107	svuint16_t svrevd_x(svbool_t, svuint16_t);
	7108	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_bf16_x)))
	7109	svbfloat16_t svrevd_x(svbool_t, svbfloat16_t);
	7110	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s8_x)))
	7111	svint8_t svrevd_x(svbool_t, svint8_t);
	7112	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f64_x)))
	7113	svfloat64_t svrevd_x(svbool_t, svfloat64_t);
	7114	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f32_x)))
	7115	svfloat32_t svrevd_x(svbool_t, svfloat32_t);
	7116	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f16_x)))
	7117	svfloat16_t svrevd_x(svbool_t, svfloat16_t);
	7118	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s32_x)))
	7119	svint32_t svrevd_x(svbool_t, svint32_t);
	7120	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s64_x)))
	7121	svint64_t svrevd_x(svbool_t, svint64_t);
	7122	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_mf8_x)))
	7123	svmfloat8_t svrevd_x(svbool_t, svmfloat8_t);
	7124	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s16_x)))
	7125	svint16_t svrevd_x(svbool_t, svint16_t);
	7126	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u8_z)))
	7127	svuint8_t svrevd_z(svbool_t, svuint8_t);
	7128	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u32_z)))
	7129	svuint32_t svrevd_z(svbool_t, svuint32_t);
	7130	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u64_z)))
	7131	svuint64_t svrevd_z(svbool_t, svuint64_t);
	7132	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u16_z)))
	7133	svuint16_t svrevd_z(svbool_t, svuint16_t);
	7134	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_bf16_z)))
	7135	svbfloat16_t svrevd_z(svbool_t, svbfloat16_t);
	7136	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s8_z)))
	7137	svint8_t svrevd_z(svbool_t, svint8_t);
	7138	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f64_z)))
	7139	svfloat64_t svrevd_z(svbool_t, svfloat64_t);
	7140	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f32_z)))
	7141	svfloat32_t svrevd_z(svbool_t, svfloat32_t);
	7142	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f16_z)))
	7143	svfloat16_t svrevd_z(svbool_t, svfloat16_t);
	7144	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s32_z)))
	7145	svint32_t svrevd_z(svbool_t, svint32_t);
	7146	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s64_z)))
	7147	svint64_t svrevd_z(svbool_t, svint64_t);
	7148	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_mf8_z)))
	7149	svmfloat8_t svrevd_z(svbool_t, svmfloat8_t);
	7150	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s16_z)))
	7151	svint16_t svrevd_z(svbool_t, svint16_t);
	7152	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlslb_f32)))
	7153	svfloat32_t svbfmlslb_f32(svfloat32_t, svbfloat16_t, svbfloat16_t);
	7154	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlslb_lane_f32)))
	7155	svfloat32_t svbfmlslb_lane_f32(svfloat32_t, svbfloat16_t, svbfloat16_t, uint64_t);
	7156	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlslt_f32)))
	7157	svfloat32_t svbfmlslt_f32(svfloat32_t, svbfloat16_t, svbfloat16_t);
	7158	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlslt_lane_f32)))
	7159	svfloat32_t svbfmlslt_lane_f32(svfloat32_t, svbfloat16_t, svbfloat16_t, uint64_t);
	7160	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_f64)))
	7161	svfloat64_t svclamp_f64(svfloat64_t, svfloat64_t, svfloat64_t);
	7162	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_f32)))
	7163	svfloat32_t svclamp_f32(svfloat32_t, svfloat32_t, svfloat32_t);
	7164	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_f16)))
	7165	svfloat16_t svclamp_f16(svfloat16_t, svfloat16_t, svfloat16_t);
	7166	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate2_b)))
	7167	svboolx2_t svcreate2_b(svbool_t, svbool_t);
	7168	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate4_b)))
	7169	svboolx4_t svcreate4_b(svbool_t, svbool_t, svbool_t, svbool_t);
	7170	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_f32_f16)))
	7171	svfloat32_t svdot_f32_f16(svfloat32_t, svfloat16_t, svfloat16_t);
	7172	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_s32_s16)))
	7173	svint32_t svdot_s32_s16(svint32_t, svint16_t, svint16_t);
	7174	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_u32_u16)))
	7175	svuint32_t svdot_u32_u16(svuint32_t, svuint16_t, svuint16_t);
	7176	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_lane_f32_f16)))
	7177	svfloat32_t svdot_lane_f32_f16(svfloat32_t, svfloat16_t, svfloat16_t, uint64_t);
	7178	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_lane_s32_s16)))
	7179	svint32_t svdot_lane_s32_s16(svint32_t, svint16_t, svint16_t, uint64_t);
	7180	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_lane_u32_u16)))
	7181	svuint32_t svdot_lane_u32_u16(svuint32_t, svuint16_t, svuint16_t, uint64_t);
	7182	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget2_b)))
	7183	svbool_t svget2_b(svboolx2_t, uint64_t);
	7184	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget4_b)))
	7185	svbool_t svget4_b(svboolx4_t, uint64_t);
	7186	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svqcvtn_s16_s32_x2)))
	7187	svint16_t svqcvtn_s16_s32_x2(svint32x2_t);
	7188	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svqcvtn_u16_s32_x2)))
	7189	svuint16_t svqcvtn_u16_s32_x2(svint32x2_t);
	7190	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svqcvtn_u16_u32_x2)))
	7191	svuint16_t svqcvtn_u16_u32_x2(svuint32x2_t);
	7192	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svqrshrn_n_s16_s32_x2)))
	7193	svint16_t svqrshrn_n_s16_s32_x2(svint32x2_t, uint64_t);
	7194	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svqrshrn_n_u16_u32_x2)))
	7195	svuint16_t svqrshrn_n_u16_u32_x2(svuint32x2_t, uint64_t);
	7196	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svqrshrun_n_u16_s32_x2)))
	7197	svuint16_t svqrshrun_n_u16_s32_x2(svint32x2_t, uint64_t);
	7198	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset2_b)))
	7199	svboolx2_t svset2_b(svboolx2_t, uint64_t, svbool_t);
	7200	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset4_b)))
	7201	svboolx4_t svset4_b(svboolx4_t, uint64_t, svbool_t);
	7202	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef2_b)))
	7203	svboolx2_t svundef2_b();
	7204	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef4_b)))
	7205	svboolx4_t svundef4_b();
	7206	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b8_s64_x2)))
	7207	svboolx2_t svwhilege_b8_s64_x2(int64_t, int64_t);
	7208	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b32_s64_x2)))
	7209	svboolx2_t svwhilege_b32_s64_x2(int64_t, int64_t);
	7210	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b64_s64_x2)))
	7211	svboolx2_t svwhilege_b64_s64_x2(int64_t, int64_t);
	7212	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b16_s64_x2)))
	7213	svboolx2_t svwhilege_b16_s64_x2(int64_t, int64_t);
	7214	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b8_u64_x2)))
	7215	svboolx2_t svwhilege_b8_u64_x2(uint64_t, uint64_t);
	7216	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b32_u64_x2)))
	7217	svboolx2_t svwhilege_b32_u64_x2(uint64_t, uint64_t);
	7218	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b64_u64_x2)))
	7219	svboolx2_t svwhilege_b64_u64_x2(uint64_t, uint64_t);
	7220	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b16_u64_x2)))
	7221	svboolx2_t svwhilege_b16_u64_x2(uint64_t, uint64_t);
	7222	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b8_s64_x2)))
	7223	svboolx2_t svwhilegt_b8_s64_x2(int64_t, int64_t);
	7224	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b32_s64_x2)))
	7225	svboolx2_t svwhilegt_b32_s64_x2(int64_t, int64_t);
	7226	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b64_s64_x2)))
	7227	svboolx2_t svwhilegt_b64_s64_x2(int64_t, int64_t);
	7228	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b16_s64_x2)))
	7229	svboolx2_t svwhilegt_b16_s64_x2(int64_t, int64_t);
	7230	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b8_u64_x2)))
	7231	svboolx2_t svwhilegt_b8_u64_x2(uint64_t, uint64_t);
	7232	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b32_u64_x2)))
	7233	svboolx2_t svwhilegt_b32_u64_x2(uint64_t, uint64_t);
	7234	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b64_u64_x2)))
	7235	svboolx2_t svwhilegt_b64_u64_x2(uint64_t, uint64_t);
	7236	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b16_u64_x2)))
	7237	svboolx2_t svwhilegt_b16_u64_x2(uint64_t, uint64_t);
	7238	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b8_s64_x2)))
	7239	svboolx2_t svwhilele_b8_s64_x2(int64_t, int64_t);
	7240	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b32_s64_x2)))
	7241	svboolx2_t svwhilele_b32_s64_x2(int64_t, int64_t);
	7242	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b64_s64_x2)))
	7243	svboolx2_t svwhilele_b64_s64_x2(int64_t, int64_t);
	7244	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b16_s64_x2)))
	7245	svboolx2_t svwhilele_b16_s64_x2(int64_t, int64_t);
	7246	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b8_u64_x2)))
	7247	svboolx2_t svwhilele_b8_u64_x2(uint64_t, uint64_t);
	7248	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b32_u64_x2)))
	7249	svboolx2_t svwhilele_b32_u64_x2(uint64_t, uint64_t);
	7250	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b64_u64_x2)))
	7251	svboolx2_t svwhilele_b64_u64_x2(uint64_t, uint64_t);
	7252	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b16_u64_x2)))
	7253	svboolx2_t svwhilele_b16_u64_x2(uint64_t, uint64_t);
	7254	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b8_u64_x2)))
	7255	svboolx2_t svwhilelt_b8_u64_x2(uint64_t, uint64_t);
	7256	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b32_u64_x2)))
	7257	svboolx2_t svwhilelt_b32_u64_x2(uint64_t, uint64_t);
	7258	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b64_u64_x2)))
	7259	svboolx2_t svwhilelt_b64_u64_x2(uint64_t, uint64_t);
	7260	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b16_u64_x2)))
	7261	svboolx2_t svwhilelt_b16_u64_x2(uint64_t, uint64_t);
	7262	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b8_s64_x2)))
	7263	svboolx2_t svwhilelt_b8_s64_x2(int64_t, int64_t);
	7264	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b32_s64_x2)))
	7265	svboolx2_t svwhilelt_b32_s64_x2(int64_t, int64_t);
	7266	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b64_s64_x2)))
	7267	svboolx2_t svwhilelt_b64_s64_x2(int64_t, int64_t);
	7268	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b16_s64_x2)))
	7269	svboolx2_t svwhilelt_b16_s64_x2(int64_t, int64_t);
	7270	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlslb_f32)))
	7271	svfloat32_t svbfmlslb(svfloat32_t, svbfloat16_t, svbfloat16_t);
	7272	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlslb_lane_f32)))
	7273	svfloat32_t svbfmlslb_lane(svfloat32_t, svbfloat16_t, svbfloat16_t, uint64_t);
	7274	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlslt_f32)))
	7275	svfloat32_t svbfmlslt(svfloat32_t, svbfloat16_t, svbfloat16_t);
	7276	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlslt_lane_f32)))
	7277	svfloat32_t svbfmlslt_lane(svfloat32_t, svbfloat16_t, svbfloat16_t, uint64_t);
	7278	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_f64)))
	7279	svfloat64_t svclamp(svfloat64_t, svfloat64_t, svfloat64_t);
	7280	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_f32)))
	7281	svfloat32_t svclamp(svfloat32_t, svfloat32_t, svfloat32_t);
	7282	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_f16)))
	7283	svfloat16_t svclamp(svfloat16_t, svfloat16_t, svfloat16_t);
	7284	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate2_b)))
	7285	svboolx2_t svcreate2(svbool_t, svbool_t);
	7286	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate4_b)))
	7287	svboolx4_t svcreate4(svbool_t, svbool_t, svbool_t, svbool_t);
	7288	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_f32_f16)))
	7289	svfloat32_t svdot(svfloat32_t, svfloat16_t, svfloat16_t);
	7290	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_s32_s16)))
	7291	svint32_t svdot(svint32_t, svint16_t, svint16_t);
	7292	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_u32_u16)))
	7293	svuint32_t svdot(svuint32_t, svuint16_t, svuint16_t);
	7294	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_lane_f32_f16)))
	7295	svfloat32_t svdot_lane(svfloat32_t, svfloat16_t, svfloat16_t, uint64_t);
	7296	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_lane_s32_s16)))
	7297	svint32_t svdot_lane(svint32_t, svint16_t, svint16_t, uint64_t);
	7298	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_lane_u32_u16)))
	7299	svuint32_t svdot_lane(svuint32_t, svuint16_t, svuint16_t, uint64_t);
	7300	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget2_b)))
	7301	svbool_t svget2(svboolx2_t, uint64_t);
	7302	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget4_b)))
	7303	svbool_t svget4(svboolx4_t, uint64_t);
	7304	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svqcvtn_s16_s32_x2)))
	7305	svint16_t svqcvtn_s16(svint32x2_t);
	7306	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svqcvtn_u16_s32_x2)))
	7307	svuint16_t svqcvtn_u16(svint32x2_t);
	7308	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svqcvtn_u16_u32_x2)))
	7309	svuint16_t svqcvtn_u16(svuint32x2_t);
	7310	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svqrshrn_n_s16_s32_x2)))
	7311	svint16_t svqrshrn_s16(svint32x2_t, uint64_t);
	7312	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svqrshrn_n_u16_u32_x2)))
	7313	svuint16_t svqrshrn_u16(svuint32x2_t, uint64_t);
	7314	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svqrshrun_n_u16_s32_x2)))
	7315	svuint16_t svqrshrun_u16(svint32x2_t, uint64_t);
	7316	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset2_b)))
	7317	svboolx2_t svset2(svboolx2_t, uint64_t, svbool_t);
	7318	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset4_b)))
	7319	svboolx4_t svset4(svboolx4_t, uint64_t, svbool_t);
	7320	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b8_s64_x2)))
	7321	svboolx2_t svwhilege_b8_x2(int64_t, int64_t);
	7322	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b32_s64_x2)))
	7323	svboolx2_t svwhilege_b32_x2(int64_t, int64_t);
	7324	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b64_s64_x2)))
	7325	svboolx2_t svwhilege_b64_x2(int64_t, int64_t);
	7326	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b16_s64_x2)))
	7327	svboolx2_t svwhilege_b16_x2(int64_t, int64_t);
	7328	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b8_u64_x2)))
	7329	svboolx2_t svwhilege_b8_x2(uint64_t, uint64_t);
	7330	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b32_u64_x2)))
	7331	svboolx2_t svwhilege_b32_x2(uint64_t, uint64_t);
	7332	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b64_u64_x2)))
	7333	svboolx2_t svwhilege_b64_x2(uint64_t, uint64_t);
	7334	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b16_u64_x2)))
	7335	svboolx2_t svwhilege_b16_x2(uint64_t, uint64_t);
	7336	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b8_s64_x2)))
	7337	svboolx2_t svwhilegt_b8_x2(int64_t, int64_t);
	7338	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b32_s64_x2)))
	7339	svboolx2_t svwhilegt_b32_x2(int64_t, int64_t);
	7340	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b64_s64_x2)))
	7341	svboolx2_t svwhilegt_b64_x2(int64_t, int64_t);
	7342	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b16_s64_x2)))
	7343	svboolx2_t svwhilegt_b16_x2(int64_t, int64_t);
	7344	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b8_u64_x2)))
	7345	svboolx2_t svwhilegt_b8_x2(uint64_t, uint64_t);
	7346	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b32_u64_x2)))
	7347	svboolx2_t svwhilegt_b32_x2(uint64_t, uint64_t);
	7348	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b64_u64_x2)))
	7349	svboolx2_t svwhilegt_b64_x2(uint64_t, uint64_t);
	7350	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b16_u64_x2)))
	7351	svboolx2_t svwhilegt_b16_x2(uint64_t, uint64_t);
	7352	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b8_s64_x2)))
	7353	svboolx2_t svwhilele_b8_x2(int64_t, int64_t);
	7354	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b32_s64_x2)))
	7355	svboolx2_t svwhilele_b32_x2(int64_t, int64_t);
	7356	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b64_s64_x2)))
	7357	svboolx2_t svwhilele_b64_x2(int64_t, int64_t);
	7358	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b16_s64_x2)))
	7359	svboolx2_t svwhilele_b16_x2(int64_t, int64_t);
	7360	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b8_u64_x2)))
	7361	svboolx2_t svwhilele_b8_x2(uint64_t, uint64_t);
	7362	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b32_u64_x2)))
	7363	svboolx2_t svwhilele_b32_x2(uint64_t, uint64_t);
	7364	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b64_u64_x2)))
	7365	svboolx2_t svwhilele_b64_x2(uint64_t, uint64_t);
	7366	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b16_u64_x2)))
	7367	svboolx2_t svwhilele_b16_x2(uint64_t, uint64_t);
	7368	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b8_u64_x2)))
	7369	svboolx2_t svwhilelt_b8_x2(uint64_t, uint64_t);
	7370	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b32_u64_x2)))
	7371	svboolx2_t svwhilelt_b32_x2(uint64_t, uint64_t);
	7372	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b64_u64_x2)))
	7373	svboolx2_t svwhilelt_b64_x2(uint64_t, uint64_t);
	7374	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b16_u64_x2)))
	7375	svboolx2_t svwhilelt_b16_x2(uint64_t, uint64_t);
	7376	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b8_s64_x2)))
	7377	svboolx2_t svwhilelt_b8_x2(int64_t, int64_t);
	7378	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b32_s64_x2)))
	7379	svboolx2_t svwhilelt_b32_x2(int64_t, int64_t);
	7380	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b64_s64_x2)))
	7381	svboolx2_t svwhilelt_b64_x2(int64_t, int64_t);
	7382	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b16_s64_x2)))
	7383	svboolx2_t svwhilelt_b16_x2(int64_t, int64_t);
	7384	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_u8)))
	7385	uint8x16_t svaddqv_u8(svbool_t, svuint8_t);
	7386	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_u32)))
	7387	uint32x4_t svaddqv_u32(svbool_t, svuint32_t);
	7388	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_u64)))
	7389	uint64x2_t svaddqv_u64(svbool_t, svuint64_t);
	7390	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_u16)))
	7391	uint16x8_t svaddqv_u16(svbool_t, svuint16_t);
	7392	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_s8)))
	7393	int8x16_t svaddqv_s8(svbool_t, svint8_t);
	7394	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_s32)))
	7395	int32x4_t svaddqv_s32(svbool_t, svint32_t);
	7396	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_s64)))
	7397	int64x2_t svaddqv_s64(svbool_t, svint64_t);
	7398	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_s16)))
	7399	int16x8_t svaddqv_s16(svbool_t, svint16_t);
	7400	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_f64)))
	7401	float64x2_t svaddqv_f64(svbool_t, svfloat64_t);
	7402	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_f32)))
	7403	float32x4_t svaddqv_f32(svbool_t, svfloat32_t);
	7404	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_f16)))
	7405	float16x8_t svaddqv_f16(svbool_t, svfloat16_t);
	7406	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_u8)))
	7407	uint8x16_t svandqv_u8(svbool_t, svuint8_t);
	7408	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_u32)))
	7409	uint32x4_t svandqv_u32(svbool_t, svuint32_t);
	7410	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_u64)))
	7411	uint64x2_t svandqv_u64(svbool_t, svuint64_t);
	7412	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_u16)))
	7413	uint16x8_t svandqv_u16(svbool_t, svuint16_t);
	7414	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_s8)))
	7415	int8x16_t svandqv_s8(svbool_t, svint8_t);
	7416	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_s32)))
	7417	int32x4_t svandqv_s32(svbool_t, svint32_t);
	7418	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_s64)))
	7419	int64x2_t svandqv_s64(svbool_t, svint64_t);
	7420	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_s16)))
	7421	int16x8_t svandqv_s16(svbool_t, svint16_t);
	7422	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_u8)))
	7423	svuint8_t svdup_laneq_u8(svuint8_t, uint64_t);
	7424	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_s8)))
	7425	svint8_t svdup_laneq_s8(svint8_t, uint64_t);
	7426	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_mf8)))
	7427	svmfloat8_t svdup_laneq_mf8(svmfloat8_t, uint64_t);
	7428	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_u64)))
	7429	svuint64_t svdup_laneq_u64(svuint64_t, uint64_t);
	7430	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_f64)))
	7431	svfloat64_t svdup_laneq_f64(svfloat64_t, uint64_t);
	7432	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_s64)))
	7433	svint64_t svdup_laneq_s64(svint64_t, uint64_t);
	7434	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_u16)))
	7435	svuint16_t svdup_laneq_u16(svuint16_t, uint64_t);
11184	7436	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_bf16)))
11185	7437	svbfloat16_t svdup_laneq_bf16(svbfloat16_t, uint64_t);
	7438	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_f16)))
	7439	svfloat16_t svdup_laneq_f16(svfloat16_t, uint64_t);
	7440	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_s16)))
	7441	svint16_t svdup_laneq_s16(svint16_t, uint64_t);
	7442	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_u32)))
	7443	svuint32_t svdup_laneq_u32(svuint32_t, uint64_t);
	7444	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_f32)))
	7445	svfloat32_t svdup_laneq_f32(svfloat32_t, uint64_t);
	7446	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_s32)))
	7447	svint32_t svdup_laneq_s32(svint32_t, uint64_t);
	7448	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_u8)))
	7449	uint8x16_t sveorqv_u8(svbool_t, svuint8_t);
	7450	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_u32)))
	7451	uint32x4_t sveorqv_u32(svbool_t, svuint32_t);
	7452	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_u64)))
	7453	uint64x2_t sveorqv_u64(svbool_t, svuint64_t);
	7454	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_u16)))
	7455	uint16x8_t sveorqv_u16(svbool_t, svuint16_t);
	7456	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_s8)))
	7457	int8x16_t sveorqv_s8(svbool_t, svint8_t);
	7458	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_s32)))
	7459	int32x4_t sveorqv_s32(svbool_t, svint32_t);
	7460	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_s64)))
	7461	int64x2_t sveorqv_s64(svbool_t, svint64_t);
	7462	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_s16)))
	7463	int16x8_t sveorqv_s16(svbool_t, svint16_t);
	7464	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_u8)))
	7465	svuint8_t svextq_u8(svuint8_t, svuint8_t, int32_t);
	7466	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_u32)))
	7467	svuint32_t svextq_u32(svuint32_t, svuint32_t, int32_t);
	7468	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_u64)))
	7469	svuint64_t svextq_u64(svuint64_t, svuint64_t, int32_t);
	7470	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_u16)))
	7471	svuint16_t svextq_u16(svuint16_t, svuint16_t, int32_t);
	7472	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_bf16)))
	7473	svbfloat16_t svextq_bf16(svbfloat16_t, svbfloat16_t, int32_t);
	7474	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_s8)))
	7475	svint8_t svextq_s8(svint8_t, svint8_t, int32_t);
	7476	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_f64)))
	7477	svfloat64_t svextq_f64(svfloat64_t, svfloat64_t, int32_t);
	7478	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_f32)))
	7479	svfloat32_t svextq_f32(svfloat32_t, svfloat32_t, int32_t);
	7480	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_f16)))
	7481	svfloat16_t svextq_f16(svfloat16_t, svfloat16_t, int32_t);
	7482	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_s32)))
	7483	svint32_t svextq_s32(svint32_t, svint32_t, int32_t);
	7484	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_s64)))
	7485	svint64_t svextq_s64(svint64_t, svint64_t, int32_t);
	7486	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_mf8)))
	7487	svmfloat8_t svextq_mf8(svmfloat8_t, svmfloat8_t, int32_t);
	7488	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_s16)))
	7489	svint16_t svextq_s16(svint16_t, svint16_t, int32_t);
	7490	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_u8)))
	7491	svuint8x2_t svld2q_u8(svbool_t, uint8_t const *);
	7492	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_u32)))
	7493	svuint32x2_t svld2q_u32(svbool_t, uint32_t const *);
	7494	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_u64)))
	7495	svuint64x2_t svld2q_u64(svbool_t, uint64_t const *);
	7496	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_u16)))
	7497	svuint16x2_t svld2q_u16(svbool_t, uint16_t const *);
	7498	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_bf16)))
	7499	svbfloat16x2_t svld2q_bf16(svbool_t, bfloat16_t const *);
	7500	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_s8)))
	7501	svint8x2_t svld2q_s8(svbool_t, int8_t const *);
	7502	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_f64)))
	7503	svfloat64x2_t svld2q_f64(svbool_t, float64_t const *);
	7504	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_f32)))
	7505	svfloat32x2_t svld2q_f32(svbool_t, float32_t const *);
	7506	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_f16)))
	7507	svfloat16x2_t svld2q_f16(svbool_t, float16_t const *);
	7508	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_s32)))
	7509	svint32x2_t svld2q_s32(svbool_t, int32_t const *);
	7510	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_s64)))
	7511	svint64x2_t svld2q_s64(svbool_t, int64_t const *);
	7512	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_mf8)))
	7513	svmfloat8x2_t svld2q_mf8(svbool_t, mfloat8_t const *);
	7514	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_s16)))
	7515	svint16x2_t svld2q_s16(svbool_t, int16_t const *);
	7516	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_u8)))
	7517	svuint8x2_t svld2q_vnum_u8(svbool_t, uint8_t const *, int64_t);
	7518	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_u32)))
	7519	svuint32x2_t svld2q_vnum_u32(svbool_t, uint32_t const *, int64_t);
	7520	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_u64)))
	7521	svuint64x2_t svld2q_vnum_u64(svbool_t, uint64_t const *, int64_t);
	7522	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_u16)))
	7523	svuint16x2_t svld2q_vnum_u16(svbool_t, uint16_t const *, int64_t);
	7524	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_bf16)))
	7525	svbfloat16x2_t svld2q_vnum_bf16(svbool_t, bfloat16_t const *, int64_t);
	7526	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_s8)))
	7527	svint8x2_t svld2q_vnum_s8(svbool_t, int8_t const *, int64_t);
	7528	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_f64)))
	7529	svfloat64x2_t svld2q_vnum_f64(svbool_t, float64_t const *, int64_t);
	7530	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_f32)))
	7531	svfloat32x2_t svld2q_vnum_f32(svbool_t, float32_t const *, int64_t);
	7532	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_f16)))
	7533	svfloat16x2_t svld2q_vnum_f16(svbool_t, float16_t const *, int64_t);
	7534	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_s32)))
	7535	svint32x2_t svld2q_vnum_s32(svbool_t, int32_t const *, int64_t);
	7536	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_s64)))
	7537	svint64x2_t svld2q_vnum_s64(svbool_t, int64_t const *, int64_t);
	7538	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_mf8)))
	7539	svmfloat8x2_t svld2q_vnum_mf8(svbool_t, mfloat8_t const *, int64_t);
	7540	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_s16)))
	7541	svint16x2_t svld2q_vnum_s16(svbool_t, int16_t const *, int64_t);
	7542	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_u8)))
	7543	svuint8x3_t svld3q_u8(svbool_t, uint8_t const *);
	7544	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_u32)))
	7545	svuint32x3_t svld3q_u32(svbool_t, uint32_t const *);
	7546	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_u64)))
	7547	svuint64x3_t svld3q_u64(svbool_t, uint64_t const *);
	7548	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_u16)))
	7549	svuint16x3_t svld3q_u16(svbool_t, uint16_t const *);
	7550	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_bf16)))
	7551	svbfloat16x3_t svld3q_bf16(svbool_t, bfloat16_t const *);
	7552	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_s8)))
	7553	svint8x3_t svld3q_s8(svbool_t, int8_t const *);
	7554	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_f64)))
	7555	svfloat64x3_t svld3q_f64(svbool_t, float64_t const *);
	7556	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_f32)))
	7557	svfloat32x3_t svld3q_f32(svbool_t, float32_t const *);
	7558	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_f16)))
	7559	svfloat16x3_t svld3q_f16(svbool_t, float16_t const *);
	7560	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_s32)))
	7561	svint32x3_t svld3q_s32(svbool_t, int32_t const *);
	7562	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_s64)))
	7563	svint64x3_t svld3q_s64(svbool_t, int64_t const *);
	7564	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_mf8)))
	7565	svmfloat8x3_t svld3q_mf8(svbool_t, mfloat8_t const *);
	7566	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_s16)))
	7567	svint16x3_t svld3q_s16(svbool_t, int16_t const *);
	7568	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_u8)))
	7569	svuint8x3_t svld3q_vnum_u8(svbool_t, uint8_t const *, int64_t);
	7570	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_u32)))
	7571	svuint32x3_t svld3q_vnum_u32(svbool_t, uint32_t const *, int64_t);
	7572	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_u64)))
	7573	svuint64x3_t svld3q_vnum_u64(svbool_t, uint64_t const *, int64_t);
	7574	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_u16)))
	7575	svuint16x3_t svld3q_vnum_u16(svbool_t, uint16_t const *, int64_t);
	7576	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_bf16)))
	7577	svbfloat16x3_t svld3q_vnum_bf16(svbool_t, bfloat16_t const *, int64_t);
	7578	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_s8)))
	7579	svint8x3_t svld3q_vnum_s8(svbool_t, int8_t const *, int64_t);
	7580	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_f64)))
	7581	svfloat64x3_t svld3q_vnum_f64(svbool_t, float64_t const *, int64_t);
	7582	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_f32)))
	7583	svfloat32x3_t svld3q_vnum_f32(svbool_t, float32_t const *, int64_t);
	7584	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_f16)))
	7585	svfloat16x3_t svld3q_vnum_f16(svbool_t, float16_t const *, int64_t);
	7586	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_s32)))
	7587	svint32x3_t svld3q_vnum_s32(svbool_t, int32_t const *, int64_t);
	7588	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_s64)))
	7589	svint64x3_t svld3q_vnum_s64(svbool_t, int64_t const *, int64_t);
	7590	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_mf8)))
	7591	svmfloat8x3_t svld3q_vnum_mf8(svbool_t, mfloat8_t const *, int64_t);
	7592	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_s16)))
	7593	svint16x3_t svld3q_vnum_s16(svbool_t, int16_t const *, int64_t);
	7594	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_u8)))
	7595	svuint8x4_t svld4q_u8(svbool_t, uint8_t const *);
	7596	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_u32)))
	7597	svuint32x4_t svld4q_u32(svbool_t, uint32_t const *);
	7598	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_u64)))
	7599	svuint64x4_t svld4q_u64(svbool_t, uint64_t const *);
	7600	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_u16)))
	7601	svuint16x4_t svld4q_u16(svbool_t, uint16_t const *);
	7602	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_bf16)))
	7603	svbfloat16x4_t svld4q_bf16(svbool_t, bfloat16_t const *);
	7604	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_s8)))
	7605	svint8x4_t svld4q_s8(svbool_t, int8_t const *);
	7606	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_f64)))
	7607	svfloat64x4_t svld4q_f64(svbool_t, float64_t const *);
	7608	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_f32)))
	7609	svfloat32x4_t svld4q_f32(svbool_t, float32_t const *);
	7610	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_f16)))
	7611	svfloat16x4_t svld4q_f16(svbool_t, float16_t const *);
	7612	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_s32)))
	7613	svint32x4_t svld4q_s32(svbool_t, int32_t const *);
	7614	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_s64)))
	7615	svint64x4_t svld4q_s64(svbool_t, int64_t const *);
	7616	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_mf8)))
	7617	svmfloat8x4_t svld4q_mf8(svbool_t, mfloat8_t const *);
	7618	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_s16)))
	7619	svint16x4_t svld4q_s16(svbool_t, int16_t const *);
	7620	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_u8)))
	7621	svuint8x4_t svld4q_vnum_u8(svbool_t, uint8_t const *, int64_t);
	7622	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_u32)))
	7623	svuint32x4_t svld4q_vnum_u32(svbool_t, uint32_t const *, int64_t);
	7624	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_u64)))
	7625	svuint64x4_t svld4q_vnum_u64(svbool_t, uint64_t const *, int64_t);
	7626	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_u16)))
	7627	svuint16x4_t svld4q_vnum_u16(svbool_t, uint16_t const *, int64_t);
	7628	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_bf16)))
	7629	svbfloat16x4_t svld4q_vnum_bf16(svbool_t, bfloat16_t const *, int64_t);
	7630	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_s8)))
	7631	svint8x4_t svld4q_vnum_s8(svbool_t, int8_t const *, int64_t);
	7632	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_f64)))
	7633	svfloat64x4_t svld4q_vnum_f64(svbool_t, float64_t const *, int64_t);
	7634	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_f32)))
	7635	svfloat32x4_t svld4q_vnum_f32(svbool_t, float32_t const *, int64_t);
	7636	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_f16)))
	7637	svfloat16x4_t svld4q_vnum_f16(svbool_t, float16_t const *, int64_t);
	7638	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_s32)))
	7639	svint32x4_t svld4q_vnum_s32(svbool_t, int32_t const *, int64_t);
	7640	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_s64)))
	7641	svint64x4_t svld4q_vnum_s64(svbool_t, int64_t const *, int64_t);
	7642	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_mf8)))
	7643	svmfloat8x4_t svld4q_vnum_mf8(svbool_t, mfloat8_t const *, int64_t);
	7644	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_s16)))
	7645	svint16x4_t svld4q_vnum_s16(svbool_t, int16_t const *, int64_t);
	7646	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnmqv_f64)))
	7647	float64x2_t svmaxnmqv_f64(svbool_t, svfloat64_t);
	7648	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnmqv_f32)))
	7649	float32x4_t svmaxnmqv_f32(svbool_t, svfloat32_t);
	7650	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnmqv_f16)))
	7651	float16x8_t svmaxnmqv_f16(svbool_t, svfloat16_t);
	7652	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_f64)))
	7653	float64x2_t svmaxqv_f64(svbool_t, svfloat64_t);
	7654	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_f32)))
	7655	float32x4_t svmaxqv_f32(svbool_t, svfloat32_t);
	7656	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_f16)))
	7657	float16x8_t svmaxqv_f16(svbool_t, svfloat16_t);
	7658	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_s8)))
	7659	int8x16_t svmaxqv_s8(svbool_t, svint8_t);
	7660	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_s32)))
	7661	int32x4_t svmaxqv_s32(svbool_t, svint32_t);
	7662	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_s64)))
	7663	int64x2_t svmaxqv_s64(svbool_t, svint64_t);
	7664	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_s16)))
	7665	int16x8_t svmaxqv_s16(svbool_t, svint16_t);
	7666	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_u8)))
	7667	uint8x16_t svmaxqv_u8(svbool_t, svuint8_t);
	7668	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_u32)))
	7669	uint32x4_t svmaxqv_u32(svbool_t, svuint32_t);
	7670	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_u64)))
	7671	uint64x2_t svmaxqv_u64(svbool_t, svuint64_t);
	7672	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_u16)))
	7673	uint16x8_t svmaxqv_u16(svbool_t, svuint16_t);
	7674	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnmqv_f64)))
	7675	float64x2_t svminnmqv_f64(svbool_t, svfloat64_t);
	7676	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnmqv_f32)))
	7677	float32x4_t svminnmqv_f32(svbool_t, svfloat32_t);
	7678	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnmqv_f16)))
	7679	float16x8_t svminnmqv_f16(svbool_t, svfloat16_t);
	7680	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_f64)))
	7681	float64x2_t svminqv_f64(svbool_t, svfloat64_t);
	7682	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_f32)))
	7683	float32x4_t svminqv_f32(svbool_t, svfloat32_t);
	7684	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_f16)))
	7685	float16x8_t svminqv_f16(svbool_t, svfloat16_t);
	7686	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_s8)))
	7687	int8x16_t svminqv_s8(svbool_t, svint8_t);
	7688	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_s32)))
	7689	int32x4_t svminqv_s32(svbool_t, svint32_t);
	7690	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_s64)))
	7691	int64x2_t svminqv_s64(svbool_t, svint64_t);
	7692	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_s16)))
	7693	int16x8_t svminqv_s16(svbool_t, svint16_t);
	7694	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_u8)))
	7695	uint8x16_t svminqv_u8(svbool_t, svuint8_t);
	7696	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_u32)))
	7697	uint32x4_t svminqv_u32(svbool_t, svuint32_t);
	7698	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_u64)))
	7699	uint64x2_t svminqv_u64(svbool_t, svuint64_t);
	7700	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_u16)))
	7701	uint16x8_t svminqv_u16(svbool_t, svuint16_t);
	7702	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_u8)))
	7703	uint8x16_t svorqv_u8(svbool_t, svuint8_t);
	7704	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_u32)))
	7705	uint32x4_t svorqv_u32(svbool_t, svuint32_t);
	7706	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_u64)))
	7707	uint64x2_t svorqv_u64(svbool_t, svuint64_t);
	7708	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_u16)))
	7709	uint16x8_t svorqv_u16(svbool_t, svuint16_t);
	7710	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_s8)))
	7711	int8x16_t svorqv_s8(svbool_t, svint8_t);
	7712	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_s32)))
	7713	int32x4_t svorqv_s32(svbool_t, svint32_t);
	7714	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_s64)))
	7715	int64x2_t svorqv_s64(svbool_t, svint64_t);
	7716	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_s16)))
	7717	int16x8_t svorqv_s16(svbool_t, svint16_t);
	7718	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_u8)))
	7719	svbool_t svpmov_u8(svuint8_t);
	7720	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_s8)))
	7721	svbool_t svpmov_s8(svint8_t);
	7722	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_u64)))
	7723	svbool_t svpmov_u64(svuint64_t);
	7724	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_s64)))
	7725	svbool_t svpmov_s64(svint64_t);
	7726	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_u16)))
	7727	svbool_t svpmov_u16(svuint16_t);
	7728	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_s16)))
	7729	svbool_t svpmov_s16(svint16_t);
	7730	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_u32)))
	7731	svbool_t svpmov_u32(svuint32_t);
	7732	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_s32)))
	7733	svbool_t svpmov_s32(svint32_t);
	7734	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_u8)))
	7735	svbool_t svpmov_lane_u8(svuint8_t, uint64_t);
	7736	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_s8)))
	7737	svbool_t svpmov_lane_s8(svint8_t, uint64_t);
	7738	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_u64)))
	7739	svbool_t svpmov_lane_u64(svuint64_t, uint64_t);
	7740	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_s64)))
	7741	svbool_t svpmov_lane_s64(svint64_t, uint64_t);
	7742	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_u16)))
	7743	svbool_t svpmov_lane_u16(svuint16_t, uint64_t);
	7744	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_s16)))
	7745	svbool_t svpmov_lane_s16(svint16_t, uint64_t);
	7746	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_u32)))
	7747	svbool_t svpmov_lane_u32(svuint32_t, uint64_t);
	7748	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_s32)))
	7749	svbool_t svpmov_lane_s32(svint32_t, uint64_t);
	7750	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_u64_m)))
	7751	svuint64_t svpmov_lane_u64_m(svuint64_t, svbool_t, uint64_t);
	7752	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_s64_m)))
	7753	svint64_t svpmov_lane_s64_m(svint64_t, svbool_t, uint64_t);
	7754	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_u16_m)))
	7755	svuint16_t svpmov_lane_u16_m(svuint16_t, svbool_t, uint64_t);
	7756	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_s16_m)))
	7757	svint16_t svpmov_lane_s16_m(svint16_t, svbool_t, uint64_t);
	7758	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_u32_m)))
	7759	svuint32_t svpmov_lane_u32_m(svuint32_t, svbool_t, uint64_t);
	7760	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_s32_m)))
	7761	svint32_t svpmov_lane_s32_m(svint32_t, svbool_t, uint64_t);
	7762	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_u8_z)))
	7763	svuint8_t svpmov_u8_z(svbool_t);
	7764	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_s8_z)))
	7765	svint8_t svpmov_s8_z(svbool_t);
	7766	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_u64_z)))
	7767	svuint64_t svpmov_u64_z(svbool_t);
	7768	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_s64_z)))
	7769	svint64_t svpmov_s64_z(svbool_t);
	7770	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_u16_z)))
	7771	svuint16_t svpmov_u16_z(svbool_t);
	7772	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_s16_z)))
	7773	svint16_t svpmov_s16_z(svbool_t);
	7774	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_u32_z)))
	7775	svuint32_t svpmov_u32_z(svbool_t);
	7776	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_s32_z)))
	7777	svint32_t svpmov_s32_z(svbool_t);
	7778	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_u8)))
	7779	void svst2q_u8(svbool_t, uint8_t const *, svuint8x2_t);
	7780	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_u32)))
	7781	void svst2q_u32(svbool_t, uint32_t const *, svuint32x2_t);
	7782	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_u64)))
	7783	void svst2q_u64(svbool_t, uint64_t const *, svuint64x2_t);
	7784	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_u16)))
	7785	void svst2q_u16(svbool_t, uint16_t const *, svuint16x2_t);
	7786	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_bf16)))
	7787	void svst2q_bf16(svbool_t, bfloat16_t const *, svbfloat16x2_t);
	7788	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_s8)))
	7789	void svst2q_s8(svbool_t, int8_t const *, svint8x2_t);
	7790	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_f64)))
	7791	void svst2q_f64(svbool_t, float64_t const *, svfloat64x2_t);
	7792	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_f32)))
	7793	void svst2q_f32(svbool_t, float32_t const *, svfloat32x2_t);
	7794	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_f16)))
	7795	void svst2q_f16(svbool_t, float16_t const *, svfloat16x2_t);
	7796	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_s32)))
	7797	void svst2q_s32(svbool_t, int32_t const *, svint32x2_t);
	7798	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_s64)))
	7799	void svst2q_s64(svbool_t, int64_t const *, svint64x2_t);
	7800	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_mf8)))
	7801	void svst2q_mf8(svbool_t, mfloat8_t const *, svmfloat8x2_t);
	7802	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_s16)))
	7803	void svst2q_s16(svbool_t, int16_t const *, svint16x2_t);
	7804	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_u8)))
	7805	void svst2q_vnum_u8(svbool_t, uint8_t const *, int64_t, svuint8x2_t);
	7806	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_u32)))
	7807	void svst2q_vnum_u32(svbool_t, uint32_t const *, int64_t, svuint32x2_t);
	7808	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_u64)))
	7809	void svst2q_vnum_u64(svbool_t, uint64_t const *, int64_t, svuint64x2_t);
	7810	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_u16)))
	7811	void svst2q_vnum_u16(svbool_t, uint16_t const *, int64_t, svuint16x2_t);
	7812	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_bf16)))
	7813	void svst2q_vnum_bf16(svbool_t, bfloat16_t const *, int64_t, svbfloat16x2_t);
	7814	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_s8)))
	7815	void svst2q_vnum_s8(svbool_t, int8_t const *, int64_t, svint8x2_t);
	7816	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_f64)))
	7817	void svst2q_vnum_f64(svbool_t, float64_t const *, int64_t, svfloat64x2_t);
	7818	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_f32)))
	7819	void svst2q_vnum_f32(svbool_t, float32_t const *, int64_t, svfloat32x2_t);
	7820	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_f16)))
	7821	void svst2q_vnum_f16(svbool_t, float16_t const *, int64_t, svfloat16x2_t);
	7822	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_s32)))
	7823	void svst2q_vnum_s32(svbool_t, int32_t const *, int64_t, svint32x2_t);
	7824	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_s64)))
	7825	void svst2q_vnum_s64(svbool_t, int64_t const *, int64_t, svint64x2_t);
	7826	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_mf8)))
	7827	void svst2q_vnum_mf8(svbool_t, mfloat8_t const *, int64_t, svmfloat8x2_t);
	7828	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_s16)))
	7829	void svst2q_vnum_s16(svbool_t, int16_t const *, int64_t, svint16x2_t);
	7830	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_u8)))
	7831	void svst3q_u8(svbool_t, uint8_t const *, svuint8x3_t);
	7832	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_u32)))
	7833	void svst3q_u32(svbool_t, uint32_t const *, svuint32x3_t);
	7834	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_u64)))
	7835	void svst3q_u64(svbool_t, uint64_t const *, svuint64x3_t);
	7836	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_u16)))
	7837	void svst3q_u16(svbool_t, uint16_t const *, svuint16x3_t);
	7838	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_bf16)))
	7839	void svst3q_bf16(svbool_t, bfloat16_t const *, svbfloat16x3_t);
	7840	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_s8)))
	7841	void svst3q_s8(svbool_t, int8_t const *, svint8x3_t);
	7842	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_f64)))
	7843	void svst3q_f64(svbool_t, float64_t const *, svfloat64x3_t);
	7844	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_f32)))
	7845	void svst3q_f32(svbool_t, float32_t const *, svfloat32x3_t);
	7846	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_f16)))
	7847	void svst3q_f16(svbool_t, float16_t const *, svfloat16x3_t);
	7848	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_s32)))
	7849	void svst3q_s32(svbool_t, int32_t const *, svint32x3_t);
	7850	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_s64)))
	7851	void svst3q_s64(svbool_t, int64_t const *, svint64x3_t);
	7852	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_mf8)))
	7853	void svst3q_mf8(svbool_t, mfloat8_t const *, svmfloat8x3_t);
	7854	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_s16)))
	7855	void svst3q_s16(svbool_t, int16_t const *, svint16x3_t);
	7856	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_u8)))
	7857	void svst3q_vnum_u8(svbool_t, uint8_t const *, int64_t, svuint8x3_t);
	7858	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_u32)))
	7859	void svst3q_vnum_u32(svbool_t, uint32_t const *, int64_t, svuint32x3_t);
	7860	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_u64)))
	7861	void svst3q_vnum_u64(svbool_t, uint64_t const *, int64_t, svuint64x3_t);
	7862	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_u16)))
	7863	void svst3q_vnum_u16(svbool_t, uint16_t const *, int64_t, svuint16x3_t);
	7864	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_bf16)))
	7865	void svst3q_vnum_bf16(svbool_t, bfloat16_t const *, int64_t, svbfloat16x3_t);
	7866	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_s8)))
	7867	void svst3q_vnum_s8(svbool_t, int8_t const *, int64_t, svint8x3_t);
	7868	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_f64)))
	7869	void svst3q_vnum_f64(svbool_t, float64_t const *, int64_t, svfloat64x3_t);
	7870	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_f32)))
	7871	void svst3q_vnum_f32(svbool_t, float32_t const *, int64_t, svfloat32x3_t);
	7872	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_f16)))
	7873	void svst3q_vnum_f16(svbool_t, float16_t const *, int64_t, svfloat16x3_t);
	7874	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_s32)))
	7875	void svst3q_vnum_s32(svbool_t, int32_t const *, int64_t, svint32x3_t);
	7876	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_s64)))
	7877	void svst3q_vnum_s64(svbool_t, int64_t const *, int64_t, svint64x3_t);
	7878	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_mf8)))
	7879	void svst3q_vnum_mf8(svbool_t, mfloat8_t const *, int64_t, svmfloat8x3_t);
	7880	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_s16)))
	7881	void svst3q_vnum_s16(svbool_t, int16_t const *, int64_t, svint16x3_t);
	7882	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_u8)))
	7883	void svst4q_u8(svbool_t, uint8_t const *, svuint8x4_t);
	7884	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_u32)))
	7885	void svst4q_u32(svbool_t, uint32_t const *, svuint32x4_t);
	7886	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_u64)))
	7887	void svst4q_u64(svbool_t, uint64_t const *, svuint64x4_t);
	7888	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_u16)))
	7889	void svst4q_u16(svbool_t, uint16_t const *, svuint16x4_t);
	7890	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_bf16)))
	7891	void svst4q_bf16(svbool_t, bfloat16_t const *, svbfloat16x4_t);
	7892	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_s8)))
	7893	void svst4q_s8(svbool_t, int8_t const *, svint8x4_t);
	7894	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_f64)))
	7895	void svst4q_f64(svbool_t, float64_t const *, svfloat64x4_t);
	7896	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_f32)))
	7897	void svst4q_f32(svbool_t, float32_t const *, svfloat32x4_t);
	7898	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_f16)))
	7899	void svst4q_f16(svbool_t, float16_t const *, svfloat16x4_t);
	7900	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_s32)))
	7901	void svst4q_s32(svbool_t, int32_t const *, svint32x4_t);
	7902	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_s64)))
	7903	void svst4q_s64(svbool_t, int64_t const *, svint64x4_t);
	7904	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_mf8)))
	7905	void svst4q_mf8(svbool_t, mfloat8_t const *, svmfloat8x4_t);
	7906	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_s16)))
	7907	void svst4q_s16(svbool_t, int16_t const *, svint16x4_t);
	7908	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_u8)))
	7909	void svst4q_vnum_u8(svbool_t, uint8_t const *, int64_t, svuint8x4_t);
	7910	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_u32)))
	7911	void svst4q_vnum_u32(svbool_t, uint32_t const *, int64_t, svuint32x4_t);
	7912	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_u64)))
	7913	void svst4q_vnum_u64(svbool_t, uint64_t const *, int64_t, svuint64x4_t);
	7914	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_u16)))
	7915	void svst4q_vnum_u16(svbool_t, uint16_t const *, int64_t, svuint16x4_t);
	7916	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_bf16)))
	7917	void svst4q_vnum_bf16(svbool_t, bfloat16_t const *, int64_t, svbfloat16x4_t);
	7918	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_s8)))
	7919	void svst4q_vnum_s8(svbool_t, int8_t const *, int64_t, svint8x4_t);
	7920	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_f64)))
	7921	void svst4q_vnum_f64(svbool_t, float64_t const *, int64_t, svfloat64x4_t);
	7922	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_f32)))
	7923	void svst4q_vnum_f32(svbool_t, float32_t const *, int64_t, svfloat32x4_t);
	7924	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_f16)))
	7925	void svst4q_vnum_f16(svbool_t, float16_t const *, int64_t, svfloat16x4_t);
	7926	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_s32)))
	7927	void svst4q_vnum_s32(svbool_t, int32_t const *, int64_t, svint32x4_t);
	7928	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_s64)))
	7929	void svst4q_vnum_s64(svbool_t, int64_t const *, int64_t, svint64x4_t);
	7930	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_mf8)))
	7931	void svst4q_vnum_mf8(svbool_t, mfloat8_t const *, int64_t, svmfloat8x4_t);
	7932	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_s16)))
	7933	void svst4q_vnum_s16(svbool_t, int16_t const *, int64_t, svint16x4_t);
	7934	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_u8)))
	7935	svuint8_t svtblq_u8(svuint8_t, svuint8_t);
	7936	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_u32)))
	7937	svuint32_t svtblq_u32(svuint32_t, svuint32_t);
	7938	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_u64)))
	7939	svuint64_t svtblq_u64(svuint64_t, svuint64_t);
	7940	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_u16)))
	7941	svuint16_t svtblq_u16(svuint16_t, svuint16_t);
	7942	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_bf16)))
	7943	svbfloat16_t svtblq_bf16(svbfloat16_t, svuint16_t);
	7944	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_s8)))
	7945	svint8_t svtblq_s8(svint8_t, svuint8_t);
	7946	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_f64)))
	7947	svfloat64_t svtblq_f64(svfloat64_t, svuint64_t);
	7948	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_f32)))
	7949	svfloat32_t svtblq_f32(svfloat32_t, svuint32_t);
	7950	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_f16)))
	7951	svfloat16_t svtblq_f16(svfloat16_t, svuint16_t);
	7952	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_s32)))
	7953	svint32_t svtblq_s32(svint32_t, svuint32_t);
	7954	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_s64)))
	7955	svint64_t svtblq_s64(svint64_t, svuint64_t);
	7956	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_mf8)))
	7957	svmfloat8_t svtblq_mf8(svmfloat8_t, svuint8_t);
	7958	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_s16)))
	7959	svint16_t svtblq_s16(svint16_t, svuint16_t);
	7960	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_u8)))
	7961	svuint8_t svtbxq_u8(svuint8_t, svuint8_t, svuint8_t);
	7962	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_u32)))
	7963	svuint32_t svtbxq_u32(svuint32_t, svuint32_t, svuint32_t);
	7964	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_u64)))
	7965	svuint64_t svtbxq_u64(svuint64_t, svuint64_t, svuint64_t);
	7966	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_u16)))
	7967	svuint16_t svtbxq_u16(svuint16_t, svuint16_t, svuint16_t);
	7968	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_bf16)))
	7969	svbfloat16_t svtbxq_bf16(svbfloat16_t, svbfloat16_t, svuint16_t);
	7970	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_s8)))
	7971	svint8_t svtbxq_s8(svint8_t, svint8_t, svuint8_t);
	7972	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_f64)))
	7973	svfloat64_t svtbxq_f64(svfloat64_t, svfloat64_t, svuint64_t);
	7974	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_f32)))
	7975	svfloat32_t svtbxq_f32(svfloat32_t, svfloat32_t, svuint32_t);
	7976	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_f16)))
	7977	svfloat16_t svtbxq_f16(svfloat16_t, svfloat16_t, svuint16_t);
	7978	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_s32)))
	7979	svint32_t svtbxq_s32(svint32_t, svint32_t, svuint32_t);
	7980	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_s64)))
	7981	svint64_t svtbxq_s64(svint64_t, svint64_t, svuint64_t);
	7982	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_mf8)))
	7983	svmfloat8_t svtbxq_mf8(svmfloat8_t, svmfloat8_t, svuint8_t);
	7984	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_s16)))
	7985	svint16_t svtbxq_s16(svint16_t, svint16_t, svuint16_t);
	7986	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_u8)))
	7987	svuint8_t svuzpq1_u8(svuint8_t, svuint8_t);
	7988	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_u32)))
	7989	svuint32_t svuzpq1_u32(svuint32_t, svuint32_t);
	7990	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_u64)))
	7991	svuint64_t svuzpq1_u64(svuint64_t, svuint64_t);
	7992	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_u16)))
	7993	svuint16_t svuzpq1_u16(svuint16_t, svuint16_t);
	7994	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_bf16)))
	7995	svbfloat16_t svuzpq1_bf16(svbfloat16_t, svbfloat16_t);
	7996	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_s8)))
	7997	svint8_t svuzpq1_s8(svint8_t, svint8_t);
	7998	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_f64)))
	7999	svfloat64_t svuzpq1_f64(svfloat64_t, svfloat64_t);
	8000	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_f32)))
	8001	svfloat32_t svuzpq1_f32(svfloat32_t, svfloat32_t);
	8002	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_f16)))
	8003	svfloat16_t svuzpq1_f16(svfloat16_t, svfloat16_t);
	8004	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_s32)))
	8005	svint32_t svuzpq1_s32(svint32_t, svint32_t);
	8006	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_s64)))
	8007	svint64_t svuzpq1_s64(svint64_t, svint64_t);
	8008	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_mf8)))
	8009	svmfloat8_t svuzpq1_mf8(svmfloat8_t, svmfloat8_t);
	8010	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_s16)))
	8011	svint16_t svuzpq1_s16(svint16_t, svint16_t);
	8012	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_u8)))
	8013	svuint8_t svuzpq2_u8(svuint8_t, svuint8_t);
	8014	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_u32)))
	8015	svuint32_t svuzpq2_u32(svuint32_t, svuint32_t);
	8016	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_u64)))
	8017	svuint64_t svuzpq2_u64(svuint64_t, svuint64_t);
	8018	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_u16)))
	8019	svuint16_t svuzpq2_u16(svuint16_t, svuint16_t);
	8020	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_bf16)))
	8021	svbfloat16_t svuzpq2_bf16(svbfloat16_t, svbfloat16_t);
	8022	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_s8)))
	8023	svint8_t svuzpq2_s8(svint8_t, svint8_t);
	8024	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_f64)))
	8025	svfloat64_t svuzpq2_f64(svfloat64_t, svfloat64_t);
	8026	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_f32)))
	8027	svfloat32_t svuzpq2_f32(svfloat32_t, svfloat32_t);
	8028	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_f16)))
	8029	svfloat16_t svuzpq2_f16(svfloat16_t, svfloat16_t);
	8030	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_s32)))
	8031	svint32_t svuzpq2_s32(svint32_t, svint32_t);
	8032	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_s64)))
	8033	svint64_t svuzpq2_s64(svint64_t, svint64_t);
	8034	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_mf8)))
	8035	svmfloat8_t svuzpq2_mf8(svmfloat8_t, svmfloat8_t);
	8036	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_s16)))
	8037	svint16_t svuzpq2_s16(svint16_t, svint16_t);
	8038	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_u8)))
	8039	svuint8_t svzipq1_u8(svuint8_t, svuint8_t);
	8040	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_u32)))
	8041	svuint32_t svzipq1_u32(svuint32_t, svuint32_t);
	8042	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_u64)))
	8043	svuint64_t svzipq1_u64(svuint64_t, svuint64_t);
	8044	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_u16)))
	8045	svuint16_t svzipq1_u16(svuint16_t, svuint16_t);
	8046	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_bf16)))
	8047	svbfloat16_t svzipq1_bf16(svbfloat16_t, svbfloat16_t);
	8048	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_s8)))
	8049	svint8_t svzipq1_s8(svint8_t, svint8_t);
	8050	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_f64)))
	8051	svfloat64_t svzipq1_f64(svfloat64_t, svfloat64_t);
	8052	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_f32)))
	8053	svfloat32_t svzipq1_f32(svfloat32_t, svfloat32_t);
	8054	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_f16)))
	8055	svfloat16_t svzipq1_f16(svfloat16_t, svfloat16_t);
	8056	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_s32)))
	8057	svint32_t svzipq1_s32(svint32_t, svint32_t);
	8058	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_s64)))
	8059	svint64_t svzipq1_s64(svint64_t, svint64_t);
	8060	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_mf8)))
	8061	svmfloat8_t svzipq1_mf8(svmfloat8_t, svmfloat8_t);
	8062	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_s16)))
	8063	svint16_t svzipq1_s16(svint16_t, svint16_t);
	8064	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_u8)))
	8065	svuint8_t svzipq2_u8(svuint8_t, svuint8_t);
	8066	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_u32)))
	8067	svuint32_t svzipq2_u32(svuint32_t, svuint32_t);
	8068	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_u64)))
	8069	svuint64_t svzipq2_u64(svuint64_t, svuint64_t);
	8070	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_u16)))
	8071	svuint16_t svzipq2_u16(svuint16_t, svuint16_t);
	8072	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_bf16)))
	8073	svbfloat16_t svzipq2_bf16(svbfloat16_t, svbfloat16_t);
	8074	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_s8)))
	8075	svint8_t svzipq2_s8(svint8_t, svint8_t);
	8076	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_f64)))
	8077	svfloat64_t svzipq2_f64(svfloat64_t, svfloat64_t);
	8078	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_f32)))
	8079	svfloat32_t svzipq2_f32(svfloat32_t, svfloat32_t);
	8080	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_f16)))
	8081	svfloat16_t svzipq2_f16(svfloat16_t, svfloat16_t);
	8082	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_s32)))
	8083	svint32_t svzipq2_s32(svint32_t, svint32_t);
	8084	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_s64)))
	8085	svint64_t svzipq2_s64(svint64_t, svint64_t);
	8086	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_mf8)))
	8087	svmfloat8_t svzipq2_mf8(svmfloat8_t, svmfloat8_t);
	8088	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_s16)))
	8089	svint16_t svzipq2_s16(svint16_t, svint16_t);
	8090	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_u8)))
	8091	uint8x16_t svaddqv(svbool_t, svuint8_t);
	8092	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_u32)))
	8093	uint32x4_t svaddqv(svbool_t, svuint32_t);
	8094	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_u64)))
	8095	uint64x2_t svaddqv(svbool_t, svuint64_t);
	8096	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_u16)))
	8097	uint16x8_t svaddqv(svbool_t, svuint16_t);
	8098	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_s8)))
	8099	int8x16_t svaddqv(svbool_t, svint8_t);
	8100	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_s32)))
	8101	int32x4_t svaddqv(svbool_t, svint32_t);
	8102	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_s64)))
	8103	int64x2_t svaddqv(svbool_t, svint64_t);
	8104	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_s16)))
	8105	int16x8_t svaddqv(svbool_t, svint16_t);
	8106	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_f64)))
	8107	float64x2_t svaddqv(svbool_t, svfloat64_t);
	8108	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_f32)))
	8109	float32x4_t svaddqv(svbool_t, svfloat32_t);
	8110	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaddqv_f16)))
	8111	float16x8_t svaddqv(svbool_t, svfloat16_t);
	8112	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_u8)))
	8113	uint8x16_t svandqv(svbool_t, svuint8_t);
	8114	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_u32)))
	8115	uint32x4_t svandqv(svbool_t, svuint32_t);
	8116	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_u64)))
	8117	uint64x2_t svandqv(svbool_t, svuint64_t);
	8118	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_u16)))
	8119	uint16x8_t svandqv(svbool_t, svuint16_t);
	8120	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_s8)))
	8121	int8x16_t svandqv(svbool_t, svint8_t);
	8122	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_s32)))
	8123	int32x4_t svandqv(svbool_t, svint32_t);
	8124	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_s64)))
	8125	int64x2_t svandqv(svbool_t, svint64_t);
	8126	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svandqv_s16)))
	8127	int16x8_t svandqv(svbool_t, svint16_t);
	8128	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_u8)))
	8129	svuint8_t svdup_laneq(svuint8_t, uint64_t);
	8130	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_s8)))
	8131	svint8_t svdup_laneq(svint8_t, uint64_t);
	8132	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_mf8)))
	8133	svmfloat8_t svdup_laneq(svmfloat8_t, uint64_t);
	8134	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_u64)))
	8135	svuint64_t svdup_laneq(svuint64_t, uint64_t);
	8136	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_f64)))
	8137	svfloat64_t svdup_laneq(svfloat64_t, uint64_t);
	8138	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_s64)))
	8139	svint64_t svdup_laneq(svint64_t, uint64_t);
	8140	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_u16)))
	8141	svuint16_t svdup_laneq(svuint16_t, uint64_t);
11186	8142	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_bf16)))
11187	8143	svbfloat16_t svdup_laneq(svbfloat16_t, uint64_t);
11188		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_s8)))
11189		svint8_t svclamp_s8(svint8_t, svint8_t, svint8_t);
11190		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_s32)))
11191		svint32_t svclamp_s32(svint32_t, svint32_t, svint32_t);
11192		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_s64)))
11193		svint64_t svclamp_s64(svint64_t, svint64_t, svint64_t);
11194		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_s16)))
11195		svint16_t svclamp_s16(svint16_t, svint16_t, svint16_t);
11196		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_u8)))
11197		svuint8_t svclamp_u8(svuint8_t, svuint8_t, svuint8_t);
11198		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_u32)))
11199		svuint32_t svclamp_u32(svuint32_t, svuint32_t, svuint32_t);
11200		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_u64)))
11201		svuint64_t svclamp_u64(svuint64_t, svuint64_t, svuint64_t);
11202		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_u16)))
11203		svuint16_t svclamp_u16(svuint16_t, svuint16_t, svuint16_t);
11204		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpsel_lane_b16)))
11205		svbool_t svpsel_lane_b16(svbool_t, svbool_t, uint32_t);
11206		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpsel_lane_b32)))
11207		svbool_t svpsel_lane_b32(svbool_t, svbool_t, uint32_t);
11208		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpsel_lane_b64)))
11209		svbool_t svpsel_lane_b64(svbool_t, svbool_t, uint32_t);
11210		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpsel_lane_b8)))
11211		svbool_t svpsel_lane_b8(svbool_t, svbool_t, uint32_t);
11212		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u8_m)))
11213		svuint8_t svrevd_u8_m(svuint8_t, svbool_t, svuint8_t);
11214		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u32_m)))
11215		svuint32_t svrevd_u32_m(svuint32_t, svbool_t, svuint32_t);
11216		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u64_m)))
11217		svuint64_t svrevd_u64_m(svuint64_t, svbool_t, svuint64_t);
11218		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u16_m)))
11219		svuint16_t svrevd_u16_m(svuint16_t, svbool_t, svuint16_t);
11220		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_bf16_m)))
11221		svbfloat16_t svrevd_bf16_m(svbfloat16_t, svbool_t, svbfloat16_t);
11222		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s8_m)))
11223		svint8_t svrevd_s8_m(svint8_t, svbool_t, svint8_t);
11224		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f64_m)))
11225		svfloat64_t svrevd_f64_m(svfloat64_t, svbool_t, svfloat64_t);
11226		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f32_m)))
11227		svfloat32_t svrevd_f32_m(svfloat32_t, svbool_t, svfloat32_t);
11228		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f16_m)))
11229		svfloat16_t svrevd_f16_m(svfloat16_t, svbool_t, svfloat16_t);
11230		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s32_m)))
11231		svint32_t svrevd_s32_m(svint32_t, svbool_t, svint32_t);
11232		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s64_m)))
11233		svint64_t svrevd_s64_m(svint64_t, svbool_t, svint64_t);
11234		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s16_m)))
11235		svint16_t svrevd_s16_m(svint16_t, svbool_t, svint16_t);
11236		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u8_x)))
11237		svuint8_t svrevd_u8_x(svbool_t, svuint8_t);
11238		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u32_x)))
11239		svuint32_t svrevd_u32_x(svbool_t, svuint32_t);
11240		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u64_x)))
11241		svuint64_t svrevd_u64_x(svbool_t, svuint64_t);
11242		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u16_x)))
11243		svuint16_t svrevd_u16_x(svbool_t, svuint16_t);
11244		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_bf16_x)))
11245		svbfloat16_t svrevd_bf16_x(svbool_t, svbfloat16_t);
11246		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s8_x)))
11247		svint8_t svrevd_s8_x(svbool_t, svint8_t);
11248		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f64_x)))
11249		svfloat64_t svrevd_f64_x(svbool_t, svfloat64_t);
11250		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f32_x)))
11251		svfloat32_t svrevd_f32_x(svbool_t, svfloat32_t);
11252		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f16_x)))
11253		svfloat16_t svrevd_f16_x(svbool_t, svfloat16_t);
11254		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s32_x)))
11255		svint32_t svrevd_s32_x(svbool_t, svint32_t);
11256		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s64_x)))
11257		svint64_t svrevd_s64_x(svbool_t, svint64_t);
11258		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s16_x)))
11259		svint16_t svrevd_s16_x(svbool_t, svint16_t);
11260		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u8_z)))
11261		svuint8_t svrevd_u8_z(svbool_t, svuint8_t);
11262		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u32_z)))
11263		svuint32_t svrevd_u32_z(svbool_t, svuint32_t);
11264		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u64_z)))
11265		svuint64_t svrevd_u64_z(svbool_t, svuint64_t);
11266		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u16_z)))
11267		svuint16_t svrevd_u16_z(svbool_t, svuint16_t);
11268		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_bf16_z)))
11269		svbfloat16_t svrevd_bf16_z(svbool_t, svbfloat16_t);
11270		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s8_z)))
11271		svint8_t svrevd_s8_z(svbool_t, svint8_t);
11272		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f64_z)))
11273		svfloat64_t svrevd_f64_z(svbool_t, svfloat64_t);
11274		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f32_z)))
11275		svfloat32_t svrevd_f32_z(svbool_t, svfloat32_t);
11276		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f16_z)))
11277		svfloat16_t svrevd_f16_z(svbool_t, svfloat16_t);
11278		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s32_z)))
11279		svint32_t svrevd_s32_z(svbool_t, svint32_t);
11280		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s64_z)))
11281		svint64_t svrevd_s64_z(svbool_t, svint64_t);
11282		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s16_z)))
11283		svint16_t svrevd_s16_z(svbool_t, svint16_t);
11284		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_s8)))
11285		svint8_t svclamp(svint8_t, svint8_t, svint8_t);
11286		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_s32)))
11287		svint32_t svclamp(svint32_t, svint32_t, svint32_t);
11288		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_s64)))
11289		svint64_t svclamp(svint64_t, svint64_t, svint64_t);
11290		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_s16)))
11291		svint16_t svclamp(svint16_t, svint16_t, svint16_t);
11292		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_u8)))
11293		svuint8_t svclamp(svuint8_t, svuint8_t, svuint8_t);
11294		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_u32)))
11295		svuint32_t svclamp(svuint32_t, svuint32_t, svuint32_t);
11296		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_u64)))
11297		svuint64_t svclamp(svuint64_t, svuint64_t, svuint64_t);
11298		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_u16)))
11299		svuint16_t svclamp(svuint16_t, svuint16_t, svuint16_t);
11300		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u8_m)))
11301		svuint8_t svrevd_m(svuint8_t, svbool_t, svuint8_t);
11302		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u32_m)))
11303		svuint32_t svrevd_m(svuint32_t, svbool_t, svuint32_t);
11304		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u64_m)))
11305		svuint64_t svrevd_m(svuint64_t, svbool_t, svuint64_t);
11306		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u16_m)))
11307		svuint16_t svrevd_m(svuint16_t, svbool_t, svuint16_t);
11308		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_bf16_m)))
11309		svbfloat16_t svrevd_m(svbfloat16_t, svbool_t, svbfloat16_t);
11310		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s8_m)))
11311		svint8_t svrevd_m(svint8_t, svbool_t, svint8_t);
11312		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f64_m)))
11313		svfloat64_t svrevd_m(svfloat64_t, svbool_t, svfloat64_t);
11314		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f32_m)))
11315		svfloat32_t svrevd_m(svfloat32_t, svbool_t, svfloat32_t);
11316		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f16_m)))
11317		svfloat16_t svrevd_m(svfloat16_t, svbool_t, svfloat16_t);
11318		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s32_m)))
11319		svint32_t svrevd_m(svint32_t, svbool_t, svint32_t);
11320		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s64_m)))
11321		svint64_t svrevd_m(svint64_t, svbool_t, svint64_t);
11322		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s16_m)))
11323		svint16_t svrevd_m(svint16_t, svbool_t, svint16_t);
11324		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u8_x)))
11325		svuint8_t svrevd_x(svbool_t, svuint8_t);
11326		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u32_x)))
11327		svuint32_t svrevd_x(svbool_t, svuint32_t);
11328		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u64_x)))
11329		svuint64_t svrevd_x(svbool_t, svuint64_t);
11330		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u16_x)))
11331		svuint16_t svrevd_x(svbool_t, svuint16_t);
11332		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_bf16_x)))
11333		svbfloat16_t svrevd_x(svbool_t, svbfloat16_t);
11334		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s8_x)))
11335		svint8_t svrevd_x(svbool_t, svint8_t);
11336		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f64_x)))
11337		svfloat64_t svrevd_x(svbool_t, svfloat64_t);
11338		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f32_x)))
11339		svfloat32_t svrevd_x(svbool_t, svfloat32_t);
11340		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f16_x)))
11341		svfloat16_t svrevd_x(svbool_t, svfloat16_t);
11342		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s32_x)))
11343		svint32_t svrevd_x(svbool_t, svint32_t);
11344		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s64_x)))
11345		svint64_t svrevd_x(svbool_t, svint64_t);
11346		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s16_x)))
11347		svint16_t svrevd_x(svbool_t, svint16_t);
11348		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u8_z)))
11349		svuint8_t svrevd_z(svbool_t, svuint8_t);
11350		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u32_z)))
11351		svuint32_t svrevd_z(svbool_t, svuint32_t);
11352		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u64_z)))
11353		svuint64_t svrevd_z(svbool_t, svuint64_t);
11354		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_u16_z)))
11355		svuint16_t svrevd_z(svbool_t, svuint16_t);
11356		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_bf16_z)))
11357		svbfloat16_t svrevd_z(svbool_t, svbfloat16_t);
11358		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s8_z)))
11359		svint8_t svrevd_z(svbool_t, svint8_t);
11360		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f64_z)))
11361		svfloat64_t svrevd_z(svbool_t, svfloat64_t);
11362		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f32_z)))
11363		svfloat32_t svrevd_z(svbool_t, svfloat32_t);
11364		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_f16_z)))
11365		svfloat16_t svrevd_z(svbool_t, svfloat16_t);
11366		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s32_z)))
11367		svint32_t svrevd_z(svbool_t, svint32_t);
11368		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s64_z)))
11369		svint64_t svrevd_z(svbool_t, svint64_t);
11370		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrevd_s16_z)))
11371		svint16_t svrevd_z(svbool_t, svint16_t);
11372		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlslb_f32)))
11373		svfloat32_t svbfmlslb_f32(svfloat32_t, svbfloat16_t, svbfloat16_t);
11374		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlslb_lane_f32)))
11375		svfloat32_t svbfmlslb_lane_f32(svfloat32_t, svbfloat16_t, svbfloat16_t, uint64_t);
11376		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlslt_f32)))
11377		svfloat32_t svbfmlslt_f32(svfloat32_t, svbfloat16_t, svbfloat16_t);
11378		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlslt_lane_f32)))
11379		svfloat32_t svbfmlslt_lane_f32(svfloat32_t, svbfloat16_t, svbfloat16_t, uint64_t);
11380		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_f64)))
11381		svfloat64_t svclamp_f64(svfloat64_t, svfloat64_t, svfloat64_t);
11382		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_f32)))
11383		svfloat32_t svclamp_f32(svfloat32_t, svfloat32_t, svfloat32_t);
11384		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_f16)))
11385		svfloat16_t svclamp_f16(svfloat16_t, svfloat16_t, svfloat16_t);
	8144	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_f16)))
	8145	svfloat16_t svdup_laneq(svfloat16_t, uint64_t);
	8146	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_s16)))
	8147	svint16_t svdup_laneq(svint16_t, uint64_t);
	8148	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_u32)))
	8149	svuint32_t svdup_laneq(svuint32_t, uint64_t);
	8150	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_f32)))
	8151	svfloat32_t svdup_laneq(svfloat32_t, uint64_t);
	8152	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_s32)))
	8153	svint32_t svdup_laneq(svint32_t, uint64_t);
	8154	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_u8)))
	8155	uint8x16_t sveorqv(svbool_t, svuint8_t);
	8156	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_u32)))
	8157	uint32x4_t sveorqv(svbool_t, svuint32_t);
	8158	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_u64)))
	8159	uint64x2_t sveorqv(svbool_t, svuint64_t);
	8160	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_u16)))
	8161	uint16x8_t sveorqv(svbool_t, svuint16_t);
	8162	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_s8)))
	8163	int8x16_t sveorqv(svbool_t, svint8_t);
	8164	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_s32)))
	8165	int32x4_t sveorqv(svbool_t, svint32_t);
	8166	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_s64)))
	8167	int64x2_t sveorqv(svbool_t, svint64_t);
	8168	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_sveorqv_s16)))
	8169	int16x8_t sveorqv(svbool_t, svint16_t);
	8170	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_u8)))
	8171	svuint8_t svextq(svuint8_t, svuint8_t, int32_t);
	8172	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_u32)))
	8173	svuint32_t svextq(svuint32_t, svuint32_t, int32_t);
	8174	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_u64)))
	8175	svuint64_t svextq(svuint64_t, svuint64_t, int32_t);
	8176	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_u16)))
	8177	svuint16_t svextq(svuint16_t, svuint16_t, int32_t);
	8178	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_bf16)))
	8179	svbfloat16_t svextq(svbfloat16_t, svbfloat16_t, int32_t);
	8180	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_s8)))
	8181	svint8_t svextq(svint8_t, svint8_t, int32_t);
	8182	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_f64)))
	8183	svfloat64_t svextq(svfloat64_t, svfloat64_t, int32_t);
	8184	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_f32)))
	8185	svfloat32_t svextq(svfloat32_t, svfloat32_t, int32_t);
	8186	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_f16)))
	8187	svfloat16_t svextq(svfloat16_t, svfloat16_t, int32_t);
	8188	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_s32)))
	8189	svint32_t svextq(svint32_t, svint32_t, int32_t);
	8190	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_s64)))
	8191	svint64_t svextq(svint64_t, svint64_t, int32_t);
	8192	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_mf8)))
	8193	svmfloat8_t svextq(svmfloat8_t, svmfloat8_t, int32_t);
	8194	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svextq_s16)))
	8195	svint16_t svextq(svint16_t, svint16_t, int32_t);
	8196	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_u8)))
	8197	svuint8x2_t svld2q(svbool_t, uint8_t const *);
	8198	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_u32)))
	8199	svuint32x2_t svld2q(svbool_t, uint32_t const *);
	8200	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_u64)))
	8201	svuint64x2_t svld2q(svbool_t, uint64_t const *);
	8202	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_u16)))
	8203	svuint16x2_t svld2q(svbool_t, uint16_t const *);
	8204	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_bf16)))
	8205	svbfloat16x2_t svld2q(svbool_t, bfloat16_t const *);
	8206	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_s8)))
	8207	svint8x2_t svld2q(svbool_t, int8_t const *);
	8208	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_f64)))
	8209	svfloat64x2_t svld2q(svbool_t, float64_t const *);
	8210	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_f32)))
	8211	svfloat32x2_t svld2q(svbool_t, float32_t const *);
	8212	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_f16)))
	8213	svfloat16x2_t svld2q(svbool_t, float16_t const *);
	8214	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_s32)))
	8215	svint32x2_t svld2q(svbool_t, int32_t const *);
	8216	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_s64)))
	8217	svint64x2_t svld2q(svbool_t, int64_t const *);
	8218	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_mf8)))
	8219	svmfloat8x2_t svld2q(svbool_t, mfloat8_t const *);
	8220	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_s16)))
	8221	svint16x2_t svld2q(svbool_t, int16_t const *);
	8222	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_u8)))
	8223	svuint8x2_t svld2q_vnum(svbool_t, uint8_t const *, int64_t);
	8224	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_u32)))
	8225	svuint32x2_t svld2q_vnum(svbool_t, uint32_t const *, int64_t);
	8226	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_u64)))
	8227	svuint64x2_t svld2q_vnum(svbool_t, uint64_t const *, int64_t);
	8228	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_u16)))
	8229	svuint16x2_t svld2q_vnum(svbool_t, uint16_t const *, int64_t);
	8230	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_bf16)))
	8231	svbfloat16x2_t svld2q_vnum(svbool_t, bfloat16_t const *, int64_t);
	8232	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_s8)))
	8233	svint8x2_t svld2q_vnum(svbool_t, int8_t const *, int64_t);
	8234	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_f64)))
	8235	svfloat64x2_t svld2q_vnum(svbool_t, float64_t const *, int64_t);
	8236	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_f32)))
	8237	svfloat32x2_t svld2q_vnum(svbool_t, float32_t const *, int64_t);
	8238	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_f16)))
	8239	svfloat16x2_t svld2q_vnum(svbool_t, float16_t const *, int64_t);
	8240	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_s32)))
	8241	svint32x2_t svld2q_vnum(svbool_t, int32_t const *, int64_t);
	8242	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_s64)))
	8243	svint64x2_t svld2q_vnum(svbool_t, int64_t const *, int64_t);
	8244	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_mf8)))
	8245	svmfloat8x2_t svld2q_vnum(svbool_t, mfloat8_t const *, int64_t);
	8246	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2q_vnum_s16)))
	8247	svint16x2_t svld2q_vnum(svbool_t, int16_t const *, int64_t);
	8248	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_u8)))
	8249	svuint8x3_t svld3q(svbool_t, uint8_t const *);
	8250	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_u32)))
	8251	svuint32x3_t svld3q(svbool_t, uint32_t const *);
	8252	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_u64)))
	8253	svuint64x3_t svld3q(svbool_t, uint64_t const *);
	8254	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_u16)))
	8255	svuint16x3_t svld3q(svbool_t, uint16_t const *);
	8256	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_bf16)))
	8257	svbfloat16x3_t svld3q(svbool_t, bfloat16_t const *);
	8258	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_s8)))
	8259	svint8x3_t svld3q(svbool_t, int8_t const *);
	8260	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_f64)))
	8261	svfloat64x3_t svld3q(svbool_t, float64_t const *);
	8262	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_f32)))
	8263	svfloat32x3_t svld3q(svbool_t, float32_t const *);
	8264	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_f16)))
	8265	svfloat16x3_t svld3q(svbool_t, float16_t const *);
	8266	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_s32)))
	8267	svint32x3_t svld3q(svbool_t, int32_t const *);
	8268	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_s64)))
	8269	svint64x3_t svld3q(svbool_t, int64_t const *);
	8270	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_mf8)))
	8271	svmfloat8x3_t svld3q(svbool_t, mfloat8_t const *);
	8272	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_s16)))
	8273	svint16x3_t svld3q(svbool_t, int16_t const *);
	8274	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_u8)))
	8275	svuint8x3_t svld3q_vnum(svbool_t, uint8_t const *, int64_t);
	8276	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_u32)))
	8277	svuint32x3_t svld3q_vnum(svbool_t, uint32_t const *, int64_t);
	8278	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_u64)))
	8279	svuint64x3_t svld3q_vnum(svbool_t, uint64_t const *, int64_t);
	8280	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_u16)))
	8281	svuint16x3_t svld3q_vnum(svbool_t, uint16_t const *, int64_t);
	8282	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_bf16)))
	8283	svbfloat16x3_t svld3q_vnum(svbool_t, bfloat16_t const *, int64_t);
	8284	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_s8)))
	8285	svint8x3_t svld3q_vnum(svbool_t, int8_t const *, int64_t);
	8286	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_f64)))
	8287	svfloat64x3_t svld3q_vnum(svbool_t, float64_t const *, int64_t);
	8288	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_f32)))
	8289	svfloat32x3_t svld3q_vnum(svbool_t, float32_t const *, int64_t);
	8290	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_f16)))
	8291	svfloat16x3_t svld3q_vnum(svbool_t, float16_t const *, int64_t);
	8292	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_s32)))
	8293	svint32x3_t svld3q_vnum(svbool_t, int32_t const *, int64_t);
	8294	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_s64)))
	8295	svint64x3_t svld3q_vnum(svbool_t, int64_t const *, int64_t);
	8296	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_mf8)))
	8297	svmfloat8x3_t svld3q_vnum(svbool_t, mfloat8_t const *, int64_t);
	8298	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3q_vnum_s16)))
	8299	svint16x3_t svld3q_vnum(svbool_t, int16_t const *, int64_t);
	8300	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_u8)))
	8301	svuint8x4_t svld4q(svbool_t, uint8_t const *);
	8302	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_u32)))
	8303	svuint32x4_t svld4q(svbool_t, uint32_t const *);
	8304	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_u64)))
	8305	svuint64x4_t svld4q(svbool_t, uint64_t const *);
	8306	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_u16)))
	8307	svuint16x4_t svld4q(svbool_t, uint16_t const *);
	8308	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_bf16)))
	8309	svbfloat16x4_t svld4q(svbool_t, bfloat16_t const *);
	8310	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_s8)))
	8311	svint8x4_t svld4q(svbool_t, int8_t const *);
	8312	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_f64)))
	8313	svfloat64x4_t svld4q(svbool_t, float64_t const *);
	8314	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_f32)))
	8315	svfloat32x4_t svld4q(svbool_t, float32_t const *);
	8316	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_f16)))
	8317	svfloat16x4_t svld4q(svbool_t, float16_t const *);
	8318	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_s32)))
	8319	svint32x4_t svld4q(svbool_t, int32_t const *);
	8320	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_s64)))
	8321	svint64x4_t svld4q(svbool_t, int64_t const *);
	8322	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_mf8)))
	8323	svmfloat8x4_t svld4q(svbool_t, mfloat8_t const *);
	8324	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_s16)))
	8325	svint16x4_t svld4q(svbool_t, int16_t const *);
	8326	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_u8)))
	8327	svuint8x4_t svld4q_vnum(svbool_t, uint8_t const *, int64_t);
	8328	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_u32)))
	8329	svuint32x4_t svld4q_vnum(svbool_t, uint32_t const *, int64_t);
	8330	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_u64)))
	8331	svuint64x4_t svld4q_vnum(svbool_t, uint64_t const *, int64_t);
	8332	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_u16)))
	8333	svuint16x4_t svld4q_vnum(svbool_t, uint16_t const *, int64_t);
	8334	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_bf16)))
	8335	svbfloat16x4_t svld4q_vnum(svbool_t, bfloat16_t const *, int64_t);
	8336	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_s8)))
	8337	svint8x4_t svld4q_vnum(svbool_t, int8_t const *, int64_t);
	8338	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_f64)))
	8339	svfloat64x4_t svld4q_vnum(svbool_t, float64_t const *, int64_t);
	8340	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_f32)))
	8341	svfloat32x4_t svld4q_vnum(svbool_t, float32_t const *, int64_t);
	8342	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_f16)))
	8343	svfloat16x4_t svld4q_vnum(svbool_t, float16_t const *, int64_t);
	8344	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_s32)))
	8345	svint32x4_t svld4q_vnum(svbool_t, int32_t const *, int64_t);
	8346	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_s64)))
	8347	svint64x4_t svld4q_vnum(svbool_t, int64_t const *, int64_t);
	8348	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_mf8)))
	8349	svmfloat8x4_t svld4q_vnum(svbool_t, mfloat8_t const *, int64_t);
	8350	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4q_vnum_s16)))
	8351	svint16x4_t svld4q_vnum(svbool_t, int16_t const *, int64_t);
	8352	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnmqv_f64)))
	8353	float64x2_t svmaxnmqv(svbool_t, svfloat64_t);
	8354	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnmqv_f32)))
	8355	float32x4_t svmaxnmqv(svbool_t, svfloat32_t);
	8356	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnmqv_f16)))
	8357	float16x8_t svmaxnmqv(svbool_t, svfloat16_t);
	8358	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_f64)))
	8359	float64x2_t svmaxqv(svbool_t, svfloat64_t);
	8360	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_f32)))
	8361	float32x4_t svmaxqv(svbool_t, svfloat32_t);
	8362	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_f16)))
	8363	float16x8_t svmaxqv(svbool_t, svfloat16_t);
	8364	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_s8)))
	8365	int8x16_t svmaxqv(svbool_t, svint8_t);
	8366	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_s32)))
	8367	int32x4_t svmaxqv(svbool_t, svint32_t);
	8368	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_s64)))
	8369	int64x2_t svmaxqv(svbool_t, svint64_t);
	8370	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_s16)))
	8371	int16x8_t svmaxqv(svbool_t, svint16_t);
	8372	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_u8)))
	8373	uint8x16_t svmaxqv(svbool_t, svuint8_t);
	8374	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_u32)))
	8375	uint32x4_t svmaxqv(svbool_t, svuint32_t);
	8376	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_u64)))
	8377	uint64x2_t svmaxqv(svbool_t, svuint64_t);
	8378	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxqv_u16)))
	8379	uint16x8_t svmaxqv(svbool_t, svuint16_t);
	8380	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnmqv_f64)))
	8381	float64x2_t svminnmqv(svbool_t, svfloat64_t);
	8382	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnmqv_f32)))
	8383	float32x4_t svminnmqv(svbool_t, svfloat32_t);
	8384	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnmqv_f16)))
	8385	float16x8_t svminnmqv(svbool_t, svfloat16_t);
	8386	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_f64)))
	8387	float64x2_t svminqv(svbool_t, svfloat64_t);
	8388	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_f32)))
	8389	float32x4_t svminqv(svbool_t, svfloat32_t);
	8390	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_f16)))
	8391	float16x8_t svminqv(svbool_t, svfloat16_t);
	8392	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_s8)))
	8393	int8x16_t svminqv(svbool_t, svint8_t);
	8394	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_s32)))
	8395	int32x4_t svminqv(svbool_t, svint32_t);
	8396	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_s64)))
	8397	int64x2_t svminqv(svbool_t, svint64_t);
	8398	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_s16)))
	8399	int16x8_t svminqv(svbool_t, svint16_t);
	8400	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_u8)))
	8401	uint8x16_t svminqv(svbool_t, svuint8_t);
	8402	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_u32)))
	8403	uint32x4_t svminqv(svbool_t, svuint32_t);
	8404	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_u64)))
	8405	uint64x2_t svminqv(svbool_t, svuint64_t);
	8406	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminqv_u16)))
	8407	uint16x8_t svminqv(svbool_t, svuint16_t);
	8408	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_u8)))
	8409	uint8x16_t svorqv(svbool_t, svuint8_t);
	8410	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_u32)))
	8411	uint32x4_t svorqv(svbool_t, svuint32_t);
	8412	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_u64)))
	8413	uint64x2_t svorqv(svbool_t, svuint64_t);
	8414	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_u16)))
	8415	uint16x8_t svorqv(svbool_t, svuint16_t);
	8416	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_s8)))
	8417	int8x16_t svorqv(svbool_t, svint8_t);
	8418	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_s32)))
	8419	int32x4_t svorqv(svbool_t, svint32_t);
	8420	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_s64)))
	8421	int64x2_t svorqv(svbool_t, svint64_t);
	8422	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svorqv_s16)))
	8423	int16x8_t svorqv(svbool_t, svint16_t);
	8424	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_u8)))
	8425	svbool_t svpmov(svuint8_t);
	8426	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_s8)))
	8427	svbool_t svpmov(svint8_t);
	8428	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_u64)))
	8429	svbool_t svpmov(svuint64_t);
	8430	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_s64)))
	8431	svbool_t svpmov(svint64_t);
	8432	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_u16)))
	8433	svbool_t svpmov(svuint16_t);
	8434	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_s16)))
	8435	svbool_t svpmov(svint16_t);
	8436	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_u32)))
	8437	svbool_t svpmov(svuint32_t);
	8438	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_s32)))
	8439	svbool_t svpmov(svint32_t);
	8440	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_u8)))
	8441	svbool_t svpmov_lane(svuint8_t, uint64_t);
	8442	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_s8)))
	8443	svbool_t svpmov_lane(svint8_t, uint64_t);
	8444	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_u64)))
	8445	svbool_t svpmov_lane(svuint64_t, uint64_t);
	8446	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_s64)))
	8447	svbool_t svpmov_lane(svint64_t, uint64_t);
	8448	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_u16)))
	8449	svbool_t svpmov_lane(svuint16_t, uint64_t);
	8450	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_s16)))
	8451	svbool_t svpmov_lane(svint16_t, uint64_t);
	8452	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_u32)))
	8453	svbool_t svpmov_lane(svuint32_t, uint64_t);
	8454	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_s32)))
	8455	svbool_t svpmov_lane(svint32_t, uint64_t);
	8456	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_u64_m)))
	8457	svuint64_t svpmov_lane_m(svuint64_t, svbool_t, uint64_t);
	8458	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_s64_m)))
	8459	svint64_t svpmov_lane_m(svint64_t, svbool_t, uint64_t);
	8460	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_u16_m)))
	8461	svuint16_t svpmov_lane_m(svuint16_t, svbool_t, uint64_t);
	8462	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_s16_m)))
	8463	svint16_t svpmov_lane_m(svint16_t, svbool_t, uint64_t);
	8464	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_u32_m)))
	8465	svuint32_t svpmov_lane_m(svuint32_t, svbool_t, uint64_t);
	8466	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmov_lane_s32_m)))
	8467	svint32_t svpmov_lane_m(svint32_t, svbool_t, uint64_t);
	8468	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_u8)))
	8469	void svst2q(svbool_t, uint8_t const *, svuint8x2_t);
	8470	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_u32)))
	8471	void svst2q(svbool_t, uint32_t const *, svuint32x2_t);
	8472	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_u64)))
	8473	void svst2q(svbool_t, uint64_t const *, svuint64x2_t);
	8474	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_u16)))
	8475	void svst2q(svbool_t, uint16_t const *, svuint16x2_t);
	8476	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_bf16)))
	8477	void svst2q(svbool_t, bfloat16_t const *, svbfloat16x2_t);
	8478	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_s8)))
	8479	void svst2q(svbool_t, int8_t const *, svint8x2_t);
	8480	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_f64)))
	8481	void svst2q(svbool_t, float64_t const *, svfloat64x2_t);
	8482	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_f32)))
	8483	void svst2q(svbool_t, float32_t const *, svfloat32x2_t);
	8484	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_f16)))
	8485	void svst2q(svbool_t, float16_t const *, svfloat16x2_t);
	8486	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_s32)))
	8487	void svst2q(svbool_t, int32_t const *, svint32x2_t);
	8488	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_s64)))
	8489	void svst2q(svbool_t, int64_t const *, svint64x2_t);
	8490	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_mf8)))
	8491	void svst2q(svbool_t, mfloat8_t const *, svmfloat8x2_t);
	8492	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_s16)))
	8493	void svst2q(svbool_t, int16_t const *, svint16x2_t);
	8494	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_u8)))
	8495	void svst2q_vnum(svbool_t, uint8_t const *, int64_t, svuint8x2_t);
	8496	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_u32)))
	8497	void svst2q_vnum(svbool_t, uint32_t const *, int64_t, svuint32x2_t);
	8498	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_u64)))
	8499	void svst2q_vnum(svbool_t, uint64_t const *, int64_t, svuint64x2_t);
	8500	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_u16)))
	8501	void svst2q_vnum(svbool_t, uint16_t const *, int64_t, svuint16x2_t);
	8502	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_bf16)))
	8503	void svst2q_vnum(svbool_t, bfloat16_t const *, int64_t, svbfloat16x2_t);
	8504	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_s8)))
	8505	void svst2q_vnum(svbool_t, int8_t const *, int64_t, svint8x2_t);
	8506	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_f64)))
	8507	void svst2q_vnum(svbool_t, float64_t const *, int64_t, svfloat64x2_t);
	8508	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_f32)))
	8509	void svst2q_vnum(svbool_t, float32_t const *, int64_t, svfloat32x2_t);
	8510	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_f16)))
	8511	void svst2q_vnum(svbool_t, float16_t const *, int64_t, svfloat16x2_t);
	8512	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_s32)))
	8513	void svst2q_vnum(svbool_t, int32_t const *, int64_t, svint32x2_t);
	8514	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_s64)))
	8515	void svst2q_vnum(svbool_t, int64_t const *, int64_t, svint64x2_t);
	8516	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_mf8)))
	8517	void svst2q_vnum(svbool_t, mfloat8_t const *, int64_t, svmfloat8x2_t);
	8518	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2q_vnum_s16)))
	8519	void svst2q_vnum(svbool_t, int16_t const *, int64_t, svint16x2_t);
	8520	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_u8)))
	8521	void svst3q(svbool_t, uint8_t const *, svuint8x3_t);
	8522	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_u32)))
	8523	void svst3q(svbool_t, uint32_t const *, svuint32x3_t);
	8524	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_u64)))
	8525	void svst3q(svbool_t, uint64_t const *, svuint64x3_t);
	8526	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_u16)))
	8527	void svst3q(svbool_t, uint16_t const *, svuint16x3_t);
	8528	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_bf16)))
	8529	void svst3q(svbool_t, bfloat16_t const *, svbfloat16x3_t);
	8530	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_s8)))
	8531	void svst3q(svbool_t, int8_t const *, svint8x3_t);
	8532	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_f64)))
	8533	void svst3q(svbool_t, float64_t const *, svfloat64x3_t);
	8534	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_f32)))
	8535	void svst3q(svbool_t, float32_t const *, svfloat32x3_t);
	8536	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_f16)))
	8537	void svst3q(svbool_t, float16_t const *, svfloat16x3_t);
	8538	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_s32)))
	8539	void svst3q(svbool_t, int32_t const *, svint32x3_t);
	8540	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_s64)))
	8541	void svst3q(svbool_t, int64_t const *, svint64x3_t);
	8542	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_mf8)))
	8543	void svst3q(svbool_t, mfloat8_t const *, svmfloat8x3_t);
	8544	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_s16)))
	8545	void svst3q(svbool_t, int16_t const *, svint16x3_t);
	8546	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_u8)))
	8547	void svst3q_vnum(svbool_t, uint8_t const *, int64_t, svuint8x3_t);
	8548	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_u32)))
	8549	void svst3q_vnum(svbool_t, uint32_t const *, int64_t, svuint32x3_t);
	8550	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_u64)))
	8551	void svst3q_vnum(svbool_t, uint64_t const *, int64_t, svuint64x3_t);
	8552	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_u16)))
	8553	void svst3q_vnum(svbool_t, uint16_t const *, int64_t, svuint16x3_t);
	8554	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_bf16)))
	8555	void svst3q_vnum(svbool_t, bfloat16_t const *, int64_t, svbfloat16x3_t);
	8556	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_s8)))
	8557	void svst3q_vnum(svbool_t, int8_t const *, int64_t, svint8x3_t);
	8558	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_f64)))
	8559	void svst3q_vnum(svbool_t, float64_t const *, int64_t, svfloat64x3_t);
	8560	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_f32)))
	8561	void svst3q_vnum(svbool_t, float32_t const *, int64_t, svfloat32x3_t);
	8562	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_f16)))
	8563	void svst3q_vnum(svbool_t, float16_t const *, int64_t, svfloat16x3_t);
	8564	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_s32)))
	8565	void svst3q_vnum(svbool_t, int32_t const *, int64_t, svint32x3_t);
	8566	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_s64)))
	8567	void svst3q_vnum(svbool_t, int64_t const *, int64_t, svint64x3_t);
	8568	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_mf8)))
	8569	void svst3q_vnum(svbool_t, mfloat8_t const *, int64_t, svmfloat8x3_t);
	8570	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3q_vnum_s16)))
	8571	void svst3q_vnum(svbool_t, int16_t const *, int64_t, svint16x3_t);
	8572	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_u8)))
	8573	void svst4q(svbool_t, uint8_t const *, svuint8x4_t);
	8574	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_u32)))
	8575	void svst4q(svbool_t, uint32_t const *, svuint32x4_t);
	8576	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_u64)))
	8577	void svst4q(svbool_t, uint64_t const *, svuint64x4_t);
	8578	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_u16)))
	8579	void svst4q(svbool_t, uint16_t const *, svuint16x4_t);
	8580	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_bf16)))
	8581	void svst4q(svbool_t, bfloat16_t const *, svbfloat16x4_t);
	8582	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_s8)))
	8583	void svst4q(svbool_t, int8_t const *, svint8x4_t);
	8584	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_f64)))
	8585	void svst4q(svbool_t, float64_t const *, svfloat64x4_t);
	8586	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_f32)))
	8587	void svst4q(svbool_t, float32_t const *, svfloat32x4_t);
	8588	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_f16)))
	8589	void svst4q(svbool_t, float16_t const *, svfloat16x4_t);
	8590	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_s32)))
	8591	void svst4q(svbool_t, int32_t const *, svint32x4_t);
	8592	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_s64)))
	8593	void svst4q(svbool_t, int64_t const *, svint64x4_t);
	8594	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_mf8)))
	8595	void svst4q(svbool_t, mfloat8_t const *, svmfloat8x4_t);
	8596	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_s16)))
	8597	void svst4q(svbool_t, int16_t const *, svint16x4_t);
	8598	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_u8)))
	8599	void svst4q_vnum(svbool_t, uint8_t const *, int64_t, svuint8x4_t);
	8600	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_u32)))
	8601	void svst4q_vnum(svbool_t, uint32_t const *, int64_t, svuint32x4_t);
	8602	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_u64)))
	8603	void svst4q_vnum(svbool_t, uint64_t const *, int64_t, svuint64x4_t);
	8604	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_u16)))
	8605	void svst4q_vnum(svbool_t, uint16_t const *, int64_t, svuint16x4_t);
	8606	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_bf16)))
	8607	void svst4q_vnum(svbool_t, bfloat16_t const *, int64_t, svbfloat16x4_t);
	8608	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_s8)))
	8609	void svst4q_vnum(svbool_t, int8_t const *, int64_t, svint8x4_t);
	8610	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_f64)))
	8611	void svst4q_vnum(svbool_t, float64_t const *, int64_t, svfloat64x4_t);
	8612	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_f32)))
	8613	void svst4q_vnum(svbool_t, float32_t const *, int64_t, svfloat32x4_t);
	8614	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_f16)))
	8615	void svst4q_vnum(svbool_t, float16_t const *, int64_t, svfloat16x4_t);
	8616	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_s32)))
	8617	void svst4q_vnum(svbool_t, int32_t const *, int64_t, svint32x4_t);
	8618	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_s64)))
	8619	void svst4q_vnum(svbool_t, int64_t const *, int64_t, svint64x4_t);
	8620	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_mf8)))
	8621	void svst4q_vnum(svbool_t, mfloat8_t const *, int64_t, svmfloat8x4_t);
	8622	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4q_vnum_s16)))
	8623	void svst4q_vnum(svbool_t, int16_t const *, int64_t, svint16x4_t);
	8624	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_u8)))
	8625	svuint8_t svtblq(svuint8_t, svuint8_t);
	8626	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_u32)))
	8627	svuint32_t svtblq(svuint32_t, svuint32_t);
	8628	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_u64)))
	8629	svuint64_t svtblq(svuint64_t, svuint64_t);
	8630	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_u16)))
	8631	svuint16_t svtblq(svuint16_t, svuint16_t);
	8632	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_bf16)))
	8633	svbfloat16_t svtblq(svbfloat16_t, svuint16_t);
	8634	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_s8)))
	8635	svint8_t svtblq(svint8_t, svuint8_t);
	8636	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_f64)))
	8637	svfloat64_t svtblq(svfloat64_t, svuint64_t);
	8638	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_f32)))
	8639	svfloat32_t svtblq(svfloat32_t, svuint32_t);
	8640	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_f16)))
	8641	svfloat16_t svtblq(svfloat16_t, svuint16_t);
	8642	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_s32)))
	8643	svint32_t svtblq(svint32_t, svuint32_t);
	8644	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_s64)))
	8645	svint64_t svtblq(svint64_t, svuint64_t);
	8646	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_mf8)))
	8647	svmfloat8_t svtblq(svmfloat8_t, svuint8_t);
	8648	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtblq_s16)))
	8649	svint16_t svtblq(svint16_t, svuint16_t);
	8650	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_u8)))
	8651	svuint8_t svtbxq(svuint8_t, svuint8_t, svuint8_t);
	8652	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_u32)))
	8653	svuint32_t svtbxq(svuint32_t, svuint32_t, svuint32_t);
	8654	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_u64)))
	8655	svuint64_t svtbxq(svuint64_t, svuint64_t, svuint64_t);
	8656	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_u16)))
	8657	svuint16_t svtbxq(svuint16_t, svuint16_t, svuint16_t);
	8658	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_bf16)))
	8659	svbfloat16_t svtbxq(svbfloat16_t, svbfloat16_t, svuint16_t);
	8660	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_s8)))
	8661	svint8_t svtbxq(svint8_t, svint8_t, svuint8_t);
	8662	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_f64)))
	8663	svfloat64_t svtbxq(svfloat64_t, svfloat64_t, svuint64_t);
	8664	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_f32)))
	8665	svfloat32_t svtbxq(svfloat32_t, svfloat32_t, svuint32_t);
	8666	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_f16)))
	8667	svfloat16_t svtbxq(svfloat16_t, svfloat16_t, svuint16_t);
	8668	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_s32)))
	8669	svint32_t svtbxq(svint32_t, svint32_t, svuint32_t);
	8670	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_s64)))
	8671	svint64_t svtbxq(svint64_t, svint64_t, svuint64_t);
	8672	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_mf8)))
	8673	svmfloat8_t svtbxq(svmfloat8_t, svmfloat8_t, svuint8_t);
	8674	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbxq_s16)))
	8675	svint16_t svtbxq(svint16_t, svint16_t, svuint16_t);
	8676	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_u8)))
	8677	svuint8_t svuzpq1(svuint8_t, svuint8_t);
	8678	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_u32)))
	8679	svuint32_t svuzpq1(svuint32_t, svuint32_t);
	8680	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_u64)))
	8681	svuint64_t svuzpq1(svuint64_t, svuint64_t);
	8682	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_u16)))
	8683	svuint16_t svuzpq1(svuint16_t, svuint16_t);
	8684	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_bf16)))
	8685	svbfloat16_t svuzpq1(svbfloat16_t, svbfloat16_t);
	8686	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_s8)))
	8687	svint8_t svuzpq1(svint8_t, svint8_t);
	8688	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_f64)))
	8689	svfloat64_t svuzpq1(svfloat64_t, svfloat64_t);
	8690	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_f32)))
	8691	svfloat32_t svuzpq1(svfloat32_t, svfloat32_t);
	8692	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_f16)))
	8693	svfloat16_t svuzpq1(svfloat16_t, svfloat16_t);
	8694	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_s32)))
	8695	svint32_t svuzpq1(svint32_t, svint32_t);
	8696	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_s64)))
	8697	svint64_t svuzpq1(svint64_t, svint64_t);
	8698	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_mf8)))
	8699	svmfloat8_t svuzpq1(svmfloat8_t, svmfloat8_t);
	8700	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq1_s16)))
	8701	svint16_t svuzpq1(svint16_t, svint16_t);
	8702	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_u8)))
	8703	svuint8_t svuzpq2(svuint8_t, svuint8_t);
	8704	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_u32)))
	8705	svuint32_t svuzpq2(svuint32_t, svuint32_t);
	8706	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_u64)))
	8707	svuint64_t svuzpq2(svuint64_t, svuint64_t);
	8708	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_u16)))
	8709	svuint16_t svuzpq2(svuint16_t, svuint16_t);
	8710	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_bf16)))
	8711	svbfloat16_t svuzpq2(svbfloat16_t, svbfloat16_t);
	8712	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_s8)))
	8713	svint8_t svuzpq2(svint8_t, svint8_t);
	8714	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_f64)))
	8715	svfloat64_t svuzpq2(svfloat64_t, svfloat64_t);
	8716	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_f32)))
	8717	svfloat32_t svuzpq2(svfloat32_t, svfloat32_t);
	8718	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_f16)))
	8719	svfloat16_t svuzpq2(svfloat16_t, svfloat16_t);
	8720	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_s32)))
	8721	svint32_t svuzpq2(svint32_t, svint32_t);
	8722	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_s64)))
	8723	svint64_t svuzpq2(svint64_t, svint64_t);
	8724	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_mf8)))
	8725	svmfloat8_t svuzpq2(svmfloat8_t, svmfloat8_t);
	8726	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzpq2_s16)))
	8727	svint16_t svuzpq2(svint16_t, svint16_t);
	8728	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_u8)))
	8729	svuint8_t svzipq1(svuint8_t, svuint8_t);
	8730	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_u32)))
	8731	svuint32_t svzipq1(svuint32_t, svuint32_t);
	8732	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_u64)))
	8733	svuint64_t svzipq1(svuint64_t, svuint64_t);
	8734	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_u16)))
	8735	svuint16_t svzipq1(svuint16_t, svuint16_t);
	8736	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_bf16)))
	8737	svbfloat16_t svzipq1(svbfloat16_t, svbfloat16_t);
	8738	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_s8)))
	8739	svint8_t svzipq1(svint8_t, svint8_t);
	8740	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_f64)))
	8741	svfloat64_t svzipq1(svfloat64_t, svfloat64_t);
	8742	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_f32)))
	8743	svfloat32_t svzipq1(svfloat32_t, svfloat32_t);
	8744	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_f16)))
	8745	svfloat16_t svzipq1(svfloat16_t, svfloat16_t);
	8746	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_s32)))
	8747	svint32_t svzipq1(svint32_t, svint32_t);
	8748	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_s64)))
	8749	svint64_t svzipq1(svint64_t, svint64_t);
	8750	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_mf8)))
	8751	svmfloat8_t svzipq1(svmfloat8_t, svmfloat8_t);
	8752	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq1_s16)))
	8753	svint16_t svzipq1(svint16_t, svint16_t);
	8754	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_u8)))
	8755	svuint8_t svzipq2(svuint8_t, svuint8_t);
	8756	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_u32)))
	8757	svuint32_t svzipq2(svuint32_t, svuint32_t);
	8758	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_u64)))
	8759	svuint64_t svzipq2(svuint64_t, svuint64_t);
	8760	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_u16)))
	8761	svuint16_t svzipq2(svuint16_t, svuint16_t);
	8762	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_bf16)))
	8763	svbfloat16_t svzipq2(svbfloat16_t, svbfloat16_t);
	8764	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_s8)))
	8765	svint8_t svzipq2(svint8_t, svint8_t);
	8766	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_f64)))
	8767	svfloat64_t svzipq2(svfloat64_t, svfloat64_t);
	8768	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_f32)))
	8769	svfloat32_t svzipq2(svfloat32_t, svfloat32_t);
	8770	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_f16)))
	8771	svfloat16_t svzipq2(svfloat16_t, svfloat16_t);
	8772	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_s32)))
	8773	svint32_t svzipq2(svint32_t, svint32_t);
	8774	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_s64)))
	8775	svint64_t svzipq2(svint64_t, svint64_t);
	8776	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_mf8)))
	8777	svmfloat8_t svzipq2(svmfloat8_t, svmfloat8_t);
	8778	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzipq2_s16)))
	8779	svint16_t svzipq2(svint16_t, svint16_t);
	8780	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmmla_f32)))
	8781	svfloat32_t svbfmmla_f32(svfloat32_t, svbfloat16_t, svbfloat16_t);
	8782	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmmla_f32)))
	8783	svfloat32_t svbfmmla(svfloat32_t, svbfloat16_t, svbfloat16_t);
	8784	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfdot_n_f32)))
	8785	svfloat32_t svbfdot_n_f32(svfloat32_t, svbfloat16_t, bfloat16_t);
	8786	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfdot_f32)))
	8787	svfloat32_t svbfdot_f32(svfloat32_t, svbfloat16_t, svbfloat16_t);
	8788	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfdot_lane_f32)))
	8789	svfloat32_t svbfdot_lane_f32(svfloat32_t, svbfloat16_t, svbfloat16_t, uint64_t);
	8790	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlalb_n_f32)))
	8791	svfloat32_t svbfmlalb_n_f32(svfloat32_t, svbfloat16_t, bfloat16_t);
	8792	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlalb_f32)))
	8793	svfloat32_t svbfmlalb_f32(svfloat32_t, svbfloat16_t, svbfloat16_t);
	8794	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlalb_lane_f32)))
	8795	svfloat32_t svbfmlalb_lane_f32(svfloat32_t, svbfloat16_t, svbfloat16_t, uint64_t);
	8796	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlalt_n_f32)))
	8797	svfloat32_t svbfmlalt_n_f32(svfloat32_t, svbfloat16_t, bfloat16_t);
	8798	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlalt_f32)))
	8799	svfloat32_t svbfmlalt_f32(svfloat32_t, svbfloat16_t, svbfloat16_t);
	8800	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlalt_lane_f32)))
	8801	svfloat32_t svbfmlalt_lane_f32(svfloat32_t, svbfloat16_t, svbfloat16_t, uint64_t);
	8802	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvt_bf16_f32_m)))
	8803	svbfloat16_t svcvt_bf16_f32_m(svbfloat16_t, svbool_t, svfloat32_t);
	8804	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvt_bf16_f32_x)))
	8805	svbfloat16_t svcvt_bf16_f32_x(svbool_t, svfloat32_t);
	8806	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvt_bf16_f32_z)))
	8807	svbfloat16_t svcvt_bf16_f32_z(svbool_t, svfloat32_t);
	8808	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtnt_bf16_f32_m)))
	8809	svbfloat16_t svcvtnt_bf16_f32_m(svbfloat16_t, svbool_t, svfloat32_t);
	8810	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfdot_n_f32)))
	8811	svfloat32_t svbfdot(svfloat32_t, svbfloat16_t, bfloat16_t);
	8812	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfdot_f32)))
	8813	svfloat32_t svbfdot(svfloat32_t, svbfloat16_t, svbfloat16_t);
	8814	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfdot_lane_f32)))
	8815	svfloat32_t svbfdot_lane(svfloat32_t, svbfloat16_t, svbfloat16_t, uint64_t);
	8816	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlalb_n_f32)))
	8817	svfloat32_t svbfmlalb(svfloat32_t, svbfloat16_t, bfloat16_t);
	8818	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlalb_f32)))
	8819	svfloat32_t svbfmlalb(svfloat32_t, svbfloat16_t, svbfloat16_t);
	8820	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlalb_lane_f32)))
	8821	svfloat32_t svbfmlalb_lane(svfloat32_t, svbfloat16_t, svbfloat16_t, uint64_t);
	8822	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlalt_n_f32)))
	8823	svfloat32_t svbfmlalt(svfloat32_t, svbfloat16_t, bfloat16_t);
	8824	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlalt_f32)))
	8825	svfloat32_t svbfmlalt(svfloat32_t, svbfloat16_t, svbfloat16_t);
	8826	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlalt_lane_f32)))
	8827	svfloat32_t svbfmlalt_lane(svfloat32_t, svbfloat16_t, svbfloat16_t, uint64_t);
	8828	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvt_bf16_f32_m)))
	8829	svbfloat16_t svcvt_bf16_m(svbfloat16_t, svbool_t, svfloat32_t);
	8830	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvt_bf16_f32_x)))
	8831	svbfloat16_t svcvt_bf16_x(svbool_t, svfloat32_t);
	8832	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvt_bf16_f32_z)))
	8833	svbfloat16_t svcvt_bf16_z(svbool_t, svfloat32_t);
	8834	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtnt_bf16_f32_m)))
	8835	svbfloat16_t svcvtnt_bf16_m(svbfloat16_t, svbool_t, svfloat32_t);
	8836	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmmla_f32)))
	8837	svfloat32_t svmmla_f32(svfloat32_t, svfloat32_t, svfloat32_t);
	8838	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmmla_f32)))
	8839	svfloat32_t svmmla(svfloat32_t, svfloat32_t, svfloat32_t);
	8840	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_u8)))
	8841	svuint8_t svld1ro_u8(svbool_t, uint8_t const *);
	8842	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_u32)))
	8843	svuint32_t svld1ro_u32(svbool_t, uint32_t const *);
	8844	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_u64)))
	8845	svuint64_t svld1ro_u64(svbool_t, uint64_t const *);
	8846	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_u16)))
	8847	svuint16_t svld1ro_u16(svbool_t, uint16_t const *);
	8848	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_bf16)))
	8849	svbfloat16_t svld1ro_bf16(svbool_t, bfloat16_t const *);
	8850	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_s8)))
	8851	svint8_t svld1ro_s8(svbool_t, int8_t const *);
	8852	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_f64)))
	8853	svfloat64_t svld1ro_f64(svbool_t, float64_t const *);
	8854	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_f32)))
	8855	svfloat32_t svld1ro_f32(svbool_t, float32_t const *);
	8856	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_f16)))
	8857	svfloat16_t svld1ro_f16(svbool_t, float16_t const *);
	8858	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_s32)))
	8859	svint32_t svld1ro_s32(svbool_t, int32_t const *);
	8860	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_s64)))
	8861	svint64_t svld1ro_s64(svbool_t, int64_t const *);
	8862	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_mf8)))
	8863	svmfloat8_t svld1ro_mf8(svbool_t, mfloat8_t const *);
	8864	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_s16)))
	8865	svint16_t svld1ro_s16(svbool_t, int16_t const *);
	8866	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmmla_f64)))
	8867	svfloat64_t svmmla_f64(svfloat64_t, svfloat64_t, svfloat64_t);
	8868	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_u8)))
	8869	svuint8_t svtrn1q_u8(svuint8_t, svuint8_t);
	8870	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_u32)))
	8871	svuint32_t svtrn1q_u32(svuint32_t, svuint32_t);
	8872	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_u64)))
	8873	svuint64_t svtrn1q_u64(svuint64_t, svuint64_t);
	8874	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_u16)))
	8875	svuint16_t svtrn1q_u16(svuint16_t, svuint16_t);
	8876	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_bf16)))
	8877	svbfloat16_t svtrn1q_bf16(svbfloat16_t, svbfloat16_t);
	8878	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_s8)))
	8879	svint8_t svtrn1q_s8(svint8_t, svint8_t);
	8880	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_f64)))
	8881	svfloat64_t svtrn1q_f64(svfloat64_t, svfloat64_t);
	8882	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_f32)))
	8883	svfloat32_t svtrn1q_f32(svfloat32_t, svfloat32_t);
	8884	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_f16)))
	8885	svfloat16_t svtrn1q_f16(svfloat16_t, svfloat16_t);
	8886	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_s32)))
	8887	svint32_t svtrn1q_s32(svint32_t, svint32_t);
	8888	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_s64)))
	8889	svint64_t svtrn1q_s64(svint64_t, svint64_t);
	8890	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_s16)))
	8891	svint16_t svtrn1q_s16(svint16_t, svint16_t);
	8892	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_u8)))
	8893	svuint8_t svtrn2q_u8(svuint8_t, svuint8_t);
	8894	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_u32)))
	8895	svuint32_t svtrn2q_u32(svuint32_t, svuint32_t);
	8896	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_u64)))
	8897	svuint64_t svtrn2q_u64(svuint64_t, svuint64_t);
	8898	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_u16)))
	8899	svuint16_t svtrn2q_u16(svuint16_t, svuint16_t);
	8900	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_bf16)))
	8901	svbfloat16_t svtrn2q_bf16(svbfloat16_t, svbfloat16_t);
	8902	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_s8)))
	8903	svint8_t svtrn2q_s8(svint8_t, svint8_t);
	8904	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_f64)))
	8905	svfloat64_t svtrn2q_f64(svfloat64_t, svfloat64_t);
	8906	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_f32)))
	8907	svfloat32_t svtrn2q_f32(svfloat32_t, svfloat32_t);
	8908	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_f16)))
	8909	svfloat16_t svtrn2q_f16(svfloat16_t, svfloat16_t);
	8910	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_s32)))
	8911	svint32_t svtrn2q_s32(svint32_t, svint32_t);
	8912	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_s64)))
	8913	svint64_t svtrn2q_s64(svint64_t, svint64_t);
	8914	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_s16)))
	8915	svint16_t svtrn2q_s16(svint16_t, svint16_t);
	8916	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_u8)))
	8917	svuint8_t svuzp1q_u8(svuint8_t, svuint8_t);
	8918	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_u32)))
	8919	svuint32_t svuzp1q_u32(svuint32_t, svuint32_t);
	8920	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_u64)))
	8921	svuint64_t svuzp1q_u64(svuint64_t, svuint64_t);
	8922	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_u16)))
	8923	svuint16_t svuzp1q_u16(svuint16_t, svuint16_t);
	8924	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_bf16)))
	8925	svbfloat16_t svuzp1q_bf16(svbfloat16_t, svbfloat16_t);
	8926	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_s8)))
	8927	svint8_t svuzp1q_s8(svint8_t, svint8_t);
	8928	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_f64)))
	8929	svfloat64_t svuzp1q_f64(svfloat64_t, svfloat64_t);
	8930	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_f32)))
	8931	svfloat32_t svuzp1q_f32(svfloat32_t, svfloat32_t);
	8932	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_f16)))
	8933	svfloat16_t svuzp1q_f16(svfloat16_t, svfloat16_t);
	8934	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_s32)))
	8935	svint32_t svuzp1q_s32(svint32_t, svint32_t);
	8936	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_s64)))
	8937	svint64_t svuzp1q_s64(svint64_t, svint64_t);
	8938	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_s16)))
	8939	svint16_t svuzp1q_s16(svint16_t, svint16_t);
	8940	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_u8)))
	8941	svuint8_t svuzp2q_u8(svuint8_t, svuint8_t);
	8942	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_u32)))
	8943	svuint32_t svuzp2q_u32(svuint32_t, svuint32_t);
	8944	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_u64)))
	8945	svuint64_t svuzp2q_u64(svuint64_t, svuint64_t);
	8946	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_u16)))
	8947	svuint16_t svuzp2q_u16(svuint16_t, svuint16_t);
	8948	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_bf16)))
	8949	svbfloat16_t svuzp2q_bf16(svbfloat16_t, svbfloat16_t);
	8950	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_s8)))
	8951	svint8_t svuzp2q_s8(svint8_t, svint8_t);
	8952	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_f64)))
	8953	svfloat64_t svuzp2q_f64(svfloat64_t, svfloat64_t);
	8954	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_f32)))
	8955	svfloat32_t svuzp2q_f32(svfloat32_t, svfloat32_t);
	8956	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_f16)))
	8957	svfloat16_t svuzp2q_f16(svfloat16_t, svfloat16_t);
	8958	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_s32)))
	8959	svint32_t svuzp2q_s32(svint32_t, svint32_t);
	8960	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_s64)))
	8961	svint64_t svuzp2q_s64(svint64_t, svint64_t);
	8962	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_s16)))
	8963	svint16_t svuzp2q_s16(svint16_t, svint16_t);
	8964	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_u8)))
	8965	svuint8_t svzip1q_u8(svuint8_t, svuint8_t);
	8966	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_u32)))
	8967	svuint32_t svzip1q_u32(svuint32_t, svuint32_t);
	8968	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_u64)))
	8969	svuint64_t svzip1q_u64(svuint64_t, svuint64_t);
	8970	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_u16)))
	8971	svuint16_t svzip1q_u16(svuint16_t, svuint16_t);
	8972	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_bf16)))
	8973	svbfloat16_t svzip1q_bf16(svbfloat16_t, svbfloat16_t);
	8974	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_s8)))
	8975	svint8_t svzip1q_s8(svint8_t, svint8_t);
	8976	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_f64)))
	8977	svfloat64_t svzip1q_f64(svfloat64_t, svfloat64_t);
	8978	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_f32)))
	8979	svfloat32_t svzip1q_f32(svfloat32_t, svfloat32_t);
	8980	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_f16)))
	8981	svfloat16_t svzip1q_f16(svfloat16_t, svfloat16_t);
	8982	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_s32)))
	8983	svint32_t svzip1q_s32(svint32_t, svint32_t);
	8984	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_s64)))
	8985	svint64_t svzip1q_s64(svint64_t, svint64_t);
	8986	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_s16)))
	8987	svint16_t svzip1q_s16(svint16_t, svint16_t);
	8988	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_u8)))
	8989	svuint8_t svzip2q_u8(svuint8_t, svuint8_t);
	8990	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_u32)))
	8991	svuint32_t svzip2q_u32(svuint32_t, svuint32_t);
	8992	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_u64)))
	8993	svuint64_t svzip2q_u64(svuint64_t, svuint64_t);
	8994	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_u16)))
	8995	svuint16_t svzip2q_u16(svuint16_t, svuint16_t);
	8996	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_bf16)))
	8997	svbfloat16_t svzip2q_bf16(svbfloat16_t, svbfloat16_t);
	8998	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_s8)))
	8999	svint8_t svzip2q_s8(svint8_t, svint8_t);
	9000	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_f64)))
	9001	svfloat64_t svzip2q_f64(svfloat64_t, svfloat64_t);
	9002	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_f32)))
	9003	svfloat32_t svzip2q_f32(svfloat32_t, svfloat32_t);
	9004	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_f16)))
	9005	svfloat16_t svzip2q_f16(svfloat16_t, svfloat16_t);
	9006	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_s32)))
	9007	svint32_t svzip2q_s32(svint32_t, svint32_t);
	9008	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_s64)))
	9009	svint64_t svzip2q_s64(svint64_t, svint64_t);
	9010	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_s16)))
	9011	svint16_t svzip2q_s16(svint16_t, svint16_t);
	9012	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_u8)))
	9013	svuint8_t svld1ro(svbool_t, uint8_t const *);
	9014	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_u32)))
	9015	svuint32_t svld1ro(svbool_t, uint32_t const *);
	9016	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_u64)))
	9017	svuint64_t svld1ro(svbool_t, uint64_t const *);
	9018	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_u16)))
	9019	svuint16_t svld1ro(svbool_t, uint16_t const *);
	9020	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_bf16)))
	9021	svbfloat16_t svld1ro(svbool_t, bfloat16_t const *);
	9022	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_s8)))
	9023	svint8_t svld1ro(svbool_t, int8_t const *);
	9024	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_f64)))
	9025	svfloat64_t svld1ro(svbool_t, float64_t const *);
	9026	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_f32)))
	9027	svfloat32_t svld1ro(svbool_t, float32_t const *);
	9028	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_f16)))
	9029	svfloat16_t svld1ro(svbool_t, float16_t const *);
	9030	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_s32)))
	9031	svint32_t svld1ro(svbool_t, int32_t const *);
	9032	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_s64)))
	9033	svint64_t svld1ro(svbool_t, int64_t const *);
	9034	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_mf8)))
	9035	svmfloat8_t svld1ro(svbool_t, mfloat8_t const *);
	9036	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1ro_s16)))
	9037	svint16_t svld1ro(svbool_t, int16_t const *);
	9038	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmmla_f64)))
	9039	svfloat64_t svmmla(svfloat64_t, svfloat64_t, svfloat64_t);
	9040	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_u8)))
	9041	svuint8_t svtrn1q(svuint8_t, svuint8_t);
	9042	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_u32)))
	9043	svuint32_t svtrn1q(svuint32_t, svuint32_t);
	9044	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_u64)))
	9045	svuint64_t svtrn1q(svuint64_t, svuint64_t);
	9046	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_u16)))
	9047	svuint16_t svtrn1q(svuint16_t, svuint16_t);
	9048	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_bf16)))
	9049	svbfloat16_t svtrn1q(svbfloat16_t, svbfloat16_t);
	9050	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_s8)))
	9051	svint8_t svtrn1q(svint8_t, svint8_t);
	9052	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_f64)))
	9053	svfloat64_t svtrn1q(svfloat64_t, svfloat64_t);
	9054	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_f32)))
	9055	svfloat32_t svtrn1q(svfloat32_t, svfloat32_t);
	9056	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_f16)))
	9057	svfloat16_t svtrn1q(svfloat16_t, svfloat16_t);
	9058	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_s32)))
	9059	svint32_t svtrn1q(svint32_t, svint32_t);
	9060	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_s64)))
	9061	svint64_t svtrn1q(svint64_t, svint64_t);
	9062	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1q_s16)))
	9063	svint16_t svtrn1q(svint16_t, svint16_t);
	9064	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_u8)))
	9065	svuint8_t svtrn2q(svuint8_t, svuint8_t);
	9066	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_u32)))
	9067	svuint32_t svtrn2q(svuint32_t, svuint32_t);
	9068	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_u64)))
	9069	svuint64_t svtrn2q(svuint64_t, svuint64_t);
	9070	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_u16)))
	9071	svuint16_t svtrn2q(svuint16_t, svuint16_t);
	9072	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_bf16)))
	9073	svbfloat16_t svtrn2q(svbfloat16_t, svbfloat16_t);
	9074	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_s8)))
	9075	svint8_t svtrn2q(svint8_t, svint8_t);
	9076	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_f64)))
	9077	svfloat64_t svtrn2q(svfloat64_t, svfloat64_t);
	9078	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_f32)))
	9079	svfloat32_t svtrn2q(svfloat32_t, svfloat32_t);
	9080	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_f16)))
	9081	svfloat16_t svtrn2q(svfloat16_t, svfloat16_t);
	9082	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_s32)))
	9083	svint32_t svtrn2q(svint32_t, svint32_t);
	9084	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_s64)))
	9085	svint64_t svtrn2q(svint64_t, svint64_t);
	9086	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2q_s16)))
	9087	svint16_t svtrn2q(svint16_t, svint16_t);
	9088	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_u8)))
	9089	svuint8_t svuzp1q(svuint8_t, svuint8_t);
	9090	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_u32)))
	9091	svuint32_t svuzp1q(svuint32_t, svuint32_t);
	9092	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_u64)))
	9093	svuint64_t svuzp1q(svuint64_t, svuint64_t);
	9094	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_u16)))
	9095	svuint16_t svuzp1q(svuint16_t, svuint16_t);
	9096	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_bf16)))
	9097	svbfloat16_t svuzp1q(svbfloat16_t, svbfloat16_t);
	9098	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_s8)))
	9099	svint8_t svuzp1q(svint8_t, svint8_t);
	9100	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_f64)))
	9101	svfloat64_t svuzp1q(svfloat64_t, svfloat64_t);
	9102	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_f32)))
	9103	svfloat32_t svuzp1q(svfloat32_t, svfloat32_t);
	9104	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_f16)))
	9105	svfloat16_t svuzp1q(svfloat16_t, svfloat16_t);
	9106	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_s32)))
	9107	svint32_t svuzp1q(svint32_t, svint32_t);
	9108	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_s64)))
	9109	svint64_t svuzp1q(svint64_t, svint64_t);
	9110	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1q_s16)))
	9111	svint16_t svuzp1q(svint16_t, svint16_t);
	9112	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_u8)))
	9113	svuint8_t svuzp2q(svuint8_t, svuint8_t);
	9114	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_u32)))
	9115	svuint32_t svuzp2q(svuint32_t, svuint32_t);
	9116	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_u64)))
	9117	svuint64_t svuzp2q(svuint64_t, svuint64_t);
	9118	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_u16)))
	9119	svuint16_t svuzp2q(svuint16_t, svuint16_t);
	9120	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_bf16)))
	9121	svbfloat16_t svuzp2q(svbfloat16_t, svbfloat16_t);
	9122	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_s8)))
	9123	svint8_t svuzp2q(svint8_t, svint8_t);
	9124	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_f64)))
	9125	svfloat64_t svuzp2q(svfloat64_t, svfloat64_t);
	9126	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_f32)))
	9127	svfloat32_t svuzp2q(svfloat32_t, svfloat32_t);
	9128	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_f16)))
	9129	svfloat16_t svuzp2q(svfloat16_t, svfloat16_t);
	9130	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_s32)))
	9131	svint32_t svuzp2q(svint32_t, svint32_t);
	9132	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_s64)))
	9133	svint64_t svuzp2q(svint64_t, svint64_t);
	9134	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2q_s16)))
	9135	svint16_t svuzp2q(svint16_t, svint16_t);
	9136	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_u8)))
	9137	svuint8_t svzip1q(svuint8_t, svuint8_t);
	9138	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_u32)))
	9139	svuint32_t svzip1q(svuint32_t, svuint32_t);
	9140	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_u64)))
	9141	svuint64_t svzip1q(svuint64_t, svuint64_t);
	9142	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_u16)))
	9143	svuint16_t svzip1q(svuint16_t, svuint16_t);
	9144	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_bf16)))
	9145	svbfloat16_t svzip1q(svbfloat16_t, svbfloat16_t);
	9146	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_s8)))
	9147	svint8_t svzip1q(svint8_t, svint8_t);
	9148	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_f64)))
	9149	svfloat64_t svzip1q(svfloat64_t, svfloat64_t);
	9150	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_f32)))
	9151	svfloat32_t svzip1q(svfloat32_t, svfloat32_t);
	9152	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_f16)))
	9153	svfloat16_t svzip1q(svfloat16_t, svfloat16_t);
	9154	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_s32)))
	9155	svint32_t svzip1q(svint32_t, svint32_t);
	9156	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_s64)))
	9157	svint64_t svzip1q(svint64_t, svint64_t);
	9158	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1q_s16)))
	9159	svint16_t svzip1q(svint16_t, svint16_t);
	9160	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_u8)))
	9161	svuint8_t svzip2q(svuint8_t, svuint8_t);
	9162	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_u32)))
	9163	svuint32_t svzip2q(svuint32_t, svuint32_t);
	9164	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_u64)))
	9165	svuint64_t svzip2q(svuint64_t, svuint64_t);
	9166	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_u16)))
	9167	svuint16_t svzip2q(svuint16_t, svuint16_t);
	9168	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_bf16)))
	9169	svbfloat16_t svzip2q(svbfloat16_t, svbfloat16_t);
	9170	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_s8)))
	9171	svint8_t svzip2q(svint8_t, svint8_t);
	9172	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_f64)))
	9173	svfloat64_t svzip2q(svfloat64_t, svfloat64_t);
	9174	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_f32)))
	9175	svfloat32_t svzip2q(svfloat32_t, svfloat32_t);
	9176	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_f16)))
	9177	svfloat16_t svzip2q(svfloat16_t, svfloat16_t);
	9178	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_s32)))
	9179	svint32_t svzip2q(svint32_t, svint32_t);
	9180	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_s64)))
	9181	svint64_t svzip2q(svint64_t, svint64_t);
	9182	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2q_s16)))
	9183	svint16_t svzip2q(svint16_t, svint16_t);
	9184	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmmla_s32)))
	9185	svint32_t svmmla_s32(svint32_t, svint8_t, svint8_t);
	9186	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmmla_u32)))
	9187	svuint32_t svmmla_u32(svuint32_t, svuint8_t, svuint8_t);
	9188	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svusmmla_s32)))
	9189	svint32_t svusmmla_s32(svint32_t, svuint8_t, svint8_t);
	9190	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmmla_s32)))
	9191	svint32_t svmmla(svint32_t, svint8_t, svint8_t);
	9192	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmmla_u32)))
	9193	svuint32_t svmmla(svuint32_t, svuint8_t, svuint8_t);
	9194	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svusmmla_s32)))
	9195	svint32_t svusmmla(svint32_t, svuint8_t, svint8_t);
	9196	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsudot_n_s32)))
	9197	svint32_t svsudot_n_s32(svint32_t, svint8_t, uint8_t);
	9198	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsudot_s32)))
	9199	svint32_t svsudot_s32(svint32_t, svint8_t, svuint8_t);
	9200	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsudot_lane_s32)))
	9201	svint32_t svsudot_lane_s32(svint32_t, svint8_t, svuint8_t, uint64_t);
	9202	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svusdot_n_s32)))
	9203	svint32_t svusdot_n_s32(svint32_t, svuint8_t, int8_t);
	9204	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svusdot_s32)))
	9205	svint32_t svusdot_s32(svint32_t, svuint8_t, svint8_t);
	9206	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svusdot_lane_s32)))
	9207	svint32_t svusdot_lane_s32(svint32_t, svuint8_t, svint8_t, uint64_t);
	9208	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsudot_n_s32)))
	9209	svint32_t svsudot(svint32_t, svint8_t, uint8_t);
	9210	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsudot_s32)))
	9211	svint32_t svsudot(svint32_t, svint8_t, svuint8_t);
	9212	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsudot_lane_s32)))
	9213	svint32_t svsudot_lane(svint32_t, svint8_t, svuint8_t, uint64_t);
	9214	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svusdot_n_s32)))
	9215	svint32_t svusdot(svint32_t, svuint8_t, int8_t);
	9216	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svusdot_s32)))
	9217	svint32_t svusdot(svint32_t, svuint8_t, svint8_t);
	9218	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svusdot_lane_s32)))
	9219	svint32_t svusdot_lane(svint32_t, svuint8_t, svint8_t, uint64_t);
	9220	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrax1_u64)))
	9221	svuint64_t svrax1_u64(svuint64_t, svuint64_t);
	9222	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrax1_s64)))
	9223	svint64_t svrax1_s64(svint64_t, svint64_t);
	9224	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrax1_u64)))
	9225	svuint64_t svrax1(svuint64_t, svuint64_t);
	9226	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrax1_s64)))
	9227	svint64_t svrax1(svint64_t, svint64_t);
	9228	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svhistcnt_u32_z)))
	9229	svuint32_t svhistcnt_u32_z(svbool_t, svuint32_t, svuint32_t);
	9230	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svhistcnt_u64_z)))
	9231	svuint64_t svhistcnt_u64_z(svbool_t, svuint64_t, svuint64_t);
	9232	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svhistcnt_s32_z)))
	9233	svuint32_t svhistcnt_s32_z(svbool_t, svint32_t, svint32_t);
	9234	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svhistcnt_s64_z)))
	9235	svuint64_t svhistcnt_s64_z(svbool_t, svint64_t, svint64_t);
	9236	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svhistseg_u8)))
	9237	svuint8_t svhistseg_u8(svuint8_t, svuint8_t);
	9238	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svhistseg_s8)))
	9239	svuint8_t svhistseg_s8(svint8_t, svint8_t);
	9240	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_index_u32)))
	9241	svuint32_t svldnt1_gather_u32base_index_u32(svbool_t, svuint32_t, int64_t);
	9242	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_index_u64)))
	9243	svuint64_t svldnt1_gather_u64base_index_u64(svbool_t, svuint64_t, int64_t);
	9244	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_index_f64)))
	9245	svfloat64_t svldnt1_gather_u64base_index_f64(svbool_t, svuint64_t, int64_t);
	9246	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_index_f32)))
	9247	svfloat32_t svldnt1_gather_u32base_index_f32(svbool_t, svuint32_t, int64_t);
	9248	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_index_s32)))
	9249	svint32_t svldnt1_gather_u32base_index_s32(svbool_t, svuint32_t, int64_t);
	9250	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_index_s64)))
	9251	svint64_t svldnt1_gather_u64base_index_s64(svbool_t, svuint64_t, int64_t);
	9252	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_offset_u32)))
	9253	svuint32_t svldnt1_gather_u32base_offset_u32(svbool_t, svuint32_t, int64_t);
	9254	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_offset_u64)))
	9255	svuint64_t svldnt1_gather_u64base_offset_u64(svbool_t, svuint64_t, int64_t);
	9256	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_offset_f64)))
	9257	svfloat64_t svldnt1_gather_u64base_offset_f64(svbool_t, svuint64_t, int64_t);
	9258	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_offset_f32)))
	9259	svfloat32_t svldnt1_gather_u32base_offset_f32(svbool_t, svuint32_t, int64_t);
	9260	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_offset_s32)))
	9261	svint32_t svldnt1_gather_u32base_offset_s32(svbool_t, svuint32_t, int64_t);
	9262	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_offset_s64)))
	9263	svint64_t svldnt1_gather_u64base_offset_s64(svbool_t, svuint64_t, int64_t);
	9264	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_u32)))
	9265	svuint32_t svldnt1_gather_u32base_u32(svbool_t, svuint32_t);
	9266	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_u64)))
	9267	svuint64_t svldnt1_gather_u64base_u64(svbool_t, svuint64_t);
	9268	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_f64)))
	9269	svfloat64_t svldnt1_gather_u64base_f64(svbool_t, svuint64_t);
	9270	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_f32)))
	9271	svfloat32_t svldnt1_gather_u32base_f32(svbool_t, svuint32_t);
	9272	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_s32)))
	9273	svint32_t svldnt1_gather_u32base_s32(svbool_t, svuint32_t);
	9274	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_s64)))
	9275	svint64_t svldnt1_gather_u64base_s64(svbool_t, svuint64_t);
	9276	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_s64index_u64)))
	9277	svuint64_t svldnt1_gather_s64index_u64(svbool_t, uint64_t const *, svint64_t);
	9278	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_s64index_f64)))
	9279	svfloat64_t svldnt1_gather_s64index_f64(svbool_t, float64_t const *, svint64_t);
	9280	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_s64index_s64)))
	9281	svint64_t svldnt1_gather_s64index_s64(svbool_t, int64_t const *, svint64_t);
	9282	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64index_u64)))
	9283	svuint64_t svldnt1_gather_u64index_u64(svbool_t, uint64_t const *, svuint64_t);
	9284	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64index_f64)))
	9285	svfloat64_t svldnt1_gather_u64index_f64(svbool_t, float64_t const *, svuint64_t);
	9286	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64index_s64)))
	9287	svint64_t svldnt1_gather_u64index_s64(svbool_t, int64_t const *, svuint64_t);
	9288	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32offset_u32)))
	9289	svuint32_t svldnt1_gather_u32offset_u32(svbool_t, uint32_t const *, svuint32_t);
	9290	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32offset_f32)))
	9291	svfloat32_t svldnt1_gather_u32offset_f32(svbool_t, float32_t const *, svuint32_t);
	9292	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32offset_s32)))
	9293	svint32_t svldnt1_gather_u32offset_s32(svbool_t, int32_t const *, svuint32_t);
	9294	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_s64offset_u64)))
	9295	svuint64_t svldnt1_gather_s64offset_u64(svbool_t, uint64_t const *, svint64_t);
	9296	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_s64offset_f64)))
	9297	svfloat64_t svldnt1_gather_s64offset_f64(svbool_t, float64_t const *, svint64_t);
	9298	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_s64offset_s64)))
	9299	svint64_t svldnt1_gather_s64offset_s64(svbool_t, int64_t const *, svint64_t);
	9300	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64offset_u64)))
	9301	svuint64_t svldnt1_gather_u64offset_u64(svbool_t, uint64_t const *, svuint64_t);
	9302	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64offset_f64)))
	9303	svfloat64_t svldnt1_gather_u64offset_f64(svbool_t, float64_t const *, svuint64_t);
	9304	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64offset_s64)))
	9305	svint64_t svldnt1_gather_u64offset_s64(svbool_t, int64_t const *, svuint64_t);
	9306	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u32base_offset_u32)))
	9307	svuint32_t svldnt1sb_gather_u32base_offset_u32(svbool_t, svuint32_t, int64_t);
	9308	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u64base_offset_u64)))
	9309	svuint64_t svldnt1sb_gather_u64base_offset_u64(svbool_t, svuint64_t, int64_t);
	9310	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u32base_offset_s32)))
	9311	svint32_t svldnt1sb_gather_u32base_offset_s32(svbool_t, svuint32_t, int64_t);
	9312	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u64base_offset_s64)))
	9313	svint64_t svldnt1sb_gather_u64base_offset_s64(svbool_t, svuint64_t, int64_t);
	9314	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u32base_u32)))
	9315	svuint32_t svldnt1sb_gather_u32base_u32(svbool_t, svuint32_t);
	9316	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u64base_u64)))
	9317	svuint64_t svldnt1sb_gather_u64base_u64(svbool_t, svuint64_t);
	9318	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u32base_s32)))
	9319	svint32_t svldnt1sb_gather_u32base_s32(svbool_t, svuint32_t);
	9320	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u64base_s64)))
	9321	svint64_t svldnt1sb_gather_u64base_s64(svbool_t, svuint64_t);
	9322	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u32offset_u32)))
	9323	svuint32_t svldnt1sb_gather_u32offset_u32(svbool_t, int8_t const *, svuint32_t);
	9324	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u32offset_s32)))
	9325	svint32_t svldnt1sb_gather_u32offset_s32(svbool_t, int8_t const *, svuint32_t);
	9326	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_s64offset_u64)))
	9327	svuint64_t svldnt1sb_gather_s64offset_u64(svbool_t, int8_t const *, svint64_t);
	9328	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_s64offset_s64)))
	9329	svint64_t svldnt1sb_gather_s64offset_s64(svbool_t, int8_t const *, svint64_t);
	9330	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u64offset_u64)))
	9331	svuint64_t svldnt1sb_gather_u64offset_u64(svbool_t, int8_t const *, svuint64_t);
	9332	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u64offset_s64)))
	9333	svint64_t svldnt1sb_gather_u64offset_s64(svbool_t, int8_t const *, svuint64_t);
	9334	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32base_index_u32)))
	9335	svuint32_t svldnt1sh_gather_u32base_index_u32(svbool_t, svuint32_t, int64_t);
	9336	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64base_index_u64)))
	9337	svuint64_t svldnt1sh_gather_u64base_index_u64(svbool_t, svuint64_t, int64_t);
	9338	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32base_index_s32)))
	9339	svint32_t svldnt1sh_gather_u32base_index_s32(svbool_t, svuint32_t, int64_t);
	9340	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64base_index_s64)))
	9341	svint64_t svldnt1sh_gather_u64base_index_s64(svbool_t, svuint64_t, int64_t);
	9342	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32base_offset_u32)))
	9343	svuint32_t svldnt1sh_gather_u32base_offset_u32(svbool_t, svuint32_t, int64_t);
	9344	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64base_offset_u64)))
	9345	svuint64_t svldnt1sh_gather_u64base_offset_u64(svbool_t, svuint64_t, int64_t);
	9346	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32base_offset_s32)))
	9347	svint32_t svldnt1sh_gather_u32base_offset_s32(svbool_t, svuint32_t, int64_t);
	9348	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64base_offset_s64)))
	9349	svint64_t svldnt1sh_gather_u64base_offset_s64(svbool_t, svuint64_t, int64_t);
	9350	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32base_u32)))
	9351	svuint32_t svldnt1sh_gather_u32base_u32(svbool_t, svuint32_t);
	9352	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64base_u64)))
	9353	svuint64_t svldnt1sh_gather_u64base_u64(svbool_t, svuint64_t);
	9354	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32base_s32)))
	9355	svint32_t svldnt1sh_gather_u32base_s32(svbool_t, svuint32_t);
	9356	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64base_s64)))
	9357	svint64_t svldnt1sh_gather_u64base_s64(svbool_t, svuint64_t);
	9358	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_s64index_u64)))
	9359	svuint64_t svldnt1sh_gather_s64index_u64(svbool_t, int16_t const *, svint64_t);
	9360	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_s64index_s64)))
	9361	svint64_t svldnt1sh_gather_s64index_s64(svbool_t, int16_t const *, svint64_t);
	9362	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64index_u64)))
	9363	svuint64_t svldnt1sh_gather_u64index_u64(svbool_t, int16_t const *, svuint64_t);
	9364	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64index_s64)))
	9365	svint64_t svldnt1sh_gather_u64index_s64(svbool_t, int16_t const *, svuint64_t);
	9366	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32offset_u32)))
	9367	svuint32_t svldnt1sh_gather_u32offset_u32(svbool_t, int16_t const *, svuint32_t);
	9368	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32offset_s32)))
	9369	svint32_t svldnt1sh_gather_u32offset_s32(svbool_t, int16_t const *, svuint32_t);
	9370	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_s64offset_u64)))
	9371	svuint64_t svldnt1sh_gather_s64offset_u64(svbool_t, int16_t const *, svint64_t);
	9372	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_s64offset_s64)))
	9373	svint64_t svldnt1sh_gather_s64offset_s64(svbool_t, int16_t const *, svint64_t);
	9374	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64offset_u64)))
	9375	svuint64_t svldnt1sh_gather_u64offset_u64(svbool_t, int16_t const *, svuint64_t);
	9376	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64offset_s64)))
	9377	svint64_t svldnt1sh_gather_u64offset_s64(svbool_t, int16_t const *, svuint64_t);
	9378	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64base_index_u64)))
	9379	svuint64_t svldnt1sw_gather_u64base_index_u64(svbool_t, svuint64_t, int64_t);
	9380	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64base_index_s64)))
	9381	svint64_t svldnt1sw_gather_u64base_index_s64(svbool_t, svuint64_t, int64_t);
	9382	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64base_offset_u64)))
	9383	svuint64_t svldnt1sw_gather_u64base_offset_u64(svbool_t, svuint64_t, int64_t);
	9384	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64base_offset_s64)))
	9385	svint64_t svldnt1sw_gather_u64base_offset_s64(svbool_t, svuint64_t, int64_t);
	9386	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64base_u64)))
	9387	svuint64_t svldnt1sw_gather_u64base_u64(svbool_t, svuint64_t);
	9388	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64base_s64)))
	9389	svint64_t svldnt1sw_gather_u64base_s64(svbool_t, svuint64_t);
	9390	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_s64index_u64)))
	9391	svuint64_t svldnt1sw_gather_s64index_u64(svbool_t, int32_t const *, svint64_t);
	9392	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_s64index_s64)))
	9393	svint64_t svldnt1sw_gather_s64index_s64(svbool_t, int32_t const *, svint64_t);
	9394	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64index_u64)))
	9395	svuint64_t svldnt1sw_gather_u64index_u64(svbool_t, int32_t const *, svuint64_t);
	9396	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64index_s64)))
	9397	svint64_t svldnt1sw_gather_u64index_s64(svbool_t, int32_t const *, svuint64_t);
	9398	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_s64offset_u64)))
	9399	svuint64_t svldnt1sw_gather_s64offset_u64(svbool_t, int32_t const *, svint64_t);
	9400	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_s64offset_s64)))
	9401	svint64_t svldnt1sw_gather_s64offset_s64(svbool_t, int32_t const *, svint64_t);
	9402	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64offset_u64)))
	9403	svuint64_t svldnt1sw_gather_u64offset_u64(svbool_t, int32_t const *, svuint64_t);
	9404	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64offset_s64)))
	9405	svint64_t svldnt1sw_gather_u64offset_s64(svbool_t, int32_t const *, svuint64_t);
	9406	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u32base_offset_u32)))
	9407	svuint32_t svldnt1ub_gather_u32base_offset_u32(svbool_t, svuint32_t, int64_t);
	9408	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u64base_offset_u64)))
	9409	svuint64_t svldnt1ub_gather_u64base_offset_u64(svbool_t, svuint64_t, int64_t);
	9410	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u32base_offset_s32)))
	9411	svint32_t svldnt1ub_gather_u32base_offset_s32(svbool_t, svuint32_t, int64_t);
	9412	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u64base_offset_s64)))
	9413	svint64_t svldnt1ub_gather_u64base_offset_s64(svbool_t, svuint64_t, int64_t);
	9414	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u32base_u32)))
	9415	svuint32_t svldnt1ub_gather_u32base_u32(svbool_t, svuint32_t);
	9416	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u64base_u64)))
	9417	svuint64_t svldnt1ub_gather_u64base_u64(svbool_t, svuint64_t);
	9418	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u32base_s32)))
	9419	svint32_t svldnt1ub_gather_u32base_s32(svbool_t, svuint32_t);
	9420	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u64base_s64)))
	9421	svint64_t svldnt1ub_gather_u64base_s64(svbool_t, svuint64_t);
	9422	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u32offset_u32)))
	9423	svuint32_t svldnt1ub_gather_u32offset_u32(svbool_t, uint8_t const *, svuint32_t);
	9424	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u32offset_s32)))
	9425	svint32_t svldnt1ub_gather_u32offset_s32(svbool_t, uint8_t const *, svuint32_t);
	9426	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_s64offset_u64)))
	9427	svuint64_t svldnt1ub_gather_s64offset_u64(svbool_t, uint8_t const *, svint64_t);
	9428	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_s64offset_s64)))
	9429	svint64_t svldnt1ub_gather_s64offset_s64(svbool_t, uint8_t const *, svint64_t);
	9430	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u64offset_u64)))
	9431	svuint64_t svldnt1ub_gather_u64offset_u64(svbool_t, uint8_t const *, svuint64_t);
	9432	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u64offset_s64)))
	9433	svint64_t svldnt1ub_gather_u64offset_s64(svbool_t, uint8_t const *, svuint64_t);
	9434	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32base_index_u32)))
	9435	svuint32_t svldnt1uh_gather_u32base_index_u32(svbool_t, svuint32_t, int64_t);
	9436	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64base_index_u64)))
	9437	svuint64_t svldnt1uh_gather_u64base_index_u64(svbool_t, svuint64_t, int64_t);
	9438	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32base_index_s32)))
	9439	svint32_t svldnt1uh_gather_u32base_index_s32(svbool_t, svuint32_t, int64_t);
	9440	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64base_index_s64)))
	9441	svint64_t svldnt1uh_gather_u64base_index_s64(svbool_t, svuint64_t, int64_t);
	9442	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32base_offset_u32)))
	9443	svuint32_t svldnt1uh_gather_u32base_offset_u32(svbool_t, svuint32_t, int64_t);
	9444	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64base_offset_u64)))
	9445	svuint64_t svldnt1uh_gather_u64base_offset_u64(svbool_t, svuint64_t, int64_t);
	9446	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32base_offset_s32)))
	9447	svint32_t svldnt1uh_gather_u32base_offset_s32(svbool_t, svuint32_t, int64_t);
	9448	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64base_offset_s64)))
	9449	svint64_t svldnt1uh_gather_u64base_offset_s64(svbool_t, svuint64_t, int64_t);
	9450	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32base_u32)))
	9451	svuint32_t svldnt1uh_gather_u32base_u32(svbool_t, svuint32_t);
	9452	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64base_u64)))
	9453	svuint64_t svldnt1uh_gather_u64base_u64(svbool_t, svuint64_t);
	9454	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32base_s32)))
	9455	svint32_t svldnt1uh_gather_u32base_s32(svbool_t, svuint32_t);
	9456	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64base_s64)))
	9457	svint64_t svldnt1uh_gather_u64base_s64(svbool_t, svuint64_t);
	9458	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_s64index_u64)))
	9459	svuint64_t svldnt1uh_gather_s64index_u64(svbool_t, uint16_t const *, svint64_t);
	9460	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_s64index_s64)))
	9461	svint64_t svldnt1uh_gather_s64index_s64(svbool_t, uint16_t const *, svint64_t);
	9462	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64index_u64)))
	9463	svuint64_t svldnt1uh_gather_u64index_u64(svbool_t, uint16_t const *, svuint64_t);
	9464	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64index_s64)))
	9465	svint64_t svldnt1uh_gather_u64index_s64(svbool_t, uint16_t const *, svuint64_t);
	9466	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32offset_u32)))
	9467	svuint32_t svldnt1uh_gather_u32offset_u32(svbool_t, uint16_t const *, svuint32_t);
	9468	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32offset_s32)))
	9469	svint32_t svldnt1uh_gather_u32offset_s32(svbool_t, uint16_t const *, svuint32_t);
	9470	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_s64offset_u64)))
	9471	svuint64_t svldnt1uh_gather_s64offset_u64(svbool_t, uint16_t const *, svint64_t);
	9472	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_s64offset_s64)))
	9473	svint64_t svldnt1uh_gather_s64offset_s64(svbool_t, uint16_t const *, svint64_t);
	9474	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64offset_u64)))
	9475	svuint64_t svldnt1uh_gather_u64offset_u64(svbool_t, uint16_t const *, svuint64_t);
	9476	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64offset_s64)))
	9477	svint64_t svldnt1uh_gather_u64offset_s64(svbool_t, uint16_t const *, svuint64_t);
	9478	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64base_index_u64)))
	9479	svuint64_t svldnt1uw_gather_u64base_index_u64(svbool_t, svuint64_t, int64_t);
	9480	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64base_index_s64)))
	9481	svint64_t svldnt1uw_gather_u64base_index_s64(svbool_t, svuint64_t, int64_t);
	9482	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64base_offset_u64)))
	9483	svuint64_t svldnt1uw_gather_u64base_offset_u64(svbool_t, svuint64_t, int64_t);
	9484	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64base_offset_s64)))
	9485	svint64_t svldnt1uw_gather_u64base_offset_s64(svbool_t, svuint64_t, int64_t);
	9486	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64base_u64)))
	9487	svuint64_t svldnt1uw_gather_u64base_u64(svbool_t, svuint64_t);
	9488	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64base_s64)))
	9489	svint64_t svldnt1uw_gather_u64base_s64(svbool_t, svuint64_t);
	9490	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_s64index_u64)))
	9491	svuint64_t svldnt1uw_gather_s64index_u64(svbool_t, uint32_t const *, svint64_t);
	9492	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_s64index_s64)))
	9493	svint64_t svldnt1uw_gather_s64index_s64(svbool_t, uint32_t const *, svint64_t);
	9494	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64index_u64)))
	9495	svuint64_t svldnt1uw_gather_u64index_u64(svbool_t, uint32_t const *, svuint64_t);
	9496	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64index_s64)))
	9497	svint64_t svldnt1uw_gather_u64index_s64(svbool_t, uint32_t const *, svuint64_t);
	9498	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_s64offset_u64)))
	9499	svuint64_t svldnt1uw_gather_s64offset_u64(svbool_t, uint32_t const *, svint64_t);
	9500	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_s64offset_s64)))
	9501	svint64_t svldnt1uw_gather_s64offset_s64(svbool_t, uint32_t const *, svint64_t);
	9502	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64offset_u64)))
	9503	svuint64_t svldnt1uw_gather_u64offset_u64(svbool_t, uint32_t const *, svuint64_t);
	9504	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64offset_s64)))
	9505	svint64_t svldnt1uw_gather_u64offset_s64(svbool_t, uint32_t const *, svuint64_t);
	9506	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmatch_u8)))
	9507	svbool_t svmatch_u8(svbool_t, svuint8_t, svuint8_t);
	9508	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmatch_u16)))
	9509	svbool_t svmatch_u16(svbool_t, svuint16_t, svuint16_t);
	9510	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmatch_s8)))
	9511	svbool_t svmatch_s8(svbool_t, svint8_t, svint8_t);
	9512	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmatch_s16)))
	9513	svbool_t svmatch_s16(svbool_t, svint16_t, svint16_t);
	9514	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svnmatch_u8)))
	9515	svbool_t svnmatch_u8(svbool_t, svuint8_t, svuint8_t);
	9516	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svnmatch_u16)))
	9517	svbool_t svnmatch_u16(svbool_t, svuint16_t, svuint16_t);
	9518	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svnmatch_s8)))
	9519	svbool_t svnmatch_s8(svbool_t, svint8_t, svint8_t);
	9520	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svnmatch_s16)))
	9521	svbool_t svnmatch_s16(svbool_t, svint16_t, svint16_t);
	9522	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_index_u32)))
	9523	void svstnt1_scatter_u32base_index_u32(svbool_t, svuint32_t, int64_t, svuint32_t);
	9524	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_index_u64)))
	9525	void svstnt1_scatter_u64base_index_u64(svbool_t, svuint64_t, int64_t, svuint64_t);
	9526	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_index_f64)))
	9527	void svstnt1_scatter_u64base_index_f64(svbool_t, svuint64_t, int64_t, svfloat64_t);
	9528	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_index_f32)))
	9529	void svstnt1_scatter_u32base_index_f32(svbool_t, svuint32_t, int64_t, svfloat32_t);
	9530	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_index_s32)))
	9531	void svstnt1_scatter_u32base_index_s32(svbool_t, svuint32_t, int64_t, svint32_t);
	9532	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_index_s64)))
	9533	void svstnt1_scatter_u64base_index_s64(svbool_t, svuint64_t, int64_t, svint64_t);
	9534	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_offset_u32)))
	9535	void svstnt1_scatter_u32base_offset_u32(svbool_t, svuint32_t, int64_t, svuint32_t);
	9536	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_offset_u64)))
	9537	void svstnt1_scatter_u64base_offset_u64(svbool_t, svuint64_t, int64_t, svuint64_t);
	9538	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_offset_f64)))
	9539	void svstnt1_scatter_u64base_offset_f64(svbool_t, svuint64_t, int64_t, svfloat64_t);
	9540	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_offset_f32)))
	9541	void svstnt1_scatter_u32base_offset_f32(svbool_t, svuint32_t, int64_t, svfloat32_t);
	9542	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_offset_s32)))
	9543	void svstnt1_scatter_u32base_offset_s32(svbool_t, svuint32_t, int64_t, svint32_t);
	9544	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_offset_s64)))
	9545	void svstnt1_scatter_u64base_offset_s64(svbool_t, svuint64_t, int64_t, svint64_t);
	9546	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_u32)))
	9547	void svstnt1_scatter_u32base_u32(svbool_t, svuint32_t, svuint32_t);
	9548	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_u64)))
	9549	void svstnt1_scatter_u64base_u64(svbool_t, svuint64_t, svuint64_t);
	9550	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_f64)))
	9551	void svstnt1_scatter_u64base_f64(svbool_t, svuint64_t, svfloat64_t);
	9552	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_f32)))
	9553	void svstnt1_scatter_u32base_f32(svbool_t, svuint32_t, svfloat32_t);
	9554	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_s32)))
	9555	void svstnt1_scatter_u32base_s32(svbool_t, svuint32_t, svint32_t);
	9556	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_s64)))
	9557	void svstnt1_scatter_u64base_s64(svbool_t, svuint64_t, svint64_t);
	9558	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_s64index_u64)))
	9559	void svstnt1_scatter_s64index_u64(svbool_t, uint64_t *, svint64_t, svuint64_t);
	9560	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_s64index_f64)))
	9561	void svstnt1_scatter_s64index_f64(svbool_t, float64_t *, svint64_t, svfloat64_t);
	9562	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_s64index_s64)))
	9563	void svstnt1_scatter_s64index_s64(svbool_t, int64_t *, svint64_t, svint64_t);
	9564	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64index_u64)))
	9565	void svstnt1_scatter_u64index_u64(svbool_t, uint64_t *, svuint64_t, svuint64_t);
	9566	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64index_f64)))
	9567	void svstnt1_scatter_u64index_f64(svbool_t, float64_t *, svuint64_t, svfloat64_t);
	9568	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64index_s64)))
	9569	void svstnt1_scatter_u64index_s64(svbool_t, int64_t *, svuint64_t, svint64_t);
	9570	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32offset_u32)))
	9571	void svstnt1_scatter_u32offset_u32(svbool_t, uint32_t *, svuint32_t, svuint32_t);
	9572	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32offset_f32)))
	9573	void svstnt1_scatter_u32offset_f32(svbool_t, float32_t *, svuint32_t, svfloat32_t);
	9574	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32offset_s32)))
	9575	void svstnt1_scatter_u32offset_s32(svbool_t, int32_t *, svuint32_t, svint32_t);
	9576	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_s64offset_u64)))
	9577	void svstnt1_scatter_s64offset_u64(svbool_t, uint64_t *, svint64_t, svuint64_t);
	9578	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_s64offset_f64)))
	9579	void svstnt1_scatter_s64offset_f64(svbool_t, float64_t *, svint64_t, svfloat64_t);
	9580	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_s64offset_s64)))
	9581	void svstnt1_scatter_s64offset_s64(svbool_t, int64_t *, svint64_t, svint64_t);
	9582	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64offset_u64)))
	9583	void svstnt1_scatter_u64offset_u64(svbool_t, uint64_t *, svuint64_t, svuint64_t);
	9584	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64offset_f64)))
	9585	void svstnt1_scatter_u64offset_f64(svbool_t, float64_t *, svuint64_t, svfloat64_t);
	9586	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64offset_s64)))
	9587	void svstnt1_scatter_u64offset_s64(svbool_t, int64_t *, svuint64_t, svint64_t);
	9588	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u32base_offset_u32)))
	9589	void svstnt1b_scatter_u32base_offset_u32(svbool_t, svuint32_t, int64_t, svuint32_t);
	9590	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u64base_offset_u64)))
	9591	void svstnt1b_scatter_u64base_offset_u64(svbool_t, svuint64_t, int64_t, svuint64_t);
	9592	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u32base_offset_s32)))
	9593	void svstnt1b_scatter_u32base_offset_s32(svbool_t, svuint32_t, int64_t, svint32_t);
	9594	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u64base_offset_s64)))
	9595	void svstnt1b_scatter_u64base_offset_s64(svbool_t, svuint64_t, int64_t, svint64_t);
	9596	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u32base_u32)))
	9597	void svstnt1b_scatter_u32base_u32(svbool_t, svuint32_t, svuint32_t);
	9598	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u64base_u64)))
	9599	void svstnt1b_scatter_u64base_u64(svbool_t, svuint64_t, svuint64_t);
	9600	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u32base_s32)))
	9601	void svstnt1b_scatter_u32base_s32(svbool_t, svuint32_t, svint32_t);
	9602	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u64base_s64)))
	9603	void svstnt1b_scatter_u64base_s64(svbool_t, svuint64_t, svint64_t);
	9604	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u32offset_s32)))
	9605	void svstnt1b_scatter_u32offset_s32(svbool_t, int8_t *, svuint32_t, svint32_t);
	9606	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u32offset_u32)))
	9607	void svstnt1b_scatter_u32offset_u32(svbool_t, uint8_t *, svuint32_t, svuint32_t);
	9608	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_s64offset_s64)))
	9609	void svstnt1b_scatter_s64offset_s64(svbool_t, int8_t *, svint64_t, svint64_t);
	9610	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_s64offset_u64)))
	9611	void svstnt1b_scatter_s64offset_u64(svbool_t, uint8_t *, svint64_t, svuint64_t);
	9612	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u64offset_s64)))
	9613	void svstnt1b_scatter_u64offset_s64(svbool_t, int8_t *, svuint64_t, svint64_t);
	9614	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u64offset_u64)))
	9615	void svstnt1b_scatter_u64offset_u64(svbool_t, uint8_t *, svuint64_t, svuint64_t);
	9616	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32base_index_u32)))
	9617	void svstnt1h_scatter_u32base_index_u32(svbool_t, svuint32_t, int64_t, svuint32_t);
	9618	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64base_index_u64)))
	9619	void svstnt1h_scatter_u64base_index_u64(svbool_t, svuint64_t, int64_t, svuint64_t);
	9620	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32base_index_s32)))
	9621	void svstnt1h_scatter_u32base_index_s32(svbool_t, svuint32_t, int64_t, svint32_t);
	9622	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64base_index_s64)))
	9623	void svstnt1h_scatter_u64base_index_s64(svbool_t, svuint64_t, int64_t, svint64_t);
	9624	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32base_offset_u32)))
	9625	void svstnt1h_scatter_u32base_offset_u32(svbool_t, svuint32_t, int64_t, svuint32_t);
	9626	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64base_offset_u64)))
	9627	void svstnt1h_scatter_u64base_offset_u64(svbool_t, svuint64_t, int64_t, svuint64_t);
	9628	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32base_offset_s32)))
	9629	void svstnt1h_scatter_u32base_offset_s32(svbool_t, svuint32_t, int64_t, svint32_t);
	9630	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64base_offset_s64)))
	9631	void svstnt1h_scatter_u64base_offset_s64(svbool_t, svuint64_t, int64_t, svint64_t);
	9632	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32base_u32)))
	9633	void svstnt1h_scatter_u32base_u32(svbool_t, svuint32_t, svuint32_t);
	9634	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64base_u64)))
	9635	void svstnt1h_scatter_u64base_u64(svbool_t, svuint64_t, svuint64_t);
	9636	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32base_s32)))
	9637	void svstnt1h_scatter_u32base_s32(svbool_t, svuint32_t, svint32_t);
	9638	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64base_s64)))
	9639	void svstnt1h_scatter_u64base_s64(svbool_t, svuint64_t, svint64_t);
	9640	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_s64index_s64)))
	9641	void svstnt1h_scatter_s64index_s64(svbool_t, int16_t *, svint64_t, svint64_t);
	9642	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_s64index_u64)))
	9643	void svstnt1h_scatter_s64index_u64(svbool_t, uint16_t *, svint64_t, svuint64_t);
	9644	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64index_s64)))
	9645	void svstnt1h_scatter_u64index_s64(svbool_t, int16_t *, svuint64_t, svint64_t);
	9646	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64index_u64)))
	9647	void svstnt1h_scatter_u64index_u64(svbool_t, uint16_t *, svuint64_t, svuint64_t);
	9648	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32offset_s32)))
	9649	void svstnt1h_scatter_u32offset_s32(svbool_t, int16_t *, svuint32_t, svint32_t);
	9650	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32offset_u32)))
	9651	void svstnt1h_scatter_u32offset_u32(svbool_t, uint16_t *, svuint32_t, svuint32_t);
	9652	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_s64offset_s64)))
	9653	void svstnt1h_scatter_s64offset_s64(svbool_t, int16_t *, svint64_t, svint64_t);
	9654	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_s64offset_u64)))
	9655	void svstnt1h_scatter_s64offset_u64(svbool_t, uint16_t *, svint64_t, svuint64_t);
	9656	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64offset_s64)))
	9657	void svstnt1h_scatter_u64offset_s64(svbool_t, int16_t *, svuint64_t, svint64_t);
	9658	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64offset_u64)))
	9659	void svstnt1h_scatter_u64offset_u64(svbool_t, uint16_t *, svuint64_t, svuint64_t);
	9660	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64base_index_u64)))
	9661	void svstnt1w_scatter_u64base_index_u64(svbool_t, svuint64_t, int64_t, svuint64_t);
	9662	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64base_index_s64)))
	9663	void svstnt1w_scatter_u64base_index_s64(svbool_t, svuint64_t, int64_t, svint64_t);
	9664	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64base_offset_u64)))
	9665	void svstnt1w_scatter_u64base_offset_u64(svbool_t, svuint64_t, int64_t, svuint64_t);
	9666	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64base_offset_s64)))
	9667	void svstnt1w_scatter_u64base_offset_s64(svbool_t, svuint64_t, int64_t, svint64_t);
	9668	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64base_u64)))
	9669	void svstnt1w_scatter_u64base_u64(svbool_t, svuint64_t, svuint64_t);
	9670	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64base_s64)))
	9671	void svstnt1w_scatter_u64base_s64(svbool_t, svuint64_t, svint64_t);
	9672	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_s64index_s64)))
	9673	void svstnt1w_scatter_s64index_s64(svbool_t, int32_t *, svint64_t, svint64_t);
	9674	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_s64index_u64)))
	9675	void svstnt1w_scatter_s64index_u64(svbool_t, uint32_t *, svint64_t, svuint64_t);
	9676	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64index_s64)))
	9677	void svstnt1w_scatter_u64index_s64(svbool_t, int32_t *, svuint64_t, svint64_t);
	9678	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64index_u64)))
	9679	void svstnt1w_scatter_u64index_u64(svbool_t, uint32_t *, svuint64_t, svuint64_t);
	9680	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_s64offset_s64)))
	9681	void svstnt1w_scatter_s64offset_s64(svbool_t, int32_t *, svint64_t, svint64_t);
	9682	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_s64offset_u64)))
	9683	void svstnt1w_scatter_s64offset_u64(svbool_t, uint32_t *, svint64_t, svuint64_t);
	9684	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64offset_s64)))
	9685	void svstnt1w_scatter_u64offset_s64(svbool_t, int32_t *, svuint64_t, svint64_t);
	9686	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64offset_u64)))
	9687	void svstnt1w_scatter_u64offset_u64(svbool_t, uint32_t *, svuint64_t, svuint64_t);
	9688	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svhistcnt_u32_z)))
	9689	svuint32_t svhistcnt_z(svbool_t, svuint32_t, svuint32_t);
	9690	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svhistcnt_u64_z)))
	9691	svuint64_t svhistcnt_z(svbool_t, svuint64_t, svuint64_t);
	9692	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svhistcnt_s32_z)))
	9693	svuint32_t svhistcnt_z(svbool_t, svint32_t, svint32_t);
	9694	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svhistcnt_s64_z)))
	9695	svuint64_t svhistcnt_z(svbool_t, svint64_t, svint64_t);
	9696	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svhistseg_u8)))
	9697	svuint8_t svhistseg(svuint8_t, svuint8_t);
	9698	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svhistseg_s8)))
	9699	svuint8_t svhistseg(svint8_t, svint8_t);
	9700	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_index_u32)))
	9701	svuint32_t svldnt1_gather_index_u32(svbool_t, svuint32_t, int64_t);
	9702	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_index_u64)))
	9703	svuint64_t svldnt1_gather_index_u64(svbool_t, svuint64_t, int64_t);
	9704	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_index_f64)))
	9705	svfloat64_t svldnt1_gather_index_f64(svbool_t, svuint64_t, int64_t);
	9706	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_index_f32)))
	9707	svfloat32_t svldnt1_gather_index_f32(svbool_t, svuint32_t, int64_t);
	9708	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_index_s32)))
	9709	svint32_t svldnt1_gather_index_s32(svbool_t, svuint32_t, int64_t);
	9710	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_index_s64)))
	9711	svint64_t svldnt1_gather_index_s64(svbool_t, svuint64_t, int64_t);
	9712	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_offset_u32)))
	9713	svuint32_t svldnt1_gather_offset_u32(svbool_t, svuint32_t, int64_t);
	9714	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_offset_u64)))
	9715	svuint64_t svldnt1_gather_offset_u64(svbool_t, svuint64_t, int64_t);
	9716	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_offset_f64)))
	9717	svfloat64_t svldnt1_gather_offset_f64(svbool_t, svuint64_t, int64_t);
	9718	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_offset_f32)))
	9719	svfloat32_t svldnt1_gather_offset_f32(svbool_t, svuint32_t, int64_t);
	9720	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_offset_s32)))
	9721	svint32_t svldnt1_gather_offset_s32(svbool_t, svuint32_t, int64_t);
	9722	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_offset_s64)))
	9723	svint64_t svldnt1_gather_offset_s64(svbool_t, svuint64_t, int64_t);
	9724	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_u32)))
	9725	svuint32_t svldnt1_gather_u32(svbool_t, svuint32_t);
	9726	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_u64)))
	9727	svuint64_t svldnt1_gather_u64(svbool_t, svuint64_t);
	9728	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_f64)))
	9729	svfloat64_t svldnt1_gather_f64(svbool_t, svuint64_t);
	9730	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_f32)))
	9731	svfloat32_t svldnt1_gather_f32(svbool_t, svuint32_t);
	9732	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32base_s32)))
	9733	svint32_t svldnt1_gather_s32(svbool_t, svuint32_t);
	9734	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64base_s64)))
	9735	svint64_t svldnt1_gather_s64(svbool_t, svuint64_t);
	9736	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_s64index_u64)))
	9737	svuint64_t svldnt1_gather_index(svbool_t, uint64_t const *, svint64_t);
	9738	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_s64index_f64)))
	9739	svfloat64_t svldnt1_gather_index(svbool_t, float64_t const *, svint64_t);
	9740	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_s64index_s64)))
	9741	svint64_t svldnt1_gather_index(svbool_t, int64_t const *, svint64_t);
	9742	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64index_u64)))
	9743	svuint64_t svldnt1_gather_index(svbool_t, uint64_t const *, svuint64_t);
	9744	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64index_f64)))
	9745	svfloat64_t svldnt1_gather_index(svbool_t, float64_t const *, svuint64_t);
	9746	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64index_s64)))
	9747	svint64_t svldnt1_gather_index(svbool_t, int64_t const *, svuint64_t);
	9748	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32offset_u32)))
	9749	svuint32_t svldnt1_gather_offset(svbool_t, uint32_t const *, svuint32_t);
	9750	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32offset_f32)))
	9751	svfloat32_t svldnt1_gather_offset(svbool_t, float32_t const *, svuint32_t);
	9752	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u32offset_s32)))
	9753	svint32_t svldnt1_gather_offset(svbool_t, int32_t const *, svuint32_t);
	9754	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_s64offset_u64)))
	9755	svuint64_t svldnt1_gather_offset(svbool_t, uint64_t const *, svint64_t);
	9756	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_s64offset_f64)))
	9757	svfloat64_t svldnt1_gather_offset(svbool_t, float64_t const *, svint64_t);
	9758	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_s64offset_s64)))
	9759	svint64_t svldnt1_gather_offset(svbool_t, int64_t const *, svint64_t);
	9760	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64offset_u64)))
	9761	svuint64_t svldnt1_gather_offset(svbool_t, uint64_t const *, svuint64_t);
	9762	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64offset_f64)))
	9763	svfloat64_t svldnt1_gather_offset(svbool_t, float64_t const *, svuint64_t);
	9764	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_gather_u64offset_s64)))
	9765	svint64_t svldnt1_gather_offset(svbool_t, int64_t const *, svuint64_t);
	9766	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u32base_offset_u32)))
	9767	svuint32_t svldnt1sb_gather_offset_u32(svbool_t, svuint32_t, int64_t);
	9768	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u64base_offset_u64)))
	9769	svuint64_t svldnt1sb_gather_offset_u64(svbool_t, svuint64_t, int64_t);
	9770	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u32base_offset_s32)))
	9771	svint32_t svldnt1sb_gather_offset_s32(svbool_t, svuint32_t, int64_t);
	9772	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u64base_offset_s64)))
	9773	svint64_t svldnt1sb_gather_offset_s64(svbool_t, svuint64_t, int64_t);
	9774	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u32base_u32)))
	9775	svuint32_t svldnt1sb_gather_u32(svbool_t, svuint32_t);
	9776	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u64base_u64)))
	9777	svuint64_t svldnt1sb_gather_u64(svbool_t, svuint64_t);
	9778	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u32base_s32)))
	9779	svint32_t svldnt1sb_gather_s32(svbool_t, svuint32_t);
	9780	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u64base_s64)))
	9781	svint64_t svldnt1sb_gather_s64(svbool_t, svuint64_t);
	9782	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u32offset_u32)))
	9783	svuint32_t svldnt1sb_gather_offset_u32(svbool_t, int8_t const *, svuint32_t);
	9784	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u32offset_s32)))
	9785	svint32_t svldnt1sb_gather_offset_s32(svbool_t, int8_t const *, svuint32_t);
	9786	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_s64offset_u64)))
	9787	svuint64_t svldnt1sb_gather_offset_u64(svbool_t, int8_t const *, svint64_t);
	9788	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_s64offset_s64)))
	9789	svint64_t svldnt1sb_gather_offset_s64(svbool_t, int8_t const *, svint64_t);
	9790	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u64offset_u64)))
	9791	svuint64_t svldnt1sb_gather_offset_u64(svbool_t, int8_t const *, svuint64_t);
	9792	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sb_gather_u64offset_s64)))
	9793	svint64_t svldnt1sb_gather_offset_s64(svbool_t, int8_t const *, svuint64_t);
	9794	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32base_index_u32)))
	9795	svuint32_t svldnt1sh_gather_index_u32(svbool_t, svuint32_t, int64_t);
	9796	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64base_index_u64)))
	9797	svuint64_t svldnt1sh_gather_index_u64(svbool_t, svuint64_t, int64_t);
	9798	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32base_index_s32)))
	9799	svint32_t svldnt1sh_gather_index_s32(svbool_t, svuint32_t, int64_t);
	9800	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64base_index_s64)))
	9801	svint64_t svldnt1sh_gather_index_s64(svbool_t, svuint64_t, int64_t);
	9802	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32base_offset_u32)))
	9803	svuint32_t svldnt1sh_gather_offset_u32(svbool_t, svuint32_t, int64_t);
	9804	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64base_offset_u64)))
	9805	svuint64_t svldnt1sh_gather_offset_u64(svbool_t, svuint64_t, int64_t);
	9806	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32base_offset_s32)))
	9807	svint32_t svldnt1sh_gather_offset_s32(svbool_t, svuint32_t, int64_t);
	9808	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64base_offset_s64)))
	9809	svint64_t svldnt1sh_gather_offset_s64(svbool_t, svuint64_t, int64_t);
	9810	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32base_u32)))
	9811	svuint32_t svldnt1sh_gather_u32(svbool_t, svuint32_t);
	9812	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64base_u64)))
	9813	svuint64_t svldnt1sh_gather_u64(svbool_t, svuint64_t);
	9814	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32base_s32)))
	9815	svint32_t svldnt1sh_gather_s32(svbool_t, svuint32_t);
	9816	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64base_s64)))
	9817	svint64_t svldnt1sh_gather_s64(svbool_t, svuint64_t);
	9818	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_s64index_u64)))
	9819	svuint64_t svldnt1sh_gather_index_u64(svbool_t, int16_t const *, svint64_t);
	9820	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_s64index_s64)))
	9821	svint64_t svldnt1sh_gather_index_s64(svbool_t, int16_t const *, svint64_t);
	9822	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64index_u64)))
	9823	svuint64_t svldnt1sh_gather_index_u64(svbool_t, int16_t const *, svuint64_t);
	9824	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64index_s64)))
	9825	svint64_t svldnt1sh_gather_index_s64(svbool_t, int16_t const *, svuint64_t);
	9826	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32offset_u32)))
	9827	svuint32_t svldnt1sh_gather_offset_u32(svbool_t, int16_t const *, svuint32_t);
	9828	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u32offset_s32)))
	9829	svint32_t svldnt1sh_gather_offset_s32(svbool_t, int16_t const *, svuint32_t);
	9830	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_s64offset_u64)))
	9831	svuint64_t svldnt1sh_gather_offset_u64(svbool_t, int16_t const *, svint64_t);
	9832	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_s64offset_s64)))
	9833	svint64_t svldnt1sh_gather_offset_s64(svbool_t, int16_t const *, svint64_t);
	9834	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64offset_u64)))
	9835	svuint64_t svldnt1sh_gather_offset_u64(svbool_t, int16_t const *, svuint64_t);
	9836	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sh_gather_u64offset_s64)))
	9837	svint64_t svldnt1sh_gather_offset_s64(svbool_t, int16_t const *, svuint64_t);
	9838	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64base_index_u64)))
	9839	svuint64_t svldnt1sw_gather_index_u64(svbool_t, svuint64_t, int64_t);
	9840	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64base_index_s64)))
	9841	svint64_t svldnt1sw_gather_index_s64(svbool_t, svuint64_t, int64_t);
	9842	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64base_offset_u64)))
	9843	svuint64_t svldnt1sw_gather_offset_u64(svbool_t, svuint64_t, int64_t);
	9844	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64base_offset_s64)))
	9845	svint64_t svldnt1sw_gather_offset_s64(svbool_t, svuint64_t, int64_t);
	9846	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64base_u64)))
	9847	svuint64_t svldnt1sw_gather_u64(svbool_t, svuint64_t);
	9848	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64base_s64)))
	9849	svint64_t svldnt1sw_gather_s64(svbool_t, svuint64_t);
	9850	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_s64index_u64)))
	9851	svuint64_t svldnt1sw_gather_index_u64(svbool_t, int32_t const *, svint64_t);
	9852	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_s64index_s64)))
	9853	svint64_t svldnt1sw_gather_index_s64(svbool_t, int32_t const *, svint64_t);
	9854	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64index_u64)))
	9855	svuint64_t svldnt1sw_gather_index_u64(svbool_t, int32_t const *, svuint64_t);
	9856	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64index_s64)))
	9857	svint64_t svldnt1sw_gather_index_s64(svbool_t, int32_t const *, svuint64_t);
	9858	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_s64offset_u64)))
	9859	svuint64_t svldnt1sw_gather_offset_u64(svbool_t, int32_t const *, svint64_t);
	9860	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_s64offset_s64)))
	9861	svint64_t svldnt1sw_gather_offset_s64(svbool_t, int32_t const *, svint64_t);
	9862	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64offset_u64)))
	9863	svuint64_t svldnt1sw_gather_offset_u64(svbool_t, int32_t const *, svuint64_t);
	9864	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1sw_gather_u64offset_s64)))
	9865	svint64_t svldnt1sw_gather_offset_s64(svbool_t, int32_t const *, svuint64_t);
	9866	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u32base_offset_u32)))
	9867	svuint32_t svldnt1ub_gather_offset_u32(svbool_t, svuint32_t, int64_t);
	9868	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u64base_offset_u64)))
	9869	svuint64_t svldnt1ub_gather_offset_u64(svbool_t, svuint64_t, int64_t);
	9870	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u32base_offset_s32)))
	9871	svint32_t svldnt1ub_gather_offset_s32(svbool_t, svuint32_t, int64_t);
	9872	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u64base_offset_s64)))
	9873	svint64_t svldnt1ub_gather_offset_s64(svbool_t, svuint64_t, int64_t);
	9874	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u32base_u32)))
	9875	svuint32_t svldnt1ub_gather_u32(svbool_t, svuint32_t);
	9876	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u64base_u64)))
	9877	svuint64_t svldnt1ub_gather_u64(svbool_t, svuint64_t);
	9878	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u32base_s32)))
	9879	svint32_t svldnt1ub_gather_s32(svbool_t, svuint32_t);
	9880	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u64base_s64)))
	9881	svint64_t svldnt1ub_gather_s64(svbool_t, svuint64_t);
	9882	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u32offset_u32)))
	9883	svuint32_t svldnt1ub_gather_offset_u32(svbool_t, uint8_t const *, svuint32_t);
	9884	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u32offset_s32)))
	9885	svint32_t svldnt1ub_gather_offset_s32(svbool_t, uint8_t const *, svuint32_t);
	9886	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_s64offset_u64)))
	9887	svuint64_t svldnt1ub_gather_offset_u64(svbool_t, uint8_t const *, svint64_t);
	9888	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_s64offset_s64)))
	9889	svint64_t svldnt1ub_gather_offset_s64(svbool_t, uint8_t const *, svint64_t);
	9890	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u64offset_u64)))
	9891	svuint64_t svldnt1ub_gather_offset_u64(svbool_t, uint8_t const *, svuint64_t);
	9892	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1ub_gather_u64offset_s64)))
	9893	svint64_t svldnt1ub_gather_offset_s64(svbool_t, uint8_t const *, svuint64_t);
	9894	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32base_index_u32)))
	9895	svuint32_t svldnt1uh_gather_index_u32(svbool_t, svuint32_t, int64_t);
	9896	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64base_index_u64)))
	9897	svuint64_t svldnt1uh_gather_index_u64(svbool_t, svuint64_t, int64_t);
	9898	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32base_index_s32)))
	9899	svint32_t svldnt1uh_gather_index_s32(svbool_t, svuint32_t, int64_t);
	9900	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64base_index_s64)))
	9901	svint64_t svldnt1uh_gather_index_s64(svbool_t, svuint64_t, int64_t);
	9902	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32base_offset_u32)))
	9903	svuint32_t svldnt1uh_gather_offset_u32(svbool_t, svuint32_t, int64_t);
	9904	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64base_offset_u64)))
	9905	svuint64_t svldnt1uh_gather_offset_u64(svbool_t, svuint64_t, int64_t);
	9906	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32base_offset_s32)))
	9907	svint32_t svldnt1uh_gather_offset_s32(svbool_t, svuint32_t, int64_t);
	9908	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64base_offset_s64)))
	9909	svint64_t svldnt1uh_gather_offset_s64(svbool_t, svuint64_t, int64_t);
	9910	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32base_u32)))
	9911	svuint32_t svldnt1uh_gather_u32(svbool_t, svuint32_t);
	9912	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64base_u64)))
	9913	svuint64_t svldnt1uh_gather_u64(svbool_t, svuint64_t);
	9914	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32base_s32)))
	9915	svint32_t svldnt1uh_gather_s32(svbool_t, svuint32_t);
	9916	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64base_s64)))
	9917	svint64_t svldnt1uh_gather_s64(svbool_t, svuint64_t);
	9918	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_s64index_u64)))
	9919	svuint64_t svldnt1uh_gather_index_u64(svbool_t, uint16_t const *, svint64_t);
	9920	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_s64index_s64)))
	9921	svint64_t svldnt1uh_gather_index_s64(svbool_t, uint16_t const *, svint64_t);
	9922	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64index_u64)))
	9923	svuint64_t svldnt1uh_gather_index_u64(svbool_t, uint16_t const *, svuint64_t);
	9924	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64index_s64)))
	9925	svint64_t svldnt1uh_gather_index_s64(svbool_t, uint16_t const *, svuint64_t);
	9926	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32offset_u32)))
	9927	svuint32_t svldnt1uh_gather_offset_u32(svbool_t, uint16_t const *, svuint32_t);
	9928	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u32offset_s32)))
	9929	svint32_t svldnt1uh_gather_offset_s32(svbool_t, uint16_t const *, svuint32_t);
	9930	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_s64offset_u64)))
	9931	svuint64_t svldnt1uh_gather_offset_u64(svbool_t, uint16_t const *, svint64_t);
	9932	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_s64offset_s64)))
	9933	svint64_t svldnt1uh_gather_offset_s64(svbool_t, uint16_t const *, svint64_t);
	9934	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64offset_u64)))
	9935	svuint64_t svldnt1uh_gather_offset_u64(svbool_t, uint16_t const *, svuint64_t);
	9936	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uh_gather_u64offset_s64)))
	9937	svint64_t svldnt1uh_gather_offset_s64(svbool_t, uint16_t const *, svuint64_t);
	9938	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64base_index_u64)))
	9939	svuint64_t svldnt1uw_gather_index_u64(svbool_t, svuint64_t, int64_t);
	9940	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64base_index_s64)))
	9941	svint64_t svldnt1uw_gather_index_s64(svbool_t, svuint64_t, int64_t);
	9942	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64base_offset_u64)))
	9943	svuint64_t svldnt1uw_gather_offset_u64(svbool_t, svuint64_t, int64_t);
	9944	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64base_offset_s64)))
	9945	svint64_t svldnt1uw_gather_offset_s64(svbool_t, svuint64_t, int64_t);
	9946	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64base_u64)))
	9947	svuint64_t svldnt1uw_gather_u64(svbool_t, svuint64_t);
	9948	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64base_s64)))
	9949	svint64_t svldnt1uw_gather_s64(svbool_t, svuint64_t);
	9950	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_s64index_u64)))
	9951	svuint64_t svldnt1uw_gather_index_u64(svbool_t, uint32_t const *, svint64_t);
	9952	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_s64index_s64)))
	9953	svint64_t svldnt1uw_gather_index_s64(svbool_t, uint32_t const *, svint64_t);
	9954	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64index_u64)))
	9955	svuint64_t svldnt1uw_gather_index_u64(svbool_t, uint32_t const *, svuint64_t);
	9956	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64index_s64)))
	9957	svint64_t svldnt1uw_gather_index_s64(svbool_t, uint32_t const *, svuint64_t);
	9958	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_s64offset_u64)))
	9959	svuint64_t svldnt1uw_gather_offset_u64(svbool_t, uint32_t const *, svint64_t);
	9960	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_s64offset_s64)))
	9961	svint64_t svldnt1uw_gather_offset_s64(svbool_t, uint32_t const *, svint64_t);
	9962	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64offset_u64)))
	9963	svuint64_t svldnt1uw_gather_offset_u64(svbool_t, uint32_t const *, svuint64_t);
	9964	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1uw_gather_u64offset_s64)))
	9965	svint64_t svldnt1uw_gather_offset_s64(svbool_t, uint32_t const *, svuint64_t);
	9966	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmatch_u8)))
	9967	svbool_t svmatch(svbool_t, svuint8_t, svuint8_t);
	9968	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmatch_u16)))
	9969	svbool_t svmatch(svbool_t, svuint16_t, svuint16_t);
	9970	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmatch_s8)))
	9971	svbool_t svmatch(svbool_t, svint8_t, svint8_t);
	9972	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmatch_s16)))
	9973	svbool_t svmatch(svbool_t, svint16_t, svint16_t);
	9974	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svnmatch_u8)))
	9975	svbool_t svnmatch(svbool_t, svuint8_t, svuint8_t);
	9976	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svnmatch_u16)))
	9977	svbool_t svnmatch(svbool_t, svuint16_t, svuint16_t);
	9978	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svnmatch_s8)))
	9979	svbool_t svnmatch(svbool_t, svint8_t, svint8_t);
	9980	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svnmatch_s16)))
	9981	svbool_t svnmatch(svbool_t, svint16_t, svint16_t);
	9982	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_index_u32)))
	9983	void svstnt1_scatter_index(svbool_t, svuint32_t, int64_t, svuint32_t);
	9984	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_index_u64)))
	9985	void svstnt1_scatter_index(svbool_t, svuint64_t, int64_t, svuint64_t);
	9986	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_index_f64)))
	9987	void svstnt1_scatter_index(svbool_t, svuint64_t, int64_t, svfloat64_t);
	9988	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_index_f32)))
	9989	void svstnt1_scatter_index(svbool_t, svuint32_t, int64_t, svfloat32_t);
	9990	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_index_s32)))
	9991	void svstnt1_scatter_index(svbool_t, svuint32_t, int64_t, svint32_t);
	9992	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_index_s64)))
	9993	void svstnt1_scatter_index(svbool_t, svuint64_t, int64_t, svint64_t);
	9994	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_offset_u32)))
	9995	void svstnt1_scatter_offset(svbool_t, svuint32_t, int64_t, svuint32_t);
	9996	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_offset_u64)))
	9997	void svstnt1_scatter_offset(svbool_t, svuint64_t, int64_t, svuint64_t);
	9998	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_offset_f64)))
	9999	void svstnt1_scatter_offset(svbool_t, svuint64_t, int64_t, svfloat64_t);
	10000	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_offset_f32)))
	10001	void svstnt1_scatter_offset(svbool_t, svuint32_t, int64_t, svfloat32_t);
	10002	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_offset_s32)))
	10003	void svstnt1_scatter_offset(svbool_t, svuint32_t, int64_t, svint32_t);
	10004	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_offset_s64)))
	10005	void svstnt1_scatter_offset(svbool_t, svuint64_t, int64_t, svint64_t);
	10006	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_u32)))
	10007	void svstnt1_scatter(svbool_t, svuint32_t, svuint32_t);
	10008	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_u64)))
	10009	void svstnt1_scatter(svbool_t, svuint64_t, svuint64_t);
	10010	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_f64)))
	10011	void svstnt1_scatter(svbool_t, svuint64_t, svfloat64_t);
	10012	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_f32)))
	10013	void svstnt1_scatter(svbool_t, svuint32_t, svfloat32_t);
	10014	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32base_s32)))
	10015	void svstnt1_scatter(svbool_t, svuint32_t, svint32_t);
	10016	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64base_s64)))
	10017	void svstnt1_scatter(svbool_t, svuint64_t, svint64_t);
	10018	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_s64index_u64)))
	10019	void svstnt1_scatter_index(svbool_t, uint64_t *, svint64_t, svuint64_t);
	10020	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_s64index_f64)))
	10021	void svstnt1_scatter_index(svbool_t, float64_t *, svint64_t, svfloat64_t);
	10022	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_s64index_s64)))
	10023	void svstnt1_scatter_index(svbool_t, int64_t *, svint64_t, svint64_t);
	10024	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64index_u64)))
	10025	void svstnt1_scatter_index(svbool_t, uint64_t *, svuint64_t, svuint64_t);
	10026	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64index_f64)))
	10027	void svstnt1_scatter_index(svbool_t, float64_t *, svuint64_t, svfloat64_t);
	10028	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64index_s64)))
	10029	void svstnt1_scatter_index(svbool_t, int64_t *, svuint64_t, svint64_t);
	10030	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32offset_u32)))
	10031	void svstnt1_scatter_offset(svbool_t, uint32_t *, svuint32_t, svuint32_t);
	10032	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32offset_f32)))
	10033	void svstnt1_scatter_offset(svbool_t, float32_t *, svuint32_t, svfloat32_t);
	10034	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u32offset_s32)))
	10035	void svstnt1_scatter_offset(svbool_t, int32_t *, svuint32_t, svint32_t);
	10036	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_s64offset_u64)))
	10037	void svstnt1_scatter_offset(svbool_t, uint64_t *, svint64_t, svuint64_t);
	10038	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_s64offset_f64)))
	10039	void svstnt1_scatter_offset(svbool_t, float64_t *, svint64_t, svfloat64_t);
	10040	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_s64offset_s64)))
	10041	void svstnt1_scatter_offset(svbool_t, int64_t *, svint64_t, svint64_t);
	10042	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64offset_u64)))
	10043	void svstnt1_scatter_offset(svbool_t, uint64_t *, svuint64_t, svuint64_t);
	10044	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64offset_f64)))
	10045	void svstnt1_scatter_offset(svbool_t, float64_t *, svuint64_t, svfloat64_t);
	10046	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_scatter_u64offset_s64)))
	10047	void svstnt1_scatter_offset(svbool_t, int64_t *, svuint64_t, svint64_t);
	10048	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u32base_offset_u32)))
	10049	void svstnt1b_scatter_offset(svbool_t, svuint32_t, int64_t, svuint32_t);
	10050	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u64base_offset_u64)))
	10051	void svstnt1b_scatter_offset(svbool_t, svuint64_t, int64_t, svuint64_t);
	10052	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u32base_offset_s32)))
	10053	void svstnt1b_scatter_offset(svbool_t, svuint32_t, int64_t, svint32_t);
	10054	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u64base_offset_s64)))
	10055	void svstnt1b_scatter_offset(svbool_t, svuint64_t, int64_t, svint64_t);
	10056	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u32base_u32)))
	10057	void svstnt1b_scatter(svbool_t, svuint32_t, svuint32_t);
	10058	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u64base_u64)))
	10059	void svstnt1b_scatter(svbool_t, svuint64_t, svuint64_t);
	10060	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u32base_s32)))
	10061	void svstnt1b_scatter(svbool_t, svuint32_t, svint32_t);
	10062	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u64base_s64)))
	10063	void svstnt1b_scatter(svbool_t, svuint64_t, svint64_t);
	10064	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u32offset_s32)))
	10065	void svstnt1b_scatter_offset(svbool_t, int8_t *, svuint32_t, svint32_t);
	10066	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u32offset_u32)))
	10067	void svstnt1b_scatter_offset(svbool_t, uint8_t *, svuint32_t, svuint32_t);
	10068	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_s64offset_s64)))
	10069	void svstnt1b_scatter_offset(svbool_t, int8_t *, svint64_t, svint64_t);
	10070	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_s64offset_u64)))
	10071	void svstnt1b_scatter_offset(svbool_t, uint8_t *, svint64_t, svuint64_t);
	10072	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u64offset_s64)))
	10073	void svstnt1b_scatter_offset(svbool_t, int8_t *, svuint64_t, svint64_t);
	10074	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1b_scatter_u64offset_u64)))
	10075	void svstnt1b_scatter_offset(svbool_t, uint8_t *, svuint64_t, svuint64_t);
	10076	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32base_index_u32)))
	10077	void svstnt1h_scatter_index(svbool_t, svuint32_t, int64_t, svuint32_t);
	10078	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64base_index_u64)))
	10079	void svstnt1h_scatter_index(svbool_t, svuint64_t, int64_t, svuint64_t);
	10080	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32base_index_s32)))
	10081	void svstnt1h_scatter_index(svbool_t, svuint32_t, int64_t, svint32_t);
	10082	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64base_index_s64)))
	10083	void svstnt1h_scatter_index(svbool_t, svuint64_t, int64_t, svint64_t);
	10084	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32base_offset_u32)))
	10085	void svstnt1h_scatter_offset(svbool_t, svuint32_t, int64_t, svuint32_t);
	10086	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64base_offset_u64)))
	10087	void svstnt1h_scatter_offset(svbool_t, svuint64_t, int64_t, svuint64_t);
	10088	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32base_offset_s32)))
	10089	void svstnt1h_scatter_offset(svbool_t, svuint32_t, int64_t, svint32_t);
	10090	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64base_offset_s64)))
	10091	void svstnt1h_scatter_offset(svbool_t, svuint64_t, int64_t, svint64_t);
	10092	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32base_u32)))
	10093	void svstnt1h_scatter(svbool_t, svuint32_t, svuint32_t);
	10094	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64base_u64)))
	10095	void svstnt1h_scatter(svbool_t, svuint64_t, svuint64_t);
	10096	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32base_s32)))
	10097	void svstnt1h_scatter(svbool_t, svuint32_t, svint32_t);
	10098	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64base_s64)))
	10099	void svstnt1h_scatter(svbool_t, svuint64_t, svint64_t);
	10100	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_s64index_s64)))
	10101	void svstnt1h_scatter_index(svbool_t, int16_t *, svint64_t, svint64_t);
	10102	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_s64index_u64)))
	10103	void svstnt1h_scatter_index(svbool_t, uint16_t *, svint64_t, svuint64_t);
	10104	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64index_s64)))
	10105	void svstnt1h_scatter_index(svbool_t, int16_t *, svuint64_t, svint64_t);
	10106	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64index_u64)))
	10107	void svstnt1h_scatter_index(svbool_t, uint16_t *, svuint64_t, svuint64_t);
	10108	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32offset_s32)))
	10109	void svstnt1h_scatter_offset(svbool_t, int16_t *, svuint32_t, svint32_t);
	10110	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u32offset_u32)))
	10111	void svstnt1h_scatter_offset(svbool_t, uint16_t *, svuint32_t, svuint32_t);
	10112	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_s64offset_s64)))
	10113	void svstnt1h_scatter_offset(svbool_t, int16_t *, svint64_t, svint64_t);
	10114	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_s64offset_u64)))
	10115	void svstnt1h_scatter_offset(svbool_t, uint16_t *, svint64_t, svuint64_t);
	10116	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64offset_s64)))
	10117	void svstnt1h_scatter_offset(svbool_t, int16_t *, svuint64_t, svint64_t);
	10118	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1h_scatter_u64offset_u64)))
	10119	void svstnt1h_scatter_offset(svbool_t, uint16_t *, svuint64_t, svuint64_t);
	10120	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64base_index_u64)))
	10121	void svstnt1w_scatter_index(svbool_t, svuint64_t, int64_t, svuint64_t);
	10122	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64base_index_s64)))
	10123	void svstnt1w_scatter_index(svbool_t, svuint64_t, int64_t, svint64_t);
	10124	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64base_offset_u64)))
	10125	void svstnt1w_scatter_offset(svbool_t, svuint64_t, int64_t, svuint64_t);
	10126	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64base_offset_s64)))
	10127	void svstnt1w_scatter_offset(svbool_t, svuint64_t, int64_t, svint64_t);
	10128	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64base_u64)))
	10129	void svstnt1w_scatter(svbool_t, svuint64_t, svuint64_t);
	10130	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64base_s64)))
	10131	void svstnt1w_scatter(svbool_t, svuint64_t, svint64_t);
	10132	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_s64index_s64)))
	10133	void svstnt1w_scatter_index(svbool_t, int32_t *, svint64_t, svint64_t);
	10134	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_s64index_u64)))
	10135	void svstnt1w_scatter_index(svbool_t, uint32_t *, svint64_t, svuint64_t);
	10136	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64index_s64)))
	10137	void svstnt1w_scatter_index(svbool_t, int32_t *, svuint64_t, svint64_t);
	10138	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64index_u64)))
	10139	void svstnt1w_scatter_index(svbool_t, uint32_t *, svuint64_t, svuint64_t);
	10140	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_s64offset_s64)))
	10141	void svstnt1w_scatter_offset(svbool_t, int32_t *, svint64_t, svint64_t);
	10142	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_s64offset_u64)))
	10143	void svstnt1w_scatter_offset(svbool_t, uint32_t *, svint64_t, svuint64_t);
	10144	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64offset_s64)))
	10145	void svstnt1w_scatter_offset(svbool_t, int32_t *, svuint64_t, svint64_t);
	10146	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1w_scatter_u64offset_u64)))
	10147	void svstnt1w_scatter_offset(svbool_t, uint32_t *, svuint64_t, svuint64_t);
	10148	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f64_m)))
	10149	svfloat64_t svamax_n_f64_m(svbool_t, svfloat64_t, float64_t);
	10150	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f32_m)))
	10151	svfloat32_t svamax_n_f32_m(svbool_t, svfloat32_t, float32_t);
	10152	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f16_m)))
	10153	svfloat16_t svamax_n_f16_m(svbool_t, svfloat16_t, float16_t);
	10154	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f64_x)))
	10155	svfloat64_t svamax_n_f64_x(svbool_t, svfloat64_t, float64_t);
	10156	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f32_x)))
	10157	svfloat32_t svamax_n_f32_x(svbool_t, svfloat32_t, float32_t);
	10158	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f16_x)))
	10159	svfloat16_t svamax_n_f16_x(svbool_t, svfloat16_t, float16_t);
	10160	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f64_z)))
	10161	svfloat64_t svamax_n_f64_z(svbool_t, svfloat64_t, float64_t);
	10162	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f32_z)))
	10163	svfloat32_t svamax_n_f32_z(svbool_t, svfloat32_t, float32_t);
	10164	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f16_z)))
	10165	svfloat16_t svamax_n_f16_z(svbool_t, svfloat16_t, float16_t);
	10166	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f64_m)))
	10167	svfloat64_t svamax_f64_m(svbool_t, svfloat64_t, svfloat64_t);
	10168	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f32_m)))
	10169	svfloat32_t svamax_f32_m(svbool_t, svfloat32_t, svfloat32_t);
	10170	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f16_m)))
	10171	svfloat16_t svamax_f16_m(svbool_t, svfloat16_t, svfloat16_t);
	10172	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f64_x)))
	10173	svfloat64_t svamax_f64_x(svbool_t, svfloat64_t, svfloat64_t);
	10174	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f32_x)))
	10175	svfloat32_t svamax_f32_x(svbool_t, svfloat32_t, svfloat32_t);
	10176	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f16_x)))
	10177	svfloat16_t svamax_f16_x(svbool_t, svfloat16_t, svfloat16_t);
	10178	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f64_z)))
	10179	svfloat64_t svamax_f64_z(svbool_t, svfloat64_t, svfloat64_t);
	10180	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f32_z)))
	10181	svfloat32_t svamax_f32_z(svbool_t, svfloat32_t, svfloat32_t);
	10182	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f16_z)))
	10183	svfloat16_t svamax_f16_z(svbool_t, svfloat16_t, svfloat16_t);
	10184	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f64_m)))
	10185	svfloat64_t svamin_n_f64_m(svbool_t, svfloat64_t, float64_t);
	10186	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f32_m)))
	10187	svfloat32_t svamin_n_f32_m(svbool_t, svfloat32_t, float32_t);
	10188	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f16_m)))
	10189	svfloat16_t svamin_n_f16_m(svbool_t, svfloat16_t, float16_t);
	10190	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f64_x)))
	10191	svfloat64_t svamin_n_f64_x(svbool_t, svfloat64_t, float64_t);
	10192	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f32_x)))
	10193	svfloat32_t svamin_n_f32_x(svbool_t, svfloat32_t, float32_t);
	10194	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f16_x)))
	10195	svfloat16_t svamin_n_f16_x(svbool_t, svfloat16_t, float16_t);
	10196	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f64_z)))
	10197	svfloat64_t svamin_n_f64_z(svbool_t, svfloat64_t, float64_t);
	10198	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f32_z)))
	10199	svfloat32_t svamin_n_f32_z(svbool_t, svfloat32_t, float32_t);
	10200	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f16_z)))
	10201	svfloat16_t svamin_n_f16_z(svbool_t, svfloat16_t, float16_t);
	10202	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f64_m)))
	10203	svfloat64_t svamin_f64_m(svbool_t, svfloat64_t, svfloat64_t);
	10204	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f32_m)))
	10205	svfloat32_t svamin_f32_m(svbool_t, svfloat32_t, svfloat32_t);
	10206	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f16_m)))
	10207	svfloat16_t svamin_f16_m(svbool_t, svfloat16_t, svfloat16_t);
	10208	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f64_x)))
	10209	svfloat64_t svamin_f64_x(svbool_t, svfloat64_t, svfloat64_t);
	10210	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f32_x)))
	10211	svfloat32_t svamin_f32_x(svbool_t, svfloat32_t, svfloat32_t);
	10212	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f16_x)))
	10213	svfloat16_t svamin_f16_x(svbool_t, svfloat16_t, svfloat16_t);
	10214	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f64_z)))
	10215	svfloat64_t svamin_f64_z(svbool_t, svfloat64_t, svfloat64_t);
	10216	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f32_z)))
	10217	svfloat32_t svamin_f32_z(svbool_t, svfloat32_t, svfloat32_t);
	10218	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f16_z)))
	10219	svfloat16_t svamin_f16_z(svbool_t, svfloat16_t, svfloat16_t);
	10220	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f64_m)))
	10221	svfloat64_t svamax_m(svbool_t, svfloat64_t, float64_t);
	10222	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f32_m)))
	10223	svfloat32_t svamax_m(svbool_t, svfloat32_t, float32_t);
	10224	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f16_m)))
	10225	svfloat16_t svamax_m(svbool_t, svfloat16_t, float16_t);
	10226	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f64_x)))
	10227	svfloat64_t svamax_x(svbool_t, svfloat64_t, float64_t);
	10228	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f32_x)))
	10229	svfloat32_t svamax_x(svbool_t, svfloat32_t, float32_t);
	10230	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f16_x)))
	10231	svfloat16_t svamax_x(svbool_t, svfloat16_t, float16_t);
	10232	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f64_z)))
	10233	svfloat64_t svamax_z(svbool_t, svfloat64_t, float64_t);
	10234	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f32_z)))
	10235	svfloat32_t svamax_z(svbool_t, svfloat32_t, float32_t);
	10236	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_n_f16_z)))
	10237	svfloat16_t svamax_z(svbool_t, svfloat16_t, float16_t);
	10238	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f64_m)))
	10239	svfloat64_t svamax_m(svbool_t, svfloat64_t, svfloat64_t);
	10240	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f32_m)))
	10241	svfloat32_t svamax_m(svbool_t, svfloat32_t, svfloat32_t);
	10242	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f16_m)))
	10243	svfloat16_t svamax_m(svbool_t, svfloat16_t, svfloat16_t);
	10244	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f64_x)))
	10245	svfloat64_t svamax_x(svbool_t, svfloat64_t, svfloat64_t);
	10246	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f32_x)))
	10247	svfloat32_t svamax_x(svbool_t, svfloat32_t, svfloat32_t);
	10248	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f16_x)))
	10249	svfloat16_t svamax_x(svbool_t, svfloat16_t, svfloat16_t);
	10250	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f64_z)))
	10251	svfloat64_t svamax_z(svbool_t, svfloat64_t, svfloat64_t);
	10252	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f32_z)))
	10253	svfloat32_t svamax_z(svbool_t, svfloat32_t, svfloat32_t);
	10254	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamax_f16_z)))
	10255	svfloat16_t svamax_z(svbool_t, svfloat16_t, svfloat16_t);
	10256	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f64_m)))
	10257	svfloat64_t svamin_m(svbool_t, svfloat64_t, float64_t);
	10258	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f32_m)))
	10259	svfloat32_t svamin_m(svbool_t, svfloat32_t, float32_t);
	10260	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f16_m)))
	10261	svfloat16_t svamin_m(svbool_t, svfloat16_t, float16_t);
	10262	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f64_x)))
	10263	svfloat64_t svamin_x(svbool_t, svfloat64_t, float64_t);
	10264	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f32_x)))
	10265	svfloat32_t svamin_x(svbool_t, svfloat32_t, float32_t);
	10266	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f16_x)))
	10267	svfloat16_t svamin_x(svbool_t, svfloat16_t, float16_t);
	10268	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f64_z)))
	10269	svfloat64_t svamin_z(svbool_t, svfloat64_t, float64_t);
	10270	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f32_z)))
	10271	svfloat32_t svamin_z(svbool_t, svfloat32_t, float32_t);
	10272	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_n_f16_z)))
	10273	svfloat16_t svamin_z(svbool_t, svfloat16_t, float16_t);
	10274	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f64_m)))
	10275	svfloat64_t svamin_m(svbool_t, svfloat64_t, svfloat64_t);
	10276	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f32_m)))
	10277	svfloat32_t svamin_m(svbool_t, svfloat32_t, svfloat32_t);
	10278	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f16_m)))
	10279	svfloat16_t svamin_m(svbool_t, svfloat16_t, svfloat16_t);
	10280	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f64_x)))
	10281	svfloat64_t svamin_x(svbool_t, svfloat64_t, svfloat64_t);
	10282	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f32_x)))
	10283	svfloat32_t svamin_x(svbool_t, svfloat32_t, svfloat32_t);
	10284	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f16_x)))
	10285	svfloat16_t svamin_x(svbool_t, svfloat16_t, svfloat16_t);
	10286	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f64_z)))
	10287	svfloat64_t svamin_z(svbool_t, svfloat64_t, svfloat64_t);
	10288	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f32_z)))
	10289	svfloat32_t svamin_z(svbool_t, svfloat32_t, svfloat32_t);
	10290	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svamin_f16_z)))
	10291	svfloat16_t svamin_z(svbool_t, svfloat16_t, svfloat16_t);
	10292	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_f16_mf8_fpm)))
	10293	svfloat16_t svdot_f16_mf8_fpm(svfloat16_t, svmfloat8_t, svmfloat8_t, fpm_t);
	10294	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_n_f16_mf8_fpm)))
	10295	svfloat16_t svdot_n_f16_mf8_fpm(svfloat16_t, svmfloat8_t, mfloat8_t, fpm_t);
	10296	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_lane_f16_mf8_fpm)))
	10297	svfloat16_t svdot_lane_f16_mf8_fpm(svfloat16_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
	10298	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_f16_mf8_fpm)))
	10299	svfloat16_t svdot_fpm(svfloat16_t, svmfloat8_t, svmfloat8_t, fpm_t);
	10300	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_n_f16_mf8_fpm)))
	10301	svfloat16_t svdot_fpm(svfloat16_t, svmfloat8_t, mfloat8_t, fpm_t);
	10302	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_lane_f16_mf8_fpm)))
	10303	svfloat16_t svdot_lane_fpm(svfloat16_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
	10304	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_f32_mf8_fpm)))
	10305	svfloat32_t svdot_f32_mf8_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, fpm_t);
	10306	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_n_f32_mf8_fpm)))
	10307	svfloat32_t svdot_n_f32_mf8_fpm(svfloat32_t, svmfloat8_t, mfloat8_t, fpm_t);
	10308	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_lane_f32_mf8_fpm)))
	10309	svfloat32_t svdot_lane_f32_mf8_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
	10310	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_f32_mf8_fpm)))
	10311	svfloat32_t svdot_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, fpm_t);
	10312	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_n_f32_mf8_fpm)))
	10313	svfloat32_t svdot_fpm(svfloat32_t, svmfloat8_t, mfloat8_t, fpm_t);
	10314	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_lane_f32_mf8_fpm)))
	10315	svfloat32_t svdot_lane_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
	10316	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalb_f16_mf8_fpm)))
	10317	svfloat16_t svmlalb_f16_mf8_fpm(svfloat16_t, svmfloat8_t, svmfloat8_t, fpm_t);
	10318	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalb_n_f16_mf8_fpm)))
	10319	svfloat16_t svmlalb_n_f16_mf8_fpm(svfloat16_t, svmfloat8_t, mfloat8_t, fpm_t);
	10320	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalb_lane_f16_mf8_fpm)))
	10321	svfloat16_t svmlalb_lane_f16_mf8_fpm(svfloat16_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
	10322	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlallbb_f32_mf8_fpm)))
	10323	svfloat32_t svmlallbb_f32_mf8_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, fpm_t);
	10324	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlallbb_n_f32_mf8_fpm)))
	10325	svfloat32_t svmlallbb_n_f32_mf8_fpm(svfloat32_t, svmfloat8_t, mfloat8_t, fpm_t);
	10326	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlallbb_lane_f32_mf8_fpm)))
	10327	svfloat32_t svmlallbb_lane_f32_mf8_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
	10328	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlallbt_f32_mf8_fpm)))
	10329	svfloat32_t svmlallbt_f32_mf8_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, fpm_t);
	10330	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlallbt_n_f32_mf8_fpm)))
	10331	svfloat32_t svmlallbt_n_f32_mf8_fpm(svfloat32_t, svmfloat8_t, mfloat8_t, fpm_t);
	10332	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlallbt_lane_f32_mf8_fpm)))
	10333	svfloat32_t svmlallbt_lane_f32_mf8_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
	10334	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalltb_f32_mf8_fpm)))
	10335	svfloat32_t svmlalltb_f32_mf8_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, fpm_t);
	10336	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalltb_n_f32_mf8_fpm)))
	10337	svfloat32_t svmlalltb_n_f32_mf8_fpm(svfloat32_t, svmfloat8_t, mfloat8_t, fpm_t);
	10338	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalltb_lane_f32_mf8_fpm)))
	10339	svfloat32_t svmlalltb_lane_f32_mf8_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
	10340	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalltt_f32_mf8_fpm)))
	10341	svfloat32_t svmlalltt_f32_mf8_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, fpm_t);
	10342	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalltt_n_f32_mf8_fpm)))
	10343	svfloat32_t svmlalltt_n_f32_mf8_fpm(svfloat32_t, svmfloat8_t, mfloat8_t, fpm_t);
	10344	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalltt_lane_f32_mf8_fpm)))
	10345	svfloat32_t svmlalltt_lane_f32_mf8_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
	10346	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalt_f16_mf8_fpm)))
	10347	svfloat16_t svmlalt_f16_mf8_fpm(svfloat16_t, svmfloat8_t, svmfloat8_t, fpm_t);
	10348	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalt_n_f16_mf8_fpm)))
	10349	svfloat16_t svmlalt_n_f16_mf8_fpm(svfloat16_t, svmfloat8_t, mfloat8_t, fpm_t);
	10350	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalt_lane_f16_mf8_fpm)))
	10351	svfloat16_t svmlalt_lane_f16_mf8_fpm(svfloat16_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
	10352	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalb_f16_mf8_fpm)))
	10353	svfloat16_t svmlalb_fpm(svfloat16_t, svmfloat8_t, svmfloat8_t, fpm_t);
	10354	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalb_n_f16_mf8_fpm)))
	10355	svfloat16_t svmlalb_fpm(svfloat16_t, svmfloat8_t, mfloat8_t, fpm_t);
	10356	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalb_lane_f16_mf8_fpm)))
	10357	svfloat16_t svmlalb_lane_fpm(svfloat16_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
	10358	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlallbb_f32_mf8_fpm)))
	10359	svfloat32_t svmlallbb_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, fpm_t);
	10360	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlallbb_n_f32_mf8_fpm)))
	10361	svfloat32_t svmlallbb_fpm(svfloat32_t, svmfloat8_t, mfloat8_t, fpm_t);
	10362	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlallbb_lane_f32_mf8_fpm)))
	10363	svfloat32_t svmlallbb_lane_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
	10364	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlallbt_f32_mf8_fpm)))
	10365	svfloat32_t svmlallbt_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, fpm_t);
	10366	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlallbt_n_f32_mf8_fpm)))
	10367	svfloat32_t svmlallbt_fpm(svfloat32_t, svmfloat8_t, mfloat8_t, fpm_t);
	10368	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlallbt_lane_f32_mf8_fpm)))
	10369	svfloat32_t svmlallbt_lane_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
	10370	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalltb_f32_mf8_fpm)))
	10371	svfloat32_t svmlalltb_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, fpm_t);
	10372	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalltb_n_f32_mf8_fpm)))
	10373	svfloat32_t svmlalltb_fpm(svfloat32_t, svmfloat8_t, mfloat8_t, fpm_t);
	10374	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalltb_lane_f32_mf8_fpm)))
	10375	svfloat32_t svmlalltb_lane_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
	10376	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalltt_f32_mf8_fpm)))
	10377	svfloat32_t svmlalltt_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, fpm_t);
	10378	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalltt_n_f32_mf8_fpm)))
	10379	svfloat32_t svmlalltt_fpm(svfloat32_t, svmfloat8_t, mfloat8_t, fpm_t);
	10380	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalltt_lane_f32_mf8_fpm)))
	10381	svfloat32_t svmlalltt_lane_fpm(svfloat32_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
	10382	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalt_f16_mf8_fpm)))
	10383	svfloat16_t svmlalt_fpm(svfloat16_t, svmfloat8_t, svmfloat8_t, fpm_t);
	10384	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalt_n_f16_mf8_fpm)))
	10385	svfloat16_t svmlalt_fpm(svfloat16_t, svmfloat8_t, mfloat8_t, fpm_t);
	10386	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmlalt_lane_f16_mf8_fpm)))
	10387	svfloat16_t svmlalt_lane_fpm(svfloat16_t, svmfloat8_t, svmfloat8_t, uint64_t, fpm_t);
	10388	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvt1_bf16_mf8_fpm)))
	10389	svbfloat16_t svcvt1_bf16_mf8_fpm(svmfloat8_t, fpm_t);
	10390	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvt1_f16_mf8_fpm)))
	10391	svfloat16_t svcvt1_f16_mf8_fpm(svmfloat8_t, fpm_t);
	10392	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvt2_bf16_mf8_fpm)))
	10393	svbfloat16_t svcvt2_bf16_mf8_fpm(svmfloat8_t, fpm_t);
	10394	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvt2_f16_mf8_fpm)))
	10395	svfloat16_t svcvt2_f16_mf8_fpm(svmfloat8_t, fpm_t);
	10396	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtlt1_bf16_mf8_fpm)))
	10397	svbfloat16_t svcvtlt1_bf16_mf8_fpm(svmfloat8_t, fpm_t);
	10398	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtlt1_f16_mf8_fpm)))
	10399	svfloat16_t svcvtlt1_f16_mf8_fpm(svmfloat8_t, fpm_t);
	10400	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtlt2_bf16_mf8_fpm)))
	10401	svbfloat16_t svcvtlt2_bf16_mf8_fpm(svmfloat8_t, fpm_t);
	10402	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtlt2_f16_mf8_fpm)))
	10403	svfloat16_t svcvtlt2_f16_mf8_fpm(svmfloat8_t, fpm_t);
	10404	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtn_mf8_bf16_x2_fpm)))
	10405	svmfloat8_t svcvtn_mf8_bf16_x2_fpm(svbfloat16x2_t, fpm_t);
	10406	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtn_mf8_f16_x2_fpm)))
	10407	svmfloat8_t svcvtn_mf8_f16_x2_fpm(svfloat16x2_t, fpm_t);
	10408	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtnb_mf8_f32_x2_fpm)))
	10409	svmfloat8_t svcvtnb_mf8_f32_x2_fpm(svfloat32x2_t, fpm_t);
	10410	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtnt_mf8_f32_x2_fpm)))
	10411	svmfloat8_t svcvtnt_mf8_f32_x2_fpm(svmfloat8_t, svfloat32x2_t, fpm_t);
	10412	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvt1_bf16_mf8_fpm)))
	10413	svbfloat16_t svcvt1_bf16_fpm(svmfloat8_t, fpm_t);
	10414	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvt1_f16_mf8_fpm)))
	10415	svfloat16_t svcvt1_f16_fpm(svmfloat8_t, fpm_t);
	10416	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvt2_bf16_mf8_fpm)))
	10417	svbfloat16_t svcvt2_bf16_fpm(svmfloat8_t, fpm_t);
	10418	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvt2_f16_mf8_fpm)))
	10419	svfloat16_t svcvt2_f16_fpm(svmfloat8_t, fpm_t);
	10420	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtlt1_bf16_mf8_fpm)))
	10421	svbfloat16_t svcvtlt1_bf16_fpm(svmfloat8_t, fpm_t);
	10422	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtlt1_f16_mf8_fpm)))
	10423	svfloat16_t svcvtlt1_f16_fpm(svmfloat8_t, fpm_t);
	10424	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtlt2_bf16_mf8_fpm)))
	10425	svbfloat16_t svcvtlt2_bf16_fpm(svmfloat8_t, fpm_t);
	10426	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtlt2_f16_mf8_fpm)))
	10427	svfloat16_t svcvtlt2_f16_fpm(svmfloat8_t, fpm_t);
	10428	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtn_mf8_bf16_x2_fpm)))
	10429	svmfloat8_t svcvtn_mf8_fpm(svbfloat16x2_t, fpm_t);
	10430	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtn_mf8_f16_x2_fpm)))
	10431	svmfloat8_t svcvtn_mf8_fpm(svfloat16x2_t, fpm_t);
	10432	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtnb_mf8_f32_x2_fpm)))
	10433	svmfloat8_t svcvtnb_mf8_fpm(svfloat32x2_t, fpm_t);
	10434	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvtnt_mf8_f32_x2_fpm)))
	10435	svmfloat8_t svcvtnt_mf8_fpm(svmfloat8_t, svfloat32x2_t, fpm_t);
	10436	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti2_lane_u8)))
	10437	svuint8_t svluti2_lane_u8(svuint8_t, svuint8_t, uint64_t);
	10438	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti2_lane_s8)))
	10439	svint8_t svluti2_lane_s8(svint8_t, svuint8_t, uint64_t);
	10440	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti2_lane_u16)))
	10441	svuint16_t svluti2_lane_u16(svuint16_t, svuint8_t, uint64_t);
	10442	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti2_lane_bf16)))
	10443	svbfloat16_t svluti2_lane_bf16(svbfloat16_t, svuint8_t, uint64_t);
	10444	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti2_lane_f16)))
	10445	svfloat16_t svluti2_lane_f16(svfloat16_t, svuint8_t, uint64_t);
	10446	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti2_lane_s16)))
	10447	svint16_t svluti2_lane_s16(svint16_t, svuint8_t, uint64_t);
	10448	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_u8)))
	10449	svuint8_t svluti4_lane_u8(svuint8_t, svuint8_t, uint64_t);
	10450	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_s8)))
	10451	svint8_t svluti4_lane_s8(svint8_t, svuint8_t, uint64_t);
	10452	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_u16)))
	10453	svuint16_t svluti4_lane_u16(svuint16_t, svuint8_t, uint64_t);
	10454	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_bf16)))
	10455	svbfloat16_t svluti4_lane_bf16(svbfloat16_t, svuint8_t, uint64_t);
	10456	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_f16)))
	10457	svfloat16_t svluti4_lane_f16(svfloat16_t, svuint8_t, uint64_t);
	10458	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_s16)))
	10459	svint16_t svluti4_lane_s16(svint16_t, svuint8_t, uint64_t);
	10460	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_u16_x2)))
	10461	svuint16_t svluti4_lane_u16_x2(svuint16x2_t, svuint8_t, uint64_t);
	10462	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_bf16_x2)))
	10463	svbfloat16_t svluti4_lane_bf16_x2(svbfloat16x2_t, svuint8_t, uint64_t);
	10464	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_f16_x2)))
	10465	svfloat16_t svluti4_lane_f16_x2(svfloat16x2_t, svuint8_t, uint64_t);
	10466	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_s16_x2)))
	10467	svint16_t svluti4_lane_s16_x2(svint16x2_t, svuint8_t, uint64_t);
	10468	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti2_lane_u8)))
	10469	svuint8_t svluti2_lane(svuint8_t, svuint8_t, uint64_t);
	10470	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti2_lane_s8)))
	10471	svint8_t svluti2_lane(svint8_t, svuint8_t, uint64_t);
	10472	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti2_lane_u16)))
	10473	svuint16_t svluti2_lane(svuint16_t, svuint8_t, uint64_t);
	10474	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti2_lane_bf16)))
	10475	svbfloat16_t svluti2_lane(svbfloat16_t, svuint8_t, uint64_t);
	10476	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti2_lane_f16)))
	10477	svfloat16_t svluti2_lane(svfloat16_t, svuint8_t, uint64_t);
	10478	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti2_lane_s16)))
	10479	svint16_t svluti2_lane(svint16_t, svuint8_t, uint64_t);
	10480	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_u8)))
	10481	svuint8_t svluti4_lane(svuint8_t, svuint8_t, uint64_t);
	10482	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_s8)))
	10483	svint8_t svluti4_lane(svint8_t, svuint8_t, uint64_t);
	10484	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_u16)))
	10485	svuint16_t svluti4_lane(svuint16_t, svuint8_t, uint64_t);
	10486	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_bf16)))
	10487	svbfloat16_t svluti4_lane(svbfloat16_t, svuint8_t, uint64_t);
	10488	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_f16)))
	10489	svfloat16_t svluti4_lane(svfloat16_t, svuint8_t, uint64_t);
	10490	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_s16)))
	10491	svint16_t svluti4_lane(svint16_t, svuint8_t, uint64_t);
	10492	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_u16_x2)))
	10493	svuint16_t svluti4_lane(svuint16x2_t, svuint8_t, uint64_t);
	10494	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_bf16_x2)))
	10495	svbfloat16_t svluti4_lane(svbfloat16x2_t, svuint8_t, uint64_t);
	10496	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_f16_x2)))
	10497	svfloat16_t svluti4_lane(svfloat16x2_t, svuint8_t, uint64_t);
	10498	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svluti4_lane_s16_x2)))
	10499	svint16_t svluti4_lane(svint16x2_t, svuint8_t, uint64_t);
	10500	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaesd_u8)))
	10501	svuint8_t svaesd_u8(svuint8_t, svuint8_t);
	10502	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaese_u8)))
	10503	svuint8_t svaese_u8(svuint8_t, svuint8_t);
	10504	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaesimc_u8)))
	10505	svuint8_t svaesimc_u8(svuint8_t);
	10506	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaesmc_u8)))
	10507	svuint8_t svaesmc_u8(svuint8_t);
	10508	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmullb_pair_n_u64)))
	10509	svuint64_t svpmullb_pair_n_u64(svuint64_t, uint64_t);
	10510	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmullb_pair_u64)))
	10511	svuint64_t svpmullb_pair_u64(svuint64_t, svuint64_t);
	10512	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmullt_pair_n_u64)))
	10513	svuint64_t svpmullt_pair_n_u64(svuint64_t, uint64_t);
	10514	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmullt_pair_u64)))
	10515	svuint64_t svpmullt_pair_u64(svuint64_t, svuint64_t);
	10516	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaesd_u8)))
	10517	svuint8_t svaesd(svuint8_t, svuint8_t);
	10518	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaese_u8)))
	10519	svuint8_t svaese(svuint8_t, svuint8_t);
	10520	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaesimc_u8)))
	10521	svuint8_t svaesimc(svuint8_t);
	10522	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaesmc_u8)))
	10523	svuint8_t svaesmc(svuint8_t);
	10524	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmullb_pair_n_u64)))
	10525	svuint64_t svpmullb_pair(svuint64_t, uint64_t);
	10526	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmullb_pair_u64)))
	10527	svuint64_t svpmullb_pair(svuint64_t, svuint64_t);
	10528	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmullt_pair_n_u64)))
	10529	svuint64_t svpmullt_pair(svuint64_t, uint64_t);
	10530	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpmullt_pair_u64)))
	10531	svuint64_t svpmullt_pair(svuint64_t, svuint64_t);
	10532	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svadd_n_bf16_m)))
	10533	svbfloat16_t svadd_n_bf16_m(svbool_t, svbfloat16_t, bfloat16_t);
	10534	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svadd_n_bf16_x)))
	10535	svbfloat16_t svadd_n_bf16_x(svbool_t, svbfloat16_t, bfloat16_t);
	10536	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svadd_n_bf16_z)))
	10537	svbfloat16_t svadd_n_bf16_z(svbool_t, svbfloat16_t, bfloat16_t);
	10538	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svadd_bf16_m)))
	10539	svbfloat16_t svadd_bf16_m(svbool_t, svbfloat16_t, svbfloat16_t);
	10540	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svadd_bf16_x)))
	10541	svbfloat16_t svadd_bf16_x(svbool_t, svbfloat16_t, svbfloat16_t);
	10542	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svadd_bf16_z)))
	10543	svbfloat16_t svadd_bf16_z(svbool_t, svbfloat16_t, svbfloat16_t);
	10544	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_bf16)))
	10545	svbfloat16_t svclamp_bf16(svbfloat16_t, svbfloat16_t, svbfloat16_t);
	10546	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmax_n_bf16_m)))
	10547	svbfloat16_t svmax_n_bf16_m(svbool_t, svbfloat16_t, bfloat16_t);
	10548	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmax_n_bf16_x)))
	10549	svbfloat16_t svmax_n_bf16_x(svbool_t, svbfloat16_t, bfloat16_t);
	10550	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmax_n_bf16_z)))
	10551	svbfloat16_t svmax_n_bf16_z(svbool_t, svbfloat16_t, bfloat16_t);
	10552	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmax_bf16_m)))
	10553	svbfloat16_t svmax_bf16_m(svbool_t, svbfloat16_t, svbfloat16_t);
	10554	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmax_bf16_x)))
	10555	svbfloat16_t svmax_bf16_x(svbool_t, svbfloat16_t, svbfloat16_t);
	10556	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmax_bf16_z)))
	10557	svbfloat16_t svmax_bf16_z(svbool_t, svbfloat16_t, svbfloat16_t);
	10558	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnm_n_bf16_m)))
	10559	svbfloat16_t svmaxnm_n_bf16_m(svbool_t, svbfloat16_t, bfloat16_t);
	10560	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnm_n_bf16_x)))
	10561	svbfloat16_t svmaxnm_n_bf16_x(svbool_t, svbfloat16_t, bfloat16_t);
	10562	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnm_n_bf16_z)))
	10563	svbfloat16_t svmaxnm_n_bf16_z(svbool_t, svbfloat16_t, bfloat16_t);
	10564	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnm_bf16_m)))
	10565	svbfloat16_t svmaxnm_bf16_m(svbool_t, svbfloat16_t, svbfloat16_t);
	10566	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnm_bf16_x)))
	10567	svbfloat16_t svmaxnm_bf16_x(svbool_t, svbfloat16_t, svbfloat16_t);
	10568	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnm_bf16_z)))
	10569	svbfloat16_t svmaxnm_bf16_z(svbool_t, svbfloat16_t, svbfloat16_t);
	10570	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmin_n_bf16_m)))
	10571	svbfloat16_t svmin_n_bf16_m(svbool_t, svbfloat16_t, bfloat16_t);
	10572	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmin_n_bf16_x)))
	10573	svbfloat16_t svmin_n_bf16_x(svbool_t, svbfloat16_t, bfloat16_t);
	10574	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmin_n_bf16_z)))
	10575	svbfloat16_t svmin_n_bf16_z(svbool_t, svbfloat16_t, bfloat16_t);
	10576	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmin_bf16_m)))
	10577	svbfloat16_t svmin_bf16_m(svbool_t, svbfloat16_t, svbfloat16_t);
	10578	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmin_bf16_x)))
	10579	svbfloat16_t svmin_bf16_x(svbool_t, svbfloat16_t, svbfloat16_t);
	10580	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmin_bf16_z)))
	10581	svbfloat16_t svmin_bf16_z(svbool_t, svbfloat16_t, svbfloat16_t);
	10582	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnm_n_bf16_m)))
	10583	svbfloat16_t svminnm_n_bf16_m(svbool_t, svbfloat16_t, bfloat16_t);
	10584	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnm_n_bf16_x)))
	10585	svbfloat16_t svminnm_n_bf16_x(svbool_t, svbfloat16_t, bfloat16_t);
	10586	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnm_n_bf16_z)))
	10587	svbfloat16_t svminnm_n_bf16_z(svbool_t, svbfloat16_t, bfloat16_t);
	10588	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnm_bf16_m)))
	10589	svbfloat16_t svminnm_bf16_m(svbool_t, svbfloat16_t, svbfloat16_t);
	10590	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnm_bf16_x)))
	10591	svbfloat16_t svminnm_bf16_x(svbool_t, svbfloat16_t, svbfloat16_t);
	10592	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnm_bf16_z)))
	10593	svbfloat16_t svminnm_bf16_z(svbool_t, svbfloat16_t, svbfloat16_t);
	10594	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmla_n_bf16_m)))
	10595	svbfloat16_t svmla_n_bf16_m(svbool_t, svbfloat16_t, svbfloat16_t, bfloat16_t);
	10596	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmla_n_bf16_x)))
	10597	svbfloat16_t svmla_n_bf16_x(svbool_t, svbfloat16_t, svbfloat16_t, bfloat16_t);
	10598	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmla_n_bf16_z)))
	10599	svbfloat16_t svmla_n_bf16_z(svbool_t, svbfloat16_t, svbfloat16_t, bfloat16_t);
	10600	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmla_bf16_m)))
	10601	svbfloat16_t svmla_bf16_m(svbool_t, svbfloat16_t, svbfloat16_t, svbfloat16_t);
	10602	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmla_bf16_x)))
	10603	svbfloat16_t svmla_bf16_x(svbool_t, svbfloat16_t, svbfloat16_t, svbfloat16_t);
	10604	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmla_bf16_z)))
	10605	svbfloat16_t svmla_bf16_z(svbool_t, svbfloat16_t, svbfloat16_t, svbfloat16_t);
	10606	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmla_lane_bf16)))
	10607	svbfloat16_t svmla_lane_bf16(svbfloat16_t, svbfloat16_t, svbfloat16_t, uint64_t);
	10608	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmls_n_bf16_m)))
	10609	svbfloat16_t svmls_n_bf16_m(svbool_t, svbfloat16_t, svbfloat16_t, bfloat16_t);
	10610	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmls_n_bf16_x)))
	10611	svbfloat16_t svmls_n_bf16_x(svbool_t, svbfloat16_t, svbfloat16_t, bfloat16_t);
	10612	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmls_n_bf16_z)))
	10613	svbfloat16_t svmls_n_bf16_z(svbool_t, svbfloat16_t, svbfloat16_t, bfloat16_t);
	10614	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmls_bf16_m)))
	10615	svbfloat16_t svmls_bf16_m(svbool_t, svbfloat16_t, svbfloat16_t, svbfloat16_t);
	10616	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmls_bf16_x)))
	10617	svbfloat16_t svmls_bf16_x(svbool_t, svbfloat16_t, svbfloat16_t, svbfloat16_t);
	10618	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmls_bf16_z)))
	10619	svbfloat16_t svmls_bf16_z(svbool_t, svbfloat16_t, svbfloat16_t, svbfloat16_t);
	10620	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmls_lane_bf16)))
	10621	svbfloat16_t svmls_lane_bf16(svbfloat16_t, svbfloat16_t, svbfloat16_t, uint64_t);
	10622	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmul_n_bf16_m)))
	10623	svbfloat16_t svmul_n_bf16_m(svbool_t, svbfloat16_t, bfloat16_t);
	10624	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmul_n_bf16_x)))
	10625	svbfloat16_t svmul_n_bf16_x(svbool_t, svbfloat16_t, bfloat16_t);
	10626	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmul_n_bf16_z)))
	10627	svbfloat16_t svmul_n_bf16_z(svbool_t, svbfloat16_t, bfloat16_t);
	10628	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmul_bf16_m)))
	10629	svbfloat16_t svmul_bf16_m(svbool_t, svbfloat16_t, svbfloat16_t);
	10630	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmul_bf16_x)))
	10631	svbfloat16_t svmul_bf16_x(svbool_t, svbfloat16_t, svbfloat16_t);
	10632	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmul_bf16_z)))
	10633	svbfloat16_t svmul_bf16_z(svbool_t, svbfloat16_t, svbfloat16_t);
	10634	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmul_lane_bf16)))
	10635	svbfloat16_t svmul_lane_bf16(svbfloat16_t, svbfloat16_t, uint64_t);
	10636	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsub_n_bf16_m)))
	10637	svbfloat16_t svsub_n_bf16_m(svbool_t, svbfloat16_t, bfloat16_t);
	10638	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsub_n_bf16_x)))
	10639	svbfloat16_t svsub_n_bf16_x(svbool_t, svbfloat16_t, bfloat16_t);
	10640	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsub_n_bf16_z)))
	10641	svbfloat16_t svsub_n_bf16_z(svbool_t, svbfloat16_t, bfloat16_t);
	10642	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsub_bf16_m)))
	10643	svbfloat16_t svsub_bf16_m(svbool_t, svbfloat16_t, svbfloat16_t);
	10644	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsub_bf16_x)))
	10645	svbfloat16_t svsub_bf16_x(svbool_t, svbfloat16_t, svbfloat16_t);
	10646	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsub_bf16_z)))
	10647	svbfloat16_t svsub_bf16_z(svbool_t, svbfloat16_t, svbfloat16_t);
	10648	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svadd_n_bf16_m)))
	10649	svbfloat16_t svadd_m(svbool_t, svbfloat16_t, bfloat16_t);
	10650	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svadd_n_bf16_x)))
	10651	svbfloat16_t svadd_x(svbool_t, svbfloat16_t, bfloat16_t);
	10652	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svadd_n_bf16_z)))
	10653	svbfloat16_t svadd_z(svbool_t, svbfloat16_t, bfloat16_t);
	10654	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svadd_bf16_m)))
	10655	svbfloat16_t svadd_m(svbool_t, svbfloat16_t, svbfloat16_t);
	10656	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svadd_bf16_x)))
	10657	svbfloat16_t svadd_x(svbool_t, svbfloat16_t, svbfloat16_t);
	10658	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svadd_bf16_z)))
	10659	svbfloat16_t svadd_z(svbool_t, svbfloat16_t, svbfloat16_t);
	10660	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_bf16)))
	10661	svbfloat16_t svclamp(svbfloat16_t, svbfloat16_t, svbfloat16_t);
	10662	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmax_n_bf16_m)))
	10663	svbfloat16_t svmax_m(svbool_t, svbfloat16_t, bfloat16_t);
	10664	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmax_n_bf16_x)))
	10665	svbfloat16_t svmax_x(svbool_t, svbfloat16_t, bfloat16_t);
	10666	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmax_n_bf16_z)))
	10667	svbfloat16_t svmax_z(svbool_t, svbfloat16_t, bfloat16_t);
	10668	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmax_bf16_m)))
	10669	svbfloat16_t svmax_m(svbool_t, svbfloat16_t, svbfloat16_t);
	10670	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmax_bf16_x)))
	10671	svbfloat16_t svmax_x(svbool_t, svbfloat16_t, svbfloat16_t);
	10672	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmax_bf16_z)))
	10673	svbfloat16_t svmax_z(svbool_t, svbfloat16_t, svbfloat16_t);
	10674	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnm_n_bf16_m)))
	10675	svbfloat16_t svmaxnm_m(svbool_t, svbfloat16_t, bfloat16_t);
	10676	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnm_n_bf16_x)))
	10677	svbfloat16_t svmaxnm_x(svbool_t, svbfloat16_t, bfloat16_t);
	10678	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnm_n_bf16_z)))
	10679	svbfloat16_t svmaxnm_z(svbool_t, svbfloat16_t, bfloat16_t);
	10680	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnm_bf16_m)))
	10681	svbfloat16_t svmaxnm_m(svbool_t, svbfloat16_t, svbfloat16_t);
	10682	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnm_bf16_x)))
	10683	svbfloat16_t svmaxnm_x(svbool_t, svbfloat16_t, svbfloat16_t);
	10684	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmaxnm_bf16_z)))
	10685	svbfloat16_t svmaxnm_z(svbool_t, svbfloat16_t, svbfloat16_t);
	10686	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmin_n_bf16_m)))
	10687	svbfloat16_t svmin_m(svbool_t, svbfloat16_t, bfloat16_t);
	10688	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmin_n_bf16_x)))
	10689	svbfloat16_t svmin_x(svbool_t, svbfloat16_t, bfloat16_t);
	10690	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmin_n_bf16_z)))
	10691	svbfloat16_t svmin_z(svbool_t, svbfloat16_t, bfloat16_t);
	10692	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmin_bf16_m)))
	10693	svbfloat16_t svmin_m(svbool_t, svbfloat16_t, svbfloat16_t);
	10694	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmin_bf16_x)))
	10695	svbfloat16_t svmin_x(svbool_t, svbfloat16_t, svbfloat16_t);
	10696	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmin_bf16_z)))
	10697	svbfloat16_t svmin_z(svbool_t, svbfloat16_t, svbfloat16_t);
	10698	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnm_n_bf16_m)))
	10699	svbfloat16_t svminnm_m(svbool_t, svbfloat16_t, bfloat16_t);
	10700	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnm_n_bf16_x)))
	10701	svbfloat16_t svminnm_x(svbool_t, svbfloat16_t, bfloat16_t);
	10702	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnm_n_bf16_z)))
	10703	svbfloat16_t svminnm_z(svbool_t, svbfloat16_t, bfloat16_t);
	10704	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnm_bf16_m)))
	10705	svbfloat16_t svminnm_m(svbool_t, svbfloat16_t, svbfloat16_t);
	10706	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnm_bf16_x)))
	10707	svbfloat16_t svminnm_x(svbool_t, svbfloat16_t, svbfloat16_t);
	10708	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svminnm_bf16_z)))
	10709	svbfloat16_t svminnm_z(svbool_t, svbfloat16_t, svbfloat16_t);
	10710	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmla_n_bf16_m)))
	10711	svbfloat16_t svmla_m(svbool_t, svbfloat16_t, svbfloat16_t, bfloat16_t);
	10712	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmla_n_bf16_x)))
	10713	svbfloat16_t svmla_x(svbool_t, svbfloat16_t, svbfloat16_t, bfloat16_t);
	10714	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmla_n_bf16_z)))
	10715	svbfloat16_t svmla_z(svbool_t, svbfloat16_t, svbfloat16_t, bfloat16_t);
	10716	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmla_bf16_m)))
	10717	svbfloat16_t svmla_m(svbool_t, svbfloat16_t, svbfloat16_t, svbfloat16_t);
	10718	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmla_bf16_x)))
	10719	svbfloat16_t svmla_x(svbool_t, svbfloat16_t, svbfloat16_t, svbfloat16_t);
	10720	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmla_bf16_z)))
	10721	svbfloat16_t svmla_z(svbool_t, svbfloat16_t, svbfloat16_t, svbfloat16_t);
	10722	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmla_lane_bf16)))
	10723	svbfloat16_t svmla_lane(svbfloat16_t, svbfloat16_t, svbfloat16_t, uint64_t);
	10724	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmls_n_bf16_m)))
	10725	svbfloat16_t svmls_m(svbool_t, svbfloat16_t, svbfloat16_t, bfloat16_t);
	10726	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmls_n_bf16_x)))
	10727	svbfloat16_t svmls_x(svbool_t, svbfloat16_t, svbfloat16_t, bfloat16_t);
	10728	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmls_n_bf16_z)))
	10729	svbfloat16_t svmls_z(svbool_t, svbfloat16_t, svbfloat16_t, bfloat16_t);
	10730	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmls_bf16_m)))
	10731	svbfloat16_t svmls_m(svbool_t, svbfloat16_t, svbfloat16_t, svbfloat16_t);
	10732	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmls_bf16_x)))
	10733	svbfloat16_t svmls_x(svbool_t, svbfloat16_t, svbfloat16_t, svbfloat16_t);
	10734	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmls_bf16_z)))
	10735	svbfloat16_t svmls_z(svbool_t, svbfloat16_t, svbfloat16_t, svbfloat16_t);
	10736	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmls_lane_bf16)))
	10737	svbfloat16_t svmls_lane(svbfloat16_t, svbfloat16_t, svbfloat16_t, uint64_t);
	10738	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmul_n_bf16_m)))
	10739	svbfloat16_t svmul_m(svbool_t, svbfloat16_t, bfloat16_t);
	10740	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmul_n_bf16_x)))
	10741	svbfloat16_t svmul_x(svbool_t, svbfloat16_t, bfloat16_t);
	10742	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmul_n_bf16_z)))
	10743	svbfloat16_t svmul_z(svbool_t, svbfloat16_t, bfloat16_t);
	10744	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmul_bf16_m)))
	10745	svbfloat16_t svmul_m(svbool_t, svbfloat16_t, svbfloat16_t);
	10746	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmul_bf16_x)))
	10747	svbfloat16_t svmul_x(svbool_t, svbfloat16_t, svbfloat16_t);
	10748	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmul_bf16_z)))
	10749	svbfloat16_t svmul_z(svbool_t, svbfloat16_t, svbfloat16_t);
	10750	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svmul_lane_bf16)))
	10751	svbfloat16_t svmul_lane(svbfloat16_t, svbfloat16_t, uint64_t);
	10752	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsub_n_bf16_m)))
	10753	svbfloat16_t svsub_m(svbool_t, svbfloat16_t, bfloat16_t);
	10754	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsub_n_bf16_x)))
	10755	svbfloat16_t svsub_x(svbool_t, svbfloat16_t, bfloat16_t);
	10756	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsub_n_bf16_z)))
	10757	svbfloat16_t svsub_z(svbool_t, svbfloat16_t, bfloat16_t);
	10758	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsub_bf16_m)))
	10759	svbfloat16_t svsub_m(svbool_t, svbfloat16_t, svbfloat16_t);
	10760	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsub_bf16_x)))
	10761	svbfloat16_t svsub_x(svbool_t, svbfloat16_t, svbfloat16_t);
	10762	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsub_bf16_z)))
	10763	svbfloat16_t svsub_z(svbool_t, svbfloat16_t, svbfloat16_t);
	10764	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_n_u8)))
	10765	svuint8_t svbdep_n_u8(svuint8_t, uint8_t);
	10766	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_n_u32)))
	10767	svuint32_t svbdep_n_u32(svuint32_t, uint32_t);
	10768	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_n_u64)))
	10769	svuint64_t svbdep_n_u64(svuint64_t, uint64_t);
	10770	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_n_u16)))
	10771	svuint16_t svbdep_n_u16(svuint16_t, uint16_t);
	10772	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_u8)))
	10773	svuint8_t svbdep_u8(svuint8_t, svuint8_t);
	10774	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_u32)))
	10775	svuint32_t svbdep_u32(svuint32_t, svuint32_t);
	10776	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_u64)))
	10777	svuint64_t svbdep_u64(svuint64_t, svuint64_t);
	10778	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_u16)))
	10779	svuint16_t svbdep_u16(svuint16_t, svuint16_t);
	10780	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_n_u8)))
	10781	svuint8_t svbext_n_u8(svuint8_t, uint8_t);
	10782	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_n_u32)))
	10783	svuint32_t svbext_n_u32(svuint32_t, uint32_t);
	10784	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_n_u64)))
	10785	svuint64_t svbext_n_u64(svuint64_t, uint64_t);
	10786	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_n_u16)))
	10787	svuint16_t svbext_n_u16(svuint16_t, uint16_t);
	10788	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_u8)))
	10789	svuint8_t svbext_u8(svuint8_t, svuint8_t);
	10790	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_u32)))
	10791	svuint32_t svbext_u32(svuint32_t, svuint32_t);
	10792	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_u64)))
	10793	svuint64_t svbext_u64(svuint64_t, svuint64_t);
	10794	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_u16)))
	10795	svuint16_t svbext_u16(svuint16_t, svuint16_t);
	10796	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_n_u8)))
	10797	svuint8_t svbgrp_n_u8(svuint8_t, uint8_t);
	10798	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_n_u32)))
	10799	svuint32_t svbgrp_n_u32(svuint32_t, uint32_t);
	10800	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_n_u64)))
	10801	svuint64_t svbgrp_n_u64(svuint64_t, uint64_t);
	10802	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_n_u16)))
	10803	svuint16_t svbgrp_n_u16(svuint16_t, uint16_t);
	10804	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_u8)))
	10805	svuint8_t svbgrp_u8(svuint8_t, svuint8_t);
	10806	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_u32)))
	10807	svuint32_t svbgrp_u32(svuint32_t, svuint32_t);
	10808	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_u64)))
	10809	svuint64_t svbgrp_u64(svuint64_t, svuint64_t);
	10810	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_u16)))
	10811	svuint16_t svbgrp_u16(svuint16_t, svuint16_t);
	10812	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_n_u8)))
	10813	svuint8_t svbdep(svuint8_t, uint8_t);
	10814	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_n_u32)))
	10815	svuint32_t svbdep(svuint32_t, uint32_t);
	10816	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_n_u64)))
	10817	svuint64_t svbdep(svuint64_t, uint64_t);
	10818	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_n_u16)))
	10819	svuint16_t svbdep(svuint16_t, uint16_t);
	10820	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_u8)))
	10821	svuint8_t svbdep(svuint8_t, svuint8_t);
	10822	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_u32)))
	10823	svuint32_t svbdep(svuint32_t, svuint32_t);
	10824	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_u64)))
	10825	svuint64_t svbdep(svuint64_t, svuint64_t);
	10826	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbdep_u16)))
	10827	svuint16_t svbdep(svuint16_t, svuint16_t);
	10828	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_n_u8)))
	10829	svuint8_t svbext(svuint8_t, uint8_t);
	10830	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_n_u32)))
	10831	svuint32_t svbext(svuint32_t, uint32_t);
	10832	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_n_u64)))
	10833	svuint64_t svbext(svuint64_t, uint64_t);
	10834	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_n_u16)))
	10835	svuint16_t svbext(svuint16_t, uint16_t);
	10836	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_u8)))
	10837	svuint8_t svbext(svuint8_t, svuint8_t);
	10838	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_u32)))
	10839	svuint32_t svbext(svuint32_t, svuint32_t);
	10840	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_u64)))
	10841	svuint64_t svbext(svuint64_t, svuint64_t);
	10842	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbext_u16)))
	10843	svuint16_t svbext(svuint16_t, svuint16_t);
	10844	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_n_u8)))
	10845	svuint8_t svbgrp(svuint8_t, uint8_t);
	10846	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_n_u32)))
	10847	svuint32_t svbgrp(svuint32_t, uint32_t);
	10848	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_n_u64)))
	10849	svuint64_t svbgrp(svuint64_t, uint64_t);
	10850	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_n_u16)))
	10851	svuint16_t svbgrp(svuint16_t, uint16_t);
	10852	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_u8)))
	10853	svuint8_t svbgrp(svuint8_t, svuint8_t);
	10854	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_u32)))
	10855	svuint32_t svbgrp(svuint32_t, svuint32_t);
	10856	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_u64)))
	10857	svuint64_t svbgrp(svuint64_t, svuint64_t);
	10858	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbgrp_u16)))
	10859	svuint16_t svbgrp(svuint16_t, svuint16_t);
	10860	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsm4e_u32)))
	10861	svuint32_t svsm4e_u32(svuint32_t, svuint32_t);
	10862	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsm4ekey_u32)))
	10863	svuint32_t svsm4ekey_u32(svuint32_t, svuint32_t);
	10864	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsm4e_u32)))
	10865	svuint32_t svsm4e(svuint32_t, svuint32_t);
	10866	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsm4ekey_u32)))
	10867	svuint32_t svsm4ekey(svuint32_t, svuint32_t);
	10868	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_u32)))
	10869	svuint32_t svld1q_gather_u64base_index_u32(svbool_t, svuint64_t, int64_t);
	10870	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_u64)))
	10871	svuint64_t svld1q_gather_u64base_index_u64(svbool_t, svuint64_t, int64_t);
	10872	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_u16)))
	10873	svuint16_t svld1q_gather_u64base_index_u16(svbool_t, svuint64_t, int64_t);
	10874	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_bf16)))
	10875	svbfloat16_t svld1q_gather_u64base_index_bf16(svbool_t, svuint64_t, int64_t);
	10876	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_f64)))
	10877	svfloat64_t svld1q_gather_u64base_index_f64(svbool_t, svuint64_t, int64_t);
	10878	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_f32)))
	10879	svfloat32_t svld1q_gather_u64base_index_f32(svbool_t, svuint64_t, int64_t);
	10880	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_f16)))
	10881	svfloat16_t svld1q_gather_u64base_index_f16(svbool_t, svuint64_t, int64_t);
	10882	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_s32)))
	10883	svint32_t svld1q_gather_u64base_index_s32(svbool_t, svuint64_t, int64_t);
	10884	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_s64)))
	10885	svint64_t svld1q_gather_u64base_index_s64(svbool_t, svuint64_t, int64_t);
	10886	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_s16)))
	10887	svint16_t svld1q_gather_u64base_index_s16(svbool_t, svuint64_t, int64_t);
	10888	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_u8)))
	10889	svuint8_t svld1q_gather_u64base_offset_u8(svbool_t, svuint64_t, int64_t);
	10890	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_u32)))
	10891	svuint32_t svld1q_gather_u64base_offset_u32(svbool_t, svuint64_t, int64_t);
	10892	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_u64)))
	10893	svuint64_t svld1q_gather_u64base_offset_u64(svbool_t, svuint64_t, int64_t);
	10894	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_u16)))
	10895	svuint16_t svld1q_gather_u64base_offset_u16(svbool_t, svuint64_t, int64_t);
	10896	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_bf16)))
	10897	svbfloat16_t svld1q_gather_u64base_offset_bf16(svbool_t, svuint64_t, int64_t);
	10898	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_s8)))
	10899	svint8_t svld1q_gather_u64base_offset_s8(svbool_t, svuint64_t, int64_t);
	10900	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_f64)))
	10901	svfloat64_t svld1q_gather_u64base_offset_f64(svbool_t, svuint64_t, int64_t);
	10902	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_f32)))
	10903	svfloat32_t svld1q_gather_u64base_offset_f32(svbool_t, svuint64_t, int64_t);
	10904	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_f16)))
	10905	svfloat16_t svld1q_gather_u64base_offset_f16(svbool_t, svuint64_t, int64_t);
	10906	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_s32)))
	10907	svint32_t svld1q_gather_u64base_offset_s32(svbool_t, svuint64_t, int64_t);
	10908	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_s64)))
	10909	svint64_t svld1q_gather_u64base_offset_s64(svbool_t, svuint64_t, int64_t);
	10910	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_mf8)))
	10911	svmfloat8_t svld1q_gather_u64base_offset_mf8(svbool_t, svuint64_t, int64_t);
	10912	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_s16)))
	10913	svint16_t svld1q_gather_u64base_offset_s16(svbool_t, svuint64_t, int64_t);
	10914	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_u8)))
	10915	svuint8_t svld1q_gather_u64base_u8(svbool_t, svuint64_t);
	10916	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_u32)))
	10917	svuint32_t svld1q_gather_u64base_u32(svbool_t, svuint64_t);
	10918	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_u64)))
	10919	svuint64_t svld1q_gather_u64base_u64(svbool_t, svuint64_t);
	10920	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_u16)))
	10921	svuint16_t svld1q_gather_u64base_u16(svbool_t, svuint64_t);
	10922	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_bf16)))
	10923	svbfloat16_t svld1q_gather_u64base_bf16(svbool_t, svuint64_t);
	10924	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_s8)))
	10925	svint8_t svld1q_gather_u64base_s8(svbool_t, svuint64_t);
	10926	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_f64)))
	10927	svfloat64_t svld1q_gather_u64base_f64(svbool_t, svuint64_t);
	10928	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_f32)))
	10929	svfloat32_t svld1q_gather_u64base_f32(svbool_t, svuint64_t);
	10930	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_f16)))
	10931	svfloat16_t svld1q_gather_u64base_f16(svbool_t, svuint64_t);
	10932	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_s32)))
	10933	svint32_t svld1q_gather_u64base_s32(svbool_t, svuint64_t);
	10934	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_s64)))
	10935	svint64_t svld1q_gather_u64base_s64(svbool_t, svuint64_t);
	10936	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_mf8)))
	10937	svmfloat8_t svld1q_gather_u64base_mf8(svbool_t, svuint64_t);
	10938	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_s16)))
	10939	svint16_t svld1q_gather_u64base_s16(svbool_t, svuint64_t);
	10940	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_u32)))
	10941	svuint32_t svld1q_gather_u64index_u32(svbool_t, uint32_t const *, svuint64_t);
	10942	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_u64)))
	10943	svuint64_t svld1q_gather_u64index_u64(svbool_t, uint64_t const *, svuint64_t);
	10944	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_u16)))
	10945	svuint16_t svld1q_gather_u64index_u16(svbool_t, uint16_t const *, svuint64_t);
	10946	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_bf16)))
	10947	svbfloat16_t svld1q_gather_u64index_bf16(svbool_t, bfloat16_t const *, svuint64_t);
	10948	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_f64)))
	10949	svfloat64_t svld1q_gather_u64index_f64(svbool_t, float64_t const *, svuint64_t);
	10950	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_f32)))
	10951	svfloat32_t svld1q_gather_u64index_f32(svbool_t, float32_t const *, svuint64_t);
	10952	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_f16)))
	10953	svfloat16_t svld1q_gather_u64index_f16(svbool_t, float16_t const *, svuint64_t);
	10954	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_s32)))
	10955	svint32_t svld1q_gather_u64index_s32(svbool_t, int32_t const *, svuint64_t);
	10956	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_s64)))
	10957	svint64_t svld1q_gather_u64index_s64(svbool_t, int64_t const *, svuint64_t);
	10958	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_s16)))
	10959	svint16_t svld1q_gather_u64index_s16(svbool_t, int16_t const *, svuint64_t);
	10960	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_u8)))
	10961	svuint8_t svld1q_gather_u64offset_u8(svbool_t, uint8_t const *, svuint64_t);
	10962	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_u32)))
	10963	svuint32_t svld1q_gather_u64offset_u32(svbool_t, uint32_t const *, svuint64_t);
	10964	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_u64)))
	10965	svuint64_t svld1q_gather_u64offset_u64(svbool_t, uint64_t const *, svuint64_t);
	10966	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_u16)))
	10967	svuint16_t svld1q_gather_u64offset_u16(svbool_t, uint16_t const *, svuint64_t);
	10968	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_bf16)))
	10969	svbfloat16_t svld1q_gather_u64offset_bf16(svbool_t, bfloat16_t const *, svuint64_t);
	10970	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_s8)))
	10971	svint8_t svld1q_gather_u64offset_s8(svbool_t, int8_t const *, svuint64_t);
	10972	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_f64)))
	10973	svfloat64_t svld1q_gather_u64offset_f64(svbool_t, float64_t const *, svuint64_t);
	10974	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_f32)))
	10975	svfloat32_t svld1q_gather_u64offset_f32(svbool_t, float32_t const *, svuint64_t);
	10976	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_f16)))
	10977	svfloat16_t svld1q_gather_u64offset_f16(svbool_t, float16_t const *, svuint64_t);
	10978	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_s32)))
	10979	svint32_t svld1q_gather_u64offset_s32(svbool_t, int32_t const *, svuint64_t);
	10980	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_s64)))
	10981	svint64_t svld1q_gather_u64offset_s64(svbool_t, int64_t const *, svuint64_t);
	10982	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_mf8)))
	10983	svmfloat8_t svld1q_gather_u64offset_mf8(svbool_t, mfloat8_t const *, svuint64_t);
	10984	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_s16)))
	10985	svint16_t svld1q_gather_u64offset_s16(svbool_t, int16_t const *, svuint64_t);
	10986	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1udq_u64)))
	10987	svuint64_t svld1udq_u64(svbool_t, uint64_t const *);
	10988	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1udq_f64)))
	10989	svfloat64_t svld1udq_f64(svbool_t, float64_t const *);
	10990	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1udq_s64)))
	10991	svint64_t svld1udq_s64(svbool_t, int64_t const *);
	10992	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1udq_vnum_u64)))
	10993	svuint64_t svld1udq_vnum_u64(svbool_t, uint64_t const *, int64_t);
	10994	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1udq_vnum_f64)))
	10995	svfloat64_t svld1udq_vnum_f64(svbool_t, float64_t const *, int64_t);
	10996	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1udq_vnum_s64)))
	10997	svint64_t svld1udq_vnum_s64(svbool_t, int64_t const *, int64_t);
	10998	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1uwq_u32)))
	10999	svuint32_t svld1uwq_u32(svbool_t, uint32_t const *);
	11000	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1uwq_f32)))
	11001	svfloat32_t svld1uwq_f32(svbool_t, float32_t const *);
	11002	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1uwq_s32)))
	11003	svint32_t svld1uwq_s32(svbool_t, int32_t const *);
	11004	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1uwq_vnum_u32)))
	11005	svuint32_t svld1uwq_vnum_u32(svbool_t, uint32_t const *, int64_t);
	11006	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1uwq_vnum_f32)))
	11007	svfloat32_t svld1uwq_vnum_f32(svbool_t, float32_t const *, int64_t);
	11008	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1uwq_vnum_s32)))
	11009	svint32_t svld1uwq_vnum_s32(svbool_t, int32_t const *, int64_t);
	11010	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1dq_u64)))
	11011	void svst1dq_u64(svbool_t, uint64_t *, svuint64_t);
	11012	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1dq_f64)))
	11013	void svst1dq_f64(svbool_t, float64_t *, svfloat64_t);
	11014	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1dq_s64)))
	11015	void svst1dq_s64(svbool_t, int64_t *, svint64_t);
	11016	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1dq_vnum_u64)))
	11017	void svst1dq_vnum_u64(svbool_t, uint64_t *, int64_t, svuint64_t);
	11018	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1dq_vnum_f64)))
	11019	void svst1dq_vnum_f64(svbool_t, float64_t *, int64_t, svfloat64_t);
	11020	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1dq_vnum_s64)))
	11021	void svst1dq_vnum_s64(svbool_t, int64_t *, int64_t, svint64_t);
	11022	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_u8)))
	11023	void svst1q_scatter_u64base_u8(svbool_t, svuint64_t, svuint8_t);
	11024	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_u32)))
	11025	void svst1q_scatter_u64base_u32(svbool_t, svuint64_t, svuint32_t);
	11026	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_u64)))
	11027	void svst1q_scatter_u64base_u64(svbool_t, svuint64_t, svuint64_t);
	11028	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_u16)))
	11029	void svst1q_scatter_u64base_u16(svbool_t, svuint64_t, svuint16_t);
	11030	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_bf16)))
	11031	void svst1q_scatter_u64base_bf16(svbool_t, svuint64_t, svbfloat16_t);
	11032	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_s8)))
	11033	void svst1q_scatter_u64base_s8(svbool_t, svuint64_t, svint8_t);
	11034	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_f64)))
	11035	void svst1q_scatter_u64base_f64(svbool_t, svuint64_t, svfloat64_t);
	11036	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_f32)))
	11037	void svst1q_scatter_u64base_f32(svbool_t, svuint64_t, svfloat32_t);
	11038	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_f16)))
	11039	void svst1q_scatter_u64base_f16(svbool_t, svuint64_t, svfloat16_t);
	11040	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_s32)))
	11041	void svst1q_scatter_u64base_s32(svbool_t, svuint64_t, svint32_t);
	11042	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_s64)))
	11043	void svst1q_scatter_u64base_s64(svbool_t, svuint64_t, svint64_t);
	11044	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_mf8)))
	11045	void svst1q_scatter_u64base_mf8(svbool_t, svuint64_t, svmfloat8_t);
	11046	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_s16)))
	11047	void svst1q_scatter_u64base_s16(svbool_t, svuint64_t, svint16_t);
	11048	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_u32)))
	11049	void svst1q_scatter_u64base_index_u32(svbool_t, svuint64_t, int64_t, svuint32_t);
	11050	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_u64)))
	11051	void svst1q_scatter_u64base_index_u64(svbool_t, svuint64_t, int64_t, svuint64_t);
	11052	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_u16)))
	11053	void svst1q_scatter_u64base_index_u16(svbool_t, svuint64_t, int64_t, svuint16_t);
	11054	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_bf16)))
	11055	void svst1q_scatter_u64base_index_bf16(svbool_t, svuint64_t, int64_t, svbfloat16_t);
	11056	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_f64)))
	11057	void svst1q_scatter_u64base_index_f64(svbool_t, svuint64_t, int64_t, svfloat64_t);
	11058	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_f32)))
	11059	void svst1q_scatter_u64base_index_f32(svbool_t, svuint64_t, int64_t, svfloat32_t);
	11060	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_f16)))
	11061	void svst1q_scatter_u64base_index_f16(svbool_t, svuint64_t, int64_t, svfloat16_t);
	11062	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_s32)))
	11063	void svst1q_scatter_u64base_index_s32(svbool_t, svuint64_t, int64_t, svint32_t);
	11064	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_s64)))
	11065	void svst1q_scatter_u64base_index_s64(svbool_t, svuint64_t, int64_t, svint64_t);
	11066	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_s16)))
	11067	void svst1q_scatter_u64base_index_s16(svbool_t, svuint64_t, int64_t, svint16_t);
	11068	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_u8)))
	11069	void svst1q_scatter_u64base_offset_u8(svbool_t, svuint64_t, int64_t, svuint8_t);
	11070	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_u32)))
	11071	void svst1q_scatter_u64base_offset_u32(svbool_t, svuint64_t, int64_t, svuint32_t);
	11072	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_u64)))
	11073	void svst1q_scatter_u64base_offset_u64(svbool_t, svuint64_t, int64_t, svuint64_t);
	11074	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_u16)))
	11075	void svst1q_scatter_u64base_offset_u16(svbool_t, svuint64_t, int64_t, svuint16_t);
	11076	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_bf16)))
	11077	void svst1q_scatter_u64base_offset_bf16(svbool_t, svuint64_t, int64_t, svbfloat16_t);
	11078	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_s8)))
	11079	void svst1q_scatter_u64base_offset_s8(svbool_t, svuint64_t, int64_t, svint8_t);
	11080	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_f64)))
	11081	void svst1q_scatter_u64base_offset_f64(svbool_t, svuint64_t, int64_t, svfloat64_t);
	11082	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_f32)))
	11083	void svst1q_scatter_u64base_offset_f32(svbool_t, svuint64_t, int64_t, svfloat32_t);
	11084	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_f16)))
	11085	void svst1q_scatter_u64base_offset_f16(svbool_t, svuint64_t, int64_t, svfloat16_t);
	11086	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_s32)))
	11087	void svst1q_scatter_u64base_offset_s32(svbool_t, svuint64_t, int64_t, svint32_t);
	11088	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_s64)))
	11089	void svst1q_scatter_u64base_offset_s64(svbool_t, svuint64_t, int64_t, svint64_t);
	11090	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_mf8)))
	11091	void svst1q_scatter_u64base_offset_mf8(svbool_t, svuint64_t, int64_t, svmfloat8_t);
	11092	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_s16)))
	11093	void svst1q_scatter_u64base_offset_s16(svbool_t, svuint64_t, int64_t, svint16_t);
	11094	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_u32)))
	11095	void svst1q_scatter_s64index_u32(svbool_t, uint32_t *, svint64_t, svuint32_t);
	11096	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_u64)))
	11097	void svst1q_scatter_s64index_u64(svbool_t, uint64_t *, svint64_t, svuint64_t);
	11098	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_u16)))
	11099	void svst1q_scatter_s64index_u16(svbool_t, uint16_t *, svint64_t, svuint16_t);
	11100	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_bf16)))
	11101	void svst1q_scatter_s64index_bf16(svbool_t, bfloat16_t *, svint64_t, svbfloat16_t);
	11102	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_f64)))
	11103	void svst1q_scatter_s64index_f64(svbool_t, float64_t *, svint64_t, svfloat64_t);
	11104	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_f32)))
	11105	void svst1q_scatter_s64index_f32(svbool_t, float32_t *, svint64_t, svfloat32_t);
	11106	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_f16)))
	11107	void svst1q_scatter_s64index_f16(svbool_t, float16_t *, svint64_t, svfloat16_t);
	11108	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_s32)))
	11109	void svst1q_scatter_s64index_s32(svbool_t, int32_t *, svint64_t, svint32_t);
	11110	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_s64)))
	11111	void svst1q_scatter_s64index_s64(svbool_t, int64_t *, svint64_t, svint64_t);
	11112	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_s16)))
	11113	void svst1q_scatter_s64index_s16(svbool_t, int16_t *, svint64_t, svint16_t);
	11114	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_u32)))
	11115	void svst1q_scatter_u64index_u32(svbool_t, uint32_t *, svuint64_t, svuint32_t);
	11116	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_u64)))
	11117	void svst1q_scatter_u64index_u64(svbool_t, uint64_t *, svuint64_t, svuint64_t);
	11118	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_u16)))
	11119	void svst1q_scatter_u64index_u16(svbool_t, uint16_t *, svuint64_t, svuint16_t);
	11120	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_bf16)))
	11121	void svst1q_scatter_u64index_bf16(svbool_t, bfloat16_t *, svuint64_t, svbfloat16_t);
	11122	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_f64)))
	11123	void svst1q_scatter_u64index_f64(svbool_t, float64_t *, svuint64_t, svfloat64_t);
	11124	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_f32)))
	11125	void svst1q_scatter_u64index_f32(svbool_t, float32_t *, svuint64_t, svfloat32_t);
	11126	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_f16)))
	11127	void svst1q_scatter_u64index_f16(svbool_t, float16_t *, svuint64_t, svfloat16_t);
	11128	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_s32)))
	11129	void svst1q_scatter_u64index_s32(svbool_t, int32_t *, svuint64_t, svint32_t);
	11130	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_s64)))
	11131	void svst1q_scatter_u64index_s64(svbool_t, int64_t *, svuint64_t, svint64_t);
	11132	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_s16)))
	11133	void svst1q_scatter_u64index_s16(svbool_t, int16_t *, svuint64_t, svint16_t);
	11134	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_u8)))
	11135	void svst1q_scatter_s64offset_u8(svbool_t, uint8_t *, svint64_t, svuint8_t);
	11136	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_u32)))
	11137	void svst1q_scatter_s64offset_u32(svbool_t, uint32_t *, svint64_t, svuint32_t);
	11138	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_u64)))
	11139	void svst1q_scatter_s64offset_u64(svbool_t, uint64_t *, svint64_t, svuint64_t);
	11140	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_u16)))
	11141	void svst1q_scatter_s64offset_u16(svbool_t, uint16_t *, svint64_t, svuint16_t);
	11142	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_bf16)))
	11143	void svst1q_scatter_s64offset_bf16(svbool_t, bfloat16_t *, svint64_t, svbfloat16_t);
	11144	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_s8)))
	11145	void svst1q_scatter_s64offset_s8(svbool_t, int8_t *, svint64_t, svint8_t);
	11146	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_f64)))
	11147	void svst1q_scatter_s64offset_f64(svbool_t, float64_t *, svint64_t, svfloat64_t);
	11148	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_f32)))
	11149	void svst1q_scatter_s64offset_f32(svbool_t, float32_t *, svint64_t, svfloat32_t);
	11150	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_f16)))
	11151	void svst1q_scatter_s64offset_f16(svbool_t, float16_t *, svint64_t, svfloat16_t);
	11152	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_s32)))
	11153	void svst1q_scatter_s64offset_s32(svbool_t, int32_t *, svint64_t, svint32_t);
	11154	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_s64)))
	11155	void svst1q_scatter_s64offset_s64(svbool_t, int64_t *, svint64_t, svint64_t);
	11156	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_mf8)))
	11157	void svst1q_scatter_s64offset_mf8(svbool_t, mfloat8_t *, svint64_t, svmfloat8_t);
	11158	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_s16)))
	11159	void svst1q_scatter_s64offset_s16(svbool_t, int16_t *, svint64_t, svint16_t);
	11160	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_u8)))
	11161	void svst1q_scatter_u64offset_u8(svbool_t, uint8_t *, svuint64_t, svuint8_t);
	11162	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_u32)))
	11163	void svst1q_scatter_u64offset_u32(svbool_t, uint32_t *, svuint64_t, svuint32_t);
	11164	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_u64)))
	11165	void svst1q_scatter_u64offset_u64(svbool_t, uint64_t *, svuint64_t, svuint64_t);
	11166	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_u16)))
	11167	void svst1q_scatter_u64offset_u16(svbool_t, uint16_t *, svuint64_t, svuint16_t);
	11168	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_bf16)))
	11169	void svst1q_scatter_u64offset_bf16(svbool_t, bfloat16_t *, svuint64_t, svbfloat16_t);
	11170	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_s8)))
	11171	void svst1q_scatter_u64offset_s8(svbool_t, int8_t *, svuint64_t, svint8_t);
	11172	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_f64)))
	11173	void svst1q_scatter_u64offset_f64(svbool_t, float64_t *, svuint64_t, svfloat64_t);
	11174	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_f32)))
	11175	void svst1q_scatter_u64offset_f32(svbool_t, float32_t *, svuint64_t, svfloat32_t);
	11176	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_f16)))
	11177	void svst1q_scatter_u64offset_f16(svbool_t, float16_t *, svuint64_t, svfloat16_t);
	11178	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_s32)))
	11179	void svst1q_scatter_u64offset_s32(svbool_t, int32_t *, svuint64_t, svint32_t);
	11180	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_s64)))
	11181	void svst1q_scatter_u64offset_s64(svbool_t, int64_t *, svuint64_t, svint64_t);
	11182	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_mf8)))
	11183	void svst1q_scatter_u64offset_mf8(svbool_t, mfloat8_t *, svuint64_t, svmfloat8_t);
	11184	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_s16)))
	11185	void svst1q_scatter_u64offset_s16(svbool_t, int16_t *, svuint64_t, svint16_t);
	11186	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1wq_u32)))
	11187	void svst1wq_u32(svbool_t, uint32_t *, svuint32_t);
	11188	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1wq_f32)))
	11189	void svst1wq_f32(svbool_t, float32_t *, svfloat32_t);
	11190	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1wq_s32)))
	11191	void svst1wq_s32(svbool_t, int32_t *, svint32_t);
	11192	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1wq_vnum_u32)))
	11193	void svst1wq_vnum_u32(svbool_t, uint32_t *, int64_t, svuint32_t);
	11194	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1wq_vnum_f32)))
	11195	void svst1wq_vnum_f32(svbool_t, float32_t *, int64_t, svfloat32_t);
	11196	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1wq_vnum_s32)))
	11197	void svst1wq_vnum_s32(svbool_t, int32_t *, int64_t, svint32_t);
	11198	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_u32)))
	11199	svuint32_t svld1q_gather_index_u32(svbool_t, svuint64_t, int64_t);
	11200	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_u64)))
	11201	svuint64_t svld1q_gather_index_u64(svbool_t, svuint64_t, int64_t);
	11202	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_u16)))
	11203	svuint16_t svld1q_gather_index_u16(svbool_t, svuint64_t, int64_t);
	11204	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_bf16)))
	11205	svbfloat16_t svld1q_gather_index_bf16(svbool_t, svuint64_t, int64_t);
	11206	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_f64)))
	11207	svfloat64_t svld1q_gather_index_f64(svbool_t, svuint64_t, int64_t);
	11208	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_f32)))
	11209	svfloat32_t svld1q_gather_index_f32(svbool_t, svuint64_t, int64_t);
	11210	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_f16)))
	11211	svfloat16_t svld1q_gather_index_f16(svbool_t, svuint64_t, int64_t);
	11212	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_s32)))
	11213	svint32_t svld1q_gather_index_s32(svbool_t, svuint64_t, int64_t);
	11214	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_s64)))
	11215	svint64_t svld1q_gather_index_s64(svbool_t, svuint64_t, int64_t);
	11216	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_index_s16)))
	11217	svint16_t svld1q_gather_index_s16(svbool_t, svuint64_t, int64_t);
	11218	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_u8)))
	11219	svuint8_t svld1q_gather_offset_u8(svbool_t, svuint64_t, int64_t);
	11220	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_u32)))
	11221	svuint32_t svld1q_gather_offset_u32(svbool_t, svuint64_t, int64_t);
	11222	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_u64)))
	11223	svuint64_t svld1q_gather_offset_u64(svbool_t, svuint64_t, int64_t);
	11224	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_u16)))
	11225	svuint16_t svld1q_gather_offset_u16(svbool_t, svuint64_t, int64_t);
	11226	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_bf16)))
	11227	svbfloat16_t svld1q_gather_offset_bf16(svbool_t, svuint64_t, int64_t);
	11228	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_s8)))
	11229	svint8_t svld1q_gather_offset_s8(svbool_t, svuint64_t, int64_t);
	11230	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_f64)))
	11231	svfloat64_t svld1q_gather_offset_f64(svbool_t, svuint64_t, int64_t);
	11232	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_f32)))
	11233	svfloat32_t svld1q_gather_offset_f32(svbool_t, svuint64_t, int64_t);
	11234	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_f16)))
	11235	svfloat16_t svld1q_gather_offset_f16(svbool_t, svuint64_t, int64_t);
	11236	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_s32)))
	11237	svint32_t svld1q_gather_offset_s32(svbool_t, svuint64_t, int64_t);
	11238	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_s64)))
	11239	svint64_t svld1q_gather_offset_s64(svbool_t, svuint64_t, int64_t);
	11240	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_mf8)))
	11241	svmfloat8_t svld1q_gather_offset_mf8(svbool_t, svuint64_t, int64_t);
	11242	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_offset_s16)))
	11243	svint16_t svld1q_gather_offset_s16(svbool_t, svuint64_t, int64_t);
	11244	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_u8)))
	11245	svuint8_t svld1q_gather_u8(svbool_t, svuint64_t);
	11246	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_u32)))
	11247	svuint32_t svld1q_gather_u32(svbool_t, svuint64_t);
	11248	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_u64)))
	11249	svuint64_t svld1q_gather_u64(svbool_t, svuint64_t);
	11250	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_u16)))
	11251	svuint16_t svld1q_gather_u16(svbool_t, svuint64_t);
	11252	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_bf16)))
	11253	svbfloat16_t svld1q_gather_bf16(svbool_t, svuint64_t);
	11254	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_s8)))
	11255	svint8_t svld1q_gather_s8(svbool_t, svuint64_t);
	11256	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_f64)))
	11257	svfloat64_t svld1q_gather_f64(svbool_t, svuint64_t);
	11258	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_f32)))
	11259	svfloat32_t svld1q_gather_f32(svbool_t, svuint64_t);
	11260	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_f16)))
	11261	svfloat16_t svld1q_gather_f16(svbool_t, svuint64_t);
	11262	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_s32)))
	11263	svint32_t svld1q_gather_s32(svbool_t, svuint64_t);
	11264	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_s64)))
	11265	svint64_t svld1q_gather_s64(svbool_t, svuint64_t);
	11266	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_mf8)))
	11267	svmfloat8_t svld1q_gather_mf8(svbool_t, svuint64_t);
	11268	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64base_s16)))
	11269	svint16_t svld1q_gather_s16(svbool_t, svuint64_t);
	11270	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_u32)))
	11271	svuint32_t svld1q_gather_index(svbool_t, uint32_t const *, svuint64_t);
	11272	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_u64)))
	11273	svuint64_t svld1q_gather_index(svbool_t, uint64_t const *, svuint64_t);
	11274	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_u16)))
	11275	svuint16_t svld1q_gather_index(svbool_t, uint16_t const *, svuint64_t);
	11276	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_bf16)))
	11277	svbfloat16_t svld1q_gather_index(svbool_t, bfloat16_t const *, svuint64_t);
	11278	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_f64)))
	11279	svfloat64_t svld1q_gather_index(svbool_t, float64_t const *, svuint64_t);
	11280	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_f32)))
	11281	svfloat32_t svld1q_gather_index(svbool_t, float32_t const *, svuint64_t);
	11282	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_f16)))
	11283	svfloat16_t svld1q_gather_index(svbool_t, float16_t const *, svuint64_t);
	11284	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_s32)))
	11285	svint32_t svld1q_gather_index(svbool_t, int32_t const *, svuint64_t);
	11286	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_s64)))
	11287	svint64_t svld1q_gather_index(svbool_t, int64_t const *, svuint64_t);
	11288	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64index_s16)))
	11289	svint16_t svld1q_gather_index(svbool_t, int16_t const *, svuint64_t);
	11290	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_u8)))
	11291	svuint8_t svld1q_gather_offset(svbool_t, uint8_t const *, svuint64_t);
	11292	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_u32)))
	11293	svuint32_t svld1q_gather_offset(svbool_t, uint32_t const *, svuint64_t);
	11294	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_u64)))
	11295	svuint64_t svld1q_gather_offset(svbool_t, uint64_t const *, svuint64_t);
	11296	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_u16)))
	11297	svuint16_t svld1q_gather_offset(svbool_t, uint16_t const *, svuint64_t);
	11298	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_bf16)))
	11299	svbfloat16_t svld1q_gather_offset(svbool_t, bfloat16_t const *, svuint64_t);
	11300	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_s8)))
	11301	svint8_t svld1q_gather_offset(svbool_t, int8_t const *, svuint64_t);
	11302	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_f64)))
	11303	svfloat64_t svld1q_gather_offset(svbool_t, float64_t const *, svuint64_t);
	11304	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_f32)))
	11305	svfloat32_t svld1q_gather_offset(svbool_t, float32_t const *, svuint64_t);
	11306	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_f16)))
	11307	svfloat16_t svld1q_gather_offset(svbool_t, float16_t const *, svuint64_t);
	11308	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_s32)))
	11309	svint32_t svld1q_gather_offset(svbool_t, int32_t const *, svuint64_t);
	11310	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_s64)))
	11311	svint64_t svld1q_gather_offset(svbool_t, int64_t const *, svuint64_t);
	11312	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_mf8)))
	11313	svmfloat8_t svld1q_gather_offset(svbool_t, mfloat8_t const *, svuint64_t);
	11314	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1q_gather_u64offset_s16)))
	11315	svint16_t svld1q_gather_offset(svbool_t, int16_t const *, svuint64_t);
	11316	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1udq_u64)))
	11317	svuint64_t svld1udq(svbool_t, uint64_t const *);
	11318	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1udq_f64)))
	11319	svfloat64_t svld1udq(svbool_t, float64_t const *);
	11320	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1udq_s64)))
	11321	svint64_t svld1udq(svbool_t, int64_t const *);
	11322	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1udq_vnum_u64)))
	11323	svuint64_t svld1udq_vnum(svbool_t, uint64_t const *, int64_t);
	11324	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1udq_vnum_f64)))
	11325	svfloat64_t svld1udq_vnum(svbool_t, float64_t const *, int64_t);
	11326	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1udq_vnum_s64)))
	11327	svint64_t svld1udq_vnum(svbool_t, int64_t const *, int64_t);
	11328	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1uwq_u32)))
	11329	svuint32_t svld1uwq(svbool_t, uint32_t const *);
	11330	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1uwq_f32)))
	11331	svfloat32_t svld1uwq(svbool_t, float32_t const *);
	11332	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1uwq_s32)))
	11333	svint32_t svld1uwq(svbool_t, int32_t const *);
	11334	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1uwq_vnum_u32)))
	11335	svuint32_t svld1uwq_vnum(svbool_t, uint32_t const *, int64_t);
	11336	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1uwq_vnum_f32)))
	11337	svfloat32_t svld1uwq_vnum(svbool_t, float32_t const *, int64_t);
	11338	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1uwq_vnum_s32)))
	11339	svint32_t svld1uwq_vnum(svbool_t, int32_t const *, int64_t);
	11340	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1dq_u64)))
	11341	void svst1dq(svbool_t, uint64_t *, svuint64_t);
	11342	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1dq_f64)))
	11343	void svst1dq(svbool_t, float64_t *, svfloat64_t);
	11344	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1dq_s64)))
	11345	void svst1dq(svbool_t, int64_t *, svint64_t);
	11346	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1dq_vnum_u64)))
	11347	void svst1dq_vnum(svbool_t, uint64_t *, int64_t, svuint64_t);
	11348	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1dq_vnum_f64)))
	11349	void svst1dq_vnum(svbool_t, float64_t *, int64_t, svfloat64_t);
	11350	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1dq_vnum_s64)))
	11351	void svst1dq_vnum(svbool_t, int64_t *, int64_t, svint64_t);
	11352	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_u8)))
	11353	void svst1q_scatter(svbool_t, svuint64_t, svuint8_t);
	11354	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_u32)))
	11355	void svst1q_scatter(svbool_t, svuint64_t, svuint32_t);
	11356	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_u64)))
	11357	void svst1q_scatter(svbool_t, svuint64_t, svuint64_t);
	11358	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_u16)))
	11359	void svst1q_scatter(svbool_t, svuint64_t, svuint16_t);
	11360	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_bf16)))
	11361	void svst1q_scatter(svbool_t, svuint64_t, svbfloat16_t);
	11362	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_s8)))
	11363	void svst1q_scatter(svbool_t, svuint64_t, svint8_t);
	11364	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_f64)))
	11365	void svst1q_scatter(svbool_t, svuint64_t, svfloat64_t);
	11366	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_f32)))
	11367	void svst1q_scatter(svbool_t, svuint64_t, svfloat32_t);
	11368	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_f16)))
	11369	void svst1q_scatter(svbool_t, svuint64_t, svfloat16_t);
	11370	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_s32)))
	11371	void svst1q_scatter(svbool_t, svuint64_t, svint32_t);
	11372	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_s64)))
	11373	void svst1q_scatter(svbool_t, svuint64_t, svint64_t);
	11374	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_mf8)))
	11375	void svst1q_scatter(svbool_t, svuint64_t, svmfloat8_t);
	11376	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_s16)))
	11377	void svst1q_scatter(svbool_t, svuint64_t, svint16_t);
	11378	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_u32)))
	11379	void svst1q_scatter_index(svbool_t, svuint64_t, int64_t, svuint32_t);
	11380	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_u64)))
	11381	void svst1q_scatter_index(svbool_t, svuint64_t, int64_t, svuint64_t);
	11382	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_u16)))
	11383	void svst1q_scatter_index(svbool_t, svuint64_t, int64_t, svuint16_t);
	11384	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_bf16)))
	11385	void svst1q_scatter_index(svbool_t, svuint64_t, int64_t, svbfloat16_t);
	11386	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_f64)))
	11387	void svst1q_scatter_index(svbool_t, svuint64_t, int64_t, svfloat64_t);
	11388	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_f32)))
	11389	void svst1q_scatter_index(svbool_t, svuint64_t, int64_t, svfloat32_t);
	11390	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_f16)))
	11391	void svst1q_scatter_index(svbool_t, svuint64_t, int64_t, svfloat16_t);
	11392	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_s32)))
	11393	void svst1q_scatter_index(svbool_t, svuint64_t, int64_t, svint32_t);
	11394	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_s64)))
	11395	void svst1q_scatter_index(svbool_t, svuint64_t, int64_t, svint64_t);
	11396	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_index_s16)))
	11397	void svst1q_scatter_index(svbool_t, svuint64_t, int64_t, svint16_t);
	11398	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_u8)))
	11399	void svst1q_scatter_offset(svbool_t, svuint64_t, int64_t, svuint8_t);
	11400	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_u32)))
	11401	void svst1q_scatter_offset(svbool_t, svuint64_t, int64_t, svuint32_t);
	11402	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_u64)))
	11403	void svst1q_scatter_offset(svbool_t, svuint64_t, int64_t, svuint64_t);
	11404	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_u16)))
	11405	void svst1q_scatter_offset(svbool_t, svuint64_t, int64_t, svuint16_t);
	11406	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_bf16)))
	11407	void svst1q_scatter_offset(svbool_t, svuint64_t, int64_t, svbfloat16_t);
	11408	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_s8)))
	11409	void svst1q_scatter_offset(svbool_t, svuint64_t, int64_t, svint8_t);
	11410	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_f64)))
	11411	void svst1q_scatter_offset(svbool_t, svuint64_t, int64_t, svfloat64_t);
	11412	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_f32)))
	11413	void svst1q_scatter_offset(svbool_t, svuint64_t, int64_t, svfloat32_t);
	11414	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_f16)))
	11415	void svst1q_scatter_offset(svbool_t, svuint64_t, int64_t, svfloat16_t);
	11416	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_s32)))
	11417	void svst1q_scatter_offset(svbool_t, svuint64_t, int64_t, svint32_t);
	11418	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_s64)))
	11419	void svst1q_scatter_offset(svbool_t, svuint64_t, int64_t, svint64_t);
	11420	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_mf8)))
	11421	void svst1q_scatter_offset(svbool_t, svuint64_t, int64_t, svmfloat8_t);
	11422	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64base_offset_s16)))
	11423	void svst1q_scatter_offset(svbool_t, svuint64_t, int64_t, svint16_t);
	11424	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_u32)))
	11425	void svst1q_scatter_index(svbool_t, uint32_t *, svint64_t, svuint32_t);
	11426	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_u64)))
	11427	void svst1q_scatter_index(svbool_t, uint64_t *, svint64_t, svuint64_t);
	11428	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_u16)))
	11429	void svst1q_scatter_index(svbool_t, uint16_t *, svint64_t, svuint16_t);
	11430	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_bf16)))
	11431	void svst1q_scatter_index(svbool_t, bfloat16_t *, svint64_t, svbfloat16_t);
	11432	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_f64)))
	11433	void svst1q_scatter_index(svbool_t, float64_t *, svint64_t, svfloat64_t);
	11434	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_f32)))
	11435	void svst1q_scatter_index(svbool_t, float32_t *, svint64_t, svfloat32_t);
	11436	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_f16)))
	11437	void svst1q_scatter_index(svbool_t, float16_t *, svint64_t, svfloat16_t);
	11438	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_s32)))
	11439	void svst1q_scatter_index(svbool_t, int32_t *, svint64_t, svint32_t);
	11440	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_s64)))
	11441	void svst1q_scatter_index(svbool_t, int64_t *, svint64_t, svint64_t);
	11442	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64index_s16)))
	11443	void svst1q_scatter_index(svbool_t, int16_t *, svint64_t, svint16_t);
	11444	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_u32)))
	11445	void svst1q_scatter_index(svbool_t, uint32_t *, svuint64_t, svuint32_t);
	11446	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_u64)))
	11447	void svst1q_scatter_index(svbool_t, uint64_t *, svuint64_t, svuint64_t);
	11448	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_u16)))
	11449	void svst1q_scatter_index(svbool_t, uint16_t *, svuint64_t, svuint16_t);
	11450	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_bf16)))
	11451	void svst1q_scatter_index(svbool_t, bfloat16_t *, svuint64_t, svbfloat16_t);
	11452	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_f64)))
	11453	void svst1q_scatter_index(svbool_t, float64_t *, svuint64_t, svfloat64_t);
	11454	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_f32)))
	11455	void svst1q_scatter_index(svbool_t, float32_t *, svuint64_t, svfloat32_t);
	11456	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_f16)))
	11457	void svst1q_scatter_index(svbool_t, float16_t *, svuint64_t, svfloat16_t);
	11458	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_s32)))
	11459	void svst1q_scatter_index(svbool_t, int32_t *, svuint64_t, svint32_t);
	11460	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_s64)))
	11461	void svst1q_scatter_index(svbool_t, int64_t *, svuint64_t, svint64_t);
	11462	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64index_s16)))
	11463	void svst1q_scatter_index(svbool_t, int16_t *, svuint64_t, svint16_t);
	11464	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_u8)))
	11465	void svst1q_scatter_offset(svbool_t, uint8_t *, svint64_t, svuint8_t);
	11466	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_u32)))
	11467	void svst1q_scatter_offset(svbool_t, uint32_t *, svint64_t, svuint32_t);
	11468	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_u64)))
	11469	void svst1q_scatter_offset(svbool_t, uint64_t *, svint64_t, svuint64_t);
	11470	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_u16)))
	11471	void svst1q_scatter_offset(svbool_t, uint16_t *, svint64_t, svuint16_t);
	11472	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_bf16)))
	11473	void svst1q_scatter_offset(svbool_t, bfloat16_t *, svint64_t, svbfloat16_t);
	11474	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_s8)))
	11475	void svst1q_scatter_offset(svbool_t, int8_t *, svint64_t, svint8_t);
	11476	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_f64)))
	11477	void svst1q_scatter_offset(svbool_t, float64_t *, svint64_t, svfloat64_t);
	11478	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_f32)))
	11479	void svst1q_scatter_offset(svbool_t, float32_t *, svint64_t, svfloat32_t);
	11480	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_f16)))
	11481	void svst1q_scatter_offset(svbool_t, float16_t *, svint64_t, svfloat16_t);
	11482	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_s32)))
	11483	void svst1q_scatter_offset(svbool_t, int32_t *, svint64_t, svint32_t);
	11484	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_s64)))
	11485	void svst1q_scatter_offset(svbool_t, int64_t *, svint64_t, svint64_t);
	11486	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_mf8)))
	11487	void svst1q_scatter_offset(svbool_t, mfloat8_t *, svint64_t, svmfloat8_t);
	11488	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_s64offset_s16)))
	11489	void svst1q_scatter_offset(svbool_t, int16_t *, svint64_t, svint16_t);
	11490	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_u8)))
	11491	void svst1q_scatter_offset(svbool_t, uint8_t *, svuint64_t, svuint8_t);
	11492	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_u32)))
	11493	void svst1q_scatter_offset(svbool_t, uint32_t *, svuint64_t, svuint32_t);
	11494	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_u64)))
	11495	void svst1q_scatter_offset(svbool_t, uint64_t *, svuint64_t, svuint64_t);
	11496	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_u16)))
	11497	void svst1q_scatter_offset(svbool_t, uint16_t *, svuint64_t, svuint16_t);
	11498	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_bf16)))
	11499	void svst1q_scatter_offset(svbool_t, bfloat16_t *, svuint64_t, svbfloat16_t);
	11500	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_s8)))
	11501	void svst1q_scatter_offset(svbool_t, int8_t *, svuint64_t, svint8_t);
	11502	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_f64)))
	11503	void svst1q_scatter_offset(svbool_t, float64_t *, svuint64_t, svfloat64_t);
	11504	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_f32)))
	11505	void svst1q_scatter_offset(svbool_t, float32_t *, svuint64_t, svfloat32_t);
	11506	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_f16)))
	11507	void svst1q_scatter_offset(svbool_t, float16_t *, svuint64_t, svfloat16_t);
	11508	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_s32)))
	11509	void svst1q_scatter_offset(svbool_t, int32_t *, svuint64_t, svint32_t);
	11510	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_s64)))
	11511	void svst1q_scatter_offset(svbool_t, int64_t *, svuint64_t, svint64_t);
	11512	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_mf8)))
	11513	void svst1q_scatter_offset(svbool_t, mfloat8_t *, svuint64_t, svmfloat8_t);
	11514	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1q_scatter_u64offset_s16)))
	11515	void svst1q_scatter_offset(svbool_t, int16_t *, svuint64_t, svint16_t);
	11516	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1wq_u32)))
	11517	void svst1wq(svbool_t, uint32_t *, svuint32_t);
	11518	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1wq_f32)))
	11519	void svst1wq(svbool_t, float32_t *, svfloat32_t);
	11520	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1wq_s32)))
	11521	void svst1wq(svbool_t, int32_t *, svint32_t);
	11522	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1wq_vnum_u32)))
	11523	void svst1wq_vnum(svbool_t, uint32_t *, int64_t, svuint32_t);
	11524	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1wq_vnum_f32)))
	11525	void svst1wq_vnum(svbool_t, float32_t *, int64_t, svfloat32_t);
	11526	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1wq_vnum_s32)))
	11527	void svst1wq_vnum(svbool_t, int32_t *, int64_t, svint32_t);
11386	11528	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcntp_c8)))
11387	11529	uint64_t svcntp_c8(svcount_t, uint64_t);
11388	11530	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcntp_c32)))
...	...	@@ -11391,30 +11533,12 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcntp_c64)))
11391	11533	uint64_t svcntp_c64(svcount_t, uint64_t);
11392	11534	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcntp_c16)))
11393	11535	uint64_t svcntp_c16(svcount_t, uint64_t);
11394		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate2_b)))
11395		svboolx2_t svcreate2_b(svbool_t, svbool_t);
11396		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate4_b)))
11397		svboolx4_t svcreate4_b(svbool_t, svbool_t, svbool_t, svbool_t);
11398		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_f32_f16)))
11399		svfloat32_t svdot_f32_f16(svfloat32_t, svfloat16_t, svfloat16_t);
11400		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_s32_s16)))
11401		svint32_t svdot_s32_s16(svint32_t, svint16_t, svint16_t);
11402		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_u32_u16)))
11403		svuint32_t svdot_u32_u16(svuint32_t, svuint16_t, svuint16_t);
11404		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_lane_f32_f16)))
11405		svfloat32_t svdot_lane_f32_f16(svfloat32_t, svfloat16_t, svfloat16_t, uint64_t);
11406		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_lane_s32_s16)))
11407		svint32_t svdot_lane_s32_s16(svint32_t, svint16_t, svint16_t, uint64_t);
11408		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_lane_u32_u16)))
11409		svuint32_t svdot_lane_u32_u16(svuint32_t, svuint16_t, svuint16_t, uint64_t);
11410		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget2_b)))
11411		svbool_t svget2_b(svboolx2_t, uint64_t);
11412		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget4_b)))
11413		svbool_t svget4_b(svboolx4_t, uint64_t);
11414	11536	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_u8_x2)))
11415	11537	svuint8x2_t svld1_u8_x2(svcount_t, uint8_t const *);
11416	11538	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_s8_x2)))
11417	11539	svint8x2_t svld1_s8_x2(svcount_t, int8_t const *);
	11540	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_mf8_x2)))
	11541	svmfloat8x2_t svld1_mf8_x2(svcount_t, mfloat8_t const *);
11418	11542	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_u64_x2)))
11419	11543	svuint64x2_t svld1_u64_x2(svcount_t, uint64_t const *);
11420	11544	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_f64_x2)))
...	...	@@ -11439,6 +11563,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_u8_x4)))
11439	11563	svuint8x4_t svld1_u8_x4(svcount_t, uint8_t const *);
11440	11564	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_s8_x4)))
11441	11565	svint8x4_t svld1_s8_x4(svcount_t, int8_t const *);
	11566	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_mf8_x4)))
	11567	svmfloat8x4_t svld1_mf8_x4(svcount_t, mfloat8_t const *);
11442	11568	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_u64_x4)))
11443	11569	svuint64x4_t svld1_u64_x4(svcount_t, uint64_t const *);
11444	11570	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_f64_x4)))
...	...	@@ -11463,6 +11589,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_u8_x2)))
11463	11589	svuint8x2_t svld1_vnum_u8_x2(svcount_t, uint8_t const *, int64_t);
11464	11590	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_s8_x2)))
11465	11591	svint8x2_t svld1_vnum_s8_x2(svcount_t, int8_t const *, int64_t);
	11592	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_mf8_x2)))
	11593	svmfloat8x2_t svld1_vnum_mf8_x2(svcount_t, mfloat8_t const *, int64_t);
11466	11594	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_u64_x2)))
11467	11595	svuint64x2_t svld1_vnum_u64_x2(svcount_t, uint64_t const *, int64_t);
11468	11596	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_f64_x2)))
...	...	@@ -11487,6 +11615,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_u8_x4)))
11487	11615	svuint8x4_t svld1_vnum_u8_x4(svcount_t, uint8_t const *, int64_t);
11488	11616	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_s8_x4)))
11489	11617	svint8x4_t svld1_vnum_s8_x4(svcount_t, int8_t const *, int64_t);
	11618	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_mf8_x4)))
	11619	svmfloat8x4_t svld1_vnum_mf8_x4(svcount_t, mfloat8_t const *, int64_t);
11490	11620	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_u64_x4)))
11491	11621	svuint64x4_t svld1_vnum_u64_x4(svcount_t, uint64_t const *, int64_t);
11492	11622	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_f64_x4)))
...	...	@@ -11511,6 +11641,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_u8_x2)))
11511	11641	svuint8x2_t svldnt1_u8_x2(svcount_t, uint8_t const *);
11512	11642	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_s8_x2)))
11513	11643	svint8x2_t svldnt1_s8_x2(svcount_t, int8_t const *);
	11644	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_mf8_x2)))
	11645	svmfloat8x2_t svldnt1_mf8_x2(svcount_t, mfloat8_t const *);
11514	11646	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_u64_x2)))
11515	11647	svuint64x2_t svldnt1_u64_x2(svcount_t, uint64_t const *);
11516	11648	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_f64_x2)))
...	...	@@ -11535,6 +11667,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_u8_x4)))
11535	11667	svuint8x4_t svldnt1_u8_x4(svcount_t, uint8_t const *);
11536	11668	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_s8_x4)))
11537	11669	svint8x4_t svldnt1_s8_x4(svcount_t, int8_t const *);
	11670	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_mf8_x4)))
	11671	svmfloat8x4_t svldnt1_mf8_x4(svcount_t, mfloat8_t const *);
11538	11672	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_u64_x4)))
11539	11673	svuint64x4_t svldnt1_u64_x4(svcount_t, uint64_t const *);
11540	11674	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_f64_x4)))
...	...	@@ -11559,6 +11693,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_u8_x2))
11559	11693	svuint8x2_t svldnt1_vnum_u8_x2(svcount_t, uint8_t const *, int64_t);
11560	11694	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_s8_x2)))
11561	11695	svint8x2_t svldnt1_vnum_s8_x2(svcount_t, int8_t const *, int64_t);
	11696	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_mf8_x2)))
	11697	svmfloat8x2_t svldnt1_vnum_mf8_x2(svcount_t, mfloat8_t const *, int64_t);
11562	11698	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_u64_x2)))
11563	11699	svuint64x2_t svldnt1_vnum_u64_x2(svcount_t, uint64_t const *, int64_t);
11564	11700	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_f64_x2)))
...	...	@@ -11583,6 +11719,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_u8_x4))
11583	11719	svuint8x4_t svldnt1_vnum_u8_x4(svcount_t, uint8_t const *, int64_t);
11584	11720	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_s8_x4)))
11585	11721	svint8x4_t svldnt1_vnum_s8_x4(svcount_t, int8_t const *, int64_t);
	11722	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_mf8_x4)))
	11723	svmfloat8x4_t svldnt1_vnum_mf8_x4(svcount_t, mfloat8_t const *, int64_t);
11586	11724	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_u64_x4)))
11587	11725	svuint64x4_t svldnt1_vnum_u64_x4(svcount_t, uint64_t const *, int64_t);
11588	11726	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_f64_x4)))
...	...	@@ -11631,36 +11769,22 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svpsel_lane_c8)))
11631	11769	svcount_t svpsel_lane_c8(svcount_t, svbool_t, uint32_t);
11632	11770	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svptrue_c8)))
11633	11771	svcount_t svptrue_c8(void);
11634		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svptrue_c32)))
11635		svcount_t svptrue_c32(void);
11636		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svptrue_c64)))
11637		svcount_t svptrue_c64(void);
11638		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svptrue_c16)))
11639		svcount_t svptrue_c16(void);
11640		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svqcvtn_s16_s32_x2)))
11641		svint16_t svqcvtn_s16_s32_x2(svint32x2_t);
11642		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svqcvtn_u16_s32_x2)))
11643		svuint16_t svqcvtn_u16_s32_x2(svint32x2_t);
11644		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svqcvtn_u16_u32_x2)))
11645		svuint16_t svqcvtn_u16_u32_x2(svuint32x2_t);
11646		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svqrshrn_n_s16_s32_x2)))
11647		svint16_t svqrshrn_n_s16_s32_x2(svint32x2_t, uint64_t);
11648		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svqrshrn_n_u16_u32_x2)))
11649		svuint16_t svqrshrn_n_u16_u32_x2(svuint32x2_t, uint64_t);
11650		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svqrshrun_n_u16_s32_x2)))
11651		svuint16_t svqrshrun_n_u16_s32_x2(svint32x2_t, uint64_t);
	11772	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svptrue_c32)))
	11773	svcount_t svptrue_c32(void);
	11774	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svptrue_c64)))
	11775	svcount_t svptrue_c64(void);
	11776	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svptrue_c16)))
	11777	svcount_t svptrue_c16(void);
11652	11778	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svreinterpret_b)))
11653	11779	svbool_t svreinterpret_b(svcount_t);
11654	11780	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svreinterpret_c)))
11655	11781	svcount_t svreinterpret_c(svbool_t);
11656		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset2_b)))
11657		svboolx2_t svset2_b(svboolx2_t, uint64_t, svbool_t);
11658		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset4_b)))
11659		svboolx4_t svset4_b(svboolx4_t, uint64_t, svbool_t);
11660	11782	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_u8_x2)))
11661	11783	void svst1_u8_x2(svcount_t, uint8_t *, svuint8x2_t);
11662	11784	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_s8_x2)))
11663	11785	void svst1_s8_x2(svcount_t, int8_t *, svint8x2_t);
	11786	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_mf8_x2)))
	11787	void svst1_mf8_x2(svcount_t, mfloat8_t *, svmfloat8x2_t);
11664	11788	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_u64_x2)))
11665	11789	void svst1_u64_x2(svcount_t, uint64_t *, svuint64x2_t);
11666	11790	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_f64_x2)))
...	...	@@ -11685,6 +11809,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_u8_x4)))
11685	11809	void svst1_u8_x4(svcount_t, uint8_t *, svuint8x4_t);
11686	11810	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_s8_x4)))
11687	11811	void svst1_s8_x4(svcount_t, int8_t *, svint8x4_t);
	11812	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_mf8_x4)))
	11813	void svst1_mf8_x4(svcount_t, mfloat8_t *, svmfloat8x4_t);
11688	11814	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_u64_x4)))
11689	11815	void svst1_u64_x4(svcount_t, uint64_t *, svuint64x4_t);
11690	11816	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_f64_x4)))
...	...	@@ -11709,6 +11835,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_u8_x2)))
11709	11835	void svst1_vnum_u8_x2(svcount_t, uint8_t *, int64_t, svuint8x2_t);
11710	11836	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_s8_x2)))
11711	11837	void svst1_vnum_s8_x2(svcount_t, int8_t *, int64_t, svint8x2_t);
	11838	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_mf8_x2)))
	11839	void svst1_vnum_mf8_x2(svcount_t, mfloat8_t *, int64_t, svmfloat8x2_t);
11712	11840	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_u64_x2)))
11713	11841	void svst1_vnum_u64_x2(svcount_t, uint64_t *, int64_t, svuint64x2_t);
11714	11842	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_f64_x2)))
...	...	@@ -11733,6 +11861,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_u8_x4)))
11733	11861	void svst1_vnum_u8_x4(svcount_t, uint8_t *, int64_t, svuint8x4_t);
11734	11862	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_s8_x4)))
11735	11863	void svst1_vnum_s8_x4(svcount_t, int8_t *, int64_t, svint8x4_t);
	11864	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_mf8_x4)))
	11865	void svst1_vnum_mf8_x4(svcount_t, mfloat8_t *, int64_t, svmfloat8x4_t);
11736	11866	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_u64_x4)))
11737	11867	void svst1_vnum_u64_x4(svcount_t, uint64_t *, int64_t, svuint64x4_t);
11738	11868	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_f64_x4)))
...	...	@@ -11757,6 +11887,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_u8_x2)))
11757	11887	void svstnt1_u8_x2(svcount_t, uint8_t *, svuint8x2_t);
11758	11888	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_s8_x2)))
11759	11889	void svstnt1_s8_x2(svcount_t, int8_t *, svint8x2_t);
	11890	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_mf8_x2)))
	11891	void svstnt1_mf8_x2(svcount_t, mfloat8_t *, svmfloat8x2_t);
11760	11892	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_u64_x2)))
11761	11893	void svstnt1_u64_x2(svcount_t, uint64_t *, svuint64x2_t);
11762	11894	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_f64_x2)))
...	...	@@ -11781,6 +11913,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_u8_x4)))
11781	11913	void svstnt1_u8_x4(svcount_t, uint8_t *, svuint8x4_t);
11782	11914	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_s8_x4)))
11783	11915	void svstnt1_s8_x4(svcount_t, int8_t *, svint8x4_t);
	11916	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_mf8_x4)))
	11917	void svstnt1_mf8_x4(svcount_t, mfloat8_t *, svmfloat8x4_t);
11784	11918	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_u64_x4)))
11785	11919	void svstnt1_u64_x4(svcount_t, uint64_t *, svuint64x4_t);
11786	11920	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_f64_x4)))
...	...	@@ -11805,6 +11939,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_u8_x2))
11805	11939	void svstnt1_vnum_u8_x2(svcount_t, uint8_t *, int64_t, svuint8x2_t);
11806	11940	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_s8_x2)))
11807	11941	void svstnt1_vnum_s8_x2(svcount_t, int8_t *, int64_t, svint8x2_t);
	11942	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_mf8_x2)))
	11943	void svstnt1_vnum_mf8_x2(svcount_t, mfloat8_t *, int64_t, svmfloat8x2_t);
11808	11944	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_u64_x2)))
11809	11945	void svstnt1_vnum_u64_x2(svcount_t, uint64_t *, int64_t, svuint64x2_t);
11810	11946	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_f64_x2)))
...	...	@@ -11829,6 +11965,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_u8_x4))
11829	11965	void svstnt1_vnum_u8_x4(svcount_t, uint8_t *, int64_t, svuint8x4_t);
11830	11966	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_s8_x4)))
11831	11967	void svstnt1_vnum_s8_x4(svcount_t, int8_t *, int64_t, svint8x4_t);
	11968	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_mf8_x4)))
	11969	void svstnt1_vnum_mf8_x4(svcount_t, mfloat8_t *, int64_t, svmfloat8x4_t);
11832	11970	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_u64_x4)))
11833	11971	void svstnt1_vnum_u64_x4(svcount_t, uint64_t *, int64_t, svuint64x4_t);
11834	11972	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_f64_x4)))
...	...	@@ -11849,10 +11987,6 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_f32_x4)
11849	11987	void svstnt1_vnum_f32_x4(svcount_t, float32_t *, int64_t, svfloat32x4_t);
11850	11988	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_s32_x4)))
11851	11989	void svstnt1_vnum_s32_x4(svcount_t, int32_t *, int64_t, svint32x4_t);
11852		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef2_b)))
11853		svboolx2_t svundef2_b();
11854		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef4_b)))
11855		svboolx4_t svundef4_b();
11856	11990	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_c8_s64)))
11857	11991	svcount_t svwhilege_c8_s64(int64_t, int64_t, uint64_t);
11858	11992	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_c32_s64)))
...	...	@@ -11869,22 +12003,6 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_c64_u64)))
11869	12003	svcount_t svwhilege_c64_u64(uint64_t, uint64_t, uint64_t);
11870	12004	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_c16_u64)))
11871	12005	svcount_t svwhilege_c16_u64(uint64_t, uint64_t, uint64_t);
11872		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b8_s64_x2)))
11873		svboolx2_t svwhilege_b8_s64_x2(int64_t, int64_t);
11874		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b32_s64_x2)))
11875		svboolx2_t svwhilege_b32_s64_x2(int64_t, int64_t);
11876		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b64_s64_x2)))
11877		svboolx2_t svwhilege_b64_s64_x2(int64_t, int64_t);
11878		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b16_s64_x2)))
11879		svboolx2_t svwhilege_b16_s64_x2(int64_t, int64_t);
11880		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b8_u64_x2)))
11881		svboolx2_t svwhilege_b8_u64_x2(uint64_t, uint64_t);
11882		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b32_u64_x2)))
11883		svboolx2_t svwhilege_b32_u64_x2(uint64_t, uint64_t);
11884		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b64_u64_x2)))
11885		svboolx2_t svwhilege_b64_u64_x2(uint64_t, uint64_t);
11886		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b16_u64_x2)))
11887		svboolx2_t svwhilege_b16_u64_x2(uint64_t, uint64_t);
11888	12006	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_c8_s64)))
11889	12007	svcount_t svwhilegt_c8_s64(int64_t, int64_t, uint64_t);
11890	12008	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_c32_s64)))
...	...	@@ -11901,22 +12019,6 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_c64_u64)))
11901	12019	svcount_t svwhilegt_c64_u64(uint64_t, uint64_t, uint64_t);
11902	12020	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_c16_u64)))
11903	12021	svcount_t svwhilegt_c16_u64(uint64_t, uint64_t, uint64_t);
11904		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b8_s64_x2)))
11905		svboolx2_t svwhilegt_b8_s64_x2(int64_t, int64_t);
11906		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b32_s64_x2)))
11907		svboolx2_t svwhilegt_b32_s64_x2(int64_t, int64_t);
11908		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b64_s64_x2)))
11909		svboolx2_t svwhilegt_b64_s64_x2(int64_t, int64_t);
11910		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b16_s64_x2)))
11911		svboolx2_t svwhilegt_b16_s64_x2(int64_t, int64_t);
11912		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b8_u64_x2)))
11913		svboolx2_t svwhilegt_b8_u64_x2(uint64_t, uint64_t);
11914		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b32_u64_x2)))
11915		svboolx2_t svwhilegt_b32_u64_x2(uint64_t, uint64_t);
11916		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b64_u64_x2)))
11917		svboolx2_t svwhilegt_b64_u64_x2(uint64_t, uint64_t);
11918		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b16_u64_x2)))
11919		svboolx2_t svwhilegt_b16_u64_x2(uint64_t, uint64_t);
11920	12022	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_c8_s64)))
11921	12023	svcount_t svwhilele_c8_s64(int64_t, int64_t, uint64_t);
11922	12024	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_c32_s64)))
...	...	@@ -11933,22 +12035,6 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_c64_u64)))
11933	12035	svcount_t svwhilele_c64_u64(uint64_t, uint64_t, uint64_t);
11934	12036	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_c16_u64)))
11935	12037	svcount_t svwhilele_c16_u64(uint64_t, uint64_t, uint64_t);
11936		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b8_s64_x2)))
11937		svboolx2_t svwhilele_b8_s64_x2(int64_t, int64_t);
11938		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b32_s64_x2)))
11939		svboolx2_t svwhilele_b32_s64_x2(int64_t, int64_t);
11940		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b64_s64_x2)))
11941		svboolx2_t svwhilele_b64_s64_x2(int64_t, int64_t);
11942		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b16_s64_x2)))
11943		svboolx2_t svwhilele_b16_s64_x2(int64_t, int64_t);
11944		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b8_u64_x2)))
11945		svboolx2_t svwhilele_b8_u64_x2(uint64_t, uint64_t);
11946		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b32_u64_x2)))
11947		svboolx2_t svwhilele_b32_u64_x2(uint64_t, uint64_t);
11948		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b64_u64_x2)))
11949		svboolx2_t svwhilele_b64_u64_x2(uint64_t, uint64_t);
11950		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b16_u64_x2)))
11951		svboolx2_t svwhilele_b16_u64_x2(uint64_t, uint64_t);
11952	12038	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_c8_u64)))
11953	12039	svcount_t svwhilelt_c8_u64(uint64_t, uint64_t, uint64_t);
11954	12040	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_c32_u64)))
...	...	@@ -11965,60 +12051,12 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_c64_s64)))
11965	12051	svcount_t svwhilelt_c64_s64(int64_t, int64_t, uint64_t);
11966	12052	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_c16_s64)))
11967	12053	svcount_t svwhilelt_c16_s64(int64_t, int64_t, uint64_t);
11968		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b8_u64_x2)))
11969		svboolx2_t svwhilelt_b8_u64_x2(uint64_t, uint64_t);
11970		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b32_u64_x2)))
11971		svboolx2_t svwhilelt_b32_u64_x2(uint64_t, uint64_t);
11972		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b64_u64_x2)))
11973		svboolx2_t svwhilelt_b64_u64_x2(uint64_t, uint64_t);
11974		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b16_u64_x2)))
11975		svboolx2_t svwhilelt_b16_u64_x2(uint64_t, uint64_t);
11976		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b8_s64_x2)))
11977		svboolx2_t svwhilelt_b8_s64_x2(int64_t, int64_t);
11978		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b32_s64_x2)))
11979		svboolx2_t svwhilelt_b32_s64_x2(int64_t, int64_t);
11980		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b64_s64_x2)))
11981		svboolx2_t svwhilelt_b64_s64_x2(int64_t, int64_t);
11982		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b16_s64_x2)))
11983		svboolx2_t svwhilelt_b16_s64_x2(int64_t, int64_t);
11984		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlslb_f32)))
11985		svfloat32_t svbfmlslb(svfloat32_t, svbfloat16_t, svbfloat16_t);
11986		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlslb_lane_f32)))
11987		svfloat32_t svbfmlslb_lane(svfloat32_t, svbfloat16_t, svbfloat16_t, uint64_t);
11988		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlslt_f32)))
11989		svfloat32_t svbfmlslt(svfloat32_t, svbfloat16_t, svbfloat16_t);
11990		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svbfmlslt_lane_f32)))
11991		svfloat32_t svbfmlslt_lane(svfloat32_t, svbfloat16_t, svbfloat16_t, uint64_t);
11992		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_f64)))
11993		svfloat64_t svclamp(svfloat64_t, svfloat64_t, svfloat64_t);
11994		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_f32)))
11995		svfloat32_t svclamp(svfloat32_t, svfloat32_t, svfloat32_t);
11996		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclamp_f16)))
11997		svfloat16_t svclamp(svfloat16_t, svfloat16_t, svfloat16_t);
11998		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate2_b)))
11999		svboolx2_t svcreate2(svbool_t, svbool_t);
12000		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate4_b)))
12001		svboolx4_t svcreate4(svbool_t, svbool_t, svbool_t, svbool_t);
12002		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_f32_f16)))
12003		svfloat32_t svdot(svfloat32_t, svfloat16_t, svfloat16_t);
12004		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_s32_s16)))
12005		svint32_t svdot(svint32_t, svint16_t, svint16_t);
12006		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_u32_u16)))
12007		svuint32_t svdot(svuint32_t, svuint16_t, svuint16_t);
12008		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_lane_f32_f16)))
12009		svfloat32_t svdot_lane(svfloat32_t, svfloat16_t, svfloat16_t, uint64_t);
12010		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_lane_s32_s16)))
12011		svint32_t svdot_lane(svint32_t, svint16_t, svint16_t, uint64_t);
12012		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdot_lane_u32_u16)))
12013		svuint32_t svdot_lane(svuint32_t, svuint16_t, svuint16_t, uint64_t);
12014		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget2_b)))
12015		svbool_t svget2(svboolx2_t, uint64_t);
12016		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget4_b)))
12017		svbool_t svget4(svboolx4_t, uint64_t);
12018	12054	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_u8_x2)))
12019	12055	svuint8x2_t svld1_x2(svcount_t, uint8_t const *);
12020	12056	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_s8_x2)))
12021	12057	svint8x2_t svld1_x2(svcount_t, int8_t const *);
	12058	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_mf8_x2)))
	12059	svmfloat8x2_t svld1_x2(svcount_t, mfloat8_t const *);
12022	12060	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_u64_x2)))
12023	12061	svuint64x2_t svld1_x2(svcount_t, uint64_t const *);
12024	12062	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_f64_x2)))
...	...	@@ -12043,6 +12081,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_u8_x4)))
12043	12081	svuint8x4_t svld1_x4(svcount_t, uint8_t const *);
12044	12082	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_s8_x4)))
12045	12083	svint8x4_t svld1_x4(svcount_t, int8_t const *);
	12084	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_mf8_x4)))
	12085	svmfloat8x4_t svld1_x4(svcount_t, mfloat8_t const *);
12046	12086	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_u64_x4)))
12047	12087	svuint64x4_t svld1_x4(svcount_t, uint64_t const *);
12048	12088	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_f64_x4)))
...	...	@@ -12067,6 +12107,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_u8_x2)))
12067	12107	svuint8x2_t svld1_vnum_x2(svcount_t, uint8_t const *, int64_t);
12068	12108	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_s8_x2)))
12069	12109	svint8x2_t svld1_vnum_x2(svcount_t, int8_t const *, int64_t);
	12110	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_mf8_x2)))
	12111	svmfloat8x2_t svld1_vnum_x2(svcount_t, mfloat8_t const *, int64_t);
12070	12112	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_u64_x2)))
12071	12113	svuint64x2_t svld1_vnum_x2(svcount_t, uint64_t const *, int64_t);
12072	12114	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_f64_x2)))
...	...	@@ -12091,6 +12133,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_u8_x4)))
12091	12133	svuint8x4_t svld1_vnum_x4(svcount_t, uint8_t const *, int64_t);
12092	12134	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_s8_x4)))
12093	12135	svint8x4_t svld1_vnum_x4(svcount_t, int8_t const *, int64_t);
	12136	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_mf8_x4)))
	12137	svmfloat8x4_t svld1_vnum_x4(svcount_t, mfloat8_t const *, int64_t);
12094	12138	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_u64_x4)))
12095	12139	svuint64x4_t svld1_vnum_x4(svcount_t, uint64_t const *, int64_t);
12096	12140	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_f64_x4)))
...	...	@@ -12115,6 +12159,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_u8_x2)))
12115	12159	svuint8x2_t svldnt1_x2(svcount_t, uint8_t const *);
12116	12160	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_s8_x2)))
12117	12161	svint8x2_t svldnt1_x2(svcount_t, int8_t const *);
	12162	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_mf8_x2)))
	12163	svmfloat8x2_t svldnt1_x2(svcount_t, mfloat8_t const *);
12118	12164	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_u64_x2)))
12119	12165	svuint64x2_t svldnt1_x2(svcount_t, uint64_t const *);
12120	12166	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_f64_x2)))
...	...	@@ -12139,6 +12185,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_u8_x4)))
12139	12185	svuint8x4_t svldnt1_x4(svcount_t, uint8_t const *);
12140	12186	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_s8_x4)))
12141	12187	svint8x4_t svldnt1_x4(svcount_t, int8_t const *);
	12188	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_mf8_x4)))
	12189	svmfloat8x4_t svldnt1_x4(svcount_t, mfloat8_t const *);
12142	12190	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_u64_x4)))
12143	12191	svuint64x4_t svldnt1_x4(svcount_t, uint64_t const *);
12144	12192	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_f64_x4)))
...	...	@@ -12163,6 +12211,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_u8_x2)
12163	12211	svuint8x2_t svldnt1_vnum_x2(svcount_t, uint8_t const *, int64_t);
12164	12212	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_s8_x2)))
12165	12213	svint8x2_t svldnt1_vnum_x2(svcount_t, int8_t const *, int64_t);
	12214	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_mf8_x2)))
	12215	svmfloat8x2_t svldnt1_vnum_x2(svcount_t, mfloat8_t const *, int64_t);
12166	12216	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_u64_x2)))
12167	12217	svuint64x2_t svldnt1_vnum_x2(svcount_t, uint64_t const *, int64_t);
12168	12218	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_f64_x2)))
...	...	@@ -12187,6 +12237,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_u8_x4)
12187	12237	svuint8x4_t svldnt1_vnum_x4(svcount_t, uint8_t const *, int64_t);
12188	12238	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_s8_x4)))
12189	12239	svint8x4_t svldnt1_vnum_x4(svcount_t, int8_t const *, int64_t);
	12240	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_mf8_x4)))
	12241	svmfloat8x4_t svldnt1_vnum_x4(svcount_t, mfloat8_t const *, int64_t);
12190	12242	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_u64_x4)))
12191	12243	svuint64x4_t svldnt1_vnum_x4(svcount_t, uint64_t const *, int64_t);
12192	12244	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_f64_x4)))
...	...	@@ -12207,30 +12259,16 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_f32_x4
12207	12259	svfloat32x4_t svldnt1_vnum_x4(svcount_t, float32_t const *, int64_t);
12208	12260	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_s32_x4)))
12209	12261	svint32x4_t svldnt1_vnum_x4(svcount_t, int32_t const *, int64_t);
12210		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svqcvtn_s16_s32_x2)))
12211		svint16_t svqcvtn_s16(svint32x2_t);
12212		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svqcvtn_u16_s32_x2)))
12213		svuint16_t svqcvtn_u16(svint32x2_t);
12214		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svqcvtn_u16_u32_x2)))
12215		svuint16_t svqcvtn_u16(svuint32x2_t);
12216		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svqrshrn_n_s16_s32_x2)))
12217		svint16_t svqrshrn_s16(svint32x2_t, uint64_t);
12218		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svqrshrn_n_u16_u32_x2)))
12219		svuint16_t svqrshrn_u16(svuint32x2_t, uint64_t);
12220		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svqrshrun_n_u16_s32_x2)))
12221		svuint16_t svqrshrun_u16(svint32x2_t, uint64_t);
12222	12262	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svreinterpret_b)))
12223	12263	svbool_t svreinterpret(svcount_t);
12224	12264	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svreinterpret_c)))
12225	12265	svcount_t svreinterpret(svbool_t);
12226		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset2_b)))
12227		svboolx2_t svset2(svboolx2_t, uint64_t, svbool_t);
12228		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset4_b)))
12229		svboolx4_t svset4(svboolx4_t, uint64_t, svbool_t);
12230	12266	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_u8_x2)))
12231	12267	void svst1(svcount_t, uint8_t *, svuint8x2_t);
12232	12268	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_s8_x2)))
12233	12269	void svst1(svcount_t, int8_t *, svint8x2_t);
	12270	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_mf8_x2)))
	12271	void svst1(svcount_t, mfloat8_t *, svmfloat8x2_t);
12234	12272	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_u64_x2)))
12235	12273	void svst1(svcount_t, uint64_t *, svuint64x2_t);
12236	12274	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_f64_x2)))
...	...	@@ -12255,6 +12293,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_u8_x4)))
12255	12293	void svst1(svcount_t, uint8_t *, svuint8x4_t);
12256	12294	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_s8_x4)))
12257	12295	void svst1(svcount_t, int8_t *, svint8x4_t);
	12296	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_mf8_x4)))
	12297	void svst1(svcount_t, mfloat8_t *, svmfloat8x4_t);
12258	12298	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_u64_x4)))
12259	12299	void svst1(svcount_t, uint64_t *, svuint64x4_t);
12260	12300	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_f64_x4)))
...	...	@@ -12279,6 +12319,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_u8_x2)))
12279	12319	void svst1_vnum(svcount_t, uint8_t *, int64_t, svuint8x2_t);
12280	12320	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_s8_x2)))
12281	12321	void svst1_vnum(svcount_t, int8_t *, int64_t, svint8x2_t);
	12322	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_mf8_x2)))
	12323	void svst1_vnum(svcount_t, mfloat8_t *, int64_t, svmfloat8x2_t);
12282	12324	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_u64_x2)))
12283	12325	void svst1_vnum(svcount_t, uint64_t *, int64_t, svuint64x2_t);
12284	12326	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_f64_x2)))
...	...	@@ -12303,6 +12345,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_u8_x4)))
12303	12345	void svst1_vnum(svcount_t, uint8_t *, int64_t, svuint8x4_t);
12304	12346	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_s8_x4)))
12305	12347	void svst1_vnum(svcount_t, int8_t *, int64_t, svint8x4_t);
	12348	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_mf8_x4)))
	12349	void svst1_vnum(svcount_t, mfloat8_t *, int64_t, svmfloat8x4_t);
12306	12350	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_u64_x4)))
12307	12351	void svst1_vnum(svcount_t, uint64_t *, int64_t, svuint64x4_t);
12308	12352	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_f64_x4)))
...	...	@@ -12327,6 +12371,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_u8_x2)))
12327	12371	void svstnt1(svcount_t, uint8_t *, svuint8x2_t);
12328	12372	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_s8_x2)))
12329	12373	void svstnt1(svcount_t, int8_t *, svint8x2_t);
	12374	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_mf8_x2)))
	12375	void svstnt1(svcount_t, mfloat8_t *, svmfloat8x2_t);
12330	12376	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_u64_x2)))
12331	12377	void svstnt1(svcount_t, uint64_t *, svuint64x2_t);
12332	12378	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_f64_x2)))
...	...	@@ -12351,6 +12397,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_u8_x4)))
12351	12397	void svstnt1(svcount_t, uint8_t *, svuint8x4_t);
12352	12398	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_s8_x4)))
12353	12399	void svstnt1(svcount_t, int8_t *, svint8x4_t);
	12400	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_mf8_x4)))
	12401	void svstnt1(svcount_t, mfloat8_t *, svmfloat8x4_t);
12354	12402	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_u64_x4)))
12355	12403	void svstnt1(svcount_t, uint64_t *, svuint64x4_t);
12356	12404	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_f64_x4)))
...	...	@@ -12375,6 +12423,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_u8_x2)
12375	12423	void svstnt1_vnum(svcount_t, uint8_t *, int64_t, svuint8x2_t);
12376	12424	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_s8_x2)))
12377	12425	void svstnt1_vnum(svcount_t, int8_t *, int64_t, svint8x2_t);
	12426	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_mf8_x2)))
	12427	void svstnt1_vnum(svcount_t, mfloat8_t *, int64_t, svmfloat8x2_t);
12378	12428	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_u64_x2)))
12379	12429	void svstnt1_vnum(svcount_t, uint64_t *, int64_t, svuint64x2_t);
12380	12430	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_f64_x2)))
...	...	@@ -12399,6 +12449,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_u8_x4)
12399	12449	void svstnt1_vnum(svcount_t, uint8_t *, int64_t, svuint8x4_t);
12400	12450	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_s8_x4)))
12401	12451	void svstnt1_vnum(svcount_t, int8_t *, int64_t, svint8x4_t);
	12452	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_mf8_x4)))
	12453	void svstnt1_vnum(svcount_t, mfloat8_t *, int64_t, svmfloat8x4_t);
12402	12454	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_u64_x4)))
12403	12455	void svstnt1_vnum(svcount_t, uint64_t *, int64_t, svuint64x4_t);
12404	12456	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_f64_x4)))
...	...	@@ -12435,22 +12487,6 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_c64_u64))
12435	12487	svcount_t svwhilege_c64(uint64_t, uint64_t, uint64_t);
12436	12488	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_c16_u64)))
12437	12489	svcount_t svwhilege_c16(uint64_t, uint64_t, uint64_t);
12438		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b8_s64_x2)))
12439		svboolx2_t svwhilege_b8_x2(int64_t, int64_t);
12440		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b32_s64_x2)))
12441		svboolx2_t svwhilege_b32_x2(int64_t, int64_t);
12442		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b64_s64_x2)))
12443		svboolx2_t svwhilege_b64_x2(int64_t, int64_t);
12444		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b16_s64_x2)))
12445		svboolx2_t svwhilege_b16_x2(int64_t, int64_t);
12446		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b8_u64_x2)))
12447		svboolx2_t svwhilege_b8_x2(uint64_t, uint64_t);
12448		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b32_u64_x2)))
12449		svboolx2_t svwhilege_b32_x2(uint64_t, uint64_t);
12450		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b64_u64_x2)))
12451		svboolx2_t svwhilege_b64_x2(uint64_t, uint64_t);
12452		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilege_b16_u64_x2)))
12453		svboolx2_t svwhilege_b16_x2(uint64_t, uint64_t);
12454	12490	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_c8_s64)))
12455	12491	svcount_t svwhilegt_c8(int64_t, int64_t, uint64_t);
12456	12492	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_c32_s64)))
...	...	@@ -12467,22 +12503,6 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_c64_u64))
12467	12503	svcount_t svwhilegt_c64(uint64_t, uint64_t, uint64_t);
12468	12504	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_c16_u64)))
12469	12505	svcount_t svwhilegt_c16(uint64_t, uint64_t, uint64_t);
12470		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b8_s64_x2)))
12471		svboolx2_t svwhilegt_b8_x2(int64_t, int64_t);
12472		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b32_s64_x2)))
12473		svboolx2_t svwhilegt_b32_x2(int64_t, int64_t);
12474		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b64_s64_x2)))
12475		svboolx2_t svwhilegt_b64_x2(int64_t, int64_t);
12476		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b16_s64_x2)))
12477		svboolx2_t svwhilegt_b16_x2(int64_t, int64_t);
12478		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b8_u64_x2)))
12479		svboolx2_t svwhilegt_b8_x2(uint64_t, uint64_t);
12480		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b32_u64_x2)))
12481		svboolx2_t svwhilegt_b32_x2(uint64_t, uint64_t);
12482		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b64_u64_x2)))
12483		svboolx2_t svwhilegt_b64_x2(uint64_t, uint64_t);
12484		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilegt_b16_u64_x2)))
12485		svboolx2_t svwhilegt_b16_x2(uint64_t, uint64_t);
12486	12506	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_c8_s64)))
12487	12507	svcount_t svwhilele_c8(int64_t, int64_t, uint64_t);
12488	12508	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_c32_s64)))
...	...	@@ -12499,22 +12519,6 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_c64_u64))
12499	12519	svcount_t svwhilele_c64(uint64_t, uint64_t, uint64_t);
12500	12520	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_c16_u64)))
12501	12521	svcount_t svwhilele_c16(uint64_t, uint64_t, uint64_t);
12502		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b8_s64_x2)))
12503		svboolx2_t svwhilele_b8_x2(int64_t, int64_t);
12504		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b32_s64_x2)))
12505		svboolx2_t svwhilele_b32_x2(int64_t, int64_t);
12506		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b64_s64_x2)))
12507		svboolx2_t svwhilele_b64_x2(int64_t, int64_t);
12508		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b16_s64_x2)))
12509		svboolx2_t svwhilele_b16_x2(int64_t, int64_t);
12510		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b8_u64_x2)))
12511		svboolx2_t svwhilele_b8_x2(uint64_t, uint64_t);
12512		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b32_u64_x2)))
12513		svboolx2_t svwhilele_b32_x2(uint64_t, uint64_t);
12514		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b64_u64_x2)))
12515		svboolx2_t svwhilele_b64_x2(uint64_t, uint64_t);
12516		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilele_b16_u64_x2)))
12517		svboolx2_t svwhilele_b16_x2(uint64_t, uint64_t);
12518	12522	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_c8_u64)))
12519	12523	svcount_t svwhilelt_c8(uint64_t, uint64_t, uint64_t);
12520	12524	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_c32_u64)))
...	...	@@ -12531,70 +12535,6 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_c64_s64))
12531	12535	svcount_t svwhilelt_c64(int64_t, int64_t, uint64_t);
12532	12536	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_c16_s64)))
12533	12537	svcount_t svwhilelt_c16(int64_t, int64_t, uint64_t);
12534		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b8_u64_x2)))
12535		svboolx2_t svwhilelt_b8_x2(uint64_t, uint64_t);
12536		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b32_u64_x2)))
12537		svboolx2_t svwhilelt_b32_x2(uint64_t, uint64_t);
12538		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b64_u64_x2)))
12539		svboolx2_t svwhilelt_b64_x2(uint64_t, uint64_t);
12540		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b16_u64_x2)))
12541		svboolx2_t svwhilelt_b16_x2(uint64_t, uint64_t);
12542		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b8_s64_x2)))
12543		svboolx2_t svwhilelt_b8_x2(int64_t, int64_t);
12544		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b32_s64_x2)))
12545		svboolx2_t svwhilelt_b32_x2(int64_t, int64_t);
12546		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b64_s64_x2)))
12547		svboolx2_t svwhilelt_b64_x2(int64_t, int64_t);
12548		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilelt_b16_s64_x2)))
12549		svboolx2_t svwhilelt_b16_x2(int64_t, int64_t);
12550		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_u8)))
12551		svuint8_t svdup_laneq_u8(svuint8_t, uint64_t);
12552		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_s8)))
12553		svint8_t svdup_laneq_s8(svint8_t, uint64_t);
12554		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_mf8)))
12555		svmfloat8_t svdup_laneq_mf8(svmfloat8_t, uint64_t);
12556		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_u64)))
12557		svuint64_t svdup_laneq_u64(svuint64_t, uint64_t);
12558		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_f64)))
12559		svfloat64_t svdup_laneq_f64(svfloat64_t, uint64_t);
12560		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_s64)))
12561		svint64_t svdup_laneq_s64(svint64_t, uint64_t);
12562		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_u16)))
12563		svuint16_t svdup_laneq_u16(svuint16_t, uint64_t);
12564		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_f16)))
12565		svfloat16_t svdup_laneq_f16(svfloat16_t, uint64_t);
12566		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_s16)))
12567		svint16_t svdup_laneq_s16(svint16_t, uint64_t);
12568		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_u32)))
12569		svuint32_t svdup_laneq_u32(svuint32_t, uint64_t);
12570		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_f32)))
12571		svfloat32_t svdup_laneq_f32(svfloat32_t, uint64_t);
12572		__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_s32)))
12573		svint32_t svdup_laneq_s32(svint32_t, uint64_t);
12574		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_u8)))
12575		svuint8_t svdup_laneq(svuint8_t, uint64_t);
12576		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_s8)))
12577		svint8_t svdup_laneq(svint8_t, uint64_t);
12578		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_mf8)))
12579		svmfloat8_t svdup_laneq(svmfloat8_t, uint64_t);
12580		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_u64)))
12581		svuint64_t svdup_laneq(svuint64_t, uint64_t);
12582		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_f64)))
12583		svfloat64_t svdup_laneq(svfloat64_t, uint64_t);
12584		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_s64)))
12585		svint64_t svdup_laneq(svint64_t, uint64_t);
12586		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_u16)))
12587		svuint16_t svdup_laneq(svuint16_t, uint64_t);
12588		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_f16)))
12589		svfloat16_t svdup_laneq(svfloat16_t, uint64_t);
12590		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_s16)))
12591		svint16_t svdup_laneq(svint16_t, uint64_t);
12592		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_u32)))
12593		svuint32_t svdup_laneq(svuint32_t, uint64_t);
12594		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_f32)))
12595		svfloat32_t svdup_laneq(svfloat32_t, uint64_t);
12596		__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_laneq_s32)))
12597		svint32_t svdup_laneq(svint32_t, uint64_t);
12598	12538	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaba_n_s8)))
12599	12539	svint8_t svaba_n_s8(svint8_t, svint8_t, int8_t);
12600	12540	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svaba_n_s32)))
...	...	@@ -15641,6 +15581,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbl2_u64)))
15641	15581	svuint64_t svtbl2_u64(svuint64x2_t, svuint64_t);
15642	15582	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbl2_u16)))
15643	15583	svuint16_t svtbl2_u16(svuint16x2_t, svuint16_t);
	15584	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbl2_bf16)))
	15585	svbfloat16_t svtbl2_bf16(svbfloat16x2_t, svuint16_t);
15644	15586	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbl2_s8)))
15645	15587	svint8_t svtbl2_s8(svint8x2_t, svuint8_t);
15646	15588	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbl2_f64)))
...	...	@@ -15663,6 +15605,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbx_u64)))
15663	15605	svuint64_t svtbx_u64(svuint64_t, svuint64_t, svuint64_t);
15664	15606	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbx_u16)))
15665	15607	svuint16_t svtbx_u16(svuint16_t, svuint16_t, svuint16_t);
	15608	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbx_bf16)))
	15609	svbfloat16_t svtbx_bf16(svbfloat16_t, svbfloat16_t, svuint16_t);
15666	15610	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbx_s8)))
15667	15611	svint8_t svtbx_s8(svint8_t, svint8_t, svuint8_t);
15668	15612	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbx_f64)))
...	...	@@ -15801,6 +15745,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilerw_s64)))
15801	15745	svbool_t svwhilerw_s64(int64_t const , int64_t const );
15802	15746	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilerw_u16)))
15803	15747	svbool_t svwhilerw_u16(uint16_t const , uint16_t const );
	15748	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilerw_bf16)))
	15749	svbool_t svwhilerw_bf16(bfloat16_t const , bfloat16_t const );
15804	15750	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilerw_f16)))
15805	15751	svbool_t svwhilerw_f16(float16_t const , float16_t const );
15806	15752	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilerw_s16)))
...	...	@@ -15823,6 +15769,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilewr_s64)))
15823	15769	svbool_t svwhilewr_s64(int64_t const , int64_t const );
15824	15770	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilewr_u16)))
15825	15771	svbool_t svwhilewr_u16(uint16_t const , uint16_t const );
	15772	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilewr_bf16)))
	15773	svbool_t svwhilewr_bf16(bfloat16_t const , bfloat16_t const );
15826	15774	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilewr_f16)))
15827	15775	svbool_t svwhilewr_f16(float16_t const , float16_t const );
15828	15776	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilewr_s16)))
...	...	@@ -18895,6 +18843,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbl2_u64)))
18895	18843	svuint64_t svtbl2(svuint64x2_t, svuint64_t);
18896	18844	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbl2_u16)))
18897	18845	svuint16_t svtbl2(svuint16x2_t, svuint16_t);
	18846	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbl2_bf16)))
	18847	svbfloat16_t svtbl2(svbfloat16x2_t, svuint16_t);
18898	18848	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbl2_s8)))
18899	18849	svint8_t svtbl2(svint8x2_t, svuint8_t);
18900	18850	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbl2_f64)))
...	...	@@ -18917,6 +18867,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbx_u64)))
18917	18867	svuint64_t svtbx(svuint64_t, svuint64_t, svuint64_t);
18918	18868	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbx_u16)))
18919	18869	svuint16_t svtbx(svuint16_t, svuint16_t, svuint16_t);
	18870	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbx_bf16)))
	18871	svbfloat16_t svtbx(svbfloat16_t, svbfloat16_t, svuint16_t);
18920	18872	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbx_s8)))
18921	18873	svint8_t svtbx(svint8_t, svint8_t, svuint8_t);
18922	18874	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbx_f64)))
...	...	@@ -19055,6 +19007,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilerw_s64)))
19055	19007	svbool_t svwhilerw(int64_t const , int64_t const );
19056	19008	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilerw_u16)))
19057	19009	svbool_t svwhilerw(uint16_t const , uint16_t const );
	19010	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilerw_bf16)))
	19011	svbool_t svwhilerw(bfloat16_t const , bfloat16_t const );
19058	19012	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilerw_f16)))
19059	19013	svbool_t svwhilerw(float16_t const , float16_t const );
19060	19014	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilerw_s16)))
...	...	@@ -19077,6 +19031,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilewr_s64)))
19077	19031	svbool_t svwhilewr(int64_t const , int64_t const );
19078	19032	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilewr_u16)))
19079	19033	svbool_t svwhilewr(uint16_t const , uint16_t const );
	19034	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilewr_bf16)))
	19035	svbool_t svwhilewr(bfloat16_t const , bfloat16_t const );
19080	19036	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilewr_f16)))
19081	19037	svbool_t svwhilewr(float16_t const , float16_t const );
19082	19038	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svwhilewr_s16)))
...	...	@@ -19839,6 +19795,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclasta_n_u64)))
19839	19795	uint64_t svclasta_n_u64(svbool_t, uint64_t, svuint64_t);
19840	19796	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclasta_n_u16)))
19841	19797	uint16_t svclasta_n_u16(svbool_t, uint16_t, svuint16_t);
	19798	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclasta_n_bf16)))
	19799	bfloat16_t svclasta_n_bf16(svbool_t, bfloat16_t, svbfloat16_t);
19842	19800	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclasta_n_s8)))
19843	19801	int8_t svclasta_n_s8(svbool_t, int8_t, svint8_t);
19844	19802	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclasta_n_f64)))
...	...	@@ -19861,6 +19819,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclasta_u64)))
19861	19819	svuint64_t svclasta_u64(svbool_t, svuint64_t, svuint64_t);
19862	19820	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclasta_u16)))
19863	19821	svuint16_t svclasta_u16(svbool_t, svuint16_t, svuint16_t);
	19822	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclasta_bf16)))
	19823	svbfloat16_t svclasta_bf16(svbool_t, svbfloat16_t, svbfloat16_t);
19864	19824	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclasta_s8)))
19865	19825	svint8_t svclasta_s8(svbool_t, svint8_t, svint8_t);
19866	19826	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclasta_f64)))
...	...	@@ -19883,6 +19843,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclastb_n_u64)))
19883	19843	uint64_t svclastb_n_u64(svbool_t, uint64_t, svuint64_t);
19884	19844	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclastb_n_u16)))
19885	19845	uint16_t svclastb_n_u16(svbool_t, uint16_t, svuint16_t);
	19846	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclastb_n_bf16)))
	19847	bfloat16_t svclastb_n_bf16(svbool_t, bfloat16_t, svbfloat16_t);
19886	19848	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclastb_n_s8)))
19887	19849	int8_t svclastb_n_s8(svbool_t, int8_t, svint8_t);
19888	19850	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclastb_n_f64)))
...	...	@@ -19905,6 +19867,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclastb_u64)))
19905	19867	svuint64_t svclastb_u64(svbool_t, svuint64_t, svuint64_t);
19906	19868	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclastb_u16)))
19907	19869	svuint16_t svclastb_u16(svbool_t, svuint16_t, svuint16_t);
	19870	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclastb_bf16)))
	19871	svbfloat16_t svclastb_bf16(svbool_t, svbfloat16_t, svbfloat16_t);
19908	19872	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclastb_s8)))
19909	19873	svint8_t svclastb_s8(svbool_t, svint8_t, svint8_t);
19910	19874	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclastb_f64)))
...	...	@@ -20465,6 +20429,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_u64_m)))
20465	20429	svuint64_t svcnt_u64_m(svuint64_t, svbool_t, svuint64_t);
20466	20430	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_u16_m)))
20467	20431	svuint16_t svcnt_u16_m(svuint16_t, svbool_t, svuint16_t);
	20432	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_bf16_m)))
	20433	svuint16_t svcnt_bf16_m(svuint16_t, svbool_t, svbfloat16_t);
20468	20434	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_s8_m)))
20469	20435	svuint8_t svcnt_s8_m(svuint8_t, svbool_t, svint8_t);
20470	20436	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_f64_m)))
...	...	@@ -20487,6 +20453,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_u64_x)))
20487	20453	svuint64_t svcnt_u64_x(svbool_t, svuint64_t);
20488	20454	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_u16_x)))
20489	20455	svuint16_t svcnt_u16_x(svbool_t, svuint16_t);
	20456	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_bf16_x)))
	20457	svuint16_t svcnt_bf16_x(svbool_t, svbfloat16_t);
20490	20458	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_s8_x)))
20491	20459	svuint8_t svcnt_s8_x(svbool_t, svint8_t);
20492	20460	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_f64_x)))
...	...	@@ -20509,6 +20477,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_u64_z)))
20509	20477	svuint64_t svcnt_u64_z(svbool_t, svuint64_t);
20510	20478	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_u16_z)))
20511	20479	svuint16_t svcnt_u16_z(svbool_t, svuint16_t);
	20480	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_bf16_z)))
	20481	svuint16_t svcnt_bf16_z(svbool_t, svbfloat16_t);
20512	20482	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_s8_z)))
20513	20483	svuint8_t svcnt_s8_z(svbool_t, svint8_t);
20514	20484	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_f64_z)))
...	...	@@ -20555,6 +20525,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate2_u64)))
20555	20525	svuint64x2_t svcreate2_u64(svuint64_t, svuint64_t);
20556	20526	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate2_u16)))
20557	20527	svuint16x2_t svcreate2_u16(svuint16_t, svuint16_t);
	20528	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate2_bf16)))
	20529	svbfloat16x2_t svcreate2_bf16(svbfloat16_t, svbfloat16_t);
20558	20530	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate2_s8)))
20559	20531	svint8x2_t svcreate2_s8(svint8_t, svint8_t);
20560	20532	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate2_f64)))
...	...	@@ -20567,6 +20539,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate2_s32)))
20567	20539	svint32x2_t svcreate2_s32(svint32_t, svint32_t);
20568	20540	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate2_s64)))
20569	20541	svint64x2_t svcreate2_s64(svint64_t, svint64_t);
	20542	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate2_mf8)))
	20543	svmfloat8x2_t svcreate2_mf8(svmfloat8_t, svmfloat8_t);
20570	20544	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate2_s16)))
20571	20545	svint16x2_t svcreate2_s16(svint16_t, svint16_t);
20572	20546	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate3_u8)))
...	...	@@ -20577,6 +20551,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate3_u64)))
20577	20551	svuint64x3_t svcreate3_u64(svuint64_t, svuint64_t, svuint64_t);
20578	20552	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate3_u16)))
20579	20553	svuint16x3_t svcreate3_u16(svuint16_t, svuint16_t, svuint16_t);
	20554	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate3_bf16)))
	20555	svbfloat16x3_t svcreate3_bf16(svbfloat16_t, svbfloat16_t, svbfloat16_t);
20580	20556	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate3_s8)))
20581	20557	svint8x3_t svcreate3_s8(svint8_t, svint8_t, svint8_t);
20582	20558	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate3_f64)))
...	...	@@ -20589,6 +20565,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate3_s32)))
20589	20565	svint32x3_t svcreate3_s32(svint32_t, svint32_t, svint32_t);
20590	20566	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate3_s64)))
20591	20567	svint64x3_t svcreate3_s64(svint64_t, svint64_t, svint64_t);
	20568	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate3_mf8)))
	20569	svmfloat8x3_t svcreate3_mf8(svmfloat8_t, svmfloat8_t, svmfloat8_t);
20592	20570	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate3_s16)))
20593	20571	svint16x3_t svcreate3_s16(svint16_t, svint16_t, svint16_t);
20594	20572	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate4_u8)))
...	...	@@ -20599,6 +20577,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate4_u64)))
20599	20577	svuint64x4_t svcreate4_u64(svuint64_t, svuint64_t, svuint64_t, svuint64_t);
20600	20578	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate4_u16)))
20601	20579	svuint16x4_t svcreate4_u16(svuint16_t, svuint16_t, svuint16_t, svuint16_t);
	20580	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate4_bf16)))
	20581	svbfloat16x4_t svcreate4_bf16(svbfloat16_t, svbfloat16_t, svbfloat16_t, svbfloat16_t);
20602	20582	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate4_s8)))
20603	20583	svint8x4_t svcreate4_s8(svint8_t, svint8_t, svint8_t, svint8_t);
20604	20584	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate4_f64)))
...	...	@@ -20611,6 +20591,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate4_s32)))
20611	20591	svint32x4_t svcreate4_s32(svint32_t, svint32_t, svint32_t, svint32_t);
20612	20592	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate4_s64)))
20613	20593	svint64x4_t svcreate4_s64(svint64_t, svint64_t, svint64_t, svint64_t);
	20594	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate4_mf8)))
	20595	svmfloat8x4_t svcreate4_mf8(svmfloat8_t, svmfloat8_t, svmfloat8_t, svmfloat8_t);
20614	20596	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate4_s16)))
20615	20597	svint16x4_t svcreate4_s16(svint16_t, svint16_t, svint16_t, svint16_t);
20616	20598	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvt_f16_f32_m)))
...	...	@@ -21017,6 +20999,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_u64)))
21017	20999	svuint64_t svdup_n_u64(uint64_t);
21018	21000	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_u16)))
21019	21001	svuint16_t svdup_n_u16(uint16_t);
	21002	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_bf16)))
	21003	svbfloat16_t svdup_n_bf16(bfloat16_t);
21020	21004	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_s8)))
21021	21005	svint8_t svdup_n_s8(int8_t);
21022	21006	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_f64)))
...	...	@@ -21039,6 +21023,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_u64_m)))
21039	21023	svuint64_t svdup_n_u64_m(svuint64_t, svbool_t, uint64_t);
21040	21024	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_u16_m)))
21041	21025	svuint16_t svdup_n_u16_m(svuint16_t, svbool_t, uint16_t);
	21026	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_bf16_m)))
	21027	svbfloat16_t svdup_n_bf16_m(svbfloat16_t, svbool_t, bfloat16_t);
21042	21028	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_s8_m)))
21043	21029	svint8_t svdup_n_s8_m(svint8_t, svbool_t, int8_t);
21044	21030	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_f64_m)))
...	...	@@ -21069,6 +21055,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_u64_x)))
21069	21055	svuint64_t svdup_n_u64_x(svbool_t, uint64_t);
21070	21056	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_u16_x)))
21071	21057	svuint16_t svdup_n_u16_x(svbool_t, uint16_t);
	21058	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_bf16_x)))
	21059	svbfloat16_t svdup_n_bf16_x(svbool_t, bfloat16_t);
21072	21060	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_s8_x)))
21073	21061	svint8_t svdup_n_s8_x(svbool_t, int8_t);
21074	21062	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_f64_x)))
...	...	@@ -21091,6 +21079,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_u64_z)))
21091	21079	svuint64_t svdup_n_u64_z(svbool_t, uint64_t);
21092	21080	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_u16_z)))
21093	21081	svuint16_t svdup_n_u16_z(svbool_t, uint16_t);
	21082	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_bf16_z)))
	21083	svbfloat16_t svdup_n_bf16_z(svbool_t, bfloat16_t);
21094	21084	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_s8_z)))
21095	21085	svint8_t svdup_n_s8_z(svbool_t, int8_t);
21096	21086	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_f64_z)))
...	...	@@ -21113,6 +21103,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_lane_u64)))
21113	21103	svuint64_t svdup_lane_u64(svuint64_t, uint64_t);
21114	21104	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_lane_u16)))
21115	21105	svuint16_t svdup_lane_u16(svuint16_t, uint16_t);
	21106	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_lane_bf16)))
	21107	svbfloat16_t svdup_lane_bf16(svbfloat16_t, uint16_t);
21116	21108	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_lane_s8)))
21117	21109	svint8_t svdup_lane_s8(svint8_t, uint8_t);
21118	21110	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_lane_f64)))
...	...	@@ -21133,6 +21125,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdupq_n_s8)))
21133	21125	svint8_t svdupq_n_s8(int8_t, int8_t, int8_t, int8_t, int8_t, int8_t, int8_t, int8_t, int8_t, int8_t, int8_t, int8_t, int8_t, int8_t, int8_t, int8_t);
21134	21126	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdupq_n_u16)))
21135	21127	svuint16_t svdupq_n_u16(uint16_t, uint16_t, uint16_t, uint16_t, uint16_t, uint16_t, uint16_t, uint16_t);
	21128	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdupq_n_bf16)))
	21129	svbfloat16_t svdupq_n_bf16(bfloat16_t, bfloat16_t, bfloat16_t, bfloat16_t, bfloat16_t, bfloat16_t, bfloat16_t, bfloat16_t);
21136	21130	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdupq_n_f16)))
21137	21131	svfloat16_t svdupq_n_f16(float16_t, float16_t, float16_t, float16_t, float16_t, float16_t, float16_t, float16_t);
21138	21132	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdupq_n_s16)))
...	...	@@ -21165,6 +21159,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdupq_lane_u64)))
21165	21159	svuint64_t svdupq_lane_u64(svuint64_t, uint64_t);
21166	21160	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdupq_lane_u16)))
21167	21161	svuint16_t svdupq_lane_u16(svuint16_t, uint64_t);
	21162	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdupq_lane_bf16)))
	21163	svbfloat16_t svdupq_lane_bf16(svbfloat16_t, uint64_t);
21168	21164	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdupq_lane_s8)))
21169	21165	svint8_t svdupq_lane_s8(svint8_t, uint64_t);
21170	21166	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdupq_lane_f64)))
...	...	@@ -21301,6 +21297,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svext_u64)))
21301	21297	svuint64_t svext_u64(svuint64_t, svuint64_t, uint64_t);
21302	21298	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svext_u16)))
21303	21299	svuint16_t svext_u16(svuint16_t, svuint16_t, uint64_t);
	21300	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svext_bf16)))
	21301	svbfloat16_t svext_bf16(svbfloat16_t, svbfloat16_t, uint64_t);
21304	21302	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svext_s8)))
21305	21303	svint8_t svext_s8(svint8_t, svint8_t, uint64_t);
21306	21304	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svext_f64)))
...	...	@@ -21395,6 +21393,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget2_u64)))
21395	21393	svuint64_t svget2_u64(svuint64x2_t, uint64_t);
21396	21394	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget2_u16)))
21397	21395	svuint16_t svget2_u16(svuint16x2_t, uint64_t);
	21396	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget2_bf16)))
	21397	svbfloat16_t svget2_bf16(svbfloat16x2_t, uint64_t);
21398	21398	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget2_s8)))
21399	21399	svint8_t svget2_s8(svint8x2_t, uint64_t);
21400	21400	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget2_f64)))
...	...	@@ -21407,6 +21407,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget2_s32)))
21407	21407	svint32_t svget2_s32(svint32x2_t, uint64_t);
21408	21408	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget2_s64)))
21409	21409	svint64_t svget2_s64(svint64x2_t, uint64_t);
	21410	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget2_mf8)))
	21411	svmfloat8_t svget2_mf8(svmfloat8x2_t, uint64_t);
21410	21412	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget2_s16)))
21411	21413	svint16_t svget2_s16(svint16x2_t, uint64_t);
21412	21414	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget3_u8)))
...	...	@@ -21417,6 +21419,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget3_u64)))
21417	21419	svuint64_t svget3_u64(svuint64x3_t, uint64_t);
21418	21420	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget3_u16)))
21419	21421	svuint16_t svget3_u16(svuint16x3_t, uint64_t);
	21422	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget3_bf16)))
	21423	svbfloat16_t svget3_bf16(svbfloat16x3_t, uint64_t);
21420	21424	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget3_s8)))
21421	21425	svint8_t svget3_s8(svint8x3_t, uint64_t);
21422	21426	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget3_f64)))
...	...	@@ -21429,6 +21433,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget3_s32)))
21429	21433	svint32_t svget3_s32(svint32x3_t, uint64_t);
21430	21434	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget3_s64)))
21431	21435	svint64_t svget3_s64(svint64x3_t, uint64_t);
	21436	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget3_mf8)))
	21437	svmfloat8_t svget3_mf8(svmfloat8x3_t, uint64_t);
21432	21438	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget3_s16)))
21433	21439	svint16_t svget3_s16(svint16x3_t, uint64_t);
21434	21440	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget4_u8)))
...	...	@@ -21439,6 +21445,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget4_u64)))
21439	21445	svuint64_t svget4_u64(svuint64x4_t, uint64_t);
21440	21446	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget4_u16)))
21441	21447	svuint16_t svget4_u16(svuint16x4_t, uint64_t);
	21448	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget4_bf16)))
	21449	svbfloat16_t svget4_bf16(svbfloat16x4_t, uint64_t);
21442	21450	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget4_s8)))
21443	21451	svint8_t svget4_s8(svint8x4_t, uint64_t);
21444	21452	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget4_f64)))
...	...	@@ -21451,6 +21459,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget4_s32)))
21451	21459	svint32_t svget4_s32(svint32x4_t, uint64_t);
21452	21460	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget4_s64)))
21453	21461	svint64_t svget4_s64(svint64x4_t, uint64_t);
	21462	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget4_mf8)))
	21463	svmfloat8_t svget4_mf8(svmfloat8x4_t, uint64_t);
21454	21464	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget4_s16)))
21455	21465	svint16_t svget4_s16(svint16x4_t, uint64_t);
21456	21466	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svindex_u8)))
...	...	@@ -21477,6 +21487,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svinsr_n_u64)))
21477	21487	svuint64_t svinsr_n_u64(svuint64_t, uint64_t);
21478	21488	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svinsr_n_u16)))
21479	21489	svuint16_t svinsr_n_u16(svuint16_t, uint16_t);
	21490	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svinsr_n_bf16)))
	21491	svbfloat16_t svinsr_n_bf16(svbfloat16_t, bfloat16_t);
21480	21492	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svinsr_n_s8)))
21481	21493	svint8_t svinsr_n_s8(svint8_t, int8_t);
21482	21494	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svinsr_n_f64)))
...	...	@@ -21499,6 +21511,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlasta_u64)))
21499	21511	uint64_t svlasta_u64(svbool_t, svuint64_t);
21500	21512	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlasta_u16)))
21501	21513	uint16_t svlasta_u16(svbool_t, svuint16_t);
	21514	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlasta_bf16)))
	21515	bfloat16_t svlasta_bf16(svbool_t, svbfloat16_t);
21502	21516	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlasta_s8)))
21503	21517	int8_t svlasta_s8(svbool_t, svint8_t);
21504	21518	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlasta_f64)))
...	...	@@ -21521,6 +21535,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlastb_u64)))
21521	21535	uint64_t svlastb_u64(svbool_t, svuint64_t);
21522	21536	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlastb_u16)))
21523	21537	uint16_t svlastb_u16(svbool_t, svuint16_t);
	21538	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlastb_bf16)))
	21539	bfloat16_t svlastb_bf16(svbool_t, svbfloat16_t);
21524	21540	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlastb_s8)))
21525	21541	int8_t svlastb_s8(svbool_t, svint8_t);
21526	21542	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlastb_f64)))
...	...	@@ -21543,6 +21559,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_u64)))
21543	21559	svuint64_t svld1_u64(svbool_t, uint64_t const *);
21544	21560	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_u16)))
21545	21561	svuint16_t svld1_u16(svbool_t, uint16_t const *);
	21562	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_bf16)))
	21563	svbfloat16_t svld1_bf16(svbool_t, bfloat16_t const *);
21546	21564	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_s8)))
21547	21565	svint8_t svld1_s8(svbool_t, int8_t const *);
21548	21566	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_f64)))
...	...	@@ -21555,6 +21573,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_s32)))
21555	21573	svint32_t svld1_s32(svbool_t, int32_t const *);
21556	21574	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_s64)))
21557	21575	svint64_t svld1_s64(svbool_t, int64_t const *);
	21576	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_mf8)))
	21577	svmfloat8_t svld1_mf8(svbool_t, mfloat8_t const *);
21558	21578	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_s16)))
21559	21579	svint16_t svld1_s16(svbool_t, int16_t const *);
21560	21580	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_u8)))
...	...	@@ -21565,6 +21585,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_u64)))
21565	21585	svuint64_t svld1_vnum_u64(svbool_t, uint64_t const *, int64_t);
21566	21586	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_u16)))
21567	21587	svuint16_t svld1_vnum_u16(svbool_t, uint16_t const *, int64_t);
	21588	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_bf16)))
	21589	svbfloat16_t svld1_vnum_bf16(svbool_t, bfloat16_t const *, int64_t);
21568	21590	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_s8)))
21569	21591	svint8_t svld1_vnum_s8(svbool_t, int8_t const *, int64_t);
21570	21592	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_f64)))
...	...	@@ -21577,6 +21599,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_s32)))
21577	21599	svint32_t svld1_vnum_s32(svbool_t, int32_t const *, int64_t);
21578	21600	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_s64)))
21579	21601	svint64_t svld1_vnum_s64(svbool_t, int64_t const *, int64_t);
	21602	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_mf8)))
	21603	svmfloat8_t svld1_vnum_mf8(svbool_t, mfloat8_t const *, int64_t);
21580	21604	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_s16)))
21581	21605	svint16_t svld1_vnum_s16(svbool_t, int16_t const *, int64_t);
21582	21606	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1rq_u8)))
...	...	@@ -21587,6 +21611,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1rq_u64)))
21587	21611	svuint64_t svld1rq_u64(svbool_t, uint64_t const *);
21588	21612	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1rq_u16)))
21589	21613	svuint16_t svld1rq_u16(svbool_t, uint16_t const *);
	21614	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1rq_bf16)))
	21615	svbfloat16_t svld1rq_bf16(svbool_t, bfloat16_t const *);
21590	21616	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1rq_s8)))
21591	21617	svint8_t svld1rq_s8(svbool_t, int8_t const *);
21592	21618	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1rq_f64)))
...	...	@@ -21599,6 +21625,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1rq_s32)))
21599	21625	svint32_t svld1rq_s32(svbool_t, int32_t const *);
21600	21626	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1rq_s64)))
21601	21627	svint64_t svld1rq_s64(svbool_t, int64_t const *);
	21628	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1rq_mf8)))
	21629	svmfloat8_t svld1rq_mf8(svbool_t, mfloat8_t const *);
21602	21630	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1rq_s16)))
21603	21631	svint16_t svld1rq_s16(svbool_t, int16_t const *);
21604	21632	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1sb_vnum_u32)))
...	...	@@ -21705,6 +21733,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_u64)))
21705	21733	svuint64x2_t svld2_u64(svbool_t, uint64_t const *);
21706	21734	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_u16)))
21707	21735	svuint16x2_t svld2_u16(svbool_t, uint16_t const *);
	21736	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_bf16)))
	21737	svbfloat16x2_t svld2_bf16(svbool_t, bfloat16_t const *);
21708	21738	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_s8)))
21709	21739	svint8x2_t svld2_s8(svbool_t, int8_t const *);
21710	21740	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_f64)))
...	...	@@ -21717,6 +21747,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_s32)))
21717	21747	svint32x2_t svld2_s32(svbool_t, int32_t const *);
21718	21748	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_s64)))
21719	21749	svint64x2_t svld2_s64(svbool_t, int64_t const *);
	21750	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_mf8)))
	21751	svmfloat8x2_t svld2_mf8(svbool_t, mfloat8_t const *);
21720	21752	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_s16)))
21721	21753	svint16x2_t svld2_s16(svbool_t, int16_t const *);
21722	21754	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_vnum_u8)))
...	...	@@ -21727,6 +21759,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_vnum_u64)))
21727	21759	svuint64x2_t svld2_vnum_u64(svbool_t, uint64_t const *, int64_t);
21728	21760	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_vnum_u16)))
21729	21761	svuint16x2_t svld2_vnum_u16(svbool_t, uint16_t const *, int64_t);
	21762	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_vnum_bf16)))
	21763	svbfloat16x2_t svld2_vnum_bf16(svbool_t, bfloat16_t const *, int64_t);
21730	21764	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_vnum_s8)))
21731	21765	svint8x2_t svld2_vnum_s8(svbool_t, int8_t const *, int64_t);
21732	21766	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_vnum_f64)))
...	...	@@ -21739,6 +21773,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_vnum_s32)))
21739	21773	svint32x2_t svld2_vnum_s32(svbool_t, int32_t const *, int64_t);
21740	21774	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_vnum_s64)))
21741	21775	svint64x2_t svld2_vnum_s64(svbool_t, int64_t const *, int64_t);
	21776	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_vnum_mf8)))
	21777	svmfloat8x2_t svld2_vnum_mf8(svbool_t, mfloat8_t const *, int64_t);
21742	21778	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_vnum_s16)))
21743	21779	svint16x2_t svld2_vnum_s16(svbool_t, int16_t const *, int64_t);
21744	21780	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_u8)))
...	...	@@ -21749,6 +21785,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_u64)))
21749	21785	svuint64x3_t svld3_u64(svbool_t, uint64_t const *);
21750	21786	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_u16)))
21751	21787	svuint16x3_t svld3_u16(svbool_t, uint16_t const *);
	21788	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_bf16)))
	21789	svbfloat16x3_t svld3_bf16(svbool_t, bfloat16_t const *);
21752	21790	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_s8)))
21753	21791	svint8x3_t svld3_s8(svbool_t, int8_t const *);
21754	21792	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_f64)))
...	...	@@ -21761,6 +21799,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_s32)))
21761	21799	svint32x3_t svld3_s32(svbool_t, int32_t const *);
21762	21800	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_s64)))
21763	21801	svint64x3_t svld3_s64(svbool_t, int64_t const *);
	21802	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_mf8)))
	21803	svmfloat8x3_t svld3_mf8(svbool_t, mfloat8_t const *);
21764	21804	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_s16)))
21765	21805	svint16x3_t svld3_s16(svbool_t, int16_t const *);
21766	21806	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_vnum_u8)))
...	...	@@ -21771,6 +21811,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_vnum_u64)))
21771	21811	svuint64x3_t svld3_vnum_u64(svbool_t, uint64_t const *, int64_t);
21772	21812	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_vnum_u16)))
21773	21813	svuint16x3_t svld3_vnum_u16(svbool_t, uint16_t const *, int64_t);
	21814	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_vnum_bf16)))
	21815	svbfloat16x3_t svld3_vnum_bf16(svbool_t, bfloat16_t const *, int64_t);
21774	21816	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_vnum_s8)))
21775	21817	svint8x3_t svld3_vnum_s8(svbool_t, int8_t const *, int64_t);
21776	21818	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_vnum_f64)))
...	...	@@ -21783,6 +21825,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_vnum_s32)))
21783	21825	svint32x3_t svld3_vnum_s32(svbool_t, int32_t const *, int64_t);
21784	21826	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_vnum_s64)))
21785	21827	svint64x3_t svld3_vnum_s64(svbool_t, int64_t const *, int64_t);
	21828	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_vnum_mf8)))
	21829	svmfloat8x3_t svld3_vnum_mf8(svbool_t, mfloat8_t const *, int64_t);
21786	21830	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_vnum_s16)))
21787	21831	svint16x3_t svld3_vnum_s16(svbool_t, int16_t const *, int64_t);
21788	21832	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_u8)))
...	...	@@ -21793,6 +21837,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_u64)))
21793	21837	svuint64x4_t svld4_u64(svbool_t, uint64_t const *);
21794	21838	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_u16)))
21795	21839	svuint16x4_t svld4_u16(svbool_t, uint16_t const *);
	21840	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_bf16)))
	21841	svbfloat16x4_t svld4_bf16(svbool_t, bfloat16_t const *);
21796	21842	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_s8)))
21797	21843	svint8x4_t svld4_s8(svbool_t, int8_t const *);
21798	21844	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_f64)))
...	...	@@ -21805,6 +21851,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_s32)))
21805	21851	svint32x4_t svld4_s32(svbool_t, int32_t const *);
21806	21852	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_s64)))
21807	21853	svint64x4_t svld4_s64(svbool_t, int64_t const *);
	21854	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_mf8)))
	21855	svmfloat8x4_t svld4_mf8(svbool_t, mfloat8_t const *);
21808	21856	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_s16)))
21809	21857	svint16x4_t svld4_s16(svbool_t, int16_t const *);
21810	21858	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_vnum_u8)))
...	...	@@ -21815,6 +21863,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_vnum_u64)))
21815	21863	svuint64x4_t svld4_vnum_u64(svbool_t, uint64_t const *, int64_t);
21816	21864	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_vnum_u16)))
21817	21865	svuint16x4_t svld4_vnum_u16(svbool_t, uint16_t const *, int64_t);
	21866	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_vnum_bf16)))
	21867	svbfloat16x4_t svld4_vnum_bf16(svbool_t, bfloat16_t const *, int64_t);
21818	21868	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_vnum_s8)))
21819	21869	svint8x4_t svld4_vnum_s8(svbool_t, int8_t const *, int64_t);
21820	21870	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_vnum_f64)))
...	...	@@ -21827,6 +21877,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_vnum_s32)))
21827	21877	svint32x4_t svld4_vnum_s32(svbool_t, int32_t const *, int64_t);
21828	21878	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_vnum_s64)))
21829	21879	svint64x4_t svld4_vnum_s64(svbool_t, int64_t const *, int64_t);
	21880	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_vnum_mf8)))
	21881	svmfloat8x4_t svld4_vnum_mf8(svbool_t, mfloat8_t const *, int64_t);
21830	21882	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_vnum_s16)))
21831	21883	svint16x4_t svld4_vnum_s16(svbool_t, int16_t const *, int64_t);
21832	21884	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_u8)))
...	...	@@ -21837,6 +21889,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_u64)))
21837	21889	svuint64_t svldnt1_u64(svbool_t, uint64_t const *);
21838	21890	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_u16)))
21839	21891	svuint16_t svldnt1_u16(svbool_t, uint16_t const *);
	21892	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_bf16)))
	21893	svbfloat16_t svldnt1_bf16(svbool_t, bfloat16_t const *);
21840	21894	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_s8)))
21841	21895	svint8_t svldnt1_s8(svbool_t, int8_t const *);
21842	21896	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_f64)))
...	...	@@ -21849,6 +21903,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_s32)))
21849	21903	svint32_t svldnt1_s32(svbool_t, int32_t const *);
21850	21904	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_s64)))
21851	21905	svint64_t svldnt1_s64(svbool_t, int64_t const *);
	21906	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_mf8)))
	21907	svmfloat8_t svldnt1_mf8(svbool_t, mfloat8_t const *);
21852	21908	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_s16)))
21853	21909	svint16_t svldnt1_s16(svbool_t, int16_t const *);
21854	21910	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_u8)))
...	...	@@ -21859,6 +21915,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_u64)))
21859	21915	svuint64_t svldnt1_vnum_u64(svbool_t, uint64_t const *, int64_t);
21860	21916	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_u16)))
21861	21917	svuint16_t svldnt1_vnum_u16(svbool_t, uint16_t const *, int64_t);
	21918	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_bf16)))
	21919	svbfloat16_t svldnt1_vnum_bf16(svbool_t, bfloat16_t const *, int64_t);
21862	21920	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_s8)))
21863	21921	svint8_t svldnt1_vnum_s8(svbool_t, int8_t const *, int64_t);
21864	21922	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_f64)))
...	...	@@ -21871,6 +21929,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_s32)))
21871	21929	svint32_t svldnt1_vnum_s32(svbool_t, int32_t const *, int64_t);
21872	21930	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_s64)))
21873	21931	svint64_t svldnt1_vnum_s64(svbool_t, int64_t const *, int64_t);
	21932	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_mf8)))
	21933	svmfloat8_t svldnt1_vnum_mf8(svbool_t, mfloat8_t const *, int64_t);
21874	21934	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_s16)))
21875	21935	svint16_t svldnt1_vnum_s16(svbool_t, int16_t const *, int64_t);
21876	21936	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlen_u8)))
...	...	@@ -21881,6 +21941,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlen_u64)))
21881	21941	uint64_t svlen_u64(svuint64_t);
21882	21942	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlen_u16)))
21883	21943	uint64_t svlen_u16(svuint16_t);
	21944	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlen_bf16)))
	21945	uint64_t svlen_bf16(svbfloat16_t);
21884	21946	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlen_s8)))
21885	21947	uint64_t svlen_s8(svint8_t);
21886	21948	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlen_f64)))
...	...	@@ -24171,6 +24233,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrev_u64)))
24171	24233	svuint64_t svrev_u64(svuint64_t);
24172	24234	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrev_u16)))
24173	24235	svuint16_t svrev_u16(svuint16_t);
	24236	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrev_bf16)))
	24237	svbfloat16_t svrev_bf16(svbfloat16_t);
24174	24238	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrev_s8)))
24175	24239	svint8_t svrev_s8(svint8_t);
24176	24240	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrev_f64)))
...	...	@@ -24449,6 +24513,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsel_u64)))
24449	24513	svuint64_t svsel_u64(svbool_t, svuint64_t, svuint64_t);
24450	24514	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsel_u16)))
24451	24515	svuint16_t svsel_u16(svbool_t, svuint16_t, svuint16_t);
	24516	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsel_bf16)))
	24517	svbfloat16_t svsel_bf16(svbool_t, svbfloat16_t, svbfloat16_t);
24452	24518	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsel_s8)))
24453	24519	svint8_t svsel_s8(svbool_t, svint8_t, svint8_t);
24454	24520	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsel_f64)))
...	...	@@ -24471,6 +24537,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset2_u64)))
24471	24537	svuint64x2_t svset2_u64(svuint64x2_t, uint64_t, svuint64_t);
24472	24538	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset2_u16)))
24473	24539	svuint16x2_t svset2_u16(svuint16x2_t, uint64_t, svuint16_t);
	24540	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset2_bf16)))
	24541	svbfloat16x2_t svset2_bf16(svbfloat16x2_t, uint64_t, svbfloat16_t);
24474	24542	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset2_s8)))
24475	24543	svint8x2_t svset2_s8(svint8x2_t, uint64_t, svint8_t);
24476	24544	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset2_f64)))
...	...	@@ -24483,6 +24551,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset2_s32)))
24483	24551	svint32x2_t svset2_s32(svint32x2_t, uint64_t, svint32_t);
24484	24552	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset2_s64)))
24485	24553	svint64x2_t svset2_s64(svint64x2_t, uint64_t, svint64_t);
	24554	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset2_mf8)))
	24555	svmfloat8x2_t svset2_mf8(svmfloat8x2_t, uint64_t, svmfloat8_t);
24486	24556	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset2_s16)))
24487	24557	svint16x2_t svset2_s16(svint16x2_t, uint64_t, svint16_t);
24488	24558	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset3_u8)))
...	...	@@ -24493,6 +24563,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset3_u64)))
24493	24563	svuint64x3_t svset3_u64(svuint64x3_t, uint64_t, svuint64_t);
24494	24564	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset3_u16)))
24495	24565	svuint16x3_t svset3_u16(svuint16x3_t, uint64_t, svuint16_t);
	24566	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset3_bf16)))
	24567	svbfloat16x3_t svset3_bf16(svbfloat16x3_t, uint64_t, svbfloat16_t);
24496	24568	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset3_s8)))
24497	24569	svint8x3_t svset3_s8(svint8x3_t, uint64_t, svint8_t);
24498	24570	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset3_f64)))
...	...	@@ -24505,6 +24577,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset3_s32)))
24505	24577	svint32x3_t svset3_s32(svint32x3_t, uint64_t, svint32_t);
24506	24578	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset3_s64)))
24507	24579	svint64x3_t svset3_s64(svint64x3_t, uint64_t, svint64_t);
	24580	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset3_mf8)))
	24581	svmfloat8x3_t svset3_mf8(svmfloat8x3_t, uint64_t, svmfloat8_t);
24508	24582	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset3_s16)))
24509	24583	svint16x3_t svset3_s16(svint16x3_t, uint64_t, svint16_t);
24510	24584	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset4_u8)))
...	...	@@ -24515,6 +24589,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset4_u64)))
24515	24589	svuint64x4_t svset4_u64(svuint64x4_t, uint64_t, svuint64_t);
24516	24590	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset4_u16)))
24517	24591	svuint16x4_t svset4_u16(svuint16x4_t, uint64_t, svuint16_t);
	24592	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset4_bf16)))
	24593	svbfloat16x4_t svset4_bf16(svbfloat16x4_t, uint64_t, svbfloat16_t);
24518	24594	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset4_s8)))
24519	24595	svint8x4_t svset4_s8(svint8x4_t, uint64_t, svint8_t);
24520	24596	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset4_f64)))
...	...	@@ -24527,6 +24603,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset4_s32)))
24527	24603	svint32x4_t svset4_s32(svint32x4_t, uint64_t, svint32_t);
24528	24604	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset4_s64)))
24529	24605	svint64x4_t svset4_s64(svint64x4_t, uint64_t, svint64_t);
	24606	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset4_mf8)))
	24607	svmfloat8x4_t svset4_mf8(svmfloat8x4_t, uint64_t, svmfloat8_t);
24530	24608	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset4_s16)))
24531	24609	svint16x4_t svset4_s16(svint16x4_t, uint64_t, svint16_t);
24532	24610	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsplice_u8)))
...	...	@@ -24537,6 +24615,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsplice_u64)))
24537	24615	svuint64_t svsplice_u64(svbool_t, svuint64_t, svuint64_t);
24538	24616	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsplice_u16)))
24539	24617	svuint16_t svsplice_u16(svbool_t, svuint16_t, svuint16_t);
	24618	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsplice_bf16)))
	24619	svbfloat16_t svsplice_bf16(svbool_t, svbfloat16_t, svbfloat16_t);
24540	24620	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsplice_s8)))
24541	24621	svint8_t svsplice_s8(svbool_t, svint8_t, svint8_t);
24542	24622	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsplice_f64)))
...	...	@@ -24577,6 +24657,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_u64)))
24577	24657	void svst1_u64(svbool_t, uint64_t *, svuint64_t);
24578	24658	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_u16)))
24579	24659	void svst1_u16(svbool_t, uint16_t *, svuint16_t);
	24660	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_bf16)))
	24661	void svst1_bf16(svbool_t, bfloat16_t *, svbfloat16_t);
24580	24662	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_s8)))
24581	24663	void svst1_s8(svbool_t, int8_t *, svint8_t);
24582	24664	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_f64)))
...	...	@@ -24589,6 +24671,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_s32)))
24589	24671	void svst1_s32(svbool_t, int32_t *, svint32_t);
24590	24672	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_s64)))
24591	24673	void svst1_s64(svbool_t, int64_t *, svint64_t);
	24674	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_mf8)))
	24675	void svst1_mf8(svbool_t, mfloat8_t *, svmfloat8_t);
24592	24676	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_s16)))
24593	24677	void svst1_s16(svbool_t, int16_t *, svint16_t);
24594	24678	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_u8)))
...	...	@@ -24599,6 +24683,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_u64)))
24599	24683	void svst1_vnum_u64(svbool_t, uint64_t *, int64_t, svuint64_t);
24600	24684	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_u16)))
24601	24685	void svst1_vnum_u16(svbool_t, uint16_t *, int64_t, svuint16_t);
	24686	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_bf16)))
	24687	void svst1_vnum_bf16(svbool_t, bfloat16_t *, int64_t, svbfloat16_t);
24602	24688	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_s8)))
24603	24689	void svst1_vnum_s8(svbool_t, int8_t *, int64_t, svint8_t);
24604	24690	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_f64)))
...	...	@@ -24611,6 +24697,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_s32)))
24611	24697	void svst1_vnum_s32(svbool_t, int32_t *, int64_t, svint32_t);
24612	24698	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_s64)))
24613	24699	void svst1_vnum_s64(svbool_t, int64_t *, int64_t, svint64_t);
	24700	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_mf8)))
	24701	void svst1_vnum_mf8(svbool_t, mfloat8_t *, int64_t, svmfloat8_t);
24614	24702	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_s16)))
24615	24703	void svst1_vnum_s16(svbool_t, int16_t *, int64_t, svint16_t);
24616	24704	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1b_s32)))
...	...	@@ -24669,6 +24757,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_u64)))
24669	24757	void svst2_u64(svbool_t, uint64_t *, svuint64x2_t);
24670	24758	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_u16)))
24671	24759	void svst2_u16(svbool_t, uint16_t *, svuint16x2_t);
	24760	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_bf16)))
	24761	void svst2_bf16(svbool_t, bfloat16_t *, svbfloat16x2_t);
24672	24762	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_s8)))
24673	24763	void svst2_s8(svbool_t, int8_t *, svint8x2_t);
24674	24764	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_f64)))
...	...	@@ -24681,6 +24771,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_s32)))
24681	24771	void svst2_s32(svbool_t, int32_t *, svint32x2_t);
24682	24772	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_s64)))
24683	24773	void svst2_s64(svbool_t, int64_t *, svint64x2_t);
	24774	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_mf8)))
	24775	void svst2_mf8(svbool_t, mfloat8_t *, svmfloat8x2_t);
24684	24776	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_s16)))
24685	24777	void svst2_s16(svbool_t, int16_t *, svint16x2_t);
24686	24778	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_vnum_u8)))
...	...	@@ -24691,6 +24783,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_vnum_u64)))
24691	24783	void svst2_vnum_u64(svbool_t, uint64_t *, int64_t, svuint64x2_t);
24692	24784	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_vnum_u16)))
24693	24785	void svst2_vnum_u16(svbool_t, uint16_t *, int64_t, svuint16x2_t);
	24786	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_vnum_bf16)))
	24787	void svst2_vnum_bf16(svbool_t, bfloat16_t *, int64_t, svbfloat16x2_t);
24694	24788	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_vnum_s8)))
24695	24789	void svst2_vnum_s8(svbool_t, int8_t *, int64_t, svint8x2_t);
24696	24790	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_vnum_f64)))
...	...	@@ -24703,6 +24797,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_vnum_s32)))
24703	24797	void svst2_vnum_s32(svbool_t, int32_t *, int64_t, svint32x2_t);
24704	24798	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_vnum_s64)))
24705	24799	void svst2_vnum_s64(svbool_t, int64_t *, int64_t, svint64x2_t);
	24800	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_vnum_mf8)))
	24801	void svst2_vnum_mf8(svbool_t, mfloat8_t *, int64_t, svmfloat8x2_t);
24706	24802	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_vnum_s16)))
24707	24803	void svst2_vnum_s16(svbool_t, int16_t *, int64_t, svint16x2_t);
24708	24804	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_u8)))
...	...	@@ -24713,6 +24809,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_u64)))
24713	24809	void svst3_u64(svbool_t, uint64_t *, svuint64x3_t);
24714	24810	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_u16)))
24715	24811	void svst3_u16(svbool_t, uint16_t *, svuint16x3_t);
	24812	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_bf16)))
	24813	void svst3_bf16(svbool_t, bfloat16_t *, svbfloat16x3_t);
24716	24814	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_s8)))
24717	24815	void svst3_s8(svbool_t, int8_t *, svint8x3_t);
24718	24816	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_f64)))
...	...	@@ -24725,6 +24823,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_s32)))
24725	24823	void svst3_s32(svbool_t, int32_t *, svint32x3_t);
24726	24824	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_s64)))
24727	24825	void svst3_s64(svbool_t, int64_t *, svint64x3_t);
	24826	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_mf8)))
	24827	void svst3_mf8(svbool_t, mfloat8_t *, svmfloat8x3_t);
24728	24828	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_s16)))
24729	24829	void svst3_s16(svbool_t, int16_t *, svint16x3_t);
24730	24830	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_vnum_u8)))
...	...	@@ -24735,6 +24835,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_vnum_u64)))
24735	24835	void svst3_vnum_u64(svbool_t, uint64_t *, int64_t, svuint64x3_t);
24736	24836	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_vnum_u16)))
24737	24837	void svst3_vnum_u16(svbool_t, uint16_t *, int64_t, svuint16x3_t);
	24838	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_vnum_bf16)))
	24839	void svst3_vnum_bf16(svbool_t, bfloat16_t *, int64_t, svbfloat16x3_t);
24738	24840	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_vnum_s8)))
24739	24841	void svst3_vnum_s8(svbool_t, int8_t *, int64_t, svint8x3_t);
24740	24842	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_vnum_f64)))
...	...	@@ -24747,6 +24849,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_vnum_s32)))
24747	24849	void svst3_vnum_s32(svbool_t, int32_t *, int64_t, svint32x3_t);
24748	24850	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_vnum_s64)))
24749	24851	void svst3_vnum_s64(svbool_t, int64_t *, int64_t, svint64x3_t);
	24852	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_vnum_mf8)))
	24853	void svst3_vnum_mf8(svbool_t, mfloat8_t *, int64_t, svmfloat8x3_t);
24750	24854	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_vnum_s16)))
24751	24855	void svst3_vnum_s16(svbool_t, int16_t *, int64_t, svint16x3_t);
24752	24856	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_u8)))
...	...	@@ -24757,6 +24861,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_u64)))
24757	24861	void svst4_u64(svbool_t, uint64_t *, svuint64x4_t);
24758	24862	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_u16)))
24759	24863	void svst4_u16(svbool_t, uint16_t *, svuint16x4_t);
	24864	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_bf16)))
	24865	void svst4_bf16(svbool_t, bfloat16_t *, svbfloat16x4_t);
24760	24866	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_s8)))
24761	24867	void svst4_s8(svbool_t, int8_t *, svint8x4_t);
24762	24868	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_f64)))
...	...	@@ -24769,6 +24875,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_s32)))
24769	24875	void svst4_s32(svbool_t, int32_t *, svint32x4_t);
24770	24876	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_s64)))
24771	24877	void svst4_s64(svbool_t, int64_t *, svint64x4_t);
	24878	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_mf8)))
	24879	void svst4_mf8(svbool_t, mfloat8_t *, svmfloat8x4_t);
24772	24880	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_s16)))
24773	24881	void svst4_s16(svbool_t, int16_t *, svint16x4_t);
24774	24882	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_vnum_u8)))
...	...	@@ -24779,6 +24887,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_vnum_u64)))
24779	24887	void svst4_vnum_u64(svbool_t, uint64_t *, int64_t, svuint64x4_t);
24780	24888	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_vnum_u16)))
24781	24889	void svst4_vnum_u16(svbool_t, uint16_t *, int64_t, svuint16x4_t);
	24890	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_vnum_bf16)))
	24891	void svst4_vnum_bf16(svbool_t, bfloat16_t *, int64_t, svbfloat16x4_t);
24782	24892	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_vnum_s8)))
24783	24893	void svst4_vnum_s8(svbool_t, int8_t *, int64_t, svint8x4_t);
24784	24894	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_vnum_f64)))
...	...	@@ -24791,6 +24901,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_vnum_s32)))
24791	24901	void svst4_vnum_s32(svbool_t, int32_t *, int64_t, svint32x4_t);
24792	24902	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_vnum_s64)))
24793	24903	void svst4_vnum_s64(svbool_t, int64_t *, int64_t, svint64x4_t);
	24904	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_vnum_mf8)))
	24905	void svst4_vnum_mf8(svbool_t, mfloat8_t *, int64_t, svmfloat8x4_t);
24794	24906	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_vnum_s16)))
24795	24907	void svst4_vnum_s16(svbool_t, int16_t *, int64_t, svint16x4_t);
24796	24908	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_u8)))
...	...	@@ -24801,6 +24913,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_u64)))
24801	24913	void svstnt1_u64(svbool_t, uint64_t *, svuint64_t);
24802	24914	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_u16)))
24803	24915	void svstnt1_u16(svbool_t, uint16_t *, svuint16_t);
	24916	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_bf16)))
	24917	void svstnt1_bf16(svbool_t, bfloat16_t *, svbfloat16_t);
24804	24918	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_s8)))
24805	24919	void svstnt1_s8(svbool_t, int8_t *, svint8_t);
24806	24920	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_f64)))
...	...	@@ -24813,6 +24927,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_s32)))
24813	24927	void svstnt1_s32(svbool_t, int32_t *, svint32_t);
24814	24928	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_s64)))
24815	24929	void svstnt1_s64(svbool_t, int64_t *, svint64_t);
	24930	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_mf8)))
	24931	void svstnt1_mf8(svbool_t, mfloat8_t *, svmfloat8_t);
24816	24932	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_s16)))
24817	24933	void svstnt1_s16(svbool_t, int16_t *, svint16_t);
24818	24934	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_u8)))
...	...	@@ -24823,6 +24939,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_u64)))
24823	24939	void svstnt1_vnum_u64(svbool_t, uint64_t *, int64_t, svuint64_t);
24824	24940	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_u16)))
24825	24941	void svstnt1_vnum_u16(svbool_t, uint16_t *, int64_t, svuint16_t);
	24942	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_bf16)))
	24943	void svstnt1_vnum_bf16(svbool_t, bfloat16_t *, int64_t, svbfloat16_t);
24826	24944	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_s8)))
24827	24945	void svstnt1_vnum_s8(svbool_t, int8_t *, int64_t, svint8_t);
24828	24946	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_f64)))
...	...	@@ -24835,6 +24953,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_s32)))
24835	24953	void svstnt1_vnum_s32(svbool_t, int32_t *, int64_t, svint32_t);
24836	24954	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_s64)))
24837	24955	void svstnt1_vnum_s64(svbool_t, int64_t *, int64_t, svint64_t);
	24956	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_mf8)))
	24957	void svstnt1_vnum_mf8(svbool_t, mfloat8_t *, int64_t, svmfloat8_t);
24838	24958	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_s16)))
24839	24959	void svstnt1_vnum_s16(svbool_t, int16_t *, int64_t, svint16_t);
24840	24960	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsub_n_f64_m)))
...	...	@@ -25109,6 +25229,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbl_u64)))
25109	25229	svuint64_t svtbl_u64(svuint64_t, svuint64_t);
25110	25230	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbl_u16)))
25111	25231	svuint16_t svtbl_u16(svuint16_t, svuint16_t);
	25232	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbl_bf16)))
	25233	svbfloat16_t svtbl_bf16(svbfloat16_t, svuint16_t);
25112	25234	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbl_s8)))
25113	25235	svint8_t svtbl_s8(svint8_t, svuint8_t);
25114	25236	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbl_f64)))
...	...	@@ -25131,6 +25253,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1_u64)))
25131	25253	svuint64_t svtrn1_u64(svuint64_t, svuint64_t);
25132	25254	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1_u16)))
25133	25255	svuint16_t svtrn1_u16(svuint16_t, svuint16_t);
	25256	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1_bf16)))
	25257	svbfloat16_t svtrn1_bf16(svbfloat16_t, svbfloat16_t);
25134	25258	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1_s8)))
25135	25259	svint8_t svtrn1_s8(svint8_t, svint8_t);
25136	25260	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1_f64)))
...	...	@@ -25161,6 +25285,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2_u64)))
25161	25285	svuint64_t svtrn2_u64(svuint64_t, svuint64_t);
25162	25286	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2_u16)))
25163	25287	svuint16_t svtrn2_u16(svuint16_t, svuint16_t);
	25288	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2_bf16)))
	25289	svbfloat16_t svtrn2_bf16(svbfloat16_t, svbfloat16_t);
25164	25290	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2_s8)))
25165	25291	svint8_t svtrn2_s8(svint8_t, svint8_t);
25166	25292	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2_f64)))
...	...	@@ -25191,6 +25317,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef2_u64)))
25191	25317	svuint64x2_t svundef2_u64(void);
25192	25318	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef2_u16)))
25193	25319	svuint16x2_t svundef2_u16(void);
	25320	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef2_bf16)))
	25321	svbfloat16x2_t svundef2_bf16(void);
25194	25322	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef2_s8)))
25195	25323	svint8x2_t svundef2_s8(void);
25196	25324	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef2_f64)))
...	...	@@ -25203,6 +25331,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef2_s32)))
25203	25331	svint32x2_t svundef2_s32(void);
25204	25332	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef2_s64)))
25205	25333	svint64x2_t svundef2_s64(void);
	25334	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef2_mf8)))
	25335	svmfloat8x2_t svundef2_mf8(void);
25206	25336	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef2_s16)))
25207	25337	svint16x2_t svundef2_s16(void);
25208	25338	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef3_u8)))
...	...	@@ -25213,6 +25343,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef3_u64)))
25213	25343	svuint64x3_t svundef3_u64(void);
25214	25344	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef3_u16)))
25215	25345	svuint16x3_t svundef3_u16(void);
	25346	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef3_bf16)))
	25347	svbfloat16x3_t svundef3_bf16(void);
25216	25348	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef3_s8)))
25217	25349	svint8x3_t svundef3_s8(void);
25218	25350	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef3_f64)))
...	...	@@ -25225,6 +25357,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef3_s32)))
25225	25357	svint32x3_t svundef3_s32(void);
25226	25358	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef3_s64)))
25227	25359	svint64x3_t svundef3_s64(void);
	25360	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef3_mf8)))
	25361	svmfloat8x3_t svundef3_mf8(void);
25228	25362	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef3_s16)))
25229	25363	svint16x3_t svundef3_s16(void);
25230	25364	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef4_u8)))
...	...	@@ -25235,6 +25369,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef4_u64)))
25235	25369	svuint64x4_t svundef4_u64(void);
25236	25370	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef4_u16)))
25237	25371	svuint16x4_t svundef4_u16(void);
	25372	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef4_bf16)))
	25373	svbfloat16x4_t svundef4_bf16(void);
25238	25374	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef4_s8)))
25239	25375	svint8x4_t svundef4_s8(void);
25240	25376	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef4_f64)))
...	...	@@ -25247,6 +25383,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef4_s32)))
25247	25383	svint32x4_t svundef4_s32(void);
25248	25384	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef4_s64)))
25249	25385	svint64x4_t svundef4_s64(void);
	25386	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef4_mf8)))
	25387	svmfloat8x4_t svundef4_mf8(void);
25250	25388	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef4_s16)))
25251	25389	svint16x4_t svundef4_s16(void);
25252	25390	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef_u8)))
...	...	@@ -25257,6 +25395,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef_u64)))
25257	25395	svuint64_t svundef_u64(void);
25258	25396	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef_u16)))
25259	25397	svuint16_t svundef_u16(void);
	25398	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef_bf16)))
	25399	svbfloat16_t svundef_bf16(void);
25260	25400	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef_s8)))
25261	25401	svint8_t svundef_s8(void);
25262	25402	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef_f64)))
...	...	@@ -25269,6 +25409,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef_s32)))
25269	25409	svint32_t svundef_s32(void);
25270	25410	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef_s64)))
25271	25411	svint64_t svundef_s64(void);
	25412	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef_mf8)))
	25413	svmfloat8_t svundef_mf8(void);
25272	25414	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svundef_s16)))
25273	25415	svint16_t svundef_s16(void);
25274	25416	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svunpkhi_b)))
...	...	@@ -25307,6 +25449,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1_u64)))
25307	25449	svuint64_t svuzp1_u64(svuint64_t, svuint64_t);
25308	25450	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1_u16)))
25309	25451	svuint16_t svuzp1_u16(svuint16_t, svuint16_t);
	25452	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1_bf16)))
	25453	svbfloat16_t svuzp1_bf16(svbfloat16_t, svbfloat16_t);
25310	25454	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1_s8)))
25311	25455	svint8_t svuzp1_s8(svint8_t, svint8_t);
25312	25456	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1_f64)))
...	...	@@ -25337,6 +25481,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2_u64)))
25337	25481	svuint64_t svuzp2_u64(svuint64_t, svuint64_t);
25338	25482	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2_u16)))
25339	25483	svuint16_t svuzp2_u16(svuint16_t, svuint16_t);
	25484	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2_bf16)))
	25485	svbfloat16_t svuzp2_bf16(svbfloat16_t, svbfloat16_t);
25340	25486	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2_s8)))
25341	25487	svint8_t svuzp2_s8(svint8_t, svint8_t);
25342	25488	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2_f64)))
...	...	@@ -25431,6 +25577,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1_u64)))
25431	25577	svuint64_t svzip1_u64(svuint64_t, svuint64_t);
25432	25578	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1_u16)))
25433	25579	svuint16_t svzip1_u16(svuint16_t, svuint16_t);
	25580	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1_bf16)))
	25581	svbfloat16_t svzip1_bf16(svbfloat16_t, svbfloat16_t);
25434	25582	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1_s8)))
25435	25583	svint8_t svzip1_s8(svint8_t, svint8_t);
25436	25584	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1_f64)))
...	...	@@ -25461,6 +25609,8 @@ __ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2_u64)))
25461	25609	svuint64_t svzip2_u64(svuint64_t, svuint64_t);
25462	25610	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2_u16)))
25463	25611	svuint16_t svzip2_u16(svuint16_t, svuint16_t);
	25612	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2_bf16)))
	25613	svbfloat16_t svzip2_bf16(svbfloat16_t, svbfloat16_t);
25464	25614	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2_s8)))
25465	25615	svint8_t svzip2_s8(svint8_t, svint8_t);
25466	25616	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2_f64)))
...	...	@@ -26219,6 +26369,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclasta_n_u64)))
26219	26369	uint64_t svclasta(svbool_t, uint64_t, svuint64_t);
26220	26370	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclasta_n_u16)))
26221	26371	uint16_t svclasta(svbool_t, uint16_t, svuint16_t);
	26372	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclasta_n_bf16)))
	26373	bfloat16_t svclasta(svbool_t, bfloat16_t, svbfloat16_t);
26222	26374	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclasta_n_s8)))
26223	26375	int8_t svclasta(svbool_t, int8_t, svint8_t);
26224	26376	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclasta_n_f64)))
...	...	@@ -26241,6 +26393,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclasta_u64)))
26241	26393	svuint64_t svclasta(svbool_t, svuint64_t, svuint64_t);
26242	26394	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclasta_u16)))
26243	26395	svuint16_t svclasta(svbool_t, svuint16_t, svuint16_t);
	26396	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclasta_bf16)))
	26397	svbfloat16_t svclasta(svbool_t, svbfloat16_t, svbfloat16_t);
26244	26398	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclasta_s8)))
26245	26399	svint8_t svclasta(svbool_t, svint8_t, svint8_t);
26246	26400	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclasta_f64)))
...	...	@@ -26263,6 +26417,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclastb_n_u64)))
26263	26417	uint64_t svclastb(svbool_t, uint64_t, svuint64_t);
26264	26418	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclastb_n_u16)))
26265	26419	uint16_t svclastb(svbool_t, uint16_t, svuint16_t);
	26420	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclastb_n_bf16)))
	26421	bfloat16_t svclastb(svbool_t, bfloat16_t, svbfloat16_t);
26266	26422	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclastb_n_s8)))
26267	26423	int8_t svclastb(svbool_t, int8_t, svint8_t);
26268	26424	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclastb_n_f64)))
...	...	@@ -26285,6 +26441,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclastb_u64)))
26285	26441	svuint64_t svclastb(svbool_t, svuint64_t, svuint64_t);
26286	26442	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclastb_u16)))
26287	26443	svuint16_t svclastb(svbool_t, svuint16_t, svuint16_t);
	26444	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclastb_bf16)))
	26445	svbfloat16_t svclastb(svbool_t, svbfloat16_t, svbfloat16_t);
26288	26446	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclastb_s8)))
26289	26447	svint8_t svclastb(svbool_t, svint8_t, svint8_t);
26290	26448	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svclastb_f64)))
...	...	@@ -26845,6 +27003,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_u64_m)))
26845	27003	svuint64_t svcnt_m(svuint64_t, svbool_t, svuint64_t);
26846	27004	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_u16_m)))
26847	27005	svuint16_t svcnt_m(svuint16_t, svbool_t, svuint16_t);
	27006	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_bf16_m)))
	27007	svuint16_t svcnt_m(svuint16_t, svbool_t, svbfloat16_t);
26848	27008	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_s8_m)))
26849	27009	svuint8_t svcnt_m(svuint8_t, svbool_t, svint8_t);
26850	27010	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_f64_m)))
...	...	@@ -26867,6 +27027,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_u64_x)))
26867	27027	svuint64_t svcnt_x(svbool_t, svuint64_t);
26868	27028	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_u16_x)))
26869	27029	svuint16_t svcnt_x(svbool_t, svuint16_t);
	27030	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_bf16_x)))
	27031	svuint16_t svcnt_x(svbool_t, svbfloat16_t);
26870	27032	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_s8_x)))
26871	27033	svuint8_t svcnt_x(svbool_t, svint8_t);
26872	27034	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_f64_x)))
...	...	@@ -26889,6 +27051,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_u64_z)))
26889	27051	svuint64_t svcnt_z(svbool_t, svuint64_t);
26890	27052	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_u16_z)))
26891	27053	svuint16_t svcnt_z(svbool_t, svuint16_t);
	27054	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_bf16_z)))
	27055	svuint16_t svcnt_z(svbool_t, svbfloat16_t);
26892	27056	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_s8_z)))
26893	27057	svuint8_t svcnt_z(svbool_t, svint8_t);
26894	27058	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcnt_f64_z)))
...	...	@@ -26911,6 +27075,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate2_u64)))
26911	27075	svuint64x2_t svcreate2(svuint64_t, svuint64_t);
26912	27076	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate2_u16)))
26913	27077	svuint16x2_t svcreate2(svuint16_t, svuint16_t);
	27078	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate2_bf16)))
	27079	svbfloat16x2_t svcreate2(svbfloat16_t, svbfloat16_t);
26914	27080	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate2_s8)))
26915	27081	svint8x2_t svcreate2(svint8_t, svint8_t);
26916	27082	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate2_f64)))
...	...	@@ -26923,6 +27089,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate2_s32)))
26923	27089	svint32x2_t svcreate2(svint32_t, svint32_t);
26924	27090	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate2_s64)))
26925	27091	svint64x2_t svcreate2(svint64_t, svint64_t);
	27092	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate2_mf8)))
	27093	svmfloat8x2_t svcreate2(svmfloat8_t, svmfloat8_t);
26926	27094	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate2_s16)))
26927	27095	svint16x2_t svcreate2(svint16_t, svint16_t);
26928	27096	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate3_u8)))
...	...	@@ -26933,6 +27101,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate3_u64)))
26933	27101	svuint64x3_t svcreate3(svuint64_t, svuint64_t, svuint64_t);
26934	27102	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate3_u16)))
26935	27103	svuint16x3_t svcreate3(svuint16_t, svuint16_t, svuint16_t);
	27104	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate3_bf16)))
	27105	svbfloat16x3_t svcreate3(svbfloat16_t, svbfloat16_t, svbfloat16_t);
26936	27106	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate3_s8)))
26937	27107	svint8x3_t svcreate3(svint8_t, svint8_t, svint8_t);
26938	27108	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate3_f64)))
...	...	@@ -26945,6 +27115,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate3_s32)))
26945	27115	svint32x3_t svcreate3(svint32_t, svint32_t, svint32_t);
26946	27116	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate3_s64)))
26947	27117	svint64x3_t svcreate3(svint64_t, svint64_t, svint64_t);
	27118	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate3_mf8)))
	27119	svmfloat8x3_t svcreate3(svmfloat8_t, svmfloat8_t, svmfloat8_t);
26948	27120	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate3_s16)))
26949	27121	svint16x3_t svcreate3(svint16_t, svint16_t, svint16_t);
26950	27122	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate4_u8)))
...	...	@@ -26955,6 +27127,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate4_u64)))
26955	27127	svuint64x4_t svcreate4(svuint64_t, svuint64_t, svuint64_t, svuint64_t);
26956	27128	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate4_u16)))
26957	27129	svuint16x4_t svcreate4(svuint16_t, svuint16_t, svuint16_t, svuint16_t);
	27130	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate4_bf16)))
	27131	svbfloat16x4_t svcreate4(svbfloat16_t, svbfloat16_t, svbfloat16_t, svbfloat16_t);
26958	27132	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate4_s8)))
26959	27133	svint8x4_t svcreate4(svint8_t, svint8_t, svint8_t, svint8_t);
26960	27134	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate4_f64)))
...	...	@@ -26967,6 +27141,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate4_s32)))
26967	27141	svint32x4_t svcreate4(svint32_t, svint32_t, svint32_t, svint32_t);
26968	27142	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate4_s64)))
26969	27143	svint64x4_t svcreate4(svint64_t, svint64_t, svint64_t, svint64_t);
	27144	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate4_mf8)))
	27145	svmfloat8x4_t svcreate4(svmfloat8_t, svmfloat8_t, svmfloat8_t, svmfloat8_t);
26970	27146	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcreate4_s16)))
26971	27147	svint16x4_t svcreate4(svint16_t, svint16_t, svint16_t, svint16_t);
26972	27148	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svcvt_f16_f32_m)))
...	...	@@ -27373,6 +27549,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_u64)))
27373	27549	svuint64_t svdup_u64(uint64_t);
27374	27550	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_u16)))
27375	27551	svuint16_t svdup_u16(uint16_t);
	27552	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_bf16)))
	27553	svbfloat16_t svdup_bf16(bfloat16_t);
27376	27554	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_s8)))
27377	27555	svint8_t svdup_s8(int8_t);
27378	27556	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_f64)))
...	...	@@ -27395,6 +27573,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_u64_m)))
27395	27573	svuint64_t svdup_u64_m(svuint64_t, svbool_t, uint64_t);
27396	27574	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_u16_m)))
27397	27575	svuint16_t svdup_u16_m(svuint16_t, svbool_t, uint16_t);
	27576	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_bf16_m)))
	27577	svbfloat16_t svdup_bf16_m(svbfloat16_t, svbool_t, bfloat16_t);
27398	27578	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_s8_m)))
27399	27579	svint8_t svdup_s8_m(svint8_t, svbool_t, int8_t);
27400	27580	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_f64_m)))
...	...	@@ -27425,6 +27605,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_u64_x)))
27425	27605	svuint64_t svdup_u64_x(svbool_t, uint64_t);
27426	27606	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_u16_x)))
27427	27607	svuint16_t svdup_u16_x(svbool_t, uint16_t);
	27608	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_bf16_x)))
	27609	svbfloat16_t svdup_bf16_x(svbool_t, bfloat16_t);
27428	27610	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_s8_x)))
27429	27611	svint8_t svdup_s8_x(svbool_t, int8_t);
27430	27612	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_f64_x)))
...	...	@@ -27447,6 +27629,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_u64_z)))
27447	27629	svuint64_t svdup_u64_z(svbool_t, uint64_t);
27448	27630	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_u16_z)))
27449	27631	svuint16_t svdup_u16_z(svbool_t, uint16_t);
	27632	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_bf16_z)))
	27633	svbfloat16_t svdup_bf16_z(svbool_t, bfloat16_t);
27450	27634	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_s8_z)))
27451	27635	svint8_t svdup_s8_z(svbool_t, int8_t);
27452	27636	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_n_f64_z)))
...	...	@@ -27469,6 +27653,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_lane_u64)))
27469	27653	svuint64_t svdup_lane(svuint64_t, uint64_t);
27470	27654	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_lane_u16)))
27471	27655	svuint16_t svdup_lane(svuint16_t, uint16_t);
	27656	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_lane_bf16)))
	27657	svbfloat16_t svdup_lane(svbfloat16_t, uint16_t);
27472	27658	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_lane_s8)))
27473	27659	svint8_t svdup_lane(svint8_t, uint8_t);
27474	27660	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdup_lane_f64)))
...	...	@@ -27489,6 +27675,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdupq_n_s8)))
27489	27675	svint8_t svdupq_s8(int8_t, int8_t, int8_t, int8_t, int8_t, int8_t, int8_t, int8_t, int8_t, int8_t, int8_t, int8_t, int8_t, int8_t, int8_t, int8_t);
27490	27676	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdupq_n_u16)))
27491	27677	svuint16_t svdupq_u16(uint16_t, uint16_t, uint16_t, uint16_t, uint16_t, uint16_t, uint16_t, uint16_t);
	27678	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdupq_n_bf16)))
	27679	svbfloat16_t svdupq_bf16(bfloat16_t, bfloat16_t, bfloat16_t, bfloat16_t, bfloat16_t, bfloat16_t, bfloat16_t, bfloat16_t);
27492	27680	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdupq_n_f16)))
27493	27681	svfloat16_t svdupq_f16(float16_t, float16_t, float16_t, float16_t, float16_t, float16_t, float16_t, float16_t);
27494	27682	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdupq_n_s16)))
...	...	@@ -27521,6 +27709,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdupq_lane_u64)))
27521	27709	svuint64_t svdupq_lane(svuint64_t, uint64_t);
27522	27710	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdupq_lane_u16)))
27523	27711	svuint16_t svdupq_lane(svuint16_t, uint64_t);
	27712	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdupq_lane_bf16)))
	27713	svbfloat16_t svdupq_lane(svbfloat16_t, uint64_t);
27524	27714	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdupq_lane_s8)))
27525	27715	svint8_t svdupq_lane(svint8_t, uint64_t);
27526	27716	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svdupq_lane_f64)))
...	...	@@ -27657,6 +27847,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svext_u64)))
27657	27847	svuint64_t svext(svuint64_t, svuint64_t, uint64_t);
27658	27848	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svext_u16)))
27659	27849	svuint16_t svext(svuint16_t, svuint16_t, uint64_t);
	27850	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svext_bf16)))
	27851	svbfloat16_t svext(svbfloat16_t, svbfloat16_t, uint64_t);
27660	27852	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svext_s8)))
27661	27853	svint8_t svext(svint8_t, svint8_t, uint64_t);
27662	27854	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svext_f64)))
...	...	@@ -27751,6 +27943,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget2_u64)))
27751	27943	svuint64_t svget2(svuint64x2_t, uint64_t);
27752	27944	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget2_u16)))
27753	27945	svuint16_t svget2(svuint16x2_t, uint64_t);
	27946	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget2_bf16)))
	27947	svbfloat16_t svget2(svbfloat16x2_t, uint64_t);
27754	27948	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget2_s8)))
27755	27949	svint8_t svget2(svint8x2_t, uint64_t);
27756	27950	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget2_f64)))
...	...	@@ -27763,6 +27957,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget2_s32)))
27763	27957	svint32_t svget2(svint32x2_t, uint64_t);
27764	27958	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget2_s64)))
27765	27959	svint64_t svget2(svint64x2_t, uint64_t);
	27960	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget2_mf8)))
	27961	svmfloat8_t svget2(svmfloat8x2_t, uint64_t);
27766	27962	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget2_s16)))
27767	27963	svint16_t svget2(svint16x2_t, uint64_t);
27768	27964	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget3_u8)))
...	...	@@ -27773,6 +27969,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget3_u64)))
27773	27969	svuint64_t svget3(svuint64x3_t, uint64_t);
27774	27970	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget3_u16)))
27775	27971	svuint16_t svget3(svuint16x3_t, uint64_t);
	27972	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget3_bf16)))
	27973	svbfloat16_t svget3(svbfloat16x3_t, uint64_t);
27776	27974	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget3_s8)))
27777	27975	svint8_t svget3(svint8x3_t, uint64_t);
27778	27976	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget3_f64)))
...	...	@@ -27785,6 +27983,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget3_s32)))
27785	27983	svint32_t svget3(svint32x3_t, uint64_t);
27786	27984	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget3_s64)))
27787	27985	svint64_t svget3(svint64x3_t, uint64_t);
	27986	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget3_mf8)))
	27987	svmfloat8_t svget3(svmfloat8x3_t, uint64_t);
27788	27988	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget3_s16)))
27789	27989	svint16_t svget3(svint16x3_t, uint64_t);
27790	27990	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget4_u8)))
...	...	@@ -27795,6 +27995,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget4_u64)))
27795	27995	svuint64_t svget4(svuint64x4_t, uint64_t);
27796	27996	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget4_u16)))
27797	27997	svuint16_t svget4(svuint16x4_t, uint64_t);
	27998	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget4_bf16)))
	27999	svbfloat16_t svget4(svbfloat16x4_t, uint64_t);
27798	28000	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget4_s8)))
27799	28001	svint8_t svget4(svint8x4_t, uint64_t);
27800	28002	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget4_f64)))
...	...	@@ -27807,6 +28009,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget4_s32)))
27807	28009	svint32_t svget4(svint32x4_t, uint64_t);
27808	28010	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget4_s64)))
27809	28011	svint64_t svget4(svint64x4_t, uint64_t);
	28012	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget4_mf8)))
	28013	svmfloat8_t svget4(svmfloat8x4_t, uint64_t);
27810	28014	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svget4_s16)))
27811	28015	svint16_t svget4(svint16x4_t, uint64_t);
27812	28016	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svinsr_n_u8)))
...	...	@@ -27817,6 +28021,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svinsr_n_u64)))
27817	28021	svuint64_t svinsr(svuint64_t, uint64_t);
27818	28022	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svinsr_n_u16)))
27819	28023	svuint16_t svinsr(svuint16_t, uint16_t);
	28024	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svinsr_n_bf16)))
	28025	svbfloat16_t svinsr(svbfloat16_t, bfloat16_t);
27820	28026	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svinsr_n_s8)))
27821	28027	svint8_t svinsr(svint8_t, int8_t);
27822	28028	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svinsr_n_f64)))
...	...	@@ -27839,6 +28045,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlasta_u64)))
27839	28045	uint64_t svlasta(svbool_t, svuint64_t);
27840	28046	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlasta_u16)))
27841	28047	uint16_t svlasta(svbool_t, svuint16_t);
	28048	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlasta_bf16)))
	28049	bfloat16_t svlasta(svbool_t, svbfloat16_t);
27842	28050	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlasta_s8)))
27843	28051	int8_t svlasta(svbool_t, svint8_t);
27844	28052	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlasta_f64)))
...	...	@@ -27861,6 +28069,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlastb_u64)))
27861	28069	uint64_t svlastb(svbool_t, svuint64_t);
27862	28070	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlastb_u16)))
27863	28071	uint16_t svlastb(svbool_t, svuint16_t);
	28072	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlastb_bf16)))
	28073	bfloat16_t svlastb(svbool_t, svbfloat16_t);
27864	28074	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlastb_s8)))
27865	28075	int8_t svlastb(svbool_t, svint8_t);
27866	28076	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlastb_f64)))
...	...	@@ -27883,6 +28093,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_u64)))
27883	28093	svuint64_t svld1(svbool_t, uint64_t const *);
27884	28094	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_u16)))
27885	28095	svuint16_t svld1(svbool_t, uint16_t const *);
	28096	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_bf16)))
	28097	svbfloat16_t svld1(svbool_t, bfloat16_t const *);
27886	28098	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_s8)))
27887	28099	svint8_t svld1(svbool_t, int8_t const *);
27888	28100	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_f64)))
...	...	@@ -27895,6 +28107,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_s32)))
27895	28107	svint32_t svld1(svbool_t, int32_t const *);
27896	28108	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_s64)))
27897	28109	svint64_t svld1(svbool_t, int64_t const *);
	28110	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_mf8)))
	28111	svmfloat8_t svld1(svbool_t, mfloat8_t const *);
27898	28112	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_s16)))
27899	28113	svint16_t svld1(svbool_t, int16_t const *);
27900	28114	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_u8)))
...	...	@@ -27905,6 +28119,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_u64)))
27905	28119	svuint64_t svld1_vnum(svbool_t, uint64_t const *, int64_t);
27906	28120	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_u16)))
27907	28121	svuint16_t svld1_vnum(svbool_t, uint16_t const *, int64_t);
	28122	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_bf16)))
	28123	svbfloat16_t svld1_vnum(svbool_t, bfloat16_t const *, int64_t);
27908	28124	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_s8)))
27909	28125	svint8_t svld1_vnum(svbool_t, int8_t const *, int64_t);
27910	28126	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_f64)))
...	...	@@ -27917,6 +28133,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_s32)))
27917	28133	svint32_t svld1_vnum(svbool_t, int32_t const *, int64_t);
27918	28134	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_s64)))
27919	28135	svint64_t svld1_vnum(svbool_t, int64_t const *, int64_t);
	28136	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_mf8)))
	28137	svmfloat8_t svld1_vnum(svbool_t, mfloat8_t const *, int64_t);
27920	28138	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1_vnum_s16)))
27921	28139	svint16_t svld1_vnum(svbool_t, int16_t const *, int64_t);
27922	28140	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1rq_u8)))
...	...	@@ -27927,6 +28145,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1rq_u64)))
27927	28145	svuint64_t svld1rq(svbool_t, uint64_t const *);
27928	28146	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1rq_u16)))
27929	28147	svuint16_t svld1rq(svbool_t, uint16_t const *);
	28148	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1rq_bf16)))
	28149	svbfloat16_t svld1rq(svbool_t, bfloat16_t const *);
27930	28150	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1rq_s8)))
27931	28151	svint8_t svld1rq(svbool_t, int8_t const *);
27932	28152	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1rq_f64)))
...	...	@@ -27939,6 +28159,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1rq_s32)))
27939	28159	svint32_t svld1rq(svbool_t, int32_t const *);
27940	28160	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1rq_s64)))
27941	28161	svint64_t svld1rq(svbool_t, int64_t const *);
	28162	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1rq_mf8)))
	28163	svmfloat8_t svld1rq(svbool_t, mfloat8_t const *);
27942	28164	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld1rq_s16)))
27943	28165	svint16_t svld1rq(svbool_t, int16_t const *);
27944	28166	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_u8)))
...	...	@@ -27949,6 +28171,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_u64)))
27949	28171	svuint64x2_t svld2(svbool_t, uint64_t const *);
27950	28172	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_u16)))
27951	28173	svuint16x2_t svld2(svbool_t, uint16_t const *);
	28174	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_bf16)))
	28175	svbfloat16x2_t svld2(svbool_t, bfloat16_t const *);
27952	28176	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_s8)))
27953	28177	svint8x2_t svld2(svbool_t, int8_t const *);
27954	28178	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_f64)))
...	...	@@ -27961,6 +28185,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_s32)))
27961	28185	svint32x2_t svld2(svbool_t, int32_t const *);
27962	28186	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_s64)))
27963	28187	svint64x2_t svld2(svbool_t, int64_t const *);
	28188	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_mf8)))
	28189	svmfloat8x2_t svld2(svbool_t, mfloat8_t const *);
27964	28190	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_s16)))
27965	28191	svint16x2_t svld2(svbool_t, int16_t const *);
27966	28192	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_vnum_u8)))
...	...	@@ -27971,6 +28197,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_vnum_u64)))
27971	28197	svuint64x2_t svld2_vnum(svbool_t, uint64_t const *, int64_t);
27972	28198	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_vnum_u16)))
27973	28199	svuint16x2_t svld2_vnum(svbool_t, uint16_t const *, int64_t);
	28200	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_vnum_bf16)))
	28201	svbfloat16x2_t svld2_vnum(svbool_t, bfloat16_t const *, int64_t);
27974	28202	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_vnum_s8)))
27975	28203	svint8x2_t svld2_vnum(svbool_t, int8_t const *, int64_t);
27976	28204	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_vnum_f64)))
...	...	@@ -27983,6 +28211,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_vnum_s32)))
27983	28211	svint32x2_t svld2_vnum(svbool_t, int32_t const *, int64_t);
27984	28212	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_vnum_s64)))
27985	28213	svint64x2_t svld2_vnum(svbool_t, int64_t const *, int64_t);
	28214	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_vnum_mf8)))
	28215	svmfloat8x2_t svld2_vnum(svbool_t, mfloat8_t const *, int64_t);
27986	28216	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld2_vnum_s16)))
27987	28217	svint16x2_t svld2_vnum(svbool_t, int16_t const *, int64_t);
27988	28218	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_u8)))
...	...	@@ -27993,6 +28223,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_u64)))
27993	28223	svuint64x3_t svld3(svbool_t, uint64_t const *);
27994	28224	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_u16)))
27995	28225	svuint16x3_t svld3(svbool_t, uint16_t const *);
	28226	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_bf16)))
	28227	svbfloat16x3_t svld3(svbool_t, bfloat16_t const *);
27996	28228	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_s8)))
27997	28229	svint8x3_t svld3(svbool_t, int8_t const *);
27998	28230	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_f64)))
...	...	@@ -28005,6 +28237,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_s32)))
28005	28237	svint32x3_t svld3(svbool_t, int32_t const *);
28006	28238	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_s64)))
28007	28239	svint64x3_t svld3(svbool_t, int64_t const *);
	28240	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_mf8)))
	28241	svmfloat8x3_t svld3(svbool_t, mfloat8_t const *);
28008	28242	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_s16)))
28009	28243	svint16x3_t svld3(svbool_t, int16_t const *);
28010	28244	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_vnum_u8)))
...	...	@@ -28015,6 +28249,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_vnum_u64)))
28015	28249	svuint64x3_t svld3_vnum(svbool_t, uint64_t const *, int64_t);
28016	28250	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_vnum_u16)))
28017	28251	svuint16x3_t svld3_vnum(svbool_t, uint16_t const *, int64_t);
	28252	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_vnum_bf16)))
	28253	svbfloat16x3_t svld3_vnum(svbool_t, bfloat16_t const *, int64_t);
28018	28254	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_vnum_s8)))
28019	28255	svint8x3_t svld3_vnum(svbool_t, int8_t const *, int64_t);
28020	28256	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_vnum_f64)))
...	...	@@ -28027,6 +28263,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_vnum_s32)))
28027	28263	svint32x3_t svld3_vnum(svbool_t, int32_t const *, int64_t);
28028	28264	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_vnum_s64)))
28029	28265	svint64x3_t svld3_vnum(svbool_t, int64_t const *, int64_t);
	28266	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_vnum_mf8)))
	28267	svmfloat8x3_t svld3_vnum(svbool_t, mfloat8_t const *, int64_t);
28030	28268	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld3_vnum_s16)))
28031	28269	svint16x3_t svld3_vnum(svbool_t, int16_t const *, int64_t);
28032	28270	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_u8)))
...	...	@@ -28037,6 +28275,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_u64)))
28037	28275	svuint64x4_t svld4(svbool_t, uint64_t const *);
28038	28276	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_u16)))
28039	28277	svuint16x4_t svld4(svbool_t, uint16_t const *);
	28278	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_bf16)))
	28279	svbfloat16x4_t svld4(svbool_t, bfloat16_t const *);
28040	28280	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_s8)))
28041	28281	svint8x4_t svld4(svbool_t, int8_t const *);
28042	28282	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_f64)))
...	...	@@ -28049,6 +28289,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_s32)))
28049	28289	svint32x4_t svld4(svbool_t, int32_t const *);
28050	28290	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_s64)))
28051	28291	svint64x4_t svld4(svbool_t, int64_t const *);
	28292	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_mf8)))
	28293	svmfloat8x4_t svld4(svbool_t, mfloat8_t const *);
28052	28294	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_s16)))
28053	28295	svint16x4_t svld4(svbool_t, int16_t const *);
28054	28296	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_vnum_u8)))
...	...	@@ -28059,6 +28301,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_vnum_u64)))
28059	28301	svuint64x4_t svld4_vnum(svbool_t, uint64_t const *, int64_t);
28060	28302	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_vnum_u16)))
28061	28303	svuint16x4_t svld4_vnum(svbool_t, uint16_t const *, int64_t);
	28304	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_vnum_bf16)))
	28305	svbfloat16x4_t svld4_vnum(svbool_t, bfloat16_t const *, int64_t);
28062	28306	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_vnum_s8)))
28063	28307	svint8x4_t svld4_vnum(svbool_t, int8_t const *, int64_t);
28064	28308	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_vnum_f64)))
...	...	@@ -28071,6 +28315,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_vnum_s32)))
28071	28315	svint32x4_t svld4_vnum(svbool_t, int32_t const *, int64_t);
28072	28316	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_vnum_s64)))
28073	28317	svint64x4_t svld4_vnum(svbool_t, int64_t const *, int64_t);
	28318	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_vnum_mf8)))
	28319	svmfloat8x4_t svld4_vnum(svbool_t, mfloat8_t const *, int64_t);
28074	28320	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svld4_vnum_s16)))
28075	28321	svint16x4_t svld4_vnum(svbool_t, int16_t const *, int64_t);
28076	28322	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_u8)))
...	...	@@ -28081,6 +28327,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_u64)))
28081	28327	svuint64_t svldnt1(svbool_t, uint64_t const *);
28082	28328	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_u16)))
28083	28329	svuint16_t svldnt1(svbool_t, uint16_t const *);
	28330	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_bf16)))
	28331	svbfloat16_t svldnt1(svbool_t, bfloat16_t const *);
28084	28332	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_s8)))
28085	28333	svint8_t svldnt1(svbool_t, int8_t const *);
28086	28334	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_f64)))
...	...	@@ -28093,6 +28341,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_s32)))
28093	28341	svint32_t svldnt1(svbool_t, int32_t const *);
28094	28342	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_s64)))
28095	28343	svint64_t svldnt1(svbool_t, int64_t const *);
	28344	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_mf8)))
	28345	svmfloat8_t svldnt1(svbool_t, mfloat8_t const *);
28096	28346	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_s16)))
28097	28347	svint16_t svldnt1(svbool_t, int16_t const *);
28098	28348	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_u8)))
...	...	@@ -28103,6 +28353,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_u64)))
28103	28353	svuint64_t svldnt1_vnum(svbool_t, uint64_t const *, int64_t);
28104	28354	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_u16)))
28105	28355	svuint16_t svldnt1_vnum(svbool_t, uint16_t const *, int64_t);
	28356	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_bf16)))
	28357	svbfloat16_t svldnt1_vnum(svbool_t, bfloat16_t const *, int64_t);
28106	28358	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_s8)))
28107	28359	svint8_t svldnt1_vnum(svbool_t, int8_t const *, int64_t);
28108	28360	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_f64)))
...	...	@@ -28115,6 +28367,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_s32)))
28115	28367	svint32_t svldnt1_vnum(svbool_t, int32_t const *, int64_t);
28116	28368	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_s64)))
28117	28369	svint64_t svldnt1_vnum(svbool_t, int64_t const *, int64_t);
	28370	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_mf8)))
	28371	svmfloat8_t svldnt1_vnum(svbool_t, mfloat8_t const *, int64_t);
28118	28372	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svldnt1_vnum_s16)))
28119	28373	svint16_t svldnt1_vnum(svbool_t, int16_t const *, int64_t);
28120	28374	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlen_u8)))
...	...	@@ -28125,6 +28379,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlen_u64)))
28125	28379	uint64_t svlen(svuint64_t);
28126	28380	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlen_u16)))
28127	28381	uint64_t svlen(svuint16_t);
	28382	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlen_bf16)))
	28383	uint64_t svlen(svbfloat16_t);
28128	28384	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlen_s8)))
28129	28385	uint64_t svlen(svint8_t);
28130	28386	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svlen_f64)))
...	...	@@ -30369,6 +30625,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrev_u64)))
30369	30625	svuint64_t svrev(svuint64_t);
30370	30626	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrev_u16)))
30371	30627	svuint16_t svrev(svuint16_t);
	30628	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrev_bf16)))
	30629	svbfloat16_t svrev(svbfloat16_t);
30372	30630	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrev_s8)))
30373	30631	svint8_t svrev(svint8_t);
30374	30632	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svrev_f64)))
...	...	@@ -30639,6 +30897,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsel_u64)))
30639	30897	svuint64_t svsel(svbool_t, svuint64_t, svuint64_t);
30640	30898	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsel_u16)))
30641	30899	svuint16_t svsel(svbool_t, svuint16_t, svuint16_t);
	30900	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsel_bf16)))
	30901	svbfloat16_t svsel(svbool_t, svbfloat16_t, svbfloat16_t);
30642	30902	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsel_s8)))
30643	30903	svint8_t svsel(svbool_t, svint8_t, svint8_t);
30644	30904	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsel_f64)))
...	...	@@ -30661,6 +30921,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset2_u64)))
30661	30921	svuint64x2_t svset2(svuint64x2_t, uint64_t, svuint64_t);
30662	30922	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset2_u16)))
30663	30923	svuint16x2_t svset2(svuint16x2_t, uint64_t, svuint16_t);
	30924	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset2_bf16)))
	30925	svbfloat16x2_t svset2(svbfloat16x2_t, uint64_t, svbfloat16_t);
30664	30926	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset2_s8)))
30665	30927	svint8x2_t svset2(svint8x2_t, uint64_t, svint8_t);
30666	30928	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset2_f64)))
...	...	@@ -30673,6 +30935,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset2_s32)))
30673	30935	svint32x2_t svset2(svint32x2_t, uint64_t, svint32_t);
30674	30936	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset2_s64)))
30675	30937	svint64x2_t svset2(svint64x2_t, uint64_t, svint64_t);
	30938	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset2_mf8)))
	30939	svmfloat8x2_t svset2(svmfloat8x2_t, uint64_t, svmfloat8_t);
30676	30940	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset2_s16)))
30677	30941	svint16x2_t svset2(svint16x2_t, uint64_t, svint16_t);
30678	30942	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset3_u8)))
...	...	@@ -30683,6 +30947,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset3_u64)))
30683	30947	svuint64x3_t svset3(svuint64x3_t, uint64_t, svuint64_t);
30684	30948	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset3_u16)))
30685	30949	svuint16x3_t svset3(svuint16x3_t, uint64_t, svuint16_t);
	30950	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset3_bf16)))
	30951	svbfloat16x3_t svset3(svbfloat16x3_t, uint64_t, svbfloat16_t);
30686	30952	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset3_s8)))
30687	30953	svint8x3_t svset3(svint8x3_t, uint64_t, svint8_t);
30688	30954	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset3_f64)))
...	...	@@ -30695,6 +30961,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset3_s32)))
30695	30961	svint32x3_t svset3(svint32x3_t, uint64_t, svint32_t);
30696	30962	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset3_s64)))
30697	30963	svint64x3_t svset3(svint64x3_t, uint64_t, svint64_t);
	30964	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset3_mf8)))
	30965	svmfloat8x3_t svset3(svmfloat8x3_t, uint64_t, svmfloat8_t);
30698	30966	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset3_s16)))
30699	30967	svint16x3_t svset3(svint16x3_t, uint64_t, svint16_t);
30700	30968	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset4_u8)))
...	...	@@ -30705,6 +30973,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset4_u64)))
30705	30973	svuint64x4_t svset4(svuint64x4_t, uint64_t, svuint64_t);
30706	30974	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset4_u16)))
30707	30975	svuint16x4_t svset4(svuint16x4_t, uint64_t, svuint16_t);
	30976	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset4_bf16)))
	30977	svbfloat16x4_t svset4(svbfloat16x4_t, uint64_t, svbfloat16_t);
30708	30978	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset4_s8)))
30709	30979	svint8x4_t svset4(svint8x4_t, uint64_t, svint8_t);
30710	30980	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset4_f64)))
...	...	@@ -30717,6 +30987,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset4_s32)))
30717	30987	svint32x4_t svset4(svint32x4_t, uint64_t, svint32_t);
30718	30988	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset4_s64)))
30719	30989	svint64x4_t svset4(svint64x4_t, uint64_t, svint64_t);
	30990	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset4_mf8)))
	30991	svmfloat8x4_t svset4(svmfloat8x4_t, uint64_t, svmfloat8_t);
30720	30992	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svset4_s16)))
30721	30993	svint16x4_t svset4(svint16x4_t, uint64_t, svint16_t);
30722	30994	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsplice_u8)))
...	...	@@ -30727,6 +30999,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsplice_u64)))
30727	30999	svuint64_t svsplice(svbool_t, svuint64_t, svuint64_t);
30728	31000	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsplice_u16)))
30729	31001	svuint16_t svsplice(svbool_t, svuint16_t, svuint16_t);
	31002	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsplice_bf16)))
	31003	svbfloat16_t svsplice(svbool_t, svbfloat16_t, svbfloat16_t);
30730	31004	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsplice_s8)))
30731	31005	svint8_t svsplice(svbool_t, svint8_t, svint8_t);
30732	31006	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsplice_f64)))
...	...	@@ -30767,6 +31041,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_u64)))
30767	31041	void svst1(svbool_t, uint64_t *, svuint64_t);
30768	31042	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_u16)))
30769	31043	void svst1(svbool_t, uint16_t *, svuint16_t);
	31044	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_bf16)))
	31045	void svst1(svbool_t, bfloat16_t *, svbfloat16_t);
30770	31046	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_s8)))
30771	31047	void svst1(svbool_t, int8_t *, svint8_t);
30772	31048	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_f64)))
...	...	@@ -30779,6 +31055,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_s32)))
30779	31055	void svst1(svbool_t, int32_t *, svint32_t);
30780	31056	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_s64)))
30781	31057	void svst1(svbool_t, int64_t *, svint64_t);
	31058	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_mf8)))
	31059	void svst1(svbool_t, mfloat8_t *, svmfloat8_t);
30782	31060	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_s16)))
30783	31061	void svst1(svbool_t, int16_t *, svint16_t);
30784	31062	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_u8)))
...	...	@@ -30789,6 +31067,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_u64)))
30789	31067	void svst1_vnum(svbool_t, uint64_t *, int64_t, svuint64_t);
30790	31068	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_u16)))
30791	31069	void svst1_vnum(svbool_t, uint16_t *, int64_t, svuint16_t);
	31070	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_bf16)))
	31071	void svst1_vnum(svbool_t, bfloat16_t *, int64_t, svbfloat16_t);
30792	31072	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_s8)))
30793	31073	void svst1_vnum(svbool_t, int8_t *, int64_t, svint8_t);
30794	31074	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_f64)))
...	...	@@ -30801,6 +31081,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_s32)))
30801	31081	void svst1_vnum(svbool_t, int32_t *, int64_t, svint32_t);
30802	31082	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_s64)))
30803	31083	void svst1_vnum(svbool_t, int64_t *, int64_t, svint64_t);
	31084	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_mf8)))
	31085	void svst1_vnum(svbool_t, mfloat8_t *, int64_t, svmfloat8_t);
30804	31086	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1_vnum_s16)))
30805	31087	void svst1_vnum(svbool_t, int16_t *, int64_t, svint16_t);
30806	31088	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst1b_s32)))
...	...	@@ -30859,6 +31141,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_u64)))
30859	31141	void svst2(svbool_t, uint64_t *, svuint64x2_t);
30860	31142	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_u16)))
30861	31143	void svst2(svbool_t, uint16_t *, svuint16x2_t);
	31144	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_bf16)))
	31145	void svst2(svbool_t, bfloat16_t *, svbfloat16x2_t);
30862	31146	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_s8)))
30863	31147	void svst2(svbool_t, int8_t *, svint8x2_t);
30864	31148	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_f64)))
...	...	@@ -30871,6 +31155,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_s32)))
30871	31155	void svst2(svbool_t, int32_t *, svint32x2_t);
30872	31156	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_s64)))
30873	31157	void svst2(svbool_t, int64_t *, svint64x2_t);
	31158	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_mf8)))
	31159	void svst2(svbool_t, mfloat8_t *, svmfloat8x2_t);
30874	31160	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_s16)))
30875	31161	void svst2(svbool_t, int16_t *, svint16x2_t);
30876	31162	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_vnum_u8)))
...	...	@@ -30881,6 +31167,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_vnum_u64)))
30881	31167	void svst2_vnum(svbool_t, uint64_t *, int64_t, svuint64x2_t);
30882	31168	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_vnum_u16)))
30883	31169	void svst2_vnum(svbool_t, uint16_t *, int64_t, svuint16x2_t);
	31170	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_vnum_bf16)))
	31171	void svst2_vnum(svbool_t, bfloat16_t *, int64_t, svbfloat16x2_t);
30884	31172	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_vnum_s8)))
30885	31173	void svst2_vnum(svbool_t, int8_t *, int64_t, svint8x2_t);
30886	31174	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_vnum_f64)))
...	...	@@ -30893,6 +31181,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_vnum_s32)))
30893	31181	void svst2_vnum(svbool_t, int32_t *, int64_t, svint32x2_t);
30894	31182	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_vnum_s64)))
30895	31183	void svst2_vnum(svbool_t, int64_t *, int64_t, svint64x2_t);
	31184	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_vnum_mf8)))
	31185	void svst2_vnum(svbool_t, mfloat8_t *, int64_t, svmfloat8x2_t);
30896	31186	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst2_vnum_s16)))
30897	31187	void svst2_vnum(svbool_t, int16_t *, int64_t, svint16x2_t);
30898	31188	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_u8)))
...	...	@@ -30903,6 +31193,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_u64)))
30903	31193	void svst3(svbool_t, uint64_t *, svuint64x3_t);
30904	31194	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_u16)))
30905	31195	void svst3(svbool_t, uint16_t *, svuint16x3_t);
	31196	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_bf16)))
	31197	void svst3(svbool_t, bfloat16_t *, svbfloat16x3_t);
30906	31198	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_s8)))
30907	31199	void svst3(svbool_t, int8_t *, svint8x3_t);
30908	31200	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_f64)))
...	...	@@ -30915,6 +31207,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_s32)))
30915	31207	void svst3(svbool_t, int32_t *, svint32x3_t);
30916	31208	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_s64)))
30917	31209	void svst3(svbool_t, int64_t *, svint64x3_t);
	31210	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_mf8)))
	31211	void svst3(svbool_t, mfloat8_t *, svmfloat8x3_t);
30918	31212	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_s16)))
30919	31213	void svst3(svbool_t, int16_t *, svint16x3_t);
30920	31214	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_vnum_u8)))
...	...	@@ -30925,6 +31219,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_vnum_u64)))
30925	31219	void svst3_vnum(svbool_t, uint64_t *, int64_t, svuint64x3_t);
30926	31220	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_vnum_u16)))
30927	31221	void svst3_vnum(svbool_t, uint16_t *, int64_t, svuint16x3_t);
	31222	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_vnum_bf16)))
	31223	void svst3_vnum(svbool_t, bfloat16_t *, int64_t, svbfloat16x3_t);
30928	31224	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_vnum_s8)))
30929	31225	void svst3_vnum(svbool_t, int8_t *, int64_t, svint8x3_t);
30930	31226	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_vnum_f64)))
...	...	@@ -30937,6 +31233,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_vnum_s32)))
30937	31233	void svst3_vnum(svbool_t, int32_t *, int64_t, svint32x3_t);
30938	31234	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_vnum_s64)))
30939	31235	void svst3_vnum(svbool_t, int64_t *, int64_t, svint64x3_t);
	31236	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_vnum_mf8)))
	31237	void svst3_vnum(svbool_t, mfloat8_t *, int64_t, svmfloat8x3_t);
30940	31238	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst3_vnum_s16)))
30941	31239	void svst3_vnum(svbool_t, int16_t *, int64_t, svint16x3_t);
30942	31240	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_u8)))
...	...	@@ -30947,6 +31245,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_u64)))
30947	31245	void svst4(svbool_t, uint64_t *, svuint64x4_t);
30948	31246	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_u16)))
30949	31247	void svst4(svbool_t, uint16_t *, svuint16x4_t);
	31248	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_bf16)))
	31249	void svst4(svbool_t, bfloat16_t *, svbfloat16x4_t);
30950	31250	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_s8)))
30951	31251	void svst4(svbool_t, int8_t *, svint8x4_t);
30952	31252	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_f64)))
...	...	@@ -30959,6 +31259,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_s32)))
30959	31259	void svst4(svbool_t, int32_t *, svint32x4_t);
30960	31260	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_s64)))
30961	31261	void svst4(svbool_t, int64_t *, svint64x4_t);
	31262	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_mf8)))
	31263	void svst4(svbool_t, mfloat8_t *, svmfloat8x4_t);
30962	31264	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_s16)))
30963	31265	void svst4(svbool_t, int16_t *, svint16x4_t);
30964	31266	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_vnum_u8)))
...	...	@@ -30969,6 +31271,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_vnum_u64)))
30969	31271	void svst4_vnum(svbool_t, uint64_t *, int64_t, svuint64x4_t);
30970	31272	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_vnum_u16)))
30971	31273	void svst4_vnum(svbool_t, uint16_t *, int64_t, svuint16x4_t);
	31274	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_vnum_bf16)))
	31275	void svst4_vnum(svbool_t, bfloat16_t *, int64_t, svbfloat16x4_t);
30972	31276	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_vnum_s8)))
30973	31277	void svst4_vnum(svbool_t, int8_t *, int64_t, svint8x4_t);
30974	31278	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_vnum_f64)))
...	...	@@ -30981,6 +31285,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_vnum_s32)))
30981	31285	void svst4_vnum(svbool_t, int32_t *, int64_t, svint32x4_t);
30982	31286	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_vnum_s64)))
30983	31287	void svst4_vnum(svbool_t, int64_t *, int64_t, svint64x4_t);
	31288	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_vnum_mf8)))
	31289	void svst4_vnum(svbool_t, mfloat8_t *, int64_t, svmfloat8x4_t);
30984	31290	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svst4_vnum_s16)))
30985	31291	void svst4_vnum(svbool_t, int16_t *, int64_t, svint16x4_t);
30986	31292	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_u8)))
...	...	@@ -30991,6 +31297,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_u64)))
30991	31297	void svstnt1(svbool_t, uint64_t *, svuint64_t);
30992	31298	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_u16)))
30993	31299	void svstnt1(svbool_t, uint16_t *, svuint16_t);
	31300	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_bf16)))
	31301	void svstnt1(svbool_t, bfloat16_t *, svbfloat16_t);
30994	31302	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_s8)))
30995	31303	void svstnt1(svbool_t, int8_t *, svint8_t);
30996	31304	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_f64)))
...	...	@@ -31003,6 +31311,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_s32)))
31003	31311	void svstnt1(svbool_t, int32_t *, svint32_t);
31004	31312	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_s64)))
31005	31313	void svstnt1(svbool_t, int64_t *, svint64_t);
	31314	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_mf8)))
	31315	void svstnt1(svbool_t, mfloat8_t *, svmfloat8_t);
31006	31316	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_s16)))
31007	31317	void svstnt1(svbool_t, int16_t *, svint16_t);
31008	31318	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_u8)))
...	...	@@ -31013,6 +31323,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_u64)))
31013	31323	void svstnt1_vnum(svbool_t, uint64_t *, int64_t, svuint64_t);
31014	31324	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_u16)))
31015	31325	void svstnt1_vnum(svbool_t, uint16_t *, int64_t, svuint16_t);
	31326	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_bf16)))
	31327	void svstnt1_vnum(svbool_t, bfloat16_t *, int64_t, svbfloat16_t);
31016	31328	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_s8)))
31017	31329	void svstnt1_vnum(svbool_t, int8_t *, int64_t, svint8_t);
31018	31330	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_f64)))
...	...	@@ -31025,6 +31337,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_s32)))
31025	31337	void svstnt1_vnum(svbool_t, int32_t *, int64_t, svint32_t);
31026	31338	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_s64)))
31027	31339	void svstnt1_vnum(svbool_t, int64_t *, int64_t, svint64_t);
	31340	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_mf8)))
	31341	void svstnt1_vnum(svbool_t, mfloat8_t *, int64_t, svmfloat8_t);
31028	31342	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svstnt1_vnum_s16)))
31029	31343	void svstnt1_vnum(svbool_t, int16_t *, int64_t, svint16_t);
31030	31344	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svsub_n_f64_m)))
...	...	@@ -31299,6 +31613,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbl_u64)))
31299	31613	svuint64_t svtbl(svuint64_t, svuint64_t);
31300	31614	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbl_u16)))
31301	31615	svuint16_t svtbl(svuint16_t, svuint16_t);
	31616	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbl_bf16)))
	31617	svbfloat16_t svtbl(svbfloat16_t, svuint16_t);
31302	31618	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbl_s8)))
31303	31619	svint8_t svtbl(svint8_t, svuint8_t);
31304	31620	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtbl_f64)))
...	...	@@ -31321,6 +31637,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1_u64)))
31321	31637	svuint64_t svtrn1(svuint64_t, svuint64_t);
31322	31638	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1_u16)))
31323	31639	svuint16_t svtrn1(svuint16_t, svuint16_t);
	31640	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1_bf16)))
	31641	svbfloat16_t svtrn1(svbfloat16_t, svbfloat16_t);
31324	31642	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1_s8)))
31325	31643	svint8_t svtrn1(svint8_t, svint8_t);
31326	31644	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn1_f64)))
...	...	@@ -31343,6 +31661,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2_u64)))
31343	31661	svuint64_t svtrn2(svuint64_t, svuint64_t);
31344	31662	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2_u16)))
31345	31663	svuint16_t svtrn2(svuint16_t, svuint16_t);
	31664	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2_bf16)))
	31665	svbfloat16_t svtrn2(svbfloat16_t, svbfloat16_t);
31346	31666	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2_s8)))
31347	31667	svint8_t svtrn2(svint8_t, svint8_t);
31348	31668	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svtrn2_f64)))
...	...	@@ -31393,6 +31713,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1_u64)))
31393	31713	svuint64_t svuzp1(svuint64_t, svuint64_t);
31394	31714	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1_u16)))
31395	31715	svuint16_t svuzp1(svuint16_t, svuint16_t);
	31716	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1_bf16)))
	31717	svbfloat16_t svuzp1(svbfloat16_t, svbfloat16_t);
31396	31718	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1_s8)))
31397	31719	svint8_t svuzp1(svint8_t, svint8_t);
31398	31720	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp1_f64)))
...	...	@@ -31415,6 +31737,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2_u64)))
31415	31737	svuint64_t svuzp2(svuint64_t, svuint64_t);
31416	31738	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2_u16)))
31417	31739	svuint16_t svuzp2(svuint16_t, svuint16_t);
	31740	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2_bf16)))
	31741	svbfloat16_t svuzp2(svbfloat16_t, svbfloat16_t);
31418	31742	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2_s8)))
31419	31743	svint8_t svuzp2(svint8_t, svint8_t);
31420	31744	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svuzp2_f64)))
...	...	@@ -31501,6 +31825,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1_u64)))
31501	31825	svuint64_t svzip1(svuint64_t, svuint64_t);
31502	31826	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1_u16)))
31503	31827	svuint16_t svzip1(svuint16_t, svuint16_t);
	31828	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1_bf16)))
	31829	svbfloat16_t svzip1(svbfloat16_t, svbfloat16_t);
31504	31830	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1_s8)))
31505	31831	svint8_t svzip1(svint8_t, svint8_t);
31506	31832	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip1_f64)))
...	...	@@ -31523,6 +31849,8 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2_u64)))
31523	31849	svuint64_t svzip2(svuint64_t, svuint64_t);
31524	31850	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2_u16)))
31525	31851	svuint16_t svzip2(svuint16_t, svuint16_t);
	31852	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2_bf16)))
	31853	svbfloat16_t svzip2(svbfloat16_t, svbfloat16_t);
31526	31854	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2_s8)))
31527	31855	svint8_t svzip2(svint8_t, svint8_t);
31528	31856	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2_f64)))
...	...	@@ -31537,6 +31865,18 @@ __aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2_s64)))
31537	31865	svint64_t svzip2(svint64_t, svint64_t);
31538	31866	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svzip2_s16)))
31539	31867	svint16_t svzip2(svint16_t, svint16_t);
	31868	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svexpa_f64)))
	31869	svfloat64_t svexpa_f64(svuint64_t);
	31870	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svexpa_f32)))
	31871	svfloat32_t svexpa_f32(svuint32_t);
	31872	__ai __attribute__((__clang_arm_builtin_alias(__builtin_sve_svexpa_f16)))
	31873	svfloat16_t svexpa_f16(svuint16_t);
	31874	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svexpa_f64)))
	31875	svfloat64_t svexpa(svuint64_t);
	31876	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svexpa_f32)))
	31877	svfloat32_t svexpa(svuint32_t);
	31878	__aio __attribute__((__clang_arm_builtin_alias(__builtin_sve_svexpa_f16)))
	31879	svfloat16_t svexpa(svuint16_t);
31540	31880	#define svcvtnt_bf16_x svcvtnt_bf16_m
31541	31881	#define svcvtnt_bf16_f32_x svcvtnt_bf16_f32_m
31542	31882	#define svcvtnt_f16_x svcvtnt_f16_m

lib/include/avx10_2_512bf16intrin.h+8-8

...	...	@@ -441,8 +441,8 @@ _mm512_maskz_sqrt_pbh(__mmask32 __U, __m512bh __A) {
441	441
442	442	static __inline__ __m512bh __DEFAULT_FN_ATTRS512
443	443	_mm512_fmadd_pbh(__m512bh __A, __m512bh __B, __m512bh __C) {
444		return (__m512bh)__builtin_ia32_vfmaddnepbh512((__v32bf)__A, (__v32bf)__B,
445		(__v32bf)__C);
	444	return (__m512bh)__builtin_ia32_vfmaddbf16512((__v32bf)__A, (__v32bf)__B,
	445	(__v32bf)__C);
446	446	}
447	447
448	448	static __inline__ __m512bh __DEFAULT_FN_ATTRS512
...	...	@@ -469,8 +469,8 @@ static __inline__ __m512bh __DEFAULT_FN_ATTRS512 _mm512_maskz_fmadd_pbh(
469	469
470	470	static __inline__ __m512bh __DEFAULT_FN_ATTRS512
471	471	_mm512_fmsub_pbh(__m512bh __A, __m512bh __B, __m512bh __C) {
472		return (__m512bh)__builtin_ia32_vfmaddnepbh512((__v32bf)__A, (__v32bf)__B,
473		-(__v32bf)__C);
	472	return (__m512bh)__builtin_ia32_vfmaddbf16512((__v32bf)__A, (__v32bf)__B,
	473	-(__v32bf)__C);
474	474	}
475	475
476	476	static __inline__ __m512bh __DEFAULT_FN_ATTRS512
...	...	@@ -497,8 +497,8 @@ static __inline__ __m512bh __DEFAULT_FN_ATTRS512 _mm512_maskz_fmsub_pbh(
497	497
498	498	static __inline__ __m512bh __DEFAULT_FN_ATTRS512
499	499	_mm512_fnmadd_pbh(__m512bh __A, __m512bh __B, __m512bh __C) {
500		return (__m512bh)__builtin_ia32_vfmaddnepbh512((__v32bf)__A, -(__v32bf)__B,
501		(__v32bf)__C);
	500	return (__m512bh)__builtin_ia32_vfmaddbf16512((__v32bf)__A, -(__v32bf)__B,
	501	(__v32bf)__C);
502	502	}
503	503
504	504	static __inline__ __m512bh __DEFAULT_FN_ATTRS512 _mm512_mask_fnmadd_pbh(
...	...	@@ -527,8 +527,8 @@ static __inline__ __m512bh __DEFAULT_FN_ATTRS512 _mm512_maskz_fnmadd_pbh(
527	527
528	528	static __inline__ __m512bh __DEFAULT_FN_ATTRS512
529	529	_mm512_fnmsub_pbh(__m512bh __A, __m512bh __B, __m512bh __C) {
530		return (__m512bh)__builtin_ia32_vfmaddnepbh512((__v32bf)__A, -(__v32bf)__B,
531		-(__v32bf)__C);
	530	return (__m512bh)__builtin_ia32_vfmaddbf16512((__v32bf)__A, -(__v32bf)__B,
	531	-(__v32bf)__C);
532	532	}
533	533
534	534	static __inline__ __m512bh __DEFAULT_FN_ATTRS512 _mm512_mask_fnmsub_pbh(

lib/include/avx10_2_512convertintrin.h+27-25

...	...	@@ -78,20 +78,20 @@ _mm512_maskz_cvtbiasph_bf8(__mmask32 __U, __m512i __A, __m512h __B) {
78	78	}
79	79
80	80	static __inline__ __m256i __DEFAULT_FN_ATTRS512
81		_mm512_cvtbiassph_bf8(__m512i __A, __m512h __B) {
	81	_mm512_cvts_biasph_bf8(__m512i __A, __m512h __B) {
82	82	return (__m256i)__builtin_ia32_vcvtbiasph2bf8s_512_mask(
83	83	(__v64qi)__A, (__v32hf)__B, (__v32qi)_mm256_undefined_si256(),
84	84	(__mmask32)-1);
85	85	}
86	86
87		static __inline__ __m256i __DEFAULT_FN_ATTRS512 _mm512_mask_cvtbiassph_bf8(
	87	static __inline__ __m256i __DEFAULT_FN_ATTRS512 _mm512_mask_cvts_biasph_bf8(
88	88	__m256i __W, __mmask32 __U, __m512i __A, __m512h __B) {
89	89	return (__m256i)__builtin_ia32_vcvtbiasph2bf8s_512_mask(
90	90	(__v64qi)__A, (__v32hf)__B, (__v32qi)(__m256i)__W, (__mmask32)__U);
91	91	}
92	92
93	93	static __inline__ __m256i __DEFAULT_FN_ATTRS512
94		_mm512_maskz_cvtbiassph_bf8(__mmask32 __U, __m512i __A, __m512h __B) {
	94	_mm512_maskz_cvts_biasph_bf8(__mmask32 __U, __m512i __A, __m512h __B) {
95	95	return (__m256i)__builtin_ia32_vcvtbiasph2bf8s_512_mask(
96	96	(__v64qi)__A, (__v32hf)__B, (__v32qi)(__m256i)_mm256_setzero_si256(),
97	97	(__mmask32)__U);
...	...	@@ -118,20 +118,20 @@ _mm512_maskz_cvtbiasph_hf8(__mmask32 __U, __m512i __A, __m512h __B) {
118	118	}
119	119
120	120	static __inline__ __m256i __DEFAULT_FN_ATTRS512
121		_mm512_cvtbiassph_hf8(__m512i __A, __m512h __B) {
	121	_mm512_cvts_biasph_hf8(__m512i __A, __m512h __B) {
122	122	return (__m256i)__builtin_ia32_vcvtbiasph2hf8s_512_mask(
123	123	(__v64qi)__A, (__v32hf)__B, (__v32qi)_mm256_undefined_si256(),
124	124	(__mmask32)-1);
125	125	}
126	126
127		static __inline__ __m256i __DEFAULT_FN_ATTRS512 _mm512_mask_cvtbiassph_hf8(
	127	static __inline__ __m256i __DEFAULT_FN_ATTRS512 _mm512_mask_cvts_biasph_hf8(
128	128	__m256i __W, __mmask32 __U, __m512i __A, __m512h __B) {
129	129	return (__m256i)__builtin_ia32_vcvtbiasph2hf8s_512_mask(
130	130	(__v64qi)__A, (__v32hf)__B, (__v32qi)(__m256i)__W, (__mmask32)__U);
131	131	}
132	132
133	133	static __inline__ __m256i __DEFAULT_FN_ATTRS512
134		_mm512_maskz_cvtbiassph_hf8(__mmask32 __U, __m512i __A, __m512h __B) {
	134	_mm512_maskz_cvts_biasph_hf8(__mmask32 __U, __m512i __A, __m512h __B) {
135	135	return (__m256i)__builtin_ia32_vcvtbiasph2hf8s_512_mask(
136	136	(__v64qi)__A, (__v32hf)__B, (__v32qi)(__m256i)_mm256_setzero_si256(),
137	137	(__mmask32)__U);
...	...	@@ -157,21 +157,21 @@ _mm512_maskz_cvt2ph_bf8(__mmask64 __U, __m512h __A, __m512h __B) {
157	157	}
158	158
159	159	static __inline__ __m512i __DEFAULT_FN_ATTRS512
160		_mm512_cvts2ph_bf8(__m512h __A, __m512h __B) {
	160	_mm512_cvts_2ph_bf8(__m512h __A, __m512h __B) {
161	161	return (__m512i)__builtin_ia32_vcvt2ph2bf8s_512((__v32hf)(__A),
162	162	(__v32hf)(__B));
163	163	}
164	164
165	165	static __inline__ __m512i __DEFAULT_FN_ATTRS512
166		_mm512_mask_cvts2ph_bf8(__m512i __W, __mmask64 __U, __m512h __A, __m512h __B) {
	166	_mm512_mask_cvts_2ph_bf8(__m512i __W, __mmask64 __U, __m512h __A, __m512h __B) {
167	167	return (__m512i)__builtin_ia32_selectb_512(
168		(__mmask64)__U, (__v64qi)_mm512_cvts2ph_bf8(__A, __B), (__v64qi)__W);
	168	(__mmask64)__U, (__v64qi)_mm512_cvts_2ph_bf8(__A, __B), (__v64qi)__W);
169	169	}
170	170
171	171	static __inline__ __m512i __DEFAULT_FN_ATTRS512
172		_mm512_maskz_cvts2ph_bf8(__mmask64 __U, __m512h __A, __m512h __B) {
	172	_mm512_maskz_cvts_2ph_bf8(__mmask64 __U, __m512h __A, __m512h __B) {
173	173	return (__m512i)__builtin_ia32_selectb_512(
174		(__mmask64)__U, (__v64qi)_mm512_cvts2ph_bf8(__A, __B),
	174	(__mmask64)__U, (__v64qi)_mm512_cvts_2ph_bf8(__A, __B),
175	175	(__v64qi)(__m512i)_mm512_setzero_si512());
176	176	}
177	177
...	...	@@ -195,37 +195,37 @@ _mm512_maskz_cvt2ph_hf8(__mmask64 __U, __m512h __A, __m512h __B) {
195	195	}
196	196
197	197	static __inline__ __m512i __DEFAULT_FN_ATTRS512
198		_mm512_cvts2ph_hf8(__m512h __A, __m512h __B) {
	198	_mm512_cvts_2ph_hf8(__m512h __A, __m512h __B) {
199	199	return (__m512i)__builtin_ia32_vcvt2ph2hf8s_512((__v32hf)(__A),
200	200	(__v32hf)(__B));
201	201	}
202	202
203	203	static __inline__ __m512i __DEFAULT_FN_ATTRS512
204		_mm512_mask_cvts2ph_hf8(__m512i __W, __mmask64 __U, __m512h __A, __m512h __B) {
	204	_mm512_mask_cvts_2ph_hf8(__m512i __W, __mmask64 __U, __m512h __A, __m512h __B) {
205	205	return (__m512i)__builtin_ia32_selectb_512(
206		(__mmask64)__U, (__v64qi)_mm512_cvts2ph_hf8(__A, __B), (__v64qi)__W);
	206	(__mmask64)__U, (__v64qi)_mm512_cvts_2ph_hf8(__A, __B), (__v64qi)__W);
207	207	}
208	208
209	209	static __inline__ __m512i __DEFAULT_FN_ATTRS512
210		_mm512_maskz_cvts2ph_hf8(__mmask64 __U, __m512h __A, __m512h __B) {
	210	_mm512_maskz_cvts_2ph_hf8(__mmask64 __U, __m512h __A, __m512h __B) {
211	211	return (__m512i)__builtin_ia32_selectb_512(
212		(__mmask64)__U, (__v64qi)_mm512_cvts2ph_hf8(__A, __B),
	212	(__mmask64)__U, (__v64qi)_mm512_cvts_2ph_hf8(__A, __B),
213	213	(__v64qi)(__m512i)_mm512_setzero_si512());
214	214	}
215	215
216		static __inline__ __m512h __DEFAULT_FN_ATTRS512 _mm512_cvthf8(__m256i __A) {
	216	static __inline__ __m512h __DEFAULT_FN_ATTRS512 _mm512_cvthf8_ph(__m256i __A) {
217	217	return (__m512h)__builtin_ia32_vcvthf8_2ph512_mask(
218	218	(__v32qi)__A, (__v32hf)(__m512h)_mm512_undefined_ph(), (__mmask32)-1);
219	219	}
220	220
221	221	static __inline__ __m512h __DEFAULT_FN_ATTRS512
222		_mm512_mask_cvthf8(__m512h __W, __mmask32 __U, __m256i __A) {
	222	_mm512_mask_cvthf8_ph(__m512h __W, __mmask32 __U, __m256i __A) {
223	223	return (__m512h)__builtin_ia32_vcvthf8_2ph512_mask(
224	224	(__v32qi)__A, (__v32hf)(__m512h)__W, (__mmask32)__U);
225	225	}
226	226
227	227	static __inline__ __m512h __DEFAULT_FN_ATTRS512
228		_mm512_maskz_cvthf8(__mmask32 __U, __m256i __A) {
	228	_mm512_maskz_cvthf8_ph(__mmask32 __U, __m256i __A) {
229	229	return (__m512h)__builtin_ia32_vcvthf8_2ph512_mask(
230	230	(__v32qi)__A, (__v32hf)(__m512h)_mm512_setzero_ph(), (__mmask32)__U);
231	231	}
...	...	@@ -247,19 +247,20 @@ _mm512_maskz_cvtph_bf8(__mmask32 __U, __m512h __A) {
247	247	(__v32hf)__A, (__v32qi)(__m256i)_mm256_setzero_si256(), (__mmask32)__U);
248	248	}
249	249
250		static __inline__ __m256i __DEFAULT_FN_ATTRS512 _mm512_cvtsph_bf8(__m512h __A) {
	250	static __inline__ __m256i __DEFAULT_FN_ATTRS512
	251	_mm512_cvts_ph_bf8(__m512h __A) {
251	252	return (__m256i)__builtin_ia32_vcvtph2bf8s_512_mask(
252	253	(__v32hf)__A, (__v32qi)(__m256i)_mm256_undefined_si256(), (__mmask32)-1);
253	254	}
254	255
255	256	static __inline__ __m256i __DEFAULT_FN_ATTRS512
256		_mm512_mask_cvtsph_bf8(__m256i __W, __mmask32 __U, __m512h __A) {
	257	_mm512_mask_cvts_ph_bf8(__m256i __W, __mmask32 __U, __m512h __A) {
257	258	return (__m256i)__builtin_ia32_vcvtph2bf8s_512_mask(
258	259	(__v32hf)__A, (__v32qi)(__m256i)__W, (__mmask32)__U);
259	260	}
260	261
261	262	static __inline__ __m256i __DEFAULT_FN_ATTRS512
262		_mm512_maskz_cvtsph_bf8(__mmask32 __U, __m512h __A) {
	263	_mm512_maskz_cvts_ph_bf8(__mmask32 __U, __m512h __A) {
263	264	return (__m256i)__builtin_ia32_vcvtph2bf8s_512_mask(
264	265	(__v32hf)__A, (__v32qi)(__m256i)_mm256_setzero_si256(), (__mmask32)__U);
265	266	}
...	...	@@ -281,19 +282,20 @@ _mm512_maskz_cvtph_hf8(__mmask32 __U, __m512h __A) {
281	282	(__v32hf)__A, (__v32qi)(__m256i)_mm256_setzero_si256(), (__mmask32)__U);
282	283	}
283	284
284		static __inline__ __m256i __DEFAULT_FN_ATTRS512 _mm512_cvtsph_hf8(__m512h __A) {
	285	static __inline__ __m256i __DEFAULT_FN_ATTRS512
	286	_mm512_cvts_ph_hf8(__m512h __A) {
285	287	return (__m256i)__builtin_ia32_vcvtph2hf8s_512_mask(
286	288	(__v32hf)__A, (__v32qi)(__m256i)_mm256_undefined_si256(), (__mmask32)-1);
287	289	}
288	290
289	291	static __inline__ __m256i __DEFAULT_FN_ATTRS512
290		_mm512_mask_cvtsph_hf8(__m256i __W, __mmask32 __U, __m512h __A) {
	292	_mm512_mask_cvts_ph_hf8(__m256i __W, __mmask32 __U, __m512h __A) {
291	293	return (__m256i)__builtin_ia32_vcvtph2hf8s_512_mask(
292	294	(__v32hf)__A, (__v32qi)(__m256i)__W, (__mmask32)__U);
293	295	}
294	296
295	297	static __inline__ __m256i __DEFAULT_FN_ATTRS512
296		_mm512_maskz_cvtsph_hf8(__mmask32 __U, __m512h __A) {
	298	_mm512_maskz_cvts_ph_hf8(__mmask32 __U, __m512h __A) {
297	299	return (__m256i)__builtin_ia32_vcvtph2hf8s_512_mask(
298	300	(__v32hf)__A, (__v32qi)(__m256i)_mm256_setzero_si256(), (__mmask32)__U);
299	301	}

lib/include/avx10_2_512satcvtdsintrin.h+28-24

...	...	@@ -20,20 +20,21 @@
20	20	__min_vector_width__(512)))
21	21
22	22	// 512 bit : Double -> Int
23		static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_cvttspd_epi32(__m512d __A) {
	23	static __inline__ __m256i __DEFAULT_FN_ATTRS
	24	_mm512_cvtts_pd_epi32(__m512d __A) {
24	25	return ((__m256i)__builtin_ia32_vcvttpd2dqs512_round_mask(
25	26	(__v8df)__A, (__v8si)_mm256_undefined_si256(), (__mmask8)-1,
26	27	_MM_FROUND_CUR_DIRECTION));
27	28	}
28	29
29	30	static __inline__ __m256i __DEFAULT_FN_ATTRS
30		_mm512_mask_cvttspd_epi32(__m256i __W, __mmask8 __U, __m512d __A) {
	31	_mm512_mask_cvtts_pd_epi32(__m256i __W, __mmask8 __U, __m512d __A) {
31	32	return ((__m256i)__builtin_ia32_vcvttpd2dqs512_round_mask(
32	33	(__v8df)__A, (__v8si)__W, __U, _MM_FROUND_CUR_DIRECTION));
33	34	}
34	35
35	36	static __inline__ __m256i __DEFAULT_FN_ATTRS
36		_mm512_maskz_cvttspd_epi32(__mmask8 __U, __m512d __A) {
	37	_mm512_maskz_cvtts_pd_epi32(__mmask8 __U, __m512d __A) {
37	38	return ((__m256i)__builtin_ia32_vcvttpd2dqs512_round_mask(
38	39	(__v8df)__A, (__v8si)_mm256_setzero_si256(), __U,
39	40	_MM_FROUND_CUR_DIRECTION));
...	...	@@ -55,20 +56,21 @@ _mm512_maskz_cvttspd_epi32(__mmask8 __U, __m512d __A) {
55	56	(const int)(__R)))
56	57
57	58	// 512 bit : Double -> uInt
58		static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_cvttspd_epu32(__m512d __A) {
	59	static __inline__ __m256i __DEFAULT_FN_ATTRS
	60	_mm512_cvtts_pd_epu32(__m512d __A) {
59	61	return ((__m256i)__builtin_ia32_vcvttpd2udqs512_round_mask(
60	62	(__v8df)__A, (__v8si)_mm256_undefined_si256(), (__mmask8)-1,
61	63	_MM_FROUND_CUR_DIRECTION));
62	64	}
63	65
64	66	static __inline__ __m256i __DEFAULT_FN_ATTRS
65		_mm512_mask_cvttspd_epu32(__m256i __W, __mmask8 __U, __m512d __A) {
	67	_mm512_mask_cvtts_pd_epu32(__m256i __W, __mmask8 __U, __m512d __A) {
66	68	return ((__m256i)__builtin_ia32_vcvttpd2udqs512_round_mask(
67	69	(__v8df)__A, (__v8si)__W, __U, _MM_FROUND_CUR_DIRECTION));
68	70	}
69	71
70	72	static __inline__ __m256i __DEFAULT_FN_ATTRS
71		_mm512_maskz_cvttspd_epu32(__mmask8 __U, __m512d __A) {
	73	_mm512_maskz_cvtts_pd_epu32(__mmask8 __U, __m512d __A) {
72	74	return ((__m256i)__builtin_ia32_vcvttpd2udqs512_round_mask(
73	75	(__v8df)__A, (__v8si)_mm256_setzero_si256(), __U,
74	76	_MM_FROUND_CUR_DIRECTION));
...	...	@@ -91,18 +93,19 @@ _mm512_maskz_cvttspd_epu32(__mmask8 __U, __m512d __A) {
91	93
92	94	// 512 bit : Double -> Long
93	95
94		static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_cvttspd_epi64(__m512d __A) {
	96	static __inline__ __m512i __DEFAULT_FN_ATTRS
	97	_mm512_cvtts_pd_epi64(__m512d __A) {
95	98	return ((__m512i)__builtin_ia32_vcvttpd2qqs512_round_mask(
96	99	(__v8df)__A, (__v8di)_mm512_undefined_epi32(), (__mmask8)-1,
97	100	_MM_FROUND_CUR_DIRECTION));
98	101	}
99	102	static __inline__ __m512i __DEFAULT_FN_ATTRS
100		_mm512_mask_cvttspd_epi64(__m512i __W, __mmask8 __U, __m512d __A) {
	103	_mm512_mask_cvtts_pd_epi64(__m512i __W, __mmask8 __U, __m512d __A) {
101	104	return ((__m512i)__builtin_ia32_vcvttpd2qqs512_round_mask(
102	105	(__v8df)__A, (__v8di)__W, __U, _MM_FROUND_CUR_DIRECTION));
103	106	}
104	107	static __inline__ __m512i __DEFAULT_FN_ATTRS
105		_mm512_maskz_cvttspd_epi64(__mmask8 __U, __m512d __A) {
	108	_mm512_maskz_cvtts_pd_epi64(__mmask8 __U, __m512d __A) {
106	109	return ((__m512i)__builtin_ia32_vcvttpd2qqs512_round_mask(
107	110	(__v8df)__A, (__v8di)_mm512_setzero_si512(), __U,
108	111	_MM_FROUND_CUR_DIRECTION));
...	...	@@ -125,20 +128,21 @@ _mm512_maskz_cvttspd_epi64(__mmask8 __U, __m512d __A) {
125	128
126	129	// 512 bit : Double -> ULong
127	130
128		static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_cvttspd_epu64(__m512d __A) {
	131	static __inline__ __m512i __DEFAULT_FN_ATTRS
	132	_mm512_cvtts_pd_epu64(__m512d __A) {
129	133	return ((__m512i)__builtin_ia32_vcvttpd2uqqs512_round_mask(
130	134	(__v8df)__A, (__v8di)_mm512_undefined_epi32(), (__mmask8)-1,
131	135	_MM_FROUND_CUR_DIRECTION));
132	136	}
133	137
134	138	static __inline__ __m512i __DEFAULT_FN_ATTRS
135		_mm512_mask_cvttspd_epu64(__m512i __W, __mmask8 __U, __m512d __A) {
	139	_mm512_mask_cvtts_pd_epu64(__m512i __W, __mmask8 __U, __m512d __A) {
136	140	return ((__m512i)__builtin_ia32_vcvttpd2uqqs512_round_mask(
137	141	(__v8df)__A, (__v8di)__W, __U, _MM_FROUND_CUR_DIRECTION));
138	142	}
139	143
140	144	static __inline__ __m512i __DEFAULT_FN_ATTRS
141		_mm512_maskz_cvttspd_epu64(__mmask8 __U, __m512d __A) {
	145	_mm512_maskz_cvtts_pd_epu64(__mmask8 __U, __m512d __A) {
142	146	return ((__m512i)__builtin_ia32_vcvttpd2uqqs512_round_mask(
143	147	(__v8df)__A, (__v8di)_mm512_setzero_si512(), __U,
144	148	_MM_FROUND_CUR_DIRECTION));
...	...	@@ -160,20 +164,20 @@ _mm512_maskz_cvttspd_epu64(__mmask8 __U, __m512d __A) {
160	164	(const int)(__R)))
161	165
162	166	// 512 bit: Float -> int
163		static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_cvttsps_epi32(__m512 __A) {
	167	static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_cvtts_ps_epi32(__m512 __A) {
164	168	return ((__m512i)__builtin_ia32_vcvttps2dqs512_round_mask(
165	169	(__v16sf)(__A), (__v16si)_mm512_undefined_epi32(), (__mmask16)-1,
166	170	_MM_FROUND_CUR_DIRECTION));
167	171	}
168	172
169	173	static __inline__ __m512i __DEFAULT_FN_ATTRS
170		_mm512_mask_cvttsps_epi32(__m512i __W, __mmask16 __U, __m512 __A) {
	174	_mm512_mask_cvtts_ps_epi32(__m512i __W, __mmask16 __U, __m512 __A) {
171	175	return ((__m512i)__builtin_ia32_vcvttps2dqs512_round_mask(
172	176	(__v16sf)(__A), (__v16si)(__W), __U, _MM_FROUND_CUR_DIRECTION));
173	177	}
174	178
175	179	static __inline__ __m512i __DEFAULT_FN_ATTRS
176		_mm512_maskz_cvttsps_epi32(__mmask16 __U, __m512 __A) {
	180	_mm512_maskz_cvtts_ps_epi32(__mmask16 __U, __m512 __A) {
177	181	return ((__m512i)__builtin_ia32_vcvttps2dqs512_round_mask(
178	182	(__v16sf)(__A), (__v16si)_mm512_setzero_si512(), __U,
179	183	_MM_FROUND_CUR_DIRECTION));
...	...	@@ -195,20 +199,20 @@ _mm512_maskz_cvttsps_epi32(__mmask16 __U, __m512 __A) {
195	199	(__mmask16)(__U), (const int)(__R)))
196	200
197	201	// 512 bit: Float -> uint
198		static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_cvttsps_epu32(__m512 __A) {
	202	static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_cvtts_ps_epu32(__m512 __A) {
199	203	return ((__m512i)__builtin_ia32_vcvttps2udqs512_round_mask(
200	204	(__v16sf)(__A), (__v16si)_mm512_undefined_epi32(), (__mmask16)-1,
201	205	_MM_FROUND_CUR_DIRECTION));
202	206	}
203	207
204	208	static __inline__ __m512i __DEFAULT_FN_ATTRS
205		_mm512_mask_cvttsps_epu32(__m512i __W, __mmask16 __U, __m512 __A) {
	209	_mm512_mask_cvtts_ps_epu32(__m512i __W, __mmask16 __U, __m512 __A) {
206	210	return ((__m512i)__builtin_ia32_vcvttps2udqs512_round_mask(
207	211	(__v16sf)(__A), (__v16si)(__W), __U, _MM_FROUND_CUR_DIRECTION));
208	212	}
209	213
210	214	static __inline__ __m512i __DEFAULT_FN_ATTRS
211		_mm512_maskz_cvttsps_epu32(__mmask16 __U, __m512 __A) {
	215	_mm512_maskz_cvtts_ps_epu32(__mmask16 __U, __m512 __A) {
212	216	return ((__m512i)__builtin_ia32_vcvttps2udqs512_round_mask(
213	217	(__v16sf)(__A), (__v16si)_mm512_setzero_si512(), __U,
214	218	_MM_FROUND_CUR_DIRECTION));
...	...	@@ -230,20 +234,20 @@ _mm512_maskz_cvttsps_epu32(__mmask16 __U, __m512 __A) {
230	234	(__mmask16)(__U), (const int)(__R)))
231	235
232	236	// 512 bit : float -> long
233		static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_cvttsps_epi64(__m256 __A) {
	237	static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_cvtts_ps_epi64(__m256 __A) {
234	238	return ((__m512i)__builtin_ia32_vcvttps2qqs512_round_mask(
235	239	(__v8sf)__A, (__v8di)_mm512_undefined_epi32(), (__mmask8)-1,
236	240	_MM_FROUND_CUR_DIRECTION));
237	241	}
238	242
239	243	static __inline__ __m512i __DEFAULT_FN_ATTRS
240		_mm512_mask_cvttsps_epi64(__m512i __W, __mmask8 __U, __m256 __A) {
	244	_mm512_mask_cvtts_ps_epi64(__m512i __W, __mmask8 __U, __m256 __A) {
241	245	return ((__m512i)__builtin_ia32_vcvttps2qqs512_round_mask(
242	246	(__v8sf)__A, (__v8di)__W, __U, _MM_FROUND_CUR_DIRECTION));
243	247	}
244	248
245	249	static __inline__ __m512i __DEFAULT_FN_ATTRS
246		_mm512_maskz_cvttsps_epi64(__mmask8 __U, __m256 __A) {
	250	_mm512_maskz_cvtts_ps_epi64(__mmask8 __U, __m256 __A) {
247	251	return ((__m512i)__builtin_ia32_vcvttps2qqs512_round_mask(
248	252	(__v8sf)__A, (__v8di)_mm512_setzero_si512(), __U,
249	253	_MM_FROUND_CUR_DIRECTION));
...	...	@@ -265,20 +269,20 @@ _mm512_maskz_cvttsps_epi64(__mmask8 __U, __m256 __A) {
265	269	(const int)(__R)))
266	270
267	271	// 512 bit : float -> ulong
268		static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_cvttsps_epu64(__m256 __A) {
	272	static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_cvtts_ps_epu64(__m256 __A) {
269	273	return ((__m512i)__builtin_ia32_vcvttps2uqqs512_round_mask(
270	274	(__v8sf)__A, (__v8di)_mm512_undefined_epi32(), (__mmask8)-1,
271	275	_MM_FROUND_CUR_DIRECTION));
272	276	}
273	277
274	278	static __inline__ __m512i __DEFAULT_FN_ATTRS
275		_mm512_mask_cvttsps_epu64(__m512i __W, __mmask8 __U, __m256 __A) {
	279	_mm512_mask_cvtts_ps_epu64(__m512i __W, __mmask8 __U, __m256 __A) {
276	280	return ((__m512i)__builtin_ia32_vcvttps2uqqs512_round_mask(
277	281	(__v8sf)__A, (__v8di)__W, __U, _MM_FROUND_CUR_DIRECTION));
278	282	}
279	283
280	284	static __inline__ __m512i __DEFAULT_FN_ATTRS
281		_mm512_maskz_cvttsps_epu64(__mmask8 __U, __m256 __A) {
	285	_mm512_maskz_cvtts_ps_epu64(__mmask8 __U, __m256 __A) {
282	286	return ((__m512i)__builtin_ia32_vcvttps2uqqs512_round_mask(
283	287	(__v8sf)__A, (__v8di)_mm512_setzero_si512(), __U,
284	288	_MM_FROUND_CUR_DIRECTION));

lib/include/avx10_2_512satcvtintrin.h+84-84

...	...	@@ -14,286 +14,286 @@
14	14	#ifndef __AVX10_2_512SATCVTINTRIN_H
15	15	#define __AVX10_2_512SATCVTINTRIN_H
16	16
17		#define _mm512_ipcvtbf16_epi8(A) \
	17	#define _mm512_ipcvts_bf16_epi8(A) \
18	18	((__m512i)__builtin_ia32_vcvtbf162ibs512((__v32bf)(__m512bh)(A)))
19	19
20		#define _mm512_mask_ipcvtbf16_epi8(W, U, A) \
	20	#define _mm512_mask_ipcvts_bf16_epi8(W, U, A) \
21	21	((__m512i)__builtin_ia32_selectw_512((__mmask32)(U), \
22		(__v32hi)_mm512_ipcvtbf16_epi8(A), \
	22	(__v32hi)_mm512_ipcvts_bf16_epi8(A), \
23	23	(__v32hi)(__m512i)(W)))
24	24
25		#define _mm512_maskz_ipcvtbf16_epi8(U, A) \
	25	#define _mm512_maskz_ipcvts_bf16_epi8(U, A) \
26	26	((__m512i)__builtin_ia32_selectw_512((__mmask32)(U), \
27		(__v32hi)_mm512_ipcvtbf16_epi8(A), \
	27	(__v32hi)_mm512_ipcvts_bf16_epi8(A), \
28	28	(__v32hi)_mm512_setzero_si512()))
29	29
30		#define _mm512_ipcvtbf16_epu8(A) \
	30	#define _mm512_ipcvts_bf16_epu8(A) \
31	31	((__m512i)__builtin_ia32_vcvtbf162iubs512((__v32bf)(__m512bh)(A)))
32	32
33		#define _mm512_mask_ipcvtbf16_epu8(W, U, A) \
	33	#define _mm512_mask_ipcvts_bf16_epu8(W, U, A) \
34	34	((__m512i)__builtin_ia32_selectw_512((__mmask32)(U), \
35		(__v32hi)_mm512_ipcvtbf16_epu8(A), \
	35	(__v32hi)_mm512_ipcvts_bf16_epu8(A), \
36	36	(__v32hi)(__m512i)(W)))
37	37
38		#define _mm512_maskz_ipcvtbf16_epu8(U, A) \
	38	#define _mm512_maskz_ipcvts_bf16_epu8(U, A) \
39	39	((__m512i)__builtin_ia32_selectw_512((__mmask32)(U), \
40		(__v32hi)_mm512_ipcvtbf16_epu8(A), \
	40	(__v32hi)_mm512_ipcvts_bf16_epu8(A), \
41	41	(__v32hi)_mm512_setzero_si512()))
42	42
43		#define _mm512_ipcvttbf16_epi8(A) \
	43	#define _mm512_ipcvtts_bf16_epi8(A) \
44	44	((__m512i)__builtin_ia32_vcvttbf162ibs512((__v32bf)(__m512bh)(A)))
45	45
46		#define _mm512_mask_ipcvttbf16_epi8(W, U, A) \
	46	#define _mm512_mask_ipcvtts_bf16_epi8(W, U, A) \
47	47	((__m512i)__builtin_ia32_selectw_512((__mmask32)(U), \
48		(__v32hi)_mm512_ipcvttbf16_epi8(A), \
	48	(__v32hi)_mm512_ipcvtts_bf16_epi8(A), \
49	49	(__v32hi)(__m512i)(W)))
50	50
51		#define _mm512_maskz_ipcvttbf16_epi8(U, A) \
	51	#define _mm512_maskz_ipcvtts_bf16_epi8(U, A) \
52	52	((__m512i)__builtin_ia32_selectw_512((__mmask32)(U), \
53		(__v32hi)_mm512_ipcvttbf16_epi8(A), \
	53	(__v32hi)_mm512_ipcvtts_bf16_epi8(A), \
54	54	(__v32hi)_mm512_setzero_si512()))
55	55
56		#define _mm512_ipcvttbf16_epu8(A) \
	56	#define _mm512_ipcvtts_bf16_epu8(A) \
57	57	((__m512i)__builtin_ia32_vcvttbf162iubs512((__v32bf)(__m512bh)(A)))
58	58
59		#define _mm512_mask_ipcvttbf16_epu8(W, U, A) \
	59	#define _mm512_mask_ipcvtts_bf16_epu8(W, U, A) \
60	60	((__m512i)__builtin_ia32_selectw_512((__mmask32)(U), \
61		(__v32hi)_mm512_ipcvttbf16_epu8(A), \
	61	(__v32hi)_mm512_ipcvtts_bf16_epu8(A), \
62	62	(__v32hi)(__m512i)(W)))
63	63
64		#define _mm512_maskz_ipcvttbf16_epu8(U, A) \
	64	#define _mm512_maskz_ipcvtts_bf16_epu8(U, A) \
65	65	((__m512i)__builtin_ia32_selectw_512((__mmask32)(U), \
66		(__v32hi)_mm512_ipcvttbf16_epu8(A), \
	66	(__v32hi)_mm512_ipcvtts_bf16_epu8(A), \
67	67	(__v32hi)_mm512_setzero_si512()))
68	68
69		#define _mm512_ipcvtph_epi8(A) \
	69	#define _mm512_ipcvts_ph_epi8(A) \
70	70	((__m512i)__builtin_ia32_vcvtph2ibs512_mask( \
71		(__v32hf)(__m512h)(A), (__v32hu)_mm512_setzero_si512(), (__mmask32)-1, \
	71	(__v32hf)(__m512h)(A), (__v32hu)_mm512_setzero_si512(), (__mmask32) - 1, \
72	72	_MM_FROUND_CUR_DIRECTION))
73	73
74		#define _mm512_mask_ipcvtph_epi8(W, U, A) \
	74	#define _mm512_mask_ipcvts_ph_epi8(W, U, A) \
75	75	((__m512i)__builtin_ia32_vcvtph2ibs512_mask((__v32hf)(__m512h)(A), \
76	76	(__v32hu)(W), (__mmask32)(U), \
77	77	_MM_FROUND_CUR_DIRECTION))
78	78
79		#define _mm512_maskz_ipcvtph_epi8(U, A) \
	79	#define _mm512_maskz_ipcvts_ph_epi8(U, A) \
80	80	((__m512i)__builtin_ia32_vcvtph2ibs512_mask( \
81	81	(__v32hf)(__m512h)(A), (__v32hu)_mm512_setzero_si512(), (__mmask32)(U), \
82	82	_MM_FROUND_CUR_DIRECTION))
83	83
84		#define _mm512_ipcvt_roundph_epi8(A, R) \
	84	#define _mm512_ipcvts_roundph_epi8(A, R) \
85	85	((__m512i)__builtin_ia32_vcvtph2ibs512_mask((__v32hf)(__m512h)(A), \
86	86	(__v32hu)_mm512_setzero_si512(), \
87		(__mmask32)-1, (const int)R))
	87	(__mmask32) - 1, (const int)R))
88	88
89		#define _mm512_mask_ipcvt_roundph_epi8(W, U, A, R) \
	89	#define _mm512_mask_ipcvts_roundph_epi8(W, U, A, R) \
90	90	((__m512i)__builtin_ia32_vcvtph2ibs512_mask( \
91	91	(__v32hf)(__m512h)(A), (__v32hu)(W), (__mmask32)(U), (const int)R))
92	92
93		#define _mm512_maskz_ipcvt_roundph_epi8(U, A, R) \
	93	#define _mm512_maskz_ipcvts_roundph_epi8(U, A, R) \
94	94	((__m512i)__builtin_ia32_vcvtph2ibs512_mask((__v32hf)(__m512h)(A), \
95	95	(__v32hu)_mm512_setzero_si512(), \
96	96	(__mmask32)(U), (const int)R))
97	97
98		#define _mm512_ipcvtph_epu8(A) \
	98	#define _mm512_ipcvts_ph_epu8(A) \
99	99	((__m512i)__builtin_ia32_vcvtph2iubs512_mask( \
100		(__v32hf)(__m512h)(A), (__v32hu)_mm512_setzero_si512(), (__mmask32)-1, \
	100	(__v32hf)(__m512h)(A), (__v32hu)_mm512_setzero_si512(), (__mmask32) - 1, \
101	101	_MM_FROUND_CUR_DIRECTION))
102	102
103		#define _mm512_mask_ipcvtph_epu8(W, U, A) \
	103	#define _mm512_mask_ipcvts_ph_epu8(W, U, A) \
104	104	((__m512i)__builtin_ia32_vcvtph2iubs512_mask((__v32hf)(__m512h)(A), \
105	105	(__v32hu)(W), (__mmask32)(U), \
106	106	_MM_FROUND_CUR_DIRECTION))
107	107
108		#define _mm512_maskz_ipcvtph_epu8(U, A) \
	108	#define _mm512_maskz_ipcvts_ph_epu8(U, A) \
109	109	((__m512i)__builtin_ia32_vcvtph2iubs512_mask( \
110	110	(__v32hf)(__m512h)(A), (__v32hu)_mm512_setzero_si512(), (__mmask32)(U), \
111	111	_MM_FROUND_CUR_DIRECTION))
112	112
113		#define _mm512_ipcvt_roundph_epu8(A, R) \
	113	#define _mm512_ipcvts_roundph_epu8(A, R) \
114	114	((__m512i)__builtin_ia32_vcvtph2iubs512_mask( \
115		(__v32hf)(__m512h)(A), (__v32hu)_mm512_setzero_si512(), (__mmask32)-1, \
	115	(__v32hf)(__m512h)(A), (__v32hu)_mm512_setzero_si512(), (__mmask32) - 1, \
116	116	(const int)R))
117	117
118		#define _mm512_mask_ipcvt_roundph_epu8(W, U, A, R) \
	118	#define _mm512_mask_ipcvts_roundph_epu8(W, U, A, R) \
119	119	((__m512i)__builtin_ia32_vcvtph2iubs512_mask( \
120	120	(__v32hf)(__m512h)(A), (__v32hu)(W), (__mmask32)(U), (const int)R))
121	121
122		#define _mm512_maskz_ipcvt_roundph_epu8(U, A, R) \
	122	#define _mm512_maskz_ipcvts_roundph_epu8(U, A, R) \
123	123	((__m512i)__builtin_ia32_vcvtph2iubs512_mask( \
124	124	(__v32hf)(__m512h)(A), (__v32hu)_mm512_setzero_si512(), (__mmask32)(U), \
125	125	(const int)R))
126	126
127		#define _mm512_ipcvtps_epi8(A) \
	127	#define _mm512_ipcvts_ps_epi8(A) \
128	128	((__m512i)__builtin_ia32_vcvtps2ibs512_mask( \
129		(__v16sf)(__m512)(A), (__v16su)_mm512_setzero_si512(), (__mmask16)-1, \
	129	(__v16sf)(__m512)(A), (__v16su)_mm512_setzero_si512(), (__mmask16) - 1, \
130	130	_MM_FROUND_CUR_DIRECTION))
131	131
132		#define _mm512_mask_ipcvtps_epi8(W, U, A) \
	132	#define _mm512_mask_ipcvts_ps_epi8(W, U, A) \
133	133	((__m512i)__builtin_ia32_vcvtps2ibs512_mask((__v16sf)(__m512)(A), \
134	134	(__v16su)(W), (__mmask16)(U), \
135	135	_MM_FROUND_CUR_DIRECTION))
136	136
137		#define _mm512_maskz_ipcvtps_epi8(U, A) \
	137	#define _mm512_maskz_ipcvts_ps_epi8(U, A) \
138	138	((__m512i)__builtin_ia32_vcvtps2ibs512_mask( \
139	139	(__v16sf)(__m512)(A), (__v16su)_mm512_setzero_si512(), (__mmask16)(U), \
140	140	_MM_FROUND_CUR_DIRECTION))
141	141
142		#define _mm512_ipcvt_roundps_epi8(A, R) \
	142	#define _mm512_ipcvts_roundps_epi8(A, R) \
143	143	((__m512i)__builtin_ia32_vcvtps2ibs512_mask((__v16sf)(__m512)(A), \
144	144	(__v16su)_mm512_setzero_si512(), \
145		(__mmask16)-1, (const int)R))
	145	(__mmask16) - 1, (const int)R))
146	146
147		#define _mm512_mask_ipcvt_roundps_epi8(W, U, A, R) \
	147	#define _mm512_mask_ipcvts_roundps_epi8(W, U, A, R) \
148	148	((__m512i)__builtin_ia32_vcvtps2ibs512_mask( \
149	149	(__v16sf)(__m512)(A), (__v16su)(W), (__mmask16)(U), (const int)R))
150	150
151		#define _mm512_maskz_ipcvt_roundps_epi8(U, A, R) \
	151	#define _mm512_maskz_ipcvts_roundps_epi8(U, A, R) \
152	152	((__m512i)__builtin_ia32_vcvtps2ibs512_mask((__v16sf)(__m512)(A), \
153	153	(__v16su)_mm512_setzero_si512(), \
154	154	(__mmask16)(U), (const int)R))
155	155
156		#define _mm512_ipcvtps_epu8(A) \
	156	#define _mm512_ipcvts_ps_epu8(A) \
157	157	((__m512i)__builtin_ia32_vcvtps2iubs512_mask( \
158		(__v16sf)(__m512)(A), (__v16su)_mm512_setzero_si512(), (__mmask16)-1, \
	158	(__v16sf)(__m512)(A), (__v16su)_mm512_setzero_si512(), (__mmask16) - 1, \
159	159	_MM_FROUND_CUR_DIRECTION))
160	160
161		#define _mm512_mask_ipcvtps_epu8(W, U, A) \
	161	#define _mm512_mask_ipcvts_ps_epu8(W, U, A) \
162	162	((__m512i)__builtin_ia32_vcvtps2iubs512_mask((__v16sf)(__m512)(A), \
163	163	(__v16su)(W), (__mmask16)(U), \
164	164	_MM_FROUND_CUR_DIRECTION))
165	165
166		#define _mm512_maskz_ipcvtps_epu8(U, A) \
	166	#define _mm512_maskz_ipcvts_ps_epu8(U, A) \
167	167	((__m512i)__builtin_ia32_vcvtps2iubs512_mask( \
168	168	(__v16sf)(__m512)(A), (__v16su)_mm512_setzero_si512(), (__mmask16)(U), \
169	169	_MM_FROUND_CUR_DIRECTION))
170	170
171		#define _mm512_ipcvt_roundps_epu8(A, R) \
	171	#define _mm512_ipcvts_roundps_epu8(A, R) \
172	172	((__m512i)__builtin_ia32_vcvtps2iubs512_mask( \
173		(__v16sf)(__m512)(A), (__v16su)_mm512_setzero_si512(), (__mmask16)-1, \
	173	(__v16sf)(__m512)(A), (__v16su)_mm512_setzero_si512(), (__mmask16) - 1, \
174	174	(const int)R))
175	175
176		#define _mm512_mask_ipcvt_roundps_epu8(W, U, A, R) \
	176	#define _mm512_mask_ipcvts_roundps_epu8(W, U, A, R) \
177	177	((__m512i)__builtin_ia32_vcvtps2iubs512_mask( \
178	178	(__v16sf)(__m512)(A), (__v16su)(W), (__mmask16)(U), (const int)R))
179	179
180		#define _mm512_maskz_ipcvt_roundps_epu8(U, A, R) \
	180	#define _mm512_maskz_ipcvts_roundps_epu8(U, A, R) \
181	181	((__m512i)__builtin_ia32_vcvtps2iubs512_mask( \
182	182	(__v16sf)(__m512)(A), (__v16su)_mm512_setzero_si512(), (__mmask16)(U), \
183	183	(const int)R))
184	184
185		#define _mm512_ipcvttph_epi8(A) \
	185	#define _mm512_ipcvtts_ph_epi8(A) \
186	186	((__m512i)__builtin_ia32_vcvttph2ibs512_mask( \
187		(__v32hf)(__m512h)(A), (__v32hu)_mm512_setzero_si512(), (__mmask32)-1, \
	187	(__v32hf)(__m512h)(A), (__v32hu)_mm512_setzero_si512(), (__mmask32) - 1, \
188	188	_MM_FROUND_CUR_DIRECTION))
189	189
190		#define _mm512_mask_ipcvttph_epi8(W, U, A) \
	190	#define _mm512_mask_ipcvtts_ph_epi8(W, U, A) \
191	191	((__m512i)__builtin_ia32_vcvttph2ibs512_mask((__v32hf)(__m512h)(A), \
192	192	(__v32hu)(W), (__mmask32)(U), \
193	193	_MM_FROUND_CUR_DIRECTION))
194	194
195		#define _mm512_maskz_ipcvttph_epi8(U, A) \
	195	#define _mm512_maskz_ipcvtts_ph_epi8(U, A) \
196	196	((__m512i)__builtin_ia32_vcvttph2ibs512_mask( \
197	197	(__v32hf)(__m512h)(A), (__v32hu)_mm512_setzero_si512(), (__mmask32)(U), \
198	198	_MM_FROUND_CUR_DIRECTION))
199	199
200		#define _mm512_ipcvtt_roundph_epi8(A, S) \
	200	#define _mm512_ipcvtts_roundph_epi8(A, S) \
201	201	((__m512i)__builtin_ia32_vcvttph2ibs512_mask( \
202		(__v32hf)(__m512h)(A), (__v32hu)_mm512_setzero_si512(), (__mmask32)-1, \
	202	(__v32hf)(__m512h)(A), (__v32hu)_mm512_setzero_si512(), (__mmask32) - 1, \
203	203	S))
204	204
205		#define _mm512_mask_ipcvtt_roundph_epi8(W, U, A, S) \
	205	#define _mm512_mask_ipcvtts_roundph_epi8(W, U, A, S) \
206	206	((__m512i)__builtin_ia32_vcvttph2ibs512_mask( \
207	207	(__v32hf)(__m512h)(A), (__v32hu)(W), (__mmask32)(U), S))
208	208
209		#define _mm512_maskz_ipcvtt_roundph_epi8(U, A, S) \
	209	#define _mm512_maskz_ipcvtts_roundph_epi8(U, A, S) \
210	210	((__m512i)__builtin_ia32_vcvttph2ibs512_mask( \
211	211	(__v32hf)(__m512h)(A), (__v32hu)_mm512_setzero_si512(), (__mmask32)(U), \
212	212	S))
213	213
214		#define _mm512_ipcvttph_epu8(A) \
	214	#define _mm512_ipcvtts_ph_epu8(A) \
215	215	((__m512i)__builtin_ia32_vcvttph2iubs512_mask( \
216		(__v32hf)(__m512h)(A), (__v32hu)_mm512_setzero_si512(), (__mmask32)-1, \
	216	(__v32hf)(__m512h)(A), (__v32hu)_mm512_setzero_si512(), (__mmask32) - 1, \
217	217	_MM_FROUND_CUR_DIRECTION))
218	218
219		#define _mm512_mask_ipcvttph_epu8(W, U, A) \
	219	#define _mm512_mask_ipcvtts_ph_epu8(W, U, A) \
220	220	((__m512i)__builtin_ia32_vcvttph2iubs512_mask((__v32hf)(__m512h)(A), \
221	221	(__v32hu)(W), (__mmask32)(U), \
222	222	_MM_FROUND_CUR_DIRECTION))
223	223
224		#define _mm512_maskz_ipcvttph_epu8(U, A) \
	224	#define _mm512_maskz_ipcvtts_ph_epu8(U, A) \
225	225	((__m512i)__builtin_ia32_vcvttph2iubs512_mask( \
226	226	(__v32hf)(__m512h)(A), (__v32hu)_mm512_setzero_si512(), (__mmask32)(U), \
227	227	_MM_FROUND_CUR_DIRECTION))
228	228
229		#define _mm512_ipcvtt_roundph_epu8(A, S) \
	229	#define _mm512_ipcvtts_roundph_epu8(A, S) \
230	230	((__m512i)__builtin_ia32_vcvttph2iubs512_mask( \
231		(__v32hf)(__m512h)(A), (__v32hu)_mm512_setzero_si512(), (__mmask32)-1, \
	231	(__v32hf)(__m512h)(A), (__v32hu)_mm512_setzero_si512(), (__mmask32) - 1, \
232	232	S))
233	233
234		#define _mm512_mask_ipcvtt_roundph_epu8(W, U, A, S) \
	234	#define _mm512_mask_ipcvtts_roundph_epu8(W, U, A, S) \
235	235	((__m512i)__builtin_ia32_vcvttph2iubs512_mask( \
236	236	(__v32hf)(__m512h)(A), (__v32hu)(W), (__mmask32)(U), S))
237	237
238		#define _mm512_maskz_ipcvtt_roundph_epu8(U, A, S) \
	238	#define _mm512_maskz_ipcvtts_roundph_epu8(U, A, S) \
239	239	((__m512i)__builtin_ia32_vcvttph2iubs512_mask( \
240	240	(__v32hf)(__m512h)(A), (__v32hu)_mm512_setzero_si512(), (__mmask32)(U), \
241	241	S))
242	242
243		#define _mm512_ipcvttps_epi8(A) \
	243	#define _mm512_ipcvtts_ps_epi8(A) \
244	244	((__m512i)__builtin_ia32_vcvttps2ibs512_mask( \
245		(__v16sf)(__m512h)(A), (__v16su)_mm512_setzero_si512(), (__mmask16)-1, \
	245	(__v16sf)(__m512h)(A), (__v16su)_mm512_setzero_si512(), (__mmask16) - 1, \
246	246	_MM_FROUND_CUR_DIRECTION))
247	247
248		#define _mm512_mask_ipcvttps_epi8(W, U, A) \
	248	#define _mm512_mask_ipcvtts_ps_epi8(W, U, A) \
249	249	((__m512i)__builtin_ia32_vcvttps2ibs512_mask((__v16sf)(__m512h)(A), \
250	250	(__v16su)(W), (__mmask16)(U), \
251	251	_MM_FROUND_CUR_DIRECTION))
252	252
253		#define _mm512_maskz_ipcvttps_epi8(U, A) \
	253	#define _mm512_maskz_ipcvtts_ps_epi8(U, A) \
254	254	((__m512i)__builtin_ia32_vcvttps2ibs512_mask( \
255	255	(__v16sf)(__m512h)(A), (__v16su)_mm512_setzero_si512(), (__mmask16)(U), \
256	256	_MM_FROUND_CUR_DIRECTION))
257	257
258		#define _mm512_ipcvtt_roundps_epi8(A, S) \
	258	#define _mm512_ipcvtts_roundps_epi8(A, S) \
259	259	((__m512i)__builtin_ia32_vcvttps2ibs512_mask( \
260		(__v16sf)(__m512h)(A), (__v16su)_mm512_setzero_si512(), (__mmask16)-1, \
	260	(__v16sf)(__m512h)(A), (__v16su)_mm512_setzero_si512(), (__mmask16) - 1, \
261	261	S))
262	262
263		#define _mm512_mask_ipcvtt_roundps_epi8(W, U, A, S) \
	263	#define _mm512_mask_ipcvtts_roundps_epi8(W, U, A, S) \
264	264	((__m512i)__builtin_ia32_vcvttps2ibs512_mask( \
265	265	(__v16sf)(__m512h)(A), (__v16su)(W), (__mmask16)(U), S))
266	266
267		#define _mm512_maskz_ipcvtt_roundps_epi8(U, A, S) \
	267	#define _mm512_maskz_ipcvtts_roundps_epi8(U, A, S) \
268	268	((__m512i)__builtin_ia32_vcvttps2ibs512_mask( \
269	269	(__v16sf)(__m512h)(A), (__v16su)_mm512_setzero_si512(), (__mmask16)(U), \
270	270	S))
271	271
272		#define _mm512_ipcvttps_epu8(A) \
	272	#define _mm512_ipcvtts_ps_epu8(A) \
273	273	((__m512i)__builtin_ia32_vcvttps2iubs512_mask( \
274		(__v16sf)(__m512h)(A), (__v16su)_mm512_setzero_si512(), (__mmask16)-1, \
	274	(__v16sf)(__m512h)(A), (__v16su)_mm512_setzero_si512(), (__mmask16) - 1, \
275	275	_MM_FROUND_CUR_DIRECTION))
276	276
277		#define _mm512_mask_ipcvttps_epu8(W, U, A) \
	277	#define _mm512_mask_ipcvtts_ps_epu8(W, U, A) \
278	278	((__m512i)__builtin_ia32_vcvttps2iubs512_mask((__v16sf)(__m512h)(A), \
279	279	(__v16su)(W), (__mmask16)(U), \
280	280	_MM_FROUND_CUR_DIRECTION))
281	281
282		#define _mm512_maskz_ipcvttps_epu8(U, A) \
	282	#define _mm512_maskz_ipcvtts_ps_epu8(U, A) \
283	283	((__m512i)__builtin_ia32_vcvttps2iubs512_mask( \
284	284	(__v16sf)(__m512h)(A), (__v16su)_mm512_setzero_si512(), (__mmask16)(U), \
285	285	_MM_FROUND_CUR_DIRECTION))
286	286
287		#define _mm512_ipcvtt_roundps_epu8(A, S) \
	287	#define _mm512_ipcvtts_roundps_epu8(A, S) \
288	288	((__m512i)__builtin_ia32_vcvttps2iubs512_mask( \
289		(__v16sf)(__m512h)(A), (__v16su)_mm512_setzero_si512(), (__mmask16)-1, \
	289	(__v16sf)(__m512h)(A), (__v16su)_mm512_setzero_si512(), (__mmask16) - 1, \
290	290	S))
291	291
292		#define _mm512_mask_ipcvtt_roundps_epu8(W, U, A, S) \
	292	#define _mm512_mask_ipcvtts_roundps_epu8(W, U, A, S) \
293	293	((__m512i)__builtin_ia32_vcvttps2iubs512_mask( \
294	294	(__v16sf)(__m512h)(A), (__v16su)(W), (__mmask16)(U), S))
295	295
296		#define _mm512_maskz_ipcvtt_roundps_epu8(U, A, S) \
	296	#define _mm512_maskz_ipcvtts_roundps_epu8(U, A, S) \
297	297	((__m512i)__builtin_ia32_vcvttps2iubs512_mask( \
298	298	(__v16sf)(__m512h)(A), (__v16su)_mm512_setzero_si512(), (__mmask16)(U), \
299	299	S))

lib/include/avx10_2bf16intrin.h+16-16

...	...	@@ -852,8 +852,8 @@ _mm_maskz_sqrt_pbh(__mmask8 __U, __m128bh __A) {
852	852
853	853	static __inline__ __m256bh __DEFAULT_FN_ATTRS256
854	854	_mm256_fmadd_pbh(__m256bh __A, __m256bh __B, __m256bh __C) {
855		return (__m256bh)__builtin_ia32_vfmaddnepbh256((__v16bf)__A, (__v16bf)__B,
856		(__v16bf)__C);
	855	return (__m256bh)__builtin_ia32_vfmaddbf16256((__v16bf)__A, (__v16bf)__B,
	856	(__v16bf)__C);
857	857	}
858	858
859	859	static __inline__ __m256bh __DEFAULT_FN_ATTRS256
...	...	@@ -880,8 +880,8 @@ static __inline__ __m256bh __DEFAULT_FN_ATTRS256 _mm256_maskz_fmadd_pbh(
880	880
881	881	static __inline__ __m256bh __DEFAULT_FN_ATTRS256
882	882	_mm256_fmsub_pbh(__m256bh __A, __m256bh __B, __m256bh __C) {
883		return (__m256bh)__builtin_ia32_vfmaddnepbh256((__v16bf)__A, (__v16bf)__B,
884		-(__v16bf)__C);
	883	return (__m256bh)__builtin_ia32_vfmaddbf16256((__v16bf)__A, (__v16bf)__B,
	884	-(__v16bf)__C);
885	885	}
886	886
887	887	static __inline__ __m256bh __DEFAULT_FN_ATTRS256
...	...	@@ -908,8 +908,8 @@ static __inline__ __m256bh __DEFAULT_FN_ATTRS256 _mm256_maskz_fmsub_pbh(
908	908
909	909	static __inline__ __m256bh __DEFAULT_FN_ATTRS256
910	910	_mm256_fnmadd_pbh(__m256bh __A, __m256bh __B, __m256bh __C) {
911		return (__m256bh)__builtin_ia32_vfmaddnepbh256((__v16bf)__A, -(__v16bf)__B,
912		(__v16bf)__C);
	911	return (__m256bh)__builtin_ia32_vfmaddbf16256((__v16bf)__A, -(__v16bf)__B,
	912	(__v16bf)__C);
913	913	}
914	914
915	915	static __inline__ __m256bh __DEFAULT_FN_ATTRS256 _mm256_mask_fnmadd_pbh(
...	...	@@ -938,8 +938,8 @@ static __inline__ __m256bh __DEFAULT_FN_ATTRS256 _mm256_maskz_fnmadd_pbh(
938	938
939	939	static __inline__ __m256bh __DEFAULT_FN_ATTRS256
940	940	_mm256_fnmsub_pbh(__m256bh __A, __m256bh __B, __m256bh __C) {
941		return (__m256bh)__builtin_ia32_vfmaddnepbh256((__v16bf)__A, -(__v16bf)__B,
942		-(__v16bf)__C);
	941	return (__m256bh)__builtin_ia32_vfmaddbf16256((__v16bf)__A, -(__v16bf)__B,
	942	-(__v16bf)__C);
943	943	}
944	944
945	945	static __inline__ __m256bh __DEFAULT_FN_ATTRS256 _mm256_mask_fnmsub_pbh(
...	...	@@ -969,8 +969,8 @@ static __inline__ __m256bh __DEFAULT_FN_ATTRS256 _mm256_maskz_fnmsub_pbh(
969	969	static __inline__ __m128bh __DEFAULT_FN_ATTRS128 _mm_fmadd_pbh(__m128bh __A,
970	970	__m128bh __B,
971	971	__m128bh __C) {
972		return (__m128bh)__builtin_ia32_vfmaddnepbh128((__v8bf)__A, (__v8bf)__B,
973		(__v8bf)__C);
	972	return (__m128bh)__builtin_ia32_vfmaddbf16128((__v8bf)__A, (__v8bf)__B,
	973	(__v8bf)__C);
974	974	}
975	975
976	976	static __inline__ __m128bh __DEFAULT_FN_ATTRS128
...	...	@@ -997,8 +997,8 @@ _mm_maskz_fmadd_pbh(__mmask8 __U, __m128bh __A, __m128bh __B, __m128bh __C) {
997	997	static __inline__ __m128bh __DEFAULT_FN_ATTRS128 _mm_fmsub_pbh(__m128bh __A,
998	998	__m128bh __B,
999	999	__m128bh __C) {
1000		return (__m128bh)__builtin_ia32_vfmaddnepbh128((__v8bf)__A, (__v8bf)__B,
1001		-(__v8bf)__C);
	1000	return (__m128bh)__builtin_ia32_vfmaddbf16128((__v8bf)__A, (__v8bf)__B,
	1001	-(__v8bf)__C);
1002	1002	}
1003	1003
1004	1004	static __inline__ __m128bh __DEFAULT_FN_ATTRS128
...	...	@@ -1025,8 +1025,8 @@ _mm_maskz_fmsub_pbh(__mmask8 __U, __m128bh __A, __m128bh __B, __m128bh __C) {
1025	1025	static __inline__ __m128bh __DEFAULT_FN_ATTRS128 _mm_fnmadd_pbh(__m128bh __A,
1026	1026	__m128bh __B,
1027	1027	__m128bh __C) {
1028		return (__m128bh)__builtin_ia32_vfmaddnepbh128((__v8bf)__A, -(__v8bf)__B,
1029		(__v8bf)__C);
	1028	return (__m128bh)__builtin_ia32_vfmaddbf16128((__v8bf)__A, -(__v8bf)__B,
	1029	(__v8bf)__C);
1030	1030	}
1031	1031
1032	1032	static __inline__ __m128bh __DEFAULT_FN_ATTRS128
...	...	@@ -1053,8 +1053,8 @@ _mm_maskz_fnmadd_pbh(__mmask8 __U, __m128bh __A, __m128bh __B, __m128bh __C) {
1053	1053	static __inline__ __m128bh __DEFAULT_FN_ATTRS128 _mm_fnmsub_pbh(__m128bh __A,
1054	1054	__m128bh __B,
1055	1055	__m128bh __C) {
1056		return (__m128bh)__builtin_ia32_vfmaddnepbh128((__v8bf)__A, -(__v8bf)__B,
1057		-(__v8bf)__C);
	1056	return (__m128bh)__builtin_ia32_vfmaddbf16128((__v8bf)__A, -(__v8bf)__B,
	1057	-(__v8bf)__C);
1058	1058	}
1059	1059
1060	1060	static __inline__ __m128bh __DEFAULT_FN_ATTRS128

lib/include/avx10_2convertintrin.h+2819-152

...	...	@@ -24,71 +24,351 @@
24	24	__attribute__((__always_inline__, __nodebug__, __target__("avx10.2-256"), \
25	25	__min_vector_width__(256)))
26	26
	27	// clang-format off
	28
	29	/// Convert two 128-bit vectors, \a __A and \a __B, containing packed
	30	/// single-precision (32-bit) floating-point elements to a 128-bit vector
	31	/// containing FP16 elements.
	32	///
	33	/// \code{.operation}
	34	/// FOR i := 0 to 7
	35	/// IF i < 4
	36	/// dst.fp16[i] := convert_fp32_to_fp16(__B.fp32[i])
	37	/// ELSE
	38	/// dst.fp16[i] := convert_fp32_to_fp16(__A.fp32[i - 4])
	39	/// FI
	40	///
	41	/// ENDFOR
	42	///
	43	/// dst[MAX:128] := 0
	44	/// \endcode
	45	///
	46	/// \headerfile <immintrin.h>
	47	///
	48	/// This intrinsic corresponds to the \c VCVT2PS2PHX instruction.
	49	///
	50	/// \param __A
	51	/// A 128-bit vector of [4 x float].
	52	/// \param __B
	53	/// A 128-bit vector of [4 x float].
	54	/// \returns
	55	/// A 128-bit vector of [8 x fp16]. Lower 4 elements correspond to the
	56	/// (converted) elements from \a __B; higher order elements correspond to the
	57	/// (converted) elements from \a __A.
27	58	static __inline__ __m128h __DEFAULT_FN_ATTRS128 _mm_cvtx2ps_ph(__m128 __A,
28	59	__m128 __B) {
29	60	return (__m128h)__builtin_ia32_vcvt2ps2phx128_mask(
30	61	(__v4sf)__A, (__v4sf)__B, (__v8hf)_mm_setzero_ph(), (__mmask8)(-1));
31	62	}
32	63
	64	/// Convert two 128-bit vectors, \a __A and \a __B, containing packed
	65	/// single-precision (32-bit) floating-point elements to a 128-bit vector
	66	/// containing FP16 elements. Merging mask \a __U is used to determine if given
	67	/// element should be taken from \a __W instead.
	68	///
	69	/// \code{.operation}
	70	/// FOR i := 0 to 7
	71	/// IF __U[i]
	72	/// IF i < 4
	73	/// dst.fp16[i] := convert_fp32_to_fp16(__B.fp32[i])
	74	/// ELSE
	75	/// dst.fp16[i] := convert_fp32_to_fp16(__A.fp32[i - 4])
	76	/// FI
	77	/// ELSE
	78	/// dst.fp16[i] := __W.fp16[i]
	79	/// FI
	80	/// ENDFOR
	81	///
	82	/// dst[MAX:128] := 0
	83	/// \endcode
	84	///
	85	/// \headerfile <immintrin.h>
	86	///
	87	/// This intrinsic corresponds to the \c VCVT2PS2PHX instruction.
	88	///
	89	/// \param __W
	90	/// A 128-bit vector of [8 x fp16].
	91	/// \param __U
	92	/// A 8-bit merging mask.
	93	/// \param __A
	94	/// A 128-bit vector of [4 x float].
	95	/// \param __B
	96	/// A 128-bit vector of [4 x float].
	97	/// \returns
	98	/// A 128-bit vector of [8 x fp16]. Lower elements correspond to the
	99	/// (converted) elements from \a __B; higher order elements correspond to the
	100	/// (converted) elements from \a __A. If corresponding mask bit is not set, then
	101	/// element from \a __W is taken instead.
33	102	static __inline__ __m128h __DEFAULT_FN_ATTRS128
34	103	_mm_mask_cvtx2ps_ph(__m128h __W, __mmask8 __U, __m128 __A, __m128 __B) {
35	104	return (__m128h)__builtin_ia32_vcvt2ps2phx128_mask(
36	105	(__v4sf)__A, (__v4sf)__B, (__v8hf)__W, (__mmask8)__U);
37	106	}
38	107
	108	/// Convert two 128-bit vectors, \a __A and \a __B, containing packed
	109	/// single-precision (32-bit) floating-point elements to a 128-bit vector
	110	/// containing FP16 elements. Zeroing mask \a __U is used to determine if given
	111	/// element should be zeroed instead.
	112	///
	113	/// \code{.operation}
	114	/// FOR i := 0 to 7
	115	/// IF __U[i]
	116	/// IF i < 4
	117	/// dst.fp16[i] := convert_fp32_to_fp16(__B.fp32[i])
	118	/// ELSE
	119	/// dst.fp16[i] := convert_fp32_to_fp16(__A.fp32[i - 4])
	120	/// FI
	121	/// ELSE
	122	/// dst.fp16[i] := 0
	123	/// FI
	124	/// ENDFOR
	125	///
	126	/// dst[MAX:128] := 0
	127	/// \endcode
	128	///
	129	/// \headerfile <immintrin.h>
	130	///
	131	/// This intrinsic corresponds to the \c VCVT2PS2PHX instruction.
	132	///
	133	/// \param __U
	134	/// A 8-bit zeroing mask.
	135	/// \param __A
	136	/// A 128-bit vector of [4 x float].
	137	/// \param __B
	138	/// A 128-bit vector of [4 x float].
	139	/// \returns
	140	/// A 128-bit vector of [8 x fp16]. Lower elements correspond to the
	141	/// (converted) elements from \a __B; higher order elements correspond to the
	142	/// (converted) elements from \a __A. If corresponding mask bit is not set,
	143	/// then zero is taken instead.
39	144	static __inline__ __m128h __DEFAULT_FN_ATTRS128
40	145	_mm_maskz_cvtx2ps_ph(__mmask8 __U, __m128 __A, __m128 __B) {
41	146	return (__m128h)__builtin_ia32_vcvt2ps2phx128_mask(
42	147	(__v4sf)__A, (__v4sf)__B, (__v8hf)_mm_setzero_ph(), (__mmask8)__U);
43	148	}
44	149
	150	/// Convert two 256-bit vectors, \a __A and \a __B, containing packed
	151	/// single-precision (32-bit) floating-point elements to a 256-bit vector
	152	/// containing FP16 elements.
	153	///
	154	/// \code{.operation}
	155	/// FOR i := 0 to 15
	156	/// IF i < 8
	157	/// dst.fp16[i] := convert_fp32_to_fp16(__B.fp32[i])
	158	/// ELSE
	159	/// dst.fp16[i] := convert_fp32_to_fp16(__A.fp32[i - 8])
	160	/// FI
	161	/// ENDFOR
	162	///
	163	/// dst[MAX:256] := 0
	164	/// \endcode
	165	///
	166	/// \headerfile <immintrin.h>
	167	///
	168	/// This intrinsic corresponds to the \c VCVT2PS2PHX instruction.
	169	///
	170	/// \param __A
	171	/// A 256-bit vector of [8 x float].
	172	/// \param __B
	173	/// A 256-bit vector of [8 x float].
	174	/// \returns
	175	/// A 256-bit vector of [16 x fp16]. Lower elements correspond to the
	176	/// (converted) elements from \a __B; higher order elements correspond to the
	177	/// (converted) elements from \a __A.
45	178	static __inline__ __m256h __DEFAULT_FN_ATTRS256 _mm256_cvtx2ps_ph(__m256 __A,
46	179	__m256 __B) {
47	180	return (__m256h)__builtin_ia32_vcvt2ps2phx256_mask(
48		(__v8sf)__A, (__v8sf)__B, (__v16hf)_mm256_setzero_ph(), (__mmask16)(-1),
49		_MM_FROUND_CUR_DIRECTION);
50		}
51
	181	(__v8sf)__A, (__v8sf)__B, (__v16hf)_mm256_setzero_ph(), (__mmask16)(-1));
	182	}
	183
	184	/// Convert two 256-bit vectors, \a __A and \a __B, containing packed
	185	/// single-precision (32-bit) floating-point elements to a 256-bit vector
	186	/// containing FP16 elements. Merging mask \a __U is used to determine if given
	187	/// element should be taken from \a __W instead.
	188	///
	189	/// \code{.operation}
	190	/// FOR i := 0 to 15
	191	/// IF __U[i]
	192	/// IF i < 8
	193	/// dst.fp16[i] := convert_fp32_to_fp16(__B.fp32[i])
	194	/// ELSE
	195	/// dst.fp16[i] := convert_fp32_to_fp16(__A.fp32[i - 8])
	196	/// FI
	197	/// ELSE
	198	/// dst.fp16[i] := __W.fp16[i]
	199	/// FI
	200	/// ENDFOR
	201	///
	202	/// dst[MAX:256] := 0
	203	/// \endcode
	204	///
	205	/// \headerfile <immintrin.h>
	206	///
	207	/// This intrinsic corresponds to the \c VCVT2PS2PHX instruction.
	208	///
	209	/// \param __W
	210	/// A 256-bit vector of [16 x fp16].
	211	/// \param __U
	212	/// A 16-bit merging mask.
	213	/// \param __A
	214	/// A 256-bit vector of [8 x float].
	215	/// \param __B
	216	/// A 256-bit vector of [8 x float].
	217	/// \returns
	218	/// A 256-bit vector of [16 x fp16]. Lower elements correspond to the
	219	/// (converted) elements from \a __B; higher order elements correspond to the
	220	/// (converted) elements from \a __A. If corresponding mask bit is not set, then
	221	/// element from \a __W is taken instead.
52	222	static __inline__ __m256h __DEFAULT_FN_ATTRS256
53	223	_mm256_mask_cvtx2ps_ph(__m256h __W, __mmask16 __U, __m256 __A, __m256 __B) {
54	224	return (__m256h)__builtin_ia32_vcvt2ps2phx256_mask(
55		(__v8sf)__A, (__v8sf)__B, (__v16hf)__W, (__mmask16)__U,
56		_MM_FROUND_CUR_DIRECTION);
57		}
58
	225	(__v8sf)__A, (__v8sf)__B, (__v16hf)__W, (__mmask16)__U);
	226	}
	227
	228	/// Convert two 256-bit vectors, \a __A and \a __B, containing packed
	229	/// single-precision (32-bit) floating-point elements to a 256-bit vector
	230	/// containing FP16 elements. Zeroing mask \a __U is used to determine if given
	231	/// element should be zeroed instead.
	232	///
	233	/// \code{.operation}
	234	/// FOR i := 0 to 15
	235	/// IF __U[i]
	236	/// IF i < 8
	237	/// dst.fp16[i] := convert_fp32_to_fp16(__B.fp32[i])
	238	/// ELSE
	239	/// dst.fp16[i] := convert_fp32_to_fp16(__A.fp32[i - 8])
	240	/// FI
	241	/// ELSE
	242	/// dst.fp16[i] := 0
	243	/// FI
	244	/// ENDFOR
	245	///
	246	/// dst[MAX:256] := 0
	247	/// \endcode
	248	///
	249	/// \headerfile <immintrin.h>
	250	///
	251	/// This intrinsic corresponds to the \c VCVT2PS2PHX instruction.
	252	///
	253	/// \param __U
	254	/// A 16-bit zeroing mask.
	255	/// \param __A
	256	/// A 256-bit vector of [8 x float].
	257	/// \param __B
	258	/// A 256-bit vector of [8 x float].
	259	/// \returns
	260	/// A 256-bit vector of [16 x fp16]. Lower elements correspond to the
	261	/// (converted) elements from \a __B; higher order elements correspond to the
	262	/// (converted) elements from \a __A. If corresponding mask bit is not set,
	263	/// then zero is taken instead.
59	264	static __inline__ __m256h __DEFAULT_FN_ATTRS256
60	265	_mm256_maskz_cvtx2ps_ph(__mmask16 __U, __m256 __A, __m256 __B) {
61	266	return (__m256h)__builtin_ia32_vcvt2ps2phx256_mask(
62		(__v8sf)__A, (__v8sf)__B, (__v16hf)_mm256_setzero_ph(), (__mmask16)__U,
63		_MM_FROUND_CUR_DIRECTION);
64		}
65
66		#define _mm256_cvtx_round2ps_ph(A, B, R) \
67		((__m256h)__builtin_ia32_vcvt2ps2phx256_mask( \
68		(__v8sf)(A), (__v8sf)(B), (__v16hf)_mm256_undefined_ph(), \
69		(__mmask16)(-1), (const int)(R)))
70
71		#define _mm256_mask_cvtx_round2ps_ph(W, U, A, B, R) \
72		((__m256h)__builtin_ia32_vcvt2ps2phx256_mask( \
73		(__v8sf)(A), (__v8sf)(B), (__v16hf)(W), (__mmask16)(U), (const int)(R)))
74
75		#define _mm256_maskz_cvtx_round2ps_ph(U, A, B, R) \
76		((__m256h)__builtin_ia32_vcvt2ps2phx256_mask( \
77		(__v8sf)(A), (__v8sf)(B), (__v16hf)(_mm256_setzero_ph()), \
78		(__mmask16)(U), (const int)(R)))
79
80		static __inline__ __m128i __DEFAULT_FN_ATTRS128 _mm_cvtbiasph_bf8(__m128i __A,
81		__m128h __B) {
	267	(__v8sf)__A, (__v8sf)__B, (__v16hf)_mm256_setzero_ph(), (__mmask16)__U);
	268	}
	269
	270	/// Convert 128-bit vector \a __B containing packed FP16 floating-point elements
	271	/// to FP8 E5M2 numbers, using conversion biases stored in lower 8 bits of each
	272	/// 16-bit integer stored in \a __B.
	273	///
	274	/// \code{.operation}
	275	/// FOR i := 0 to 7
	276	/// dst.bf8[i] := convert_fp16_to_bf8_with_bias(__A.int8[2 * i], __B.fp16[i])
	277	/// ENDFOR
	278	///
	279	/// dst[MAX:64] := 0
	280	/// \endcode
	281	///
	282	/// \headerfile <immintrin.h>
	283	///
	284	/// This intrinsic corresponds to the \c VCVTBIASPH2BF8 instruction.
	285	///
	286	/// \param __A
	287	/// A 128-bit vector of [8 x int16].
	288	/// \param __B
	289	/// A 128-bit vector of [8 x fp16].
	290	/// \returns
	291	/// A 128-bit vector of [16 x bf8]. Lower elements correspond to the
	292	/// converted elements from \a __B using biases from \a __A; higher order
	293	/// elements are zeroed.
	294	static __inline__ __m128i __DEFAULT_FN_ATTRS128
	295	_mm_cvtbiasph_bf8(__m128i __A, __m128h __B) {
82	296	return (__m128i)__builtin_ia32_vcvtbiasph2bf8_128_mask(
83	297	(__v16qi)__A, (__v8hf)__B, (__v16qi)_mm_undefined_si128(), (__mmask8)-1);
84	298	}
85	299
	300	/// Convert 128-bit vector \a __B containing packed FP16 floating-point elements
	301	/// to FP8 E5M2 numbers, using conversion biases stored in lower 8 bits of each
	302	/// 16-bit integer stored in \a __B. Merging mask \a __U is used to determine if
	303	/// given element should be taken from \a __W instead.
	304	///
	305	/// \code{.operation}
	306	/// FOR i := 0 to 7
	307	/// IF __U[i]
	308	/// dst.bf8[i] := convert_fp16_to_bf8_with_bias(__A.int8[2 * i], __B.fp16[i])
	309	/// ELSE
	310	/// dst.bf8[i] := __W.bf8[i]
	311	/// FI
	312	/// ENDFOR
	313	///
	314	/// dst[MAX:64] := 0
	315	/// \endcode
	316	///
	317	/// \headerfile <immintrin.h>
	318	///
	319	/// This intrinsic corresponds to the \c VCVTBIASPH2BF8 instruction.
	320	///
	321	/// \param __W
	322	/// A 128-bit vector of [16 x bf8].
	323	/// \param __U
	324	/// A 8-bit merging mask.
	325	/// \param __A
	326	/// A 128-bit vector of [8 x int16].
	327	/// \param __B
	328	/// A 128-bit vector of [8 x fp16].
	329	/// \returns
	330	/// A 128-bit vector of [16 x bf8]. Lower elements correspond to the
	331	/// converted elements from \a __B, using biases from \a __A; higher order
	332	/// elements are zeroed. If corresponding mask bit is not set, then element
	333	/// from \a __W is taken instead.
86	334	static __inline__ __m128i __DEFAULT_FN_ATTRS128
87	335	_mm_mask_cvtbiasph_bf8(__m128i __W, __mmask8 __U, __m128i __A, __m128h __B) {
88	336	return (__m128i)__builtin_ia32_vcvtbiasph2bf8_128_mask(
89	337	(__v16qi)__A, (__v8hf)__B, (__v16qi)(__m128i)__W, (__mmask8)__U);
90	338	}
91	339
	340	/// Convert 128-bit vector \a __B containing packed FP16 floating-point elements
	341	/// to FP8 E5M2 numbers, using conversion biases stored in lower 8 bits of each
	342	/// 16-bit integer stored in \a __B. Zeroing mask \a __U is used to determine if
	343	/// given element should be zeroed instead.
	344	///
	345	/// \code{.operation}
	346	/// FOR i := 0 to 7
	347	/// IF __U[i]
	348	/// dst.bf8[i] := convert_fp16_to_bf8_with_bias(__A.int8[2 * i], __B.fp16[i])
	349	/// ELSE
	350	/// dst.bf8[i] := 0
	351	/// FI
	352	/// ENDFOR
	353	///
	354	/// dst[MAX:64] := 0
	355	/// \endcode
	356	///
	357	/// \headerfile <immintrin.h>
	358	///
	359	/// This intrinsic corresponds to the \c VCVTBIASPH2BF8 instruction.
	360	///
	361	/// \param __U
	362	/// A 8-bit zeroing mask.
	363	/// \param __A
	364	/// A 128-bit vector of [8 x int16].
	365	/// \param __B
	366	/// A 128-bit vector of [8 x fp16].
	367	/// \returns
	368	/// A 128-bit vector of [16 x bf8]. Lower elements correspond to the
	369	/// converted elements from \a __B, using biases from \a __A; higher order
	370	/// elements are zeroed. If corresponding mask bit is not set, then element
	371	/// is zeroed.
92	372	static __inline__ __m128i __DEFAULT_FN_ATTRS128
93	373	_mm_maskz_cvtbiasph_bf8(__mmask8 __U, __m128i __A, __m128h __B) {
94	374	return (__m128i)__builtin_ia32_vcvtbiasph2bf8_128_mask(
...	...	@@ -96,6 +376,29 @@ _mm_maskz_cvtbiasph_bf8(__mmask8 __U, __m128i __A, __m128h __B) {
96	376	(__mmask8)__U);
97	377	}
98	378
	379	/// Convert 256-bit vector \a __B containing packed FP16 floating-point elements
	380	/// to FP8 E5M2 numbers, using conversion biases stored in lower 8 bits of each
	381	/// 16-bit integer stored in \a __B.
	382	///
	383	/// \code{.operation}
	384	/// FOR i := 0 to 15
	385	/// dst.bf8[i] := convert_fp16_to_bf8_with_bias(__A.int8[2 * i], __B.fp16[i])
	386	/// ENDFOR
	387	///
	388	/// dst[MAX:128] := 0
	389	/// \endcode
	390	///
	391	/// \headerfile <immintrin.h>
	392	///
	393	/// This intrinsic corresponds to the \c VCVTBIASPH2BF8 instruction.
	394	///
	395	/// \param __A
	396	/// A 256-bit vector of [16 x int16].
	397	/// \param __B
	398	/// A 256-bit vector of [16 x fp16].
	399	/// \returns
	400	/// A 128-bit vector of [16 x bf8]. Elements correspond to the
	401	/// converted elements from \a __B using biases from \a __A.
99	402	static __inline__ __m128i __DEFAULT_FN_ATTRS256
100	403	_mm256_cvtbiasph_bf8(__m256i __A, __m256h __B) {
101	404	return (__m128i)__builtin_ia32_vcvtbiasph2bf8_256_mask(
...	...	@@ -103,12 +406,76 @@ _mm256_cvtbiasph_bf8(__m256i __A, __m256h __B) {
103	406	(__mmask16)-1);
104	407	}
105	408
	409	/// Convert 256-bit vector \a __B containing packed FP16 floating-point elements
	410	/// to FP8 E5M2 numbers, using conversion biases stored in lower 8 bits of each
	411	/// 16-bit integer stored in \a __B. Merging mask \a __U is used to determine if
	412	/// given element should be taken from \a __W instead.
	413	///
	414	/// \code{.operation}
	415	/// FOR i := 0 to 15
	416	/// IF __U[i]
	417	/// dst.bf8[i] := convert_fp16_to_bf8_with_bias(__A.int8[2 * i], __B.fp16[i])
	418	/// ELSE
	419	/// dst.bf8[i] := __W.bf8[i]
	420	/// FI
	421	/// ENDFOR
	422	///
	423	/// dst[MAX:128] := 0
	424	/// \endcode
	425	///
	426	/// \headerfile <immintrin.h>
	427	///
	428	/// This intrinsic corresponds to the \c VCVTBIASPH2BF8 instruction.
	429	///
	430	/// \param __W
	431	/// A 128-bit vector of [16 x bf8].
	432	/// \param __U
	433	/// A 16-bit merging mask.
	434	/// \param __A
	435	/// A 256-bit vector of [16 x int16].
	436	/// \param __B
	437	/// A 256-bit vector of [16 x fp16].
	438	/// \returns
	439	/// A 128-bit vector of [16 x bf8]. Elements correspond to the converted
	440	/// elements from \a __B, using biases from \a __A. If corresponding mask bit
	441	/// is not set, then element from \a __W is taken instead.
106	442	static __inline__ __m128i __DEFAULT_FN_ATTRS256 _mm256_mask_cvtbiasph_bf8(
107	443	__m128i __W, __mmask16 __U, __m256i __A, __m256h __B) {
108	444	return (__m128i)__builtin_ia32_vcvtbiasph2bf8_256_mask(
109	445	(__v32qi)__A, (__v16hf)__B, (__v16qi)(__m128i)__W, (__mmask16)__U);
110	446	}
111	447
	448	/// Convert 256-bit vector \a __B containing packed FP16 floating-point elements
	449	/// to FP8 E5M2 numbers, using conversion biases stored in lower 8 bits of each
	450	/// 16-bit integer stored in \a __B. Zeroing mask \a __U is used to determine if
	451	/// given element should be zeroed instead.
	452	///
	453	/// \code{.operation}
	454	/// FOR i := 0 to 15
	455	/// IF __U[i]
	456	/// dst.bf8[i] := convert_fp16_to_bf8_with_bias(__A.int8[2 * i], __B.fp16[i])
	457	/// ELSE
	458	/// dst.bf8[i] := 0
	459	/// FI
	460	/// ENDFOR
	461	///
	462	/// dst[MAX:128] := 0
	463	/// \endcode
	464	///
	465	/// \headerfile <immintrin.h>
	466	///
	467	/// This intrinsic corresponds to the \c VCVTBIASPH2BF8 instruction.
	468	///
	469	/// \param __U
	470	/// A 16-bit zeroing mask.
	471	/// \param __A
	472	/// A 256-bit vector of [16 x int16].
	473	/// \param __B
	474	/// A 256-bit vector of [16 x fp16].
	475	/// \returns
	476	/// A 128-bit vector of [16 x bf8]. Elements correspond to the converted
	477	/// elements from \a __B, using biases from \a __A. If corresponding mask bit
	478	/// is not set, then element is zeroed.
112	479	static __inline__ __m128i __DEFAULT_FN_ATTRS256
113	480	_mm256_maskz_cvtbiasph_bf8(__mmask16 __U, __m256i __A, __m256h __B) {
114	481	return (__m128i)__builtin_ia32_vcvtbiasph2bf8_256_mask(
...	...	@@ -116,57 +483,324 @@ _mm256_maskz_cvtbiasph_bf8(__mmask16 __U, __m256i __A, __m256h __B) {
116	483	(__mmask16)__U);
117	484	}
118	485
	486	/// Convert 128-bit vector \a __B containing packed FP16 floating-point elements
	487	/// to FP8 E5M2 numbers, using conversion biases stored in lower 8 bits of each
	488	/// 16-bit integer stored in \a __B. Results are saturated.
	489	///
	490	/// \code{.operation}
	491	/// FOR i := 0 to 7
	492	/// dst.bf8[i] := convert_fp16_to_bf8_with_bias_saturate(__A.int8[2 * i], __B.fp16[i])
	493	/// ENDFOR
	494	///
	495	/// dst[MAX:64] := 0
	496	/// \endcode
	497	///
	498	/// \headerfile <immintrin.h>
	499	///
	500	/// This intrinsic corresponds to the \c VCVTBIASPH2BF8 instruction.
	501	///
	502	/// \param __A
	503	/// A 128-bit vector of [8 x int16].
	504	/// \param __B
	505	/// A 128-bit vector of [8 x fp16].
	506	/// \returns
	507	/// A 128-bit vector of [16 x bf8]. Lower elements correspond to the
	508	/// converted elements from \a __B using biases from \a __A; higher order
	509	/// elements are zeroed.
119	510	static __inline__ __m128i __DEFAULT_FN_ATTRS128
120		_mm_cvtbiassph_bf8(__m128i __A, __m128h __B) {
	511	_mm_cvts_biasph_bf8(__m128i __A, __m128h __B) {
121	512	return (__m128i)__builtin_ia32_vcvtbiasph2bf8s_128_mask(
122	513	(__v16qi)__A, (__v8hf)__B, (__v16qi)_mm_undefined_si128(), (__mmask8)-1);
123	514	}
124	515
	516	/// Convert 128-bit vector \a __B containing packed FP16 floating-point elements
	517	/// to FP8 E5M2 numbers, using conversion biases stored in lower 8 bits of each
	518	/// 16-bit integer stored in \a __B. Results are saturated. Merging mask \a __U
	519	/// is used to determine if given element should be taken from \a __W instead.
	520	///
	521	/// \code{.operation}
	522	/// FOR i := 0 to 7
	523	/// IF __U[i]
	524	/// dst.bf8[i] := convert_fp16_to_bf8_with_bias_saturate(__A.int8[2 * i], __B.fp16[i])
	525	/// ELSE
	526	/// dst.bf8[i] := __W.bf8[i]
	527	/// FI
	528	/// ENDFOR
	529	///
	530	/// dst[MAX:64] := 0
	531	/// \endcode
	532	///
	533	/// \headerfile <immintrin.h>
	534	///
	535	/// This intrinsic corresponds to the \c VCVTBIASPH2BF8S instruction.
	536	///
	537	/// \param __W
	538	/// A 128-bit vector of [16 x bf8].
	539	/// \param __U
	540	/// A 8-bit merging mask.
	541	/// \param __A
	542	/// A 128-bit vector of [8 x int16].
	543	/// \param __B
	544	/// A 128-bit vector of [8 x fp16].
	545	/// \returns
	546	/// A 128-bit vector of [16 x bf8]. Lower elements correspond to the
	547	/// converted elements from \a __B, using biases from \a __A; higher order
	548	/// elements are zeroed. If corresponding mask bit is not set, then element
	549	/// from \a __W is taken instead.
	550	static __inline__ __m128i __DEFAULT_FN_ATTRS128 _mm_mask_cvts_biasph_bf8(__m128i
	551	__W, __mmask8 __U, __m128i __A, __m128h __B) { return
	552	(__m128i)__builtin_ia32_vcvtbiasph2bf8s_128_mask( (__v16qi)__A,
	553	(__v8hf)__B, (__v16qi)(__m128i)__W, (__mmask8)__U); }
	554
	555	/// Convert 128-bit vector \a __B containing packed FP16 floating-point elements
	556	/// to FP8 E5M2 numbers, using conversion biases stored in lower 8 bits of each
	557	/// 16-bit integer stored in \a __B. Results are saturated. Zeroing mask \a __U
	558	/// is used to determine if given element should be zeroed instead.
	559	///
	560	/// \code{.operation}
	561	/// FOR i := 0 to 7
	562	/// IF __U[i]
	563	/// dst.bf8[i] := convert_fp16_to_bf8_with_bias_saturate(__A.int8[2 * i], __B.fp16[i])
	564	/// ELSE
	565	/// dst.bf8[i] := 0
	566	/// FI
	567	/// ENDFOR
	568	///
	569	/// dst[MAX:64] := 0
	570	/// \endcode
	571	///
	572	/// \headerfile <immintrin.h>
	573	///
	574	/// This intrinsic corresponds to the \c VCVTBIASPH2BF8S instruction.
	575	///
	576	/// \param __U
	577	/// A 8-bit zeroing mask.
	578	/// \param __A
	579	/// A 128-bit vector of [8 x int16].
	580	/// \param __B
	581	/// A 128-bit vector of [8 x fp16].
	582	/// \returns
	583	/// A 128-bit vector of [16 x bf8]. Lower elements correspond to the
	584	/// converted elements from \a __B, using biases from \a __A; higher order
	585	/// elements are zeroed. If corresponding mask bit is not set, then element
	586	/// is zeroed.
125	587	static __inline__ __m128i __DEFAULT_FN_ATTRS128
126		_mm_mask_cvtbiassph_bf8(__m128i __W, __mmask8 __U, __m128i __A, __m128h __B) {
127		return (__m128i)__builtin_ia32_vcvtbiasph2bf8s_128_mask(
128		(__v16qi)__A, (__v8hf)__B, (__v16qi)(__m128i)__W, (__mmask8)__U);
129		}
130
131		static __inline__ __m128i __DEFAULT_FN_ATTRS128
132		_mm_maskz_cvtbiassph_bf8(__mmask8 __U, __m128i __A, __m128h __B) {
	588	_mm_maskz_cvts_biasph_bf8(__mmask8 __U, __m128i __A, __m128h __B) {
133	589	return (__m128i)__builtin_ia32_vcvtbiasph2bf8s_128_mask(
134	590	(__v16qi)__A, (__v8hf)__B, (__v16qi)(__m128i)_mm_setzero_si128(),
135	591	(__mmask8)__U);
136	592	}
137	593
	594
	595	/// Convert 256-bit vector \a __B containing packed FP16 floating-point elements
	596	/// to FP8 E5M2 numbers, using conversion biases stored in lower 8 bits of each
	597	/// 16-bit integer stored in \a __B. Results are saturated.
	598	///
	599	/// \code{.operation}
	600	/// FOR i := 0 to 15
	601	/// dst.bf8[i] := convert_fp16_to_bf8_with_bias_saturate(__A.int8[2 * i], __B.fp16[i])
	602	/// ENDFOR
	603	///
	604	/// dst[MAX:128] := 0
	605	/// \endcode
	606	///
	607	/// \headerfile <immintrin.h>
	608	///
	609	/// This intrinsic corresponds to the \c VCVTBIASPH2BF8S instruction.
	610	///
	611	/// \param __A
	612	/// A 256-bit vector of [16 x int16].
	613	/// \param __B
	614	/// A 256-bit vector of [16 x fp16].
	615	/// \returns
	616	/// A 128-bit vector of [16 x bf8]. Elements correspond to the
	617	/// converted elements from \a __B using biases from \a __A.
138	618	static __inline__ __m128i __DEFAULT_FN_ATTRS256
139		_mm256_cvtbiassph_bf8(__m256i __A, __m256h __B) {
	619	_mm256_cvts_biasph_bf8(__m256i __A, __m256h __B) {
140	620	return (__m128i)__builtin_ia32_vcvtbiasph2bf8s_256_mask(
141	621	(__v32qi)__A, (__v16hf)__B, (__v16qi)(__m128i)_mm_undefined_si128(),
142	622	(__mmask16)-1);
143	623	}
144	624
145		static __inline__ __m128i __DEFAULT_FN_ATTRS256 _mm256_mask_cvtbiassph_bf8(
	625	/// Convert 256-bit vector \a __B containing packed FP16 floating-point elements
	626	/// to FP8 E5M2 numbers, using conversion biases stored in lower 8 bits of each
	627	/// 16-bit integer stored in \a __B. Results are saturated. Merging mask \a __U
	628	/// is used to determine if given element should be taken from \a __W instead.
	629	///
	630	/// \code{.operation}
	631	/// FOR i := 0 to 15
	632	/// IF __U[i]
	633	/// dst.bf8[i] := convert_fp16_to_bf8_with_bias_saturate(__A.int8[2 * i], __B.fp16[i])
	634	/// ELSE
	635	/// dst.bf8[i] := __W.bf8[i]
	636	/// FI
	637	/// ENDFOR
	638	///
	639	/// dst[MAX:128] := 0
	640	/// \endcode
	641	///
	642	/// \headerfile <immintrin.h>
	643	///
	644	/// This intrinsic corresponds to the \c VCVTBIASPH2BF8S instruction.
	645	///
	646	/// \param __W
	647	/// A 128-bit vector of [16 x bf8].
	648	/// \param __U
	649	/// A 16-bit merging mask.
	650	/// \param __A
	651	/// A 256-bit vector of [16 x int16].
	652	/// \param __B
	653	/// A 256-bit vector of [16 x fp16].
	654	/// \returns
	655	/// A 128-bit vector of [16 x bf8]. Elements correspond to the converted
	656	/// elements from \a __B, using biases from \a __A. If corresponding mask bit
	657	/// is not set, then element from \a __W is taken instead.
	658	static __inline__ __m128i __DEFAULT_FN_ATTRS256 _mm256_mask_cvts_biasph_bf8(
146	659	__m128i __W, __mmask16 __U, __m256i __A, __m256h __B) {
147	660	return (__m128i)__builtin_ia32_vcvtbiasph2bf8s_256_mask(
148	661	(__v32qi)__A, (__v16hf)__B, (__v16qi)(__m128i)__W, (__mmask16)__U);
149	662	}
150	663
	664	/// Convert 256-bit vector \a __B containing packed FP16 floating-point elements
	665	/// to FP8 E5M2 numbers, using conversion biases stored in lower 8 bits of each
	666	/// 16-bit integer stored in \a __B. Results are saturated. Zeroing mask \a __U
	667	/// is used to determine if given element should be zeroed instead.
	668	///
	669	/// \code{.operation}
	670	/// FOR i := 0 to 15
	671	/// IF __U[i]
	672	/// dst.bf8[i] := convert_fp16_to_bf8_with_bias_saturate(__A.int8[2 * i], __B.fp16[i])
	673	/// ELSE
	674	/// dst.bf8[i] := 0
	675	/// FI
	676	/// ENDFOR
	677	///
	678	/// dst[MAX:128] := 0
	679	/// \endcode
	680	///
	681	/// \headerfile <immintrin.h>
	682	///
	683	/// This intrinsic corresponds to the \c VCVTBIASPH2BF8S instruction.
	684	///
	685	/// \param __U
	686	/// A 16-bit zeroing mask.
	687	/// \param __A
	688	/// A 256-bit vector of [16 x int16].
	689	/// \param __B
	690	/// A 256-bit vector of [16 x fp16].
	691	/// \returns
	692	/// A 128-bit vector of [16 x bf8]. Elements correspond to the converted
	693	/// elements from \a __B, using biases from \a __A. If corresponding mask bit
	694	/// is not set, then element is zeroed.
151	695	static __inline__ __m128i __DEFAULT_FN_ATTRS256
152		_mm256_maskz_cvtbiassph_bf8(__mmask16 __U, __m256i __A, __m256h __B) {
	696	_mm256_maskz_cvts_biasph_bf8(__mmask16 __U, __m256i __A, __m256h __B) {
153	697	return (__m128i)__builtin_ia32_vcvtbiasph2bf8s_256_mask(
154	698	(__v32qi)__A, (__v16hf)__B, (__v16qi)(__m128i)_mm_setzero_si128(),
155	699	(__mmask16)__U);
156	700	}
157	701
158		static __inline__ __m128i __DEFAULT_FN_ATTRS128 _mm_cvtbiasph_hf8(__m128i __A,
159		__m128h __B) {
	702	/// Convert 128-bit vector \a __B containing packed FP16 floating-point elements
	703	/// to FP8 E4M3 numbers, using conversion biases stored in lower 8 bits of each
	704	/// 16-bit integer stored in \a __B.
	705	///
	706	/// \code{.operation}
	707	/// FOR i := 0 to 7
	708	/// dst.hf8[i] := convert_fp16_to_hf8_with_bias(__A.int8[2 * i], __B.fp16[i])
	709	/// ENDFOR
	710	///
	711	/// dst[MAX:64] := 0
	712	/// \endcode
	713	///
	714	/// \headerfile <immintrin.h>
	715	///
	716	/// This intrinsic corresponds to the \c VCVTBIASPH2HF8 instruction.
	717	///
	718	/// \param __A
	719	/// A 128-bit vector of [8 x int16].
	720	/// \param __B
	721	/// A 128-bit vector of [8 x fp16].
	722	/// \returns
	723	/// A 128-bit vector of [16 x hf8]. Lower elements correspond to the
	724	/// converted elements from \a __B using biases from \a __A; higher order
	725	/// elements are zeroed.
	726	static __inline__ __m128i __DEFAULT_FN_ATTRS128
	727	_mm_cvtbiasph_hf8(__m128i __A, __m128h __B) {
160	728	return (__m128i)__builtin_ia32_vcvtbiasph2hf8_128_mask(
161	729	(__v16qi)__A, (__v8hf)__B, (__v16qi)_mm_undefined_si128(), (__mmask8)-1);
162	730	}
163	731
	732	/// Convert 128-bit vector \a __B containing packed FP16 floating-point elements
	733	/// to FP8 E4M3 numbers, using conversion biases stored in lower 8 bits of each
	734	/// 16-bit integer stored in \a __B. Merging mask \a __U is used to determine if
	735	/// given element should be taken from \a __W instead.
	736	///
	737	/// \code{.operation}
	738	/// FOR i := 0 to 7
	739	/// IF __U[i]
	740	/// dst.hf8[i] := convert_fp16_to_hf8_with_bias(__A.int8[2 * i], __B.fp16[i])
	741	/// ELSE
	742	/// dst.hf8[i] := __W.hf8[i]
	743	/// FI
	744	/// ENDFOR
	745	///
	746	/// dst[MAX:64] := 0
	747	/// \endcode
	748	///
	749	/// \headerfile <immintrin.h>
	750	///
	751	/// This intrinsic corresponds to the \c VCVTBIASPH2HF8 instruction.
	752	///
	753	/// \param __W
	754	/// A 128-bit vector of [16 x hf8].
	755	/// \param __U
	756	/// A 8-bit merging mask.
	757	/// \param __A
	758	/// A 128-bit vector of [8 x int16].
	759	/// \param __B
	760	/// A 128-bit vector of [8 x fp16].
	761	/// \returns
	762	/// A 128-bit vector of [16 x hf8]. Lower elements correspond to the
	763	/// converted elements from \a __B, using biases from \a __A; higher order
	764	/// elements are zeroed. If corresponding mask bit is not set, then element
	765	/// from \a __W is taken instead.
164	766	static __inline__ __m128i __DEFAULT_FN_ATTRS128
165	767	_mm_mask_cvtbiasph_hf8(__m128i __W, __mmask8 __U, __m128i __A, __m128h __B) {
166	768	return (__m128i)__builtin_ia32_vcvtbiasph2hf8_128_mask(
167	769	(__v16qi)__A, (__v8hf)__B, (__v16qi)(__m128i)__W, (__mmask8)__U);
168	770	}
169	771
	772	/// Convert 128-bit vector \a __B containing packed FP16 floating-point elements
	773	/// to FP8 E4M3 numbers, using conversion biases stored in lower 8 bits of each
	774	/// 16-bit integer stored in \a __B. Zeroing mask \a __U is used to determine if
	775	/// given element should be zeroed instead.
	776	///
	777	/// \code{.operation}
	778	/// FOR i := 0 to 7
	779	/// IF __U[i]
	780	/// dst.hf8[i] := convert_fp16_to_hf8_with_bias(__A.int8[2 * i], __B.fp16[i])
	781	/// ELSE
	782	/// dst.hf8[i] := 0
	783	/// FI
	784	/// ENDFOR
	785	///
	786	/// dst[MAX:64] := 0
	787	/// \endcode
	788	///
	789	/// \headerfile <immintrin.h>
	790	///
	791	/// This intrinsic corresponds to the \c VCVTBIASPH2HF8 instruction.
	792	///
	793	/// \param __U
	794	/// A 8-bit zeroing mask.
	795	/// \param __A
	796	/// A 128-bit vector of [8 x int16].
	797	/// \param __B
	798	/// A 128-bit vector of [8 x fp16].
	799	/// \returns
	800	/// A 128-bit vector of [16 x hf8]. Lower elements correspond to the
	801	/// converted elements from \a __B, using biases from \a __A; higher order
	802	/// elements are zeroed. If corresponding mask bit is not set, then element
	803	/// is zeroed.
170	804	static __inline__ __m128i __DEFAULT_FN_ATTRS128
171	805	_mm_maskz_cvtbiasph_hf8(__mmask8 __U, __m128i __A, __m128h __B) {
172	806	return (__m128i)__builtin_ia32_vcvtbiasph2hf8_128_mask(
...	...	@@ -174,6 +808,29 @@ _mm_maskz_cvtbiasph_hf8(__mmask8 __U, __m128i __A, __m128h __B) {
174	808	(__mmask8)__U);
175	809	}
176	810
	811	/// Convert 256-bit vector \a __B containing packed FP16 floating-point elements
	812	/// to FP8 E4M3 numbers, using conversion biases stored in lower 8 bits of each
	813	/// 16-bit integer stored in \a __B.
	814	///
	815	/// \code{.operation}
	816	/// FOR i := 0 to 15
	817	/// dst.hf8[i] := convert_fp16_to_hf8_with_bias(__A.int8[2 * i], __B.fp16[i])
	818	/// ENDFOR
	819	///
	820	/// dst[MAX:128] := 0
	821	/// \endcode
	822	///
	823	/// \headerfile <immintrin.h>
	824	///
	825	/// This intrinsic corresponds to the \c VCVTBIASPH2HF8 instruction.
	826	///
	827	/// \param __A
	828	/// A 256-bit vector of [16 x half].
	829	/// \param __B
	830	/// A 256-bit vector of [16 x i16].
	831	/// \returns
	832	/// A 128-bit vector of [16 x hf8]. Elements correspond to the
	833	/// converted elements from \a __B using biases from \a __A.
177	834	static __inline__ __m128i __DEFAULT_FN_ATTRS256
178	835	_mm256_cvtbiasph_hf8(__m256i __A, __m256h __B) {
179	836	return (__m128i)__builtin_ia32_vcvtbiasph2hf8_256_mask(
...	...	@@ -181,12 +838,76 @@ _mm256_cvtbiasph_hf8(__m256i __A, __m256h __B) {
181	838	(__mmask16)-1);
182	839	}
183	840
	841	/// Convert 256-bit vector \a __B containing packed FP16 floating-point elements
	842	/// to FP8 E4M3 numbers, using conversion biases stored in lower 8 bits of each
	843	/// 16-bit integer stored in \a __B. Merging mask \a __U is used to determine if
	844	/// given element should be taken from \a __W instead.
	845	///
	846	/// \code{.operation}
	847	/// FOR i := 0 to 15
	848	/// IF __U[i]
	849	/// dst.hf8[i] := convert_fp16_to_hf8_with_bias(__A.int8[2 * i], __B.fp16[i])
	850	/// ELSE
	851	/// dst.hf8[i] := __W.hf8[i]
	852	/// FI
	853	/// ENDFOR
	854	///
	855	/// dst[MAX:128] := 0
	856	/// \endcode
	857	///
	858	/// \headerfile <immintrin.h>
	859	///
	860	/// This intrinsic corresponds to the \c VCVTBIASPH2HF8 instruction.
	861	///
	862	/// \param __W
	863	/// A 128-bit vector of [16 x hf8].
	864	/// \param __U
	865	/// A 16-bit merging mask.
	866	/// \param __A
	867	/// A 256-bit vector of [16 x int16].
	868	/// \param __B
	869	/// A 256-bit vector of [16 x fp16].
	870	/// \returns
	871	/// A 128-bit vector of [16 x hf8]. Elements correspond to the converted
	872	/// elements from \a __B, using biases from \a __A. If corresponding mask bit
	873	/// is not set, then element from \a __W is taken instead.
184	874	static __inline__ __m128i __DEFAULT_FN_ATTRS256 _mm256_mask_cvtbiasph_hf8(
185	875	__m128i __W, __mmask16 __U, __m256i __A, __m256h __B) {
186	876	return (__m128i)__builtin_ia32_vcvtbiasph2hf8_256_mask(
187	877	(__v32qi)__A, (__v16hf)__B, (__v16qi)(__m128i)__W, (__mmask16)__U);
188	878	}
189	879
	880	/// Convert 256-bit vector \a __B containing packed FP16 floating-point elements
	881	/// to FP8 E4M3 numbers, using conversion biases stored in lower 8 bits of each
	882	/// 16-bit integer stored in \a __B. Zeroing mask \a __U is used to determine if
	883	/// given element should be taken zeroed instead.
	884	///
	885	/// \code{.operation}
	886	/// FOR i := 0 to 15
	887	/// IF __U[i]
	888	/// dst.hf8[i] := convert_fp16_to_hf8_with_bias(__A.int8[2 * i], __B.fp16[i])
	889	/// ELSE
	890	/// dst.hf8[i] := 0
	891	/// FI
	892	/// ENDFOR
	893	///
	894	/// dst[MAX:128] := 0
	895	/// \endcode
	896	///
	897	/// \headerfile <immintrin.h>
	898	///
	899	/// This intrinsic corresponds to the \c VCVTBIASPH2HF8 instruction.
	900	///
	901	/// \param __U
	902	/// A 16-bit zeroing mask.
	903	/// \param __A
	904	/// A 256-bit vector of [16 x half].
	905	/// \param __B
	906	/// A 256-bit vector of [16 x i16].
	907	/// \returns
	908	/// A 128-bit vector of [16 x hf8]. Elements correspond to the converted
	909	/// elements from \a __B, using biases from \a __A. If corresponding mask bit
	910	/// is not set, then element is zeroed.
190	911	static __inline__ __m128i __DEFAULT_FN_ATTRS256
191	912	_mm256_maskz_cvtbiasph_hf8(__mmask16 __U, __m256i __A, __m256h __B) {
192	913	return (__m128i)__builtin_ia32_vcvtbiasph2hf8_256_mask(
...	...	@@ -194,56 +915,335 @@ _mm256_maskz_cvtbiasph_hf8(__mmask16 __U, __m256i __A, __m256h __B) {
194	915	(__mmask16)__U);
195	916	}
196	917
	918	/// Convert 128-bit vector \a __B containing packed FP16 floating-point elements
	919	/// to FP8 E4M3 numbers, using conversion biases stored in lower 8 bits of each
	920	/// 16-bit integer stored in \a __B. Results are saturated.
	921	///
	922	/// \code{.operation}
	923	/// FOR i := 0 to 7
	924	/// dst.hf8[i] := convert_fp16_to_hf8_with_bias_saturate(__A.int8[2 * i], __B.fp16[i])
	925	/// ENDFOR
	926	///
	927	/// dst[MAX:64] := 0
	928	/// \endcode
	929	///
	930	/// \headerfile <immintrin.h>
	931	///
	932	/// This intrinsic corresponds to the \c VCVTBIASPH2HF8S`instruction.
	933	///
	934	/// \param __A
	935	/// A 128-bit vector of [8 x int16].
	936	/// \param __B
	937	/// A 128-bit vector of [8 x fp16].
	938	/// \returns
	939	/// A 128-bit vector of [16 x hf8]. Lower elements correspond to the
	940	/// converted elements from \a __B using biases from \a __A; higher order
	941	/// elements are zeroed.
197	942	static __inline__ __m128i __DEFAULT_FN_ATTRS128
198		_mm_cvtbiassph_hf8(__m128i __A, __m128h __B) {
	943	_mm_cvts_biasph_hf8(__m128i __A, __m128h __B) {
199	944	return (__m128i)__builtin_ia32_vcvtbiasph2hf8s_128_mask(
200	945	(__v16qi)__A, (__v8hf)__B, (__v16qi)_mm_undefined_si128(), (__mmask8)-1);
201	946	}
202	947
	948	/// Convert 128-bit vector \a __B containing packed FP16 floating-point elements
	949	/// to FP8 E4M3 numbers, using conversion biases stored in lower 8 bits of each
	950	/// 16-bit integer stored in \a __B. Results are saturated. Merging mask \a __U
	951	/// is used to determine if given element should be taken from \a __W instead.
	952	///
	953	/// \code{.operation}
	954	/// FOR i := 0 to 7
	955	/// IF __U[i]
	956	/// dst.hf8[i] := convert_fp16_to_hf8_with_bias_saturate(__A.int8[2 * i], __B.fp16[i])
	957	/// ELSE
	958	/// dst.hf8[i] := __W.hf8[i]
	959	/// FI
	960	/// ENDFOR
	961	///
	962	/// dst[MAX:64] := 0
	963	/// \endcode
	964	///
	965	/// \headerfile <immintrin.h>
	966	///
	967	/// This intrinsic corresponds to the \c VCVTBIASPH2HF8S instruction.
	968	///
	969	/// \param __W
	970	/// A 128-bit vector of [16 x hf8].
	971	/// \param __U
	972	/// A 8-bit merging mask.
	973	/// \param __A
	974	/// A 128-bit vector of [8 x int16].
	975	/// \param __B
	976	/// A 128-bit vector of [8 x fp16].
	977	/// \returns
	978	/// A 128-bit vector of [16 x hf8]. Lower elements correspond to the
	979	/// converted elements from \a __B, using biases from \a __A; higher order
	980	/// elements are zeroed. If corresponding mask bit is not set, then element
	981	/// from \a __W is taken instead.
203	982	static __inline__ __m128i __DEFAULT_FN_ATTRS128
204		_mm_mask_cvtbiassph_hf8(__m128i __W, __mmask8 __U, __m128i __A, __m128h __B) {
	983	_mm_mask_cvts_biasph_hf8(__m128i __W, __mmask8 __U, __m128i __A, __m128h __B) {
205	984	return (__m128i)__builtin_ia32_vcvtbiasph2hf8s_128_mask(
206	985	(__v16qi)__A, (__v8hf)__B, (__v16qi)(__m128i)__W, (__mmask8)__U);
207	986	}
208	987
	988	/// Convert 128-bit vector \a __B containing packed FP16 floating-point elements
	989	/// to FP8 E4M3 numbers, using conversion biases stored in lower 8 bits of each
	990	/// 16-bit integer stored in \a __B. Results are saturated. Zeroing mask \a __U
	991	/// is used to determine if given element should be zeroed instead.
	992	///
	993	/// \code{.operation}
	994	/// FOR i := 0 to 7
	995	/// IF __U[i]
	996	/// dst.hf8[i] := convert_fp16_to_hf8_with_bias_saturate(__A.int8[2 * i], __B.fp16[i])
	997	/// ELSE
	998	/// dst.hf8[i] := 0
	999	/// FI
	1000	/// ENDFOR
	1001	///
	1002	/// dst[MAX:64] := 0
	1003	/// \endcode
	1004	///
	1005	/// \headerfile <immintrin.h>
	1006	///
	1007	/// This intrinsic corresponds to the \c VCVTBIASPH2HF8S instruction.
	1008	///
	1009	/// \param __U
	1010	/// A 8-bit zeroing mask.
	1011	/// \param __A
	1012	/// A 128-bit vector of [8 x int16].
	1013	/// \param __B
	1014	/// A 128-bit vector of [8 x fp16].
	1015	/// \returns
	1016	/// A 128-bit vector of [16 x hf8]. Lower elements correspond to the
	1017	/// converted elements from \a __B, using biases from \a __A; higher order
	1018	/// elements are zeroed. If corresponding mask bit is not set, then element
	1019	/// is zeroed.
209	1020	static __inline__ __m128i __DEFAULT_FN_ATTRS128
210		_mm_maskz_cvtbiassph_hf8(__mmask8 __U, __m128i __A, __m128h __B) {
	1021	_mm_maskz_cvts_biasph_hf8(__mmask8 __U, __m128i __A, __m128h __B) {
211	1022	return (__m128i)__builtin_ia32_vcvtbiasph2hf8s_128_mask(
212	1023	(__v16qi)__A, (__v8hf)__B, (__v16qi)(__m128i)_mm_setzero_si128(),
213	1024	(__mmask8)__U);
214	1025	}
215	1026
	1027	/// Convert 256-bit vector \a __B containing packed FP16 floating-point elements
	1028	/// to FP8 E4M3 numbers, using conversion biases stored in lower 8 bits of each
	1029	/// 16-bit integer stored in \a __B. Results are saturated.
	1030	///
	1031	/// \code{.operation}
	1032	/// FOR i := 0 to 15
	1033	/// dst.hf8[i] := convert_fp16_to_hf8_with_bias_saturate(__A.int8[2 * i], __B.fp16[i])
	1034	/// ENDFOR
	1035	///
	1036	/// dst[MAX:128] := 0
	1037	/// \endcode
	1038	///
	1039	/// \headerfile <immintrin.h>
	1040	///
	1041	/// This intrinsic corresponds to the \c VCVTBIASPH2HF8S instruction.
	1042	///
	1043	/// \param __A
	1044	/// A 256-bit vector of [16 x int16].
	1045	/// \param __B
	1046	/// A 256-bit vector of [16 x fp16].
	1047	/// \returns
	1048	/// A 128-bit vector of [16 x hf8]. Elements correspond to the
	1049	/// converted elements from \a __B using biases from \a __A.
216	1050	static __inline__ __m128i __DEFAULT_FN_ATTRS256
217		_mm256_cvtbiassph_hf8(__m256i __A, __m256h __B) {
	1051	_mm256_cvts_biasph_hf8(__m256i __A, __m256h __B) {
218	1052	return (__m128i)__builtin_ia32_vcvtbiasph2hf8s_256_mask(
219	1053	(__v32qi)__A, (__v16hf)__B, (__v16qi)(__m128i)_mm_undefined_si128(),
220	1054	(__mmask16)-1);
221	1055	}
222	1056
223		static __inline__ __m128i __DEFAULT_FN_ATTRS256 _mm256_mask_cvtbiassph_hf8(
	1057	/// Convert 256-bit vector \a __B containing packed FP16 floating-point elements
	1058	/// to FP8 E4M3 numbers, using conversion biases stored in lower 8 bits of each
	1059	/// 16-bit integer stored in \a __B. Results are saturated. Merging mask \a __U
	1060	/// is used to determine if given element should be taken from \a __W instead.
	1061	///
	1062	/// \code{.operation}
	1063	/// FOR i := 0 to 15
	1064	/// IF __U[i]
	1065	/// dst.hf8[i] := convert_fp16_to_hf8_with_bias_saturate(__A.int8[2 * i], __B.fp16[i])
	1066	/// ELSE
	1067	/// dst.hf8[i] := __W.hf8[i]
	1068	/// FI
	1069	/// ENDFOR
	1070	///
	1071	/// dst[MAX:128] := 0
	1072	/// \endcode
	1073	///
	1074	/// \headerfile <immintrin.h>
	1075	///
	1076	/// This intrinsic corresponds to the \c VCVTBIASPH2HF8S instruction.
	1077	///
	1078	/// \param __W
	1079	/// A 128-bit vector of [16 x hf8].
	1080	/// \param __U
	1081	/// A 16-bit merging mask.
	1082	/// \param __A
	1083	/// A 256-bit vector of [16 x int16].
	1084	/// \param __B
	1085	/// A 256-bit vector of [16 x fp16].
	1086	/// \returns
	1087	/// A 128-bit vector of [16 x hf8]. Elements correspond to the converted
	1088	/// elements from \a __B, using biases from \a __A. If corresponding mask bit
	1089	/// is not set, then element from \a __W is taken instead.
	1090	static __inline__ __m128i __DEFAULT_FN_ATTRS256 _mm256_mask_cvts_biasph_hf8(
224	1091	__m128i __W, __mmask16 __U, __m256i __A, __m256h __B) {
225	1092	return (__m128i)__builtin_ia32_vcvtbiasph2hf8s_256_mask(
226	1093	(__v32qi)__A, (__v16hf)__B, (__v16qi)(__m128i)__W, (__mmask16)__U);
227	1094	}
228	1095
	1096	/// Convert 256-bit vector \a __B containing packed FP16 floating-point elements
	1097	/// to FP8 E4M3 numbers, using conversion biases stored in lower 8 bits of each
	1098	/// 16-bit integer stored in \a __B. Results are saturated. Zeroing mask \a __U
	1099	/// is used to determine if given element should be zeroed instead.
	1100	///
	1101	/// \code{.operation}
	1102	/// FOR i := 0 to 15
	1103	/// IF __U[i]
	1104	/// dst.hf8[i] := convert_fp16_to_hf8_with_bias_saturate(__A.int8[2 * i], __B.fp16[i])
	1105	/// ELSE
	1106	/// dst.hf8[i] := 0
	1107	/// FI
	1108	/// ENDFOR
	1109	///
	1110	/// dst[MAX:128] := 0
	1111	/// \endcode
	1112	///
	1113	/// \headerfile <immintrin.h>
	1114	///
	1115	/// This intrinsic corresponds to the \c VCVTBIASPH2HF8S instruction.
	1116	///
	1117	/// \param __U
	1118	/// A 16-bit zeroing mask.
	1119	/// \param __A
	1120	/// A 256-bit vector of [16 x int16].
	1121	/// \param __B
	1122	/// A 256-bit vector of [16 x fp16].
	1123	/// \returns
	1124	/// A 128-bit vector of [16 x hf8]. Elements correspond to the converted
	1125	/// elements from \a __B, using biases from \a __A. If corresponding mask bit
	1126	/// is not set, then element is zeroed.
229	1127	static __inline__ __m128i __DEFAULT_FN_ATTRS256
230		_mm256_maskz_cvtbiassph_hf8(__mmask16 __U, __m256i __A, __m256h __B) {
	1128	_mm256_maskz_cvts_biasph_hf8(__mmask16 __U, __m256i __A, __m256h __B) {
231	1129	return (__m128i)__builtin_ia32_vcvtbiasph2hf8s_256_mask(
232	1130	(__v32qi)__A, (__v16hf)__B, (__v16qi)(__m128i)_mm_setzero_si128(),
233	1131	(__mmask16)__U);
234	1132	}
235	1133
	1134	/// Convert two 128-bit vectors, \a __A and \a __B, containing packed FP16
	1135	/// floating-point elements to a 128-bit vector containing E5M2 FP8 elements.
	1136	///
	1137	/// \code{.operation}
	1138	/// FOR i := 0 to 15
	1139	/// IF i < 8
	1140	/// dst.bf8[i] := convert_fp16_to_bf8(__B.fp16[i])
	1141	/// ELSE
	1142	/// dst.bf8[i] := convert_fp16_to_bf8(__A.fp16[i - 8])
	1143	/// FI
	1144	/// ENDFOR
	1145	///
	1146	/// dst[MAX:128] := 0
	1147	/// \endcode
	1148	///
	1149	/// \headerfile <immintrin.h>
	1150	///
	1151	/// This intrinsic corresponds to the \c VCVT2PH2BF8 instruction.
	1152	///
	1153	/// \param __A
	1154	/// A 128-bit vector of [8 x fp16].
	1155	/// \param __B
	1156	/// A 128-bit vector of [8 x fp16].
	1157	/// \returns
	1158	/// A 128-bit vector of [16 x bf8]. Lower 8 elements correspond to the
	1159	/// (converted) elements from \a __B; higher order elements correspond to the
	1160	/// (converted) elements from \a __A.
236	1161	static __inline__ __m128i __DEFAULT_FN_ATTRS128 _mm_cvt2ph_bf8(__m128h __A,
237		__m128h __B) {
238		return (__m128i)__builtin_ia32_vcvt2ph2bf8_128((__v8hf)(__A), (__v8hf)(__B));
239		}
240
	1162	__m128h __B) {
	1163	return (__m128i)__builtin_ia32_vcvt2ph2bf8_128((__v8hf)(__A),
	1164	(__v8hf)(__B));
	1165	}
	1166
	1167	/// Convert two 128-bit vectors, \a __A and \a __B, containing packed FP16
	1168	/// floating-point elements to a 128-bit vector containing E5M2 FP8 elements.
	1169	/// Merging mask \a __U is used to determine if given element should be taken
	1170	/// from \a __W instead.
	1171	///
	1172	/// \code{.operation}
	1173	/// FOR i := 0 to 15
	1174	/// IF __U[i]
	1175	/// IF i < 8
	1176	/// dst.bf8[i] := convert_fp16_to_bf8(__B.fp16[i])
	1177	/// ELSE
	1178	/// dst.bf8[i] := convert_fp16_to_bf8(__A.fp16[i - 8])
	1179	/// FI
	1180	/// ELSE
	1181	/// dst.bf8[i] := __W.bf8[i]
	1182	/// FI
	1183	/// ENDFOR
	1184	///
	1185	/// dst[MAX:128] := 0
	1186	/// \endcode
	1187	///
	1188	/// \headerfile <immintrin.h>
	1189	///
	1190	/// This intrinsic corresponds to the \c VCVT2PH2BF8 instruction.
	1191	///
	1192	/// \param __W
	1193	/// A 128-bit vector of [16 x bf8].
	1194	/// \param __U
	1195	/// A 16-bit merging mask.
	1196	/// \param __A
	1197	/// A 128-bit vector of [8 x fp16].
	1198	/// \param __B
	1199	/// A 128-bit vector of [8 x fp16].
	1200	/// \returns
	1201	/// A 128-bit vector of [16 x bf8]. Lower 8 elements correspond to the
	1202	/// (converted) elements from \a __B; higher order elements correspond to the
	1203	/// (converted) elements from \a __A. If corresponding mask bit is not set, then
	1204	/// element from \a __W is taken instead.
241	1205	static __inline__ __m128i __DEFAULT_FN_ATTRS128
242	1206	_mm_mask_cvt2ph_bf8(__m128i __W, __mmask16 __U, __m128h __A, __m128h __B) {
243	1207	return (__m128i)__builtin_ia32_selectb_128(
244	1208	(__mmask16)__U, (__v16qi)_mm_cvt2ph_bf8(__A, __B), (__v16qi)__W);
245	1209	}
246	1210
	1211	/// Convert two 128-bit vectors, \a __A and \a __B, containing packed FP16
	1212	/// floating-point elements to a 128-bit vector containing E5M2 FP8 elements.
	1213	/// Zeroing mask \a __U is used to determine if given element should be zeroed
	1214	/// instead.
	1215	///
	1216	/// \code{.operation}
	1217	/// FOR i := 0 to 15
	1218	/// IF __U[i]
	1219	/// IF i < 8
	1220	/// dst.bf8[i] := convert_fp16_to_bf8(__B.fp16[i])
	1221	/// ELSE
	1222	/// dst.bf8[i] := convert_fp16_to_bf8(__A.fp16[i - 8])
	1223	/// FI
	1224	/// ELSE
	1225	/// dst.bf8[i] := 0
	1226	/// FI
	1227	/// ENDFOR
	1228	///
	1229	/// dst[MAX:128] := 0
	1230	/// \endcode
	1231	///
	1232	/// \headerfile <immintrin.h>
	1233	///
	1234	/// This intrinsic corresponds to the \c VCVT2PH2BF8 instruction.
	1235	///
	1236	/// \param __U
	1237	/// A 16-bit zeroing mask.
	1238	/// \param __A
	1239	/// A 128-bit vector of [8 x fp16].
	1240	/// \param __B
	1241	/// A 128-bit vector of [8 x fp16].
	1242	/// \returns
	1243	/// A 128-bit vector of [16 x bf8]. Lower 8 elements correspond to the
	1244	/// (converted) elements from \a __B; higher order elements correspond to the
	1245	/// (converted) elements from \a __A. If corresponding mask bit is not set, then
	1246	/// zero is taken instead.
247	1247	static __inline__ __m128i __DEFAULT_FN_ATTRS128
248	1248	_mm_maskz_cvt2ph_bf8(__mmask16 __U, __m128h __A, __m128h __B) {
249	1249	return (__m128i)__builtin_ia32_selectb_128(
...	...	@@ -251,73 +1251,481 @@ _mm_maskz_cvt2ph_bf8(__mmask16 __U, __m128h __A, __m128h __B) {
251	1251	(__v16qi)(__m128i)_mm_setzero_si128());
252	1252	}
253	1253
254		static __inline__ __m256i __DEFAULT_FN_ATTRS256 _mm256_cvt2ph_bf8(__m256h __A,
255		__m256h __B) {
256		return (__m256i)__builtin_ia32_vcvt2ph2bf8_256((__v16hf)(__A),
257		(__v16hf)(__B));
258		}
259
	1254	/// Convert two 256-bit vectors, \a __A and \a __B, containing packed FP16
	1255	/// floating-point elements to a 256-bit vector containing E5M2 FP8 elements.
	1256	///
	1257	/// \code{.operation}
	1258	/// FOR i := 0 to 31
	1259	/// IF i < 16
	1260	/// dst.bf8[i] := convert_fp16_to_bf8(__B.fp16[i])
	1261	/// ELSE
	1262	/// dst.bf8[i] := convert_fp16_to_bf8(__A.fp16[i - 16])
	1263	/// FI
	1264	/// ENDFOR
	1265	///
	1266	/// dst[MAX:256] := 0
	1267	/// \endcode
	1268	///
	1269	/// \headerfile <immintrin.h>
	1270	///
	1271	/// This intrinsic corresponds to the \c VCVT2PH2BF8 instruction.
	1272	///
	1273	/// \param __A
	1274	/// A 256-bit vector of [16 x fp16].
	1275	/// \param __B
	1276	/// A 256-bit vector of [16 x fp16].
	1277	/// \returns
	1278	/// A 256-bit vector of [32 x bf8]. Lower 16 elements correspond to the
	1279	/// (converted) elements from \a __B; higher order elements correspond to the
	1280	/// (converted) elements from \a __A.
260	1281	static __inline__ __m256i __DEFAULT_FN_ATTRS256
261		_mm256_mask_cvt2ph_bf8(__m256i __W, __mmask32 __U, __m256h __A, __m256h __B) {
	1282	_mm256_cvt2ph_bf8(__m256h __A, __m256h __B) {
	1283	return (__m256i)__builtin_ia32_vcvt2ph2bf8_256((__v16hf)(__A),
	1284	(__v16hf)(__B));
	1285	}
	1286
	1287	/// Convert two 256-bit vectors, \a __A and \a __B, containing packed FP16
	1288	/// floating-point elements to a 256-bit vector containing E5M2 FP8 elements.
	1289	/// Merging mask \a __U is used to determine if given element should be taken
	1290	/// from \a __W instead.
	1291	///
	1292	/// \code{.operation}
	1293	/// FOR i := 0 to 31
	1294	/// IF __U[i]
	1295	/// IF i < 16
	1296	/// dst.bf8[i] := convert_fp16_to_bf8(__B.fp16[i])
	1297	/// ELSE
	1298	/// dst.bf8[i] := convert_fp16_to_bf8(__A.fp16[i - 16])
	1299	/// FI
	1300	/// ELSE
	1301	/// dst.bf8[i] := __W.bf8[i]
	1302	/// FI
	1303	/// ENDFOR
	1304	///
	1305	/// dst[MAX:256] := 0
	1306	/// \endcode
	1307	///
	1308	/// \headerfile <immintrin.h>
	1309	///
	1310	/// This intrinsic corresponds to the \c VCVT2PH2BF8 instruction.
	1311	///
	1312	/// \param __W
	1313	/// A 256-bit vector of [32 x bf8].
	1314	/// \param __U
	1315	/// A 32-bit merging mask.
	1316	/// \param __A
	1317	/// A 256-bit vector of [16 x fp16].
	1318	/// \param __B
	1319	/// A 256-bit vector of [16 x fp16].
	1320	/// \returns
	1321	/// A 256-bit vector of [32 x bf8]. Lower 16 elements correspond to the
	1322	/// (converted) elements from \a __B; higher order elements correspond to the
	1323	/// (converted) elements from \a __A. If corresponding mask bit is not set, then
	1324	/// element from \a __W is taken instead.
	1325	static __inline__ __m256i __DEFAULT_FN_ATTRS256 _mm256_mask_cvt2ph_bf8(
	1326	__m256i __W, __mmask32 __U, __m256h __A, __m256h __B) {
262	1327	return (__m256i)__builtin_ia32_selectb_256(
263		(__mmask16)__U, (__v32qi)_mm256_cvt2ph_bf8(__A, __B), (__v32qi)__W);
264		}
265
	1328	(__mmask32)__U, (__v32qi)_mm256_cvt2ph_bf8(__A, __B), (__v32qi)__W);
	1329	}
	1330
	1331	/// Convert two 256-bit vectors, \a __A and \a __B, containing packed FP16
	1332	/// floating-point elements to a 256-bit vector containing E5M2 FP8 elements.
	1333	/// Zeroing mask \a __U is used to determine if given element should be zeroed
	1334	/// instead.
	1335	///
	1336	/// \code{.operation}
	1337	/// FOR i := 0 to 31
	1338	/// IF __U[i]
	1339	/// IF i < 16
	1340	/// dst.bf8[i] := convert_fp16_to_bf8(__B.fp16[i])
	1341	/// ELSE
	1342	/// dst.bf8[i] := convert_fp16_to_bf8(__A.fp16[i - 16])
	1343	/// FI
	1344	/// ELSE
	1345	/// dst.bf8[i] := 0
	1346	/// FI
	1347	/// ENDFOR
	1348	///
	1349	/// dst[MAX:256] := 0
	1350	/// \endcode
	1351	///
	1352	/// \headerfile <immintrin.h>
	1353	///
	1354	/// This intrinsic corresponds to the \c VCVT2PH2BF8 instruction.
	1355	///
	1356	/// \param __U
	1357	/// A 32-bit zeroing mask.
	1358	/// \param __A
	1359	/// A 256-bit vector of [16 x fp16].
	1360	/// \param __B
	1361	/// A 256-bit vector of [16 x fp16].
	1362	/// \returns
	1363	/// A 256-bit vector of [32 x bf8]. Lower 16 elements correspond to the
	1364	/// (converted) elements from \a __B; higher order elements correspond to the
	1365	/// (converted) elements from \a __A. If corresponding mask bit is not set,
	1366	/// zero is taken instead.
266	1367	static __inline__ __m256i __DEFAULT_FN_ATTRS256
267	1368	_mm256_maskz_cvt2ph_bf8(__mmask32 __U, __m256h __A, __m256h __B) {
268	1369	return (__m256i)__builtin_ia32_selectb_256(
269		(__mmask16)__U, (__v32qi)_mm256_cvt2ph_bf8(__A, __B),
	1370	(__mmask32)__U, (__v32qi)_mm256_cvt2ph_bf8(__A, __B),
270	1371	(__v32qi)(__m256i)_mm256_setzero_si256());
271	1372	}
272	1373
273		static __inline__ __m128i __DEFAULT_FN_ATTRS128 _mm_cvts2ph_bf8(__m128h __A,
274		__m128h __B) {
275		return (__m128i)__builtin_ia32_vcvt2ph2bf8s_128((__v8hf)(__A), (__v8hf)(__B));
276		}
277
	1374	/// Convert two 128-bit vectors, \a __A and \a __B, containing packed FP16
	1375	/// floating-point elements to a 128-bit vector containing E5M2 FP8 elements.
	1376	/// Resulting elements are saturated in case of overflow.
	1377	///
	1378	/// \code{.operation}
	1379	/// FOR i := 0 to 15
	1380	/// IF i < 8
	1381	/// dst.bf8[i] := convert_fp16_to_bf8_saturate(__B.fp16[i])
	1382	/// ELSE
	1383	/// dst.bf8[i] := convert_fp16_to_bf8_saturate(__A.fp16[i - 8])
	1384	/// FI
	1385	/// ENDFOR
	1386	///
	1387	/// dst[MAX:128] := 0
	1388	/// \endcode
	1389	///
	1390	/// \headerfile <immintrin.h>
	1391	///
	1392	/// This intrinsic corresponds to the \c VCVT2PH2BF8S instruction.
	1393	///
	1394	/// \param __A
	1395	/// A 128-bit vector of [8 x fp16].
	1396	/// \param __B
	1397	/// A 128-bit vector of [8 x fp16].
	1398	/// \returns
	1399	/// A 128-bit vector of [16 x bf8]. Lower 8 elements correspond to the
	1400	/// (converted) elements from \a __B; higher order elements correspond to the
	1401	/// (converted) elements from \a __A.
278	1402	static __inline__ __m128i __DEFAULT_FN_ATTRS128
279		_mm_mask_cvts2ph_bf8(__m128i __W, __mmask16 __U, __m128h __A, __m128h __B) {
	1403	_mm_cvts_2ph_bf8(__m128h __A, __m128h __B) {
	1404	return (__m128i)__builtin_ia32_vcvt2ph2bf8s_128((__v8hf)(__A),
	1405	(__v8hf)(__B));
	1406	}
	1407
	1408	/// Convert two 128-bit vectors, \a __A and \a __B, containing packed FP16
	1409	/// floating-point elements to a 128-bit vector containing E5M2 FP8 elements.
	1410	/// Merging mask \a __U is used to determine if given element should be taken
	1411	/// from \a __W instead. Resulting elements are saturated in case of overflow.
	1412	///
	1413	/// \code{.operation}
	1414	/// FOR i := 0 to 15
	1415	/// IF __U[i]
	1416	/// IF i < 8
	1417	/// dst.bf8[i] := convert_fp16_to_bf8_saturate(__B.fp16[i])
	1418	/// ELSE
	1419	/// dst.bf8[i] := convert_fp16_to_bf8_saturate(__A.fp16[i - 8])
	1420	/// FI
	1421	/// ELSE
	1422	/// dst.bf8[i] := __W.bf8[i]
	1423	/// FI
	1424	/// ENDFOR
	1425	///
	1426	/// dst[MAX:128] := 0
	1427	/// \endcode
	1428	///
	1429	/// \headerfile <immintrin.h>
	1430	///
	1431	/// This intrinsic corresponds to the \c VCVT2PH2BF8S instruction.
	1432	///
	1433	/// \param __W
	1434	/// A 128-bit vector of [16 x bf8].
	1435	/// \param __U
	1436	/// A 16-bit merging mask.
	1437	/// \param __A
	1438	/// A 128-bit vector of [8 x fp16].
	1439	/// \param __B
	1440	/// A 128-bit vector of [8 x fp16].
	1441	/// \returns
	1442	/// A 128-bit vector of [16 x bf8]. Lower 8 elements correspond to the
	1443	/// (converted) elements from \a __B; higher order elements correspond to the
	1444	/// (converted) elements from \a __A. If corresponding mask bit is not set, then
	1445	/// element from \a __W is taken instead.
	1446	static __inline__ __m128i __DEFAULT_FN_ATTRS128
	1447	_mm_mask_cvts_2ph_bf8(__m128i __W, __mmask16 __U, __m128h __A, __m128h __B) {
280	1448	return (__m128i)__builtin_ia32_selectb_128(
281		(__mmask16)__U, (__v16qi)_mm_cvts2ph_bf8(__A, __B), (__v16qi)__W);
282		}
283
	1449	(__mmask16)__U, (__v16qi)_mm_cvts_2ph_bf8(__A, __B), (__v16qi)__W);
	1450	}
	1451
	1452	/// Convert two 128-bit vectors, \a __A and \a __B, containing packed FP16
	1453	/// floating-point elements to a 128-bit vector containing E5M2 FP8 elements.
	1454	/// Zeroing mask \a __U is used to determine if given element should be zeroed
	1455	/// instead. Resulting elements are saturated in case of overflow.
	1456	///
	1457	/// \code{.operation}
	1458	/// FOR i := 0 to 15
	1459	/// IF __U[i]
	1460	/// IF i < 8
	1461	/// dst.bf8[i] := convert_fp16_to_bf8_saturate(__B.fp16[i])
	1462	/// ELSE
	1463	/// dst.bf8[i] := convert_fp16_to_bf8_saturate(__A.fp16[i - 8])
	1464	/// FI
	1465	/// ELSE
	1466	/// dst.bf8[i] := 0
	1467	/// FI
	1468	/// ENDFOR
	1469	///
	1470	/// dst[MAX:128] := 0
	1471	/// \endcode
	1472	///
	1473	/// \headerfile <immintrin.h>
	1474	///
	1475	/// This intrinsic corresponds to the \c VCVT2PH2BF8S instruction.
	1476	///
	1477	/// \param __U
	1478	/// A 16-bit zeroing mask.
	1479	/// \param __A
	1480	/// A 128-bit vector of [8 x fp16].
	1481	/// \param __B
	1482	/// A 128-bit vector of [8 x fp16].
	1483	/// \returns
	1484	/// A 128-bit vector of [16 x bf8]. Lower 8 elements correspond to the
	1485	/// (converted) elements from \a __B; higher order elements correspond to the
	1486	/// (converted) elements from \a __A. If corresponding mask bit is not set, then
	1487	/// zero is taken instead.
284	1488	static __inline__ __m128i __DEFAULT_FN_ATTRS128
285		_mm_maskz_cvts2ph_bf8(__mmask16 __U, __m128h __A, __m128h __B) {
	1489	_mm_maskz_cvts_2ph_bf8(__mmask16 __U, __m128h __A, __m128h __B) {
286	1490	return (__m128i)__builtin_ia32_selectb_128(
287		(__mmask16)__U, (__v16qi)_mm_cvts2ph_bf8(__A, __B),
	1491	(__mmask16)__U, (__v16qi)_mm_cvts_2ph_bf8(__A, __B),
288	1492	(__v16qi)(__m128i)_mm_setzero_si128());
289	1493	}
290	1494
	1495	/// Convert two 256-bit vectors, \a __A and \a __B, containing packed FP16
	1496	/// floating-point elements to a 256-bit vector containing E5M2 FP8 elements.
	1497	/// Resulting elements are saturated in case of overflow.
	1498	///
	1499	/// \code{.operation}
	1500	/// FOR i := 0 to 31
	1501	/// IF i < 16
	1502	/// dst.bf8[i] := convert_fp16_to_bf8_saturate(__B.fp16[i])
	1503	/// ELSE
	1504	/// dst.bf8[i] := convert_fp16_to_bf8_saturate(__A.fp16[i - 16])
	1505	/// FI
	1506	/// ENDFOR
	1507	///
	1508	/// dst[MAX:256] := 0
	1509	/// \endcode
	1510	///
	1511	/// \headerfile <immintrin.h>
	1512	///
	1513	/// This intrinsic corresponds to the \c VCVT2PH2BF8S instruction.
	1514	///
	1515	/// \param __A
	1516	/// A 256-bit vector of [16 x fp16].
	1517	/// \param __B
	1518	/// A 256-bit vector of [16 x fp16].
	1519	/// \returns
	1520	/// A 256-bit vector of [32 x bf8]. Lower 16 elements correspond to the
	1521	/// (converted) elements from \a __B; higher order elements correspond to the
	1522	/// (converted) elements from \a __A.
291	1523	static __inline__ __m256i __DEFAULT_FN_ATTRS256
292		_mm256_cvts2ph_bf8(__m256h __A, __m256h __B) {
	1524	_mm256_cvts_2ph_bf8(__m256h __A, __m256h __B) {
293	1525	return (__m256i)__builtin_ia32_vcvt2ph2bf8s_256((__v16hf)(__A),
294		(__v16hf)(__B));
295		}
296
297		static __inline__ __m256i __DEFAULT_FN_ATTRS256
298		_mm256_mask_cvts2ph_bf8(__m256i __W, __mmask32 __U, __m256h __A, __m256h __B) {
	1526	(__v16hf)(__B));
	1527	}
	1528
	1529	/// Convert two 256-bit vectors, \a __A and \a __B, containing packed FP16
	1530	/// floating-point elements to a 256-bit vector containing E5M2 FP8 elements.
	1531	/// Merging mask \a __U is used to determine if given element should be taken
	1532	/// from \a __W instead. Resulting elements are saturated in case of overflow.
	1533	///
	1534	/// \code{.operation}
	1535	/// FOR i := 0 to 31
	1536	/// IF __U[i]
	1537	/// IF i < 16
	1538	/// dst.bf8[i] := convert_fp16_to_bf8_saturate(__B.fp16[i])
	1539	/// ELSE
	1540	/// dst.bf8[i] := convert_fp16_to_bf8_saturate(__A.fp16[i - 16])
	1541	/// FI
	1542	/// ELSE
	1543	/// dst.bf8[i] := __W.bf8[i]
	1544	/// FI
	1545	/// ENDFOR
	1546	///
	1547	/// dst[MAX:256] := 0
	1548	/// \endcode
	1549	///
	1550	/// \headerfile <immintrin.h>
	1551	///
	1552	/// This intrinsic corresponds to the \c VCVT2PH2BF8S instruction.
	1553	///
	1554	/// \param __W
	1555	/// A 256-bit vector of [32 x bf8].
	1556	/// \param __U
	1557	/// A 32-bit merging mask.
	1558	/// \param __A
	1559	/// A 256-bit vector of [16 x fp16].
	1560	/// \param __B
	1561	/// A 256-bit vector of [16 x fp16].
	1562	/// \returns
	1563	/// A 256-bit vector of [32 x bf8]. Lower 16 elements correspond to the
	1564	/// (converted) elements from \a __B; higher order elements correspond to the
	1565	/// (converted) elements from \a __A. If corresponding mask bit is not set, then
	1566	/// element from \a __W is taken instead.
	1567	static __inline__ __m256i __DEFAULT_FN_ATTRS256 _mm256_mask_cvts_2ph_bf8(
	1568	__m256i __W, __mmask32 __U, __m256h __A, __m256h __B) {
299	1569	return (__m256i)__builtin_ia32_selectb_256(
300		(__mmask16)__U, (__v32qi)_mm256_cvts2ph_bf8(__A, __B), (__v32qi)__W);
301		}
302
	1570	(__mmask32)__U, (__v32qi)_mm256_cvts_2ph_bf8(__A, __B), (__v32qi)__W);
	1571	}
	1572
	1573	/// Convert two 256-bit vectors, \a __A and \a __B, containing packed FP16
	1574	/// floating-point elements to a 256-bit vector containing E5M2 FP8 elements.
	1575	/// Zeroing mask \a __U is used to determine if given element should be zeroed
	1576	/// instead. Resulting elements are saturated in case of overflow.
	1577	///
	1578	/// \code{.operation}
	1579	/// FOR i := 0 to 31
	1580	/// IF __U[i]
	1581	/// IF i < 16
	1582	/// dst.bf8[i] := convert_fp16_to_bf8_saturate(__B.fp16[i])
	1583	/// ELSE
	1584	/// dst.bf8[i] := convert_fp16_to_bf8_saturate(__A.fp16[i - 16])
	1585	/// FI
	1586	/// ELSE
	1587	/// dst.bf8[i] := 0
	1588	/// FI
	1589	/// ENDFOR
	1590	///
	1591	/// dst[MAX:256] := 0
	1592	/// \endcode
	1593	///
	1594	/// \headerfile <immintrin.h>
	1595	///
	1596	/// This intrinsic corresponds to the \c VCVT2PH2BF8S instruction.
	1597	///
	1598	/// \param __U
	1599	/// A 32-bit zeroing mask.
	1600	/// \param __A
	1601	/// A 256-bit vector of [16 x fp16].
	1602	/// \param __B
	1603	/// A 256-bit vector of [16 x fp16].
	1604	/// \returns
	1605	/// A 256-bit vector of [32 x bf8]. Lower 16 elements correspond to the
	1606	/// (converted) elements from \a __B; higher order elements correspond to the
	1607	/// (converted) elements from \a __A. If corresponding mask bit is not set,
	1608	/// zero is taken instead.
303	1609	static __inline__ __m256i __DEFAULT_FN_ATTRS256
304		_mm256_maskz_cvts2ph_bf8(__mmask32 __U, __m256h __A, __m256h __B) {
	1610	_mm256_maskz_cvts_2ph_bf8(__mmask32 __U, __m256h __A, __m256h __B) {
305	1611	return (__m256i)__builtin_ia32_selectb_256(
306		(__mmask16)__U, (__v32qi)_mm256_cvts2ph_bf8(__A, __B),
	1612	(__mmask32)__U, (__v32qi)_mm256_cvts_2ph_bf8(__A, __B),
307	1613	(__v32qi)(__m256i)_mm256_setzero_si256());
308	1614	}
309	1615
	1616	/// Convert two 128-bit vectors, \a __A and \a __B, containing packed FP16
	1617	/// floating-point elements to a 128-bit vector containing E4M3 FP8 elements.
	1618	///
	1619	/// \code{.operation}
	1620	/// FOR i := 0 to 15
	1621	/// IF i < 8
	1622	/// dst.hf8[i] := convert_fp16_to_hf8(__B.fp16[i])
	1623	/// ELSE
	1624	/// dst.hf8[i] := convert_fp16_to_hf8(__A.fp16[i - 8])
	1625	/// FI
	1626	/// ENDFOR
	1627	///
	1628	/// dst[MAX:128] := 0
	1629	/// \endcode
	1630	///
	1631	/// \headerfile <immintrin.h>
	1632	///
	1633	/// This intrinsic corresponds to the \c VCVT2PH2HF8 instruction.
	1634	///
	1635	/// \param __A
	1636	/// A 128-bit vector of [8 x fp16].
	1637	/// \param __B
	1638	/// A 128-bit vector of [8 x fp16].
	1639	/// \returns
	1640	/// A 128-bit vector of [16 x hf8]. Lower 8 elements correspond to the
	1641	/// (converted) elements from \a __B; higher order elements correspond to the
	1642	/// (converted) elements from \a __A.
310	1643	static __inline__ __m128i __DEFAULT_FN_ATTRS128 _mm_cvt2ph_hf8(__m128h __A,
311		__m128h __B) {
312		return (__m128i)__builtin_ia32_vcvt2ph2hf8_128((__v8hf)(__A), (__v8hf)(__B));
313		}
314
	1644	__m128h __B) {
	1645	return (__m128i)__builtin_ia32_vcvt2ph2hf8_128((__v8hf)(__A),
	1646	(__v8hf)(__B));
	1647	}
	1648
	1649	/// Convert two 128-bit vectors, \a __A and \a __B, containing packed FP16
	1650	/// floating-point elements to a 128-bit vector containing E4M3 FP8 elements.
	1651	/// Merging mask \a __U is used to determine if given element should be taken
	1652	/// from \a __W instead.
	1653	///
	1654	/// \code{.operation}
	1655	/// FOR i := 0 to 15
	1656	/// IF __U[i]
	1657	/// IF i < 8
	1658	/// dst.hf8[i] := convert_fp16_to_hf8(__B.fp16[i])
	1659	/// ELSE
	1660	/// dst.hf8[i] := convert_fp16_to_hf8(__A.fp16[i - 8])
	1661	/// FI
	1662	/// ELSE
	1663	/// dst.hf8[i] := __W.hf8[i]
	1664	/// FI
	1665	/// ENDFOR
	1666	///
	1667	/// dst[MAX:128] := 0
	1668	/// \endcode
	1669	///
	1670	/// \headerfile <immintrin.h>
	1671	///
	1672	/// This intrinsic corresponds to the \c VCVT2PH2HF8 instruction.
	1673	///
	1674	/// \param __W
	1675	/// A 128-bit vector of [16 x hf8].
	1676	/// \param __U
	1677	/// A 16-bit merging mask.
	1678	/// \param __A
	1679	/// A 128-bit vector of [8 x fp16].
	1680	/// \param __B
	1681	/// A 128-bit vector of [8 x fp16].
	1682	/// \returns
	1683	/// A 128-bit vector of [16 x hf8]. Lower 8 elements correspond to the
	1684	/// (converted) elements from \a __B; higher order elements correspond to the
	1685	/// (converted) elements from \a __A. If corresponding mask bit is not set, then
	1686	/// element from \a __W is taken instead.
315	1687	static __inline__ __m128i __DEFAULT_FN_ATTRS128
316	1688	_mm_mask_cvt2ph_hf8(__m128i __W, __mmask16 __U, __m128h __A, __m128h __B) {
317	1689	return (__m128i)__builtin_ia32_selectb_128(
318	1690	(__mmask16)__U, (__v16qi)_mm_cvt2ph_hf8(__A, __B), (__v16qi)__W);
319	1691	}
320	1692
	1693	/// Convert two 128-bit vectors, \a __A and \a __B, containing packed FP16
	1694	/// floating-point elements to a 128-bit vector containing E4M3 FP8 elements.
	1695	/// Zeroing mask \a __U is used to determine if given element should be zeroed
	1696	/// instead.
	1697	///
	1698	/// \code{.operation}
	1699	/// FOR i := 0 to 15
	1700	/// IF __U[i]
	1701	/// IF i < 8
	1702	/// dst.hf8[i] := convert_fp16_to_hf8(__B.fp16[i])
	1703	/// ELSE
	1704	/// dst.hf8[i] := convert_fp16_to_hf8(__A.fp16[i - 8])
	1705	/// FI
	1706	/// ELSE
	1707	/// dst.hf8[i] := 0
	1708	/// FI
	1709	/// ENDFOR
	1710	///
	1711	/// dst[MAX:128] := 0
	1712	/// \endcode
	1713	///
	1714	/// \headerfile <immintrin.h>
	1715	///
	1716	/// This intrinsic corresponds to the \c VCVT2PH2HF8 instruction.
	1717	///
	1718	/// \param __U
	1719	/// A 16-bit zeroing mask.
	1720	/// \param __A
	1721	/// A 128-bit vector of [8 x fp16].
	1722	/// \param __B
	1723	/// A 128-bit vector of [8 x fp16].
	1724	/// \returns
	1725	/// A 128-bit vector of [16 x hf8]. Lower 8 elements correspond to the
	1726	/// (converted) elements from \a __B; higher order elements correspond to the
	1727	/// (converted) elements from \a __A. If corresponding mask bit is not set, then
	1728	/// zero is taken instead.
321	1729	static __inline__ __m128i __DEFAULT_FN_ATTRS128
322	1730	_mm_maskz_cvt2ph_hf8(__mmask16 __U, __m128h __A, __m128h __B) {
323	1731	return (__m128i)__builtin_ia32_selectb_128(
...	...	@@ -325,264 +1733,1523 @@ _mm_maskz_cvt2ph_hf8(__mmask16 __U, __m128h __A, __m128h __B) {
325	1733	(__v16qi)(__m128i)_mm_setzero_si128());
326	1734	}
327	1735
328		static __inline__ __m256i __DEFAULT_FN_ATTRS256 _mm256_cvt2ph_hf8(__m256h __A,
329		__m256h __B) {
330		return (__m256i)__builtin_ia32_vcvt2ph2hf8_256((__v16hf)(__A),
331		(__v16hf)(__B));
332		}
333
	1736	/// Convert two 256-bit vectors, \a __A and \a __B, containing packed FP16
	1737	/// floating-point elements to a 256-bit vector containing E4M3 FP8 elements.
	1738	///
	1739	/// \code{.operation}
	1740	/// FOR i := 0 to 31
	1741	/// IF i < 16
	1742	/// dst.hf8[i] := convert_fp16_to_hf8(__B.fp16[i])
	1743	/// ELSE
	1744	/// dst.hf8[i] := convert_fp16_to_hf8(__A.fp16[i - 16])
	1745	/// FI
	1746	/// ENDFOR
	1747	///
	1748	/// dst[MAX:256] := 0
	1749	/// \endcode
	1750	///
	1751	/// \headerfile <immintrin.h>
	1752	///
	1753	/// This intrinsic corresponds to the \c VCVT2PH2HF8 instruction.
	1754	///
	1755	/// \param __A
	1756	/// A 256-bit vector of [16 x fp16].
	1757	/// \param __B
	1758	/// A 256-bit vector of [16 x fp16].
	1759	/// \returns
	1760	/// A 256-bit vector of [32 x hf8]. Lower 16 elements correspond to the
	1761	/// (converted) elements from \a __B; higher order elements correspond to the
	1762	/// (converted) elements from \a __A.
334	1763	static __inline__ __m256i __DEFAULT_FN_ATTRS256
335		_mm256_mask_cvt2ph_hf8(__m256i __W, __mmask32 __U, __m256h __A, __m256h __B) {
	1764	_mm256_cvt2ph_hf8(__m256h __A, __m256h __B) {
	1765	return (__m256i)__builtin_ia32_vcvt2ph2hf8_256((__v16hf)(__A),
	1766	(__v16hf)(__B));
	1767	}
	1768
	1769	/// Convert two 256-bit vectors, \a __A and \a __B, containing packed FP16
	1770	/// floating-point elements to a 256-bit vector containing E4M3 FP8 elements.
	1771	/// Merging mask \a __U is used to determine if given element should be taken
	1772	/// from \a __W instead.
	1773	///
	1774	/// \code{.operation}
	1775	/// FOR i := 0 to 31
	1776	/// IF __U[i]
	1777	/// IF i < 16
	1778	/// dst.hf8[i] := convert_fp16_to_hf8(__B.fp16[i])
	1779	/// ELSE
	1780	/// dst.hf8[i] := convert_fp16_to_hf8(__A.fp16[i - 16])
	1781	/// FI
	1782	/// ELSE
	1783	/// dst.hf8[i] := __W.hf8[i]
	1784	/// FI
	1785	/// ENDFOR
	1786	///
	1787	/// dst[MAX:256] := 0
	1788	/// \endcode
	1789	///
	1790	/// \headerfile <immintrin.h>
	1791	///
	1792	/// This intrinsic corresponds to the \c VCVT2PH2HF8 instruction.
	1793	///
	1794	/// \param __W
	1795	/// A 256-bit vector of [32 x hf8].
	1796	/// \param __U
	1797	/// A 32-bit merging mask.
	1798	/// \param __A
	1799	/// A 256-bit vector of [16 x fp16].
	1800	/// \param __B
	1801	/// A 256-bit vector of [16 x fp16].
	1802	/// \returns
	1803	/// A 256-bit vector of [32 x hf8]. Lower 16 elements correspond to the
	1804	/// (converted) elements from \a __B; higher order elements correspond to the
	1805	/// (converted) elements from \a __A. If corresponding mask bit is not set, then
	1806	/// element from \a __W is taken instead.
	1807	static __inline__ __m256i __DEFAULT_FN_ATTRS256 _mm256_mask_cvt2ph_hf8(
	1808	__m256i __W, __mmask32 __U, __m256h __A, __m256h __B) {
336	1809	return (__m256i)__builtin_ia32_selectb_256(
337		(__mmask16)__U, (__v32qi)_mm256_cvt2ph_hf8(__A, __B), (__v32qi)__W);
338		}
339
	1810	(__mmask32)__U, (__v32qi)_mm256_cvt2ph_hf8(__A, __B), (__v32qi)__W);
	1811	}
	1812
	1813	/// Convert two 256-bit vectors, \a __A and \a __B, containing packed FP16
	1814	/// floating-point elements to a 256-bit vector containing E4M3 FP8 elements.
	1815	/// Zeroing mask \a __U is used to determine if given element should be zeroed
	1816	/// instead.
	1817	///
	1818	/// \code{.operation}
	1819	/// FOR i := 0 to 31
	1820	/// IF __U[i]
	1821	/// IF i < 16
	1822	/// dst.hf8[i] := convert_fp16_to_hf8(__B.fp16[i])
	1823	/// ELSE
	1824	/// dst.hf8[i] := convert_fp16_to_hf8(__A.fp16[i - 16])
	1825	/// FI
	1826	/// ELSE
	1827	/// dst.hf8[i] := 0
	1828	/// FI
	1829	/// ENDFOR
	1830	///
	1831	/// dst[MAX:256] := 0
	1832	/// \endcode
	1833	///
	1834	/// \headerfile <immintrin.h>
	1835	///
	1836	/// This intrinsic corresponds to the \c VCVT2PH2HF8 instruction.
	1837	///
	1838	/// \param __U
	1839	/// A 32-bit zeroing mask.
	1840	/// \param __A
	1841	/// A 256-bit vector of [16 x fp16].
	1842	/// \param __B
	1843	/// A 256-bit vector of [16 x fp16].
	1844	/// \returns
	1845	/// A 256-bit vector of [32 x hf8]. Lower 16 elements correspond to the
	1846	/// (converted) elements from \a __B; higher order elements correspond to the
	1847	/// (converted) elements from \a __A. If corresponding mask bit is not set,
	1848	/// zero is taken instead.
340	1849	static __inline__ __m256i __DEFAULT_FN_ATTRS256
341	1850	_mm256_maskz_cvt2ph_hf8(__mmask32 __U, __m256h __A, __m256h __B) {
342	1851	return (__m256i)__builtin_ia32_selectb_256(
343		(__mmask16)__U, (__v32qi)_mm256_cvt2ph_hf8(__A, __B),
	1852	(__mmask32)__U, (__v32qi)_mm256_cvt2ph_hf8(__A, __B),
344	1853	(__v32qi)(__m256i)_mm256_setzero_si256());
345	1854	}
346	1855
347		static __inline__ __m128i __DEFAULT_FN_ATTRS128 _mm_cvts2ph_hf8(__m128h __A,
348		__m128h __B) {
349		return (__m128i)__builtin_ia32_vcvt2ph2hf8s_128((__v8hf)(__A), (__v8hf)(__B));
350		}
351
	1856	/// Convert two 128-bit vectors, \a __A and \a __B, containing packed FP16
	1857	/// floating-point elements to a 128-bit vector containing E4M3 FP8 elements.
	1858	/// Resulting elements are saturated in case of overflow.
	1859	///
	1860	/// \code{.operation}
	1861	/// FOR i := 0 to 15
	1862	/// IF i < 8
	1863	/// dst.hf8[i] := convert_fp16_to_hf8_saturate(__B.fp16[i])
	1864	/// ELSE
	1865	/// dst.hf8[i] := convert_fp16_to_hf8_saturate(__A.fp16[i - 8])
	1866	/// FI
	1867	/// ENDFOR
	1868	///
	1869	/// dst[MAX:128] := 0
	1870	/// \endcode
	1871	///
	1872	/// \headerfile <immintrin.h>
	1873	///
	1874	/// This intrinsic corresponds to the \c VCVT2PH2HF8S instruction.
	1875	///
	1876	/// \param __A
	1877	/// A 128-bit vector of [8 x fp16].
	1878	/// \param __B
	1879	/// A 128-bit vector of [8 x fp16].
	1880	/// \returns
	1881	/// A 128-bit vector of [16 x hf8]. Lower 8 elements correspond to the
	1882	/// (converted) elements from \a __B; higher order elements correspond to the
	1883	/// (converted) elements from \a __A.
352	1884	static __inline__ __m128i __DEFAULT_FN_ATTRS128
353		_mm_mask_cvts2ph_hf8(__m128i __W, __mmask16 __U, __m128h __A, __m128h __B) {
	1885	_mm_cvts_2ph_hf8(__m128h __A, __m128h __B) {
	1886	return (__m128i)__builtin_ia32_vcvt2ph2hf8s_128((__v8hf)(__A),
	1887	(__v8hf)(__B));
	1888	}
	1889
	1890	/// Convert two 128-bit vectors, \a __A and \a __B, containing packed FP16
	1891	/// floating-point elements to a 128-bit vector containing E4M3 FP8 elements.
	1892	/// Merging mask \a __U is used to determine if given element should be taken
	1893	/// from \a __W instead. Resulting elements are saturated in case of overflow.
	1894	///
	1895	/// \code{.operation}
	1896	/// FOR i := 0 to 15
	1897	/// IF __U[i]
	1898	/// IF i < 8
	1899	/// dst.hf8[i] := convert_fp16_to_hf8_saturate(__B.fp16[i])
	1900	/// ELSE
	1901	/// dst.hf8[i] := convert_fp16_to_hf8_saturate(__A.fp16[i - 8])
	1902	/// FI
	1903	/// ELSE
	1904	/// dst.hf8[i] := __W.hf8[i]
	1905	/// FI
	1906	/// ENDFOR
	1907	///
	1908	/// dst[MAX:128] := 0
	1909	/// \endcode
	1910	///
	1911	/// \headerfile <immintrin.h>
	1912	///
	1913	/// This intrinsic corresponds to the \c VCVT2PH2HF8S instruction.
	1914	///
	1915	/// \param __W
	1916	/// A 128-bit vector of [16 x hf8].
	1917	/// \param __U
	1918	/// A 16-bit merging mask.
	1919	/// \param __A
	1920	/// A 128-bit vector of [8 x fp16].
	1921	/// \param __B
	1922	/// A 128-bit vector of [8 x fp16].
	1923	/// \returns
	1924	/// A 128-bit vector of [16 x hf8]. Lower 8 elements correspond to the
	1925	/// (converted) elements from \a __B; higher order elements correspond to the
	1926	/// (converted) elements from \a __A. If corresponding mask bit is not set, then
	1927	/// element from \a __W is taken instead.
	1928	static __inline__ __m128i __DEFAULT_FN_ATTRS128
	1929	_mm_mask_cvts_2ph_hf8(__m128i __W, __mmask16 __U, __m128h __A, __m128h __B) {
354	1930	return (__m128i)__builtin_ia32_selectb_128(
355		(__mmask16)__U, (__v16qi)_mm_cvts2ph_hf8(__A, __B), (__v16qi)__W);
356		}
357
	1931	(__mmask16)__U, (__v16qi)_mm_cvts_2ph_hf8(__A, __B), (__v16qi)__W);
	1932	}
	1933
	1934	/// Convert two 128-bit vectors, \a __A and \a __B, containing packed FP16
	1935	/// floating-point elements to a 128-bit vector containing E4M3 FP8 elements.
	1936	/// Zeroing mask \a __U is used to determine if given element should be zeroed
	1937	/// instead. Resulting elements are saturated in case of overflow.
	1938	///
	1939	/// \code{.operation}
	1940	/// FOR i := 0 to 15
	1941	/// IF __U[i]
	1942	/// IF i < 8
	1943	/// dst.hf8[i] := convert_fp16_to_hf8_saturate(__B.fp16[i])
	1944	/// ELSE
	1945	/// dst.hf8[i] := convert_fp16_to_hf8_saturate(__A.fp16[i - 8])
	1946	/// FI
	1947	/// ELSE
	1948	/// dst.hf8[i] := 0
	1949	/// FI
	1950	/// ENDFOR
	1951	///
	1952	/// dst[MAX:128] := 0
	1953	/// \endcode
	1954	///
	1955	/// \headerfile <immintrin.h>
	1956	///
	1957	/// This intrinsic corresponds to the \c VCVT2PH2HF8S instruction.
	1958	///
	1959	/// \param __U
	1960	/// A 16-bit zeroing mask.
	1961	/// \param __A
	1962	/// A 128-bit vector of [8 x fp16].
	1963	/// \param __B
	1964	/// A 128-bit vector of [8 x fp16].
	1965	/// \returns
	1966	/// A 128-bit vector of [16 x hf8]. Lower 8 elements correspond to the
	1967	/// (converted) elements from \a __B; higher order elements correspond to the
	1968	/// (converted) elements from \a __A. If corresponding mask bit is not set, then
	1969	/// zero is taken instead.
358	1970	static __inline__ __m128i __DEFAULT_FN_ATTRS128
359		_mm_maskz_cvts2ph_hf8(__mmask16 __U, __m128h __A, __m128h __B) {
	1971	_mm_maskz_cvts_2ph_hf8(__mmask16 __U, __m128h __A, __m128h __B) {
360	1972	return (__m128i)__builtin_ia32_selectb_128(
361		(__mmask16)__U, (__v16qi)_mm_cvts2ph_hf8(__A, __B),
	1973	(__mmask16)__U, (__v16qi)_mm_cvts_2ph_hf8(__A, __B),
362	1974	(__v16qi)(__m128i)_mm_setzero_si128());
363	1975	}
364	1976
	1977	/// Convert two 256-bit vectors, \a __A and \a __B, containing packed FP16
	1978	/// floating-point elements to a 256-bit vector containing E4M3 FP8 elements.
	1979	/// Resulting elements are saturated in case of overflow.
	1980	///
	1981	/// \code{.operation}
	1982	/// FOR i := 0 to 31
	1983	/// IF i < 16
	1984	/// dst.hf8[i] := convert_fp16_to_hf8_saturate(__B.fp16[i])
	1985	/// ELSE
	1986	/// dst.hf8[i] := convert_fp16_to_hf8_saturate(__A.fp16[i - 16])
	1987	/// FI
	1988	/// ENDFOR
	1989	///
	1990	/// dst[MAX:256] := 0
	1991	/// \endcode
	1992	///
	1993	/// \headerfile <immintrin.h>
	1994	///
	1995	/// This intrinsic corresponds to the \c VCVT2PH2HF8S instruction.
	1996	///
	1997	/// \param __A
	1998	/// A 256-bit vector of [16 x fp16].
	1999	/// \param __B
	2000	/// A 256-bit vector of [16 x fp16].
	2001	/// \returns
	2002	/// A 256-bit vector of [32 x hf8]. Lower 16 elements correspond to the
	2003	/// (converted) elements from \a __B; higher order elements correspond to the
	2004	/// (converted) elements from \a __A.
365	2005	static __inline__ __m256i __DEFAULT_FN_ATTRS256
366		_mm256_cvts2ph_hf8(__m256h __A, __m256h __B) {
	2006	_mm256_cvts_2ph_hf8(__m256h __A, __m256h __B) {
367	2007	return (__m256i)__builtin_ia32_vcvt2ph2hf8s_256((__v16hf)(__A),
368		(__v16hf)(__B));
369		}
370
371		static __inline__ __m256i __DEFAULT_FN_ATTRS256
372		_mm256_mask_cvts2ph_hf8(__m256i __W, __mmask32 __U, __m256h __A, __m256h __B) {
	2008	(__v16hf)(__B));
	2009	}
	2010
	2011	/// Convert two 256-bit vectors, \a __A and \a __B, containing packed FP16
	2012	/// floating-point elements to a 256-bit vector containing E4M3 FP8 elements.
	2013	/// Merging mask \a __U is used to determine if given element should be taken
	2014	/// from \a __W instead. Resulting elements are saturated in case of overflow.
	2015	///
	2016	/// \code{.operation}
	2017	/// FOR i := 0 to 31
	2018	/// IF __U[i]
	2019	/// IF i < 16
	2020	/// dst.hf8[i] := convert_fp16_to_hf8_saturate(__B.fp16[i])
	2021	/// ELSE
	2022	/// dst.hf8[i] := convert_fp16_to_hf8_saturate(__A.fp16[i - 16])
	2023	/// FI
	2024	/// ELSE
	2025	/// dst.hf8[i] := __W.hf8[i]
	2026	/// FI
	2027	/// ENDFOR
	2028	///
	2029	/// dst[MAX:256] := 0
	2030	/// \endcode
	2031	///
	2032	/// \headerfile <immintrin.h>
	2033	///
	2034	/// This intrinsic corresponds to the \c VCVT2PH2HF8S instruction.
	2035	///
	2036	/// \param __W
	2037	/// A 256-bit vector of [32 x hf8].
	2038	/// \param __U
	2039	/// A 32-bit merging mask.
	2040	/// \param __A
	2041	/// A 256-bit vector of [16 x fp16].
	2042	/// \param __B
	2043	/// A 256-bit vector of [16 x fp16].
	2044	/// \returns
	2045	/// A 256-bit vector of [32 x hf8]. Lower 16 elements correspond to the
	2046	/// (converted) elements from \a __B; higher order elements correspond to the
	2047	/// (converted) elements from \a __A. If corresponding mask bit is not set, then
	2048	/// element from \a __W is taken instead.
	2049	static __inline__ __m256i __DEFAULT_FN_ATTRS256 _mm256_mask_cvts_2ph_hf8(
	2050	__m256i __W, __mmask32 __U, __m256h __A, __m256h __B) {
373	2051	return (__m256i)__builtin_ia32_selectb_256(
374		(__mmask16)__U, (__v32qi)_mm256_cvts2ph_hf8(__A, __B), (__v32qi)__W);
375		}
376
	2052	(__mmask32)__U, (__v32qi)_mm256_cvts_2ph_hf8(__A, __B), (__v32qi)__W);
	2053	}
	2054
	2055	/// Convert two 256-bit vectors, \a __A and \a __B, containing packed FP16
	2056	/// floating-point elements to a 256-bit vector containing E4M3 FP8 elements.
	2057	/// Zeroing mask \a __U is used to determine if given element should be zeroed
	2058	/// instead. Resulting elements are saturated in case of overflow.
	2059	///
	2060	/// \code{.operation}
	2061	/// FOR i := 0 to 31
	2062	/// IF __U[i]
	2063	/// IF i < 16
	2064	/// dst.hf8[i] := convert_fp16_to_hf8_saturate(__B.fp16[i])
	2065	/// ELSE
	2066	/// dst.hf8[i] := convert_fp16_to_hf8_saturate(__A.fp16[i - 16])
	2067	/// FI
	2068	/// ELSE
	2069	/// dst.hf8[i] := 0
	2070	/// FI
	2071	/// ENDFOR
	2072	///
	2073	/// dst[MAX:256] := 0
	2074	/// \endcode
	2075	///
	2076	/// \headerfile <immintrin.h>
	2077	///
	2078	/// This intrinsic corresponds to the \c VCVT2PH2HF8S instruction.
	2079	///
	2080	/// \param __U
	2081	/// A 32-bit zeroing mask.
	2082	/// \param __A
	2083	/// A 256-bit vector of [16 x fp16].
	2084	/// \param __B
	2085	/// A 256-bit vector of [16 x fp16].
	2086	/// \returns
	2087	/// A 256-bit vector of [32 x hf8]. Lower 16 elements correspond to the
	2088	/// (converted) elements from \a __B; higher order elements correspond to the
	2089	/// (converted) elements from \a __A. If corresponding mask bit is not set,
	2090	/// zero is taken instead.
377	2091	static __inline__ __m256i __DEFAULT_FN_ATTRS256
378		_mm256_maskz_cvts2ph_hf8(__mmask32 __U, __m256h __A, __m256h __B) {
	2092	_mm256_maskz_cvts_2ph_hf8(__mmask32 __U, __m256h __A, __m256h __B) {
379	2093	return (__m256i)__builtin_ia32_selectb_256(
380		(__mmask16)__U, (__v32qi)_mm256_cvts2ph_hf8(__A, __B),
	2094	(__mmask32)__U, (__v32qi)_mm256_cvts_2ph_hf8(__A, __B),
381	2095	(__v32qi)(__m256i)_mm256_setzero_si256());
382	2096	}
383	2097
384		static __inline__ __m128h __DEFAULT_FN_ATTRS128 _mm_cvthf8(__m128i __A) {
	2098	/// Convert 128-bit vector \a __A, containing packed FP8 E4M3 floating-point
	2099	/// elements to a 128-bit vector containing FP16 elements. The conversion is exact.
	2100	///
	2101	/// \code{.operation}
	2102	/// FOR i := 0 to 7
	2103	/// dst.fp16[i] := convert_hf8_to_fp16(__A.hf8[i])
	2104	/// ENDFOR
	2105	///
	2106	/// dst[MAX:128] := 0
	2107	/// \endcode
	2108	///
	2109	/// \headerfile <immintrin.h>
	2110	///
	2111	/// This intrinsic corresponds to the \c VCVTHF82PH instruction.
	2112	///
	2113	/// \param __A
	2114	/// A 128-bit vector of [16 x hf8].
	2115	/// \returns
	2116	/// A 128-bit vector of [8 x fp16]. Resulting elements correspond to the
	2117	/// (converted) elements from \a __A.
	2118	static __inline__ __m128h __DEFAULT_FN_ATTRS128 _mm_cvthf8_ph(__m128i __A) {
385	2119	return (__m128h)__builtin_ia32_vcvthf8_2ph128_mask(
386	2120	(__v16qi)__A, (__v8hf)(__m128h)_mm_undefined_ph(), (__mmask8)-1);
387	2121	}
388	2122
389		static __inline__ __m128h __DEFAULT_FN_ATTRS128 _mm_mask_cvthf8(__m128h __W,
390		__mmask8 __U,
391		__m128i __A) {
	2123	/// Convert 128-bit vector \a __A, containing packed FP8 E4M3 floating-point
	2124	/// elements to a 128-bit vector containing FP16 elements. The conversion is
	2125	/// exact. Merging mask \a __U is used to determine if given element should be
	2126	/// taken from \a __W instead.
	2127	///
	2128	/// \code{.operation}
	2129	/// FOR i := 0 to 7
	2130	/// IF __U[i]
	2131	/// dst.fp16[i] := convert_hf8_to_fp16(__A.hf8[i])
	2132	/// ELSE
	2133	/// dst.fp16[i] := __W.fp16[i]
	2134	/// FI
	2135	/// ENDFOR
	2136	///
	2137	/// dst[MAX:128] := 0
	2138	/// \endcode
	2139	///
	2140	/// \headerfile <immintrin.h>
	2141	///
	2142	/// This intrinsic corresponds to the \c VCVTHF82PH instruction.
	2143	///
	2144	/// \param __W
	2145	/// A 128-bit vector of [8 x fp16].
	2146	/// \param __U
	2147	/// A 8-bit merging mask.
	2148	/// \param __A
	2149	/// A 128-bit vector of [16 x hf8].
	2150	/// \returns
	2151	/// A 128-bit vector of [8 x fp16]. Resulting elements correspond to the
	2152	/// (converted) elements from \a __A. If corresponding mask bit is not set, then
	2153	/// element from \a __W is taken instead.
	2154	static __inline__ __m128h __DEFAULT_FN_ATTRS128
	2155	_mm_mask_cvthf8_ph(__m128h __W, __mmask8 __U, __m128i __A) {
392	2156	return (__m128h)__builtin_ia32_vcvthf8_2ph128_mask(
393	2157	(__v16qi)__A, (__v8hf)(__m128h)__W, (__mmask8)__U);
394	2158	}
395	2159
396		static __inline__ __m128h __DEFAULT_FN_ATTRS128 _mm_maskz_cvthf8(__mmask8 __U,
397		__m128i __A) {
	2160	/// Convert 128-bit vector \a __A, containing packed FP8 E4M3 floating-point
	2161	/// elements to a 128-bit vector containing FP16 elements. The conversion is
	2162	/// exact. Zeroing mask \a __U is used to determine if given element should be
	2163	/// zeroed instead.
	2164	///
	2165	/// \code{.operation}
	2166	/// FOR i := 0 to 7
	2167	/// IF __U[i]
	2168	/// dst.fp16[i] := convert_hf8_to_fp16(__A.hf8[i])
	2169	/// ELSE
	2170	/// dst.fp16[i] := 0
	2171	/// FI
	2172	/// ENDFOR
	2173	///
	2174	/// dst[MAX:128] := 0
	2175	/// \endcode
	2176	///
	2177	/// \headerfile <immintrin.h>
	2178	///
	2179	/// This intrinsic corresponds to the \c VCVTHF82PH instruction.
	2180	///
	2181	/// \param __U
	2182	/// A 8-bit zeroing mask.
	2183	/// \param __A
	2184	/// A 128-bit vector of [16 x hf8].
	2185	/// \returns
	2186	/// A 128-bit vector of [8 x fp16]. Resulting elements correspond to the
	2187	/// (converted) elements from \a __A. If corresponding mask bit is not set, then
	2188	/// zero is taken instead.
	2189	static __inline__ __m128h __DEFAULT_FN_ATTRS128
	2190	_mm_maskz_cvthf8_ph(__mmask8 __U, __m128i __A) {
398	2191	return (__m128h)__builtin_ia32_vcvthf8_2ph128_mask(
399	2192	(__v16qi)__A, (__v8hf)(__m128h)_mm_setzero_ph(), (__mmask8)__U);
400	2193	}
401	2194
402		static __inline__ __m256h __DEFAULT_FN_ATTRS256 _mm256_cvthf8(__m128i __A) {
	2195	/// Convert 256-bit vector \a __A, containing packed FP8 E4M3 floating-point
	2196	/// elements to a 256-bit vector containing FP16 elements. The conversion is exact.
	2197	///
	2198	/// \code{.operation}
	2199	/// FOR i := 0 to 15
	2200	/// dst.fp16[i] := convert_hf8_to_fp16(__A.hf8[i])
	2201	/// ENDFOR
	2202	///
	2203	/// dst[MAX:256] := 0
	2204	/// \endcode
	2205	///
	2206	/// \headerfile <immintrin.h>
	2207	///
	2208	/// This intrinsic corresponds to the \c VCVTHF82PH instruction.
	2209	///
	2210	/// \param __A
	2211	/// A 256-bit vector of [32 x hf8].
	2212	/// \returns
	2213	/// A 256-bit vector of [16 x fp16]. Resulting elements correspond to the
	2214	/// (converted) elements from \a __A.
	2215	static __inline__ __m256h __DEFAULT_FN_ATTRS256 _mm256_cvthf8_ph(__m128i __A) {
403	2216	return (__m256h)__builtin_ia32_vcvthf8_2ph256_mask(
404	2217	(__v16qi)__A, (__v16hf)(__m256h)_mm256_undefined_ph(), (__mmask16)-1);
405	2218	}
406	2219
	2220	/// Convert 256-bit vector \a __A, containing packed FP8 E4M3 floating-point
	2221	/// elements to a 256-bit vector containing FP16 elements. The conversion is
	2222	/// exact. Merging mask \a __U is used to determine if given element should be
	2223	/// taken from \a __W instead.
	2224	///
	2225	/// \code{.operation}
	2226	/// FOR i := 0 to 15
	2227	/// IF __U[i]
	2228	/// dst.fp16[i] := convert_hf8_to_fp16(__A.hf8[i])
	2229	/// ELSE
	2230	/// dst.fp16[i] := __W.fp16[i]
	2231	/// FI
	2232	/// ENDFOR
	2233	///
	2234	/// dst[MAX:256] := 0
	2235	/// \endcode
	2236	///
	2237	/// \headerfile <immintrin.h>
	2238	///
	2239	/// This intrinsic corresponds to the \c VCVTHF82PH instruction.
	2240	///
	2241	/// \param __W
	2242	/// A 256-bit vector of [16 x fp16].
	2243	/// \param __U
	2244	/// A 16-bit merging mask.
	2245	/// \param __A
	2246	/// A 256-bit vector of [32 x hf8].
	2247	/// \returns
	2248	/// A 256-bit vector of [16 x fp16]. Resulting elements correspond to the
	2249	/// (converted) elements from \a __A. If corresponding mask bit is not set, then
	2250	/// element from \a __W is taken instead.
407	2251	static __inline__ __m256h __DEFAULT_FN_ATTRS256
408		_mm256_mask_cvthf8(__m256h __W, __mmask16 __U, __m128i __A) {
	2252	_mm256_mask_cvthf8_ph(__m256h __W, __mmask16 __U, __m128i __A) {
409	2253	return (__m256h)__builtin_ia32_vcvthf8_2ph256_mask(
410	2254	(__v16qi)__A, (__v16hf)(__m256h)__W, (__mmask16)__U);
411	2255	}
412	2256
	2257	/// Convert 256-bit vector \a __A, containing packed FP8 E4M3 floating-point
	2258	/// elements to a 256-bit vector containing FP16 elements. The conversion is
	2259	/// exact. Zeroing mask \a __U is used to determine if given element should be
	2260	/// zeroed instead.
	2261	///
	2262	/// \code{.operation}
	2263	/// FOR i := 0 to 15
	2264	/// IF __U[i]
	2265	/// dst.fp16[i] := convert_hf8_to_fp16(__A.hf8[i])
	2266	/// ELSE
	2267	/// dst.fp16[i] := 0
	2268	/// FI
	2269	/// ENDFOR
	2270	///
	2271	/// dst[MAX:256] := 0
	2272	/// \endcode
	2273	///
	2274	/// \headerfile <immintrin.h>
	2275	///
	2276	/// This intrinsic corresponds to the \c VCVTHF82PH instruction.
	2277	///
	2278	/// \param __U
	2279	/// A 16-bit zeroing mask.
	2280	/// \param __A
	2281	/// A 256-bit vector of [32 x hf8].
	2282	/// \returns
	2283	/// A 256-bit vector of [16 x fp16]. Resulting elements correspond to the
	2284	/// (converted) elements from \a __A. If corresponding mask bit is not set, then
	2285	/// zero is taken instead.
413	2286	static __inline__ __m256h __DEFAULT_FN_ATTRS256
414		_mm256_maskz_cvthf8(__mmask16 __U, __m128i __A) {
	2287	_mm256_maskz_cvthf8_ph(__mmask16 __U, __m128i __A) {
415	2288	return (__m256h)__builtin_ia32_vcvthf8_2ph256_mask(
416	2289	(__v16qi)__A, (__v16hf)(__m256h)_mm256_setzero_ph(), (__mmask16)__U);
417	2290	}
418	2291
	2292	/// Convert 128-bit vector \a __A containing packed FP16 floating-point elements
	2293	/// to a 128-bit vector containing E5M2 FP8 elements. Upper elements of
	2294	/// resulting vector are zeroed.
	2295	///
	2296	/// \code{.operation}
	2297	/// FOR i := 0 to 7
	2298	/// dst.bf8[i] := convert_fp16_to_bf8(__A.fp16[i])
	2299	/// ENDFOR
	2300	///
	2301	/// dst[MAX:64] := 0
	2302	/// \endcode
	2303	///
	2304	/// \headerfile <immintrin.h>
	2305	///
	2306	/// This intrinsic corresponds to the \c VCVTPH2BF8 instruction.
	2307	///
	2308	/// \param __A
	2309	/// A 128-bit vector of [8 x fp16].
	2310	/// \returns
	2311	/// A 128-bit vector of [16 x bf8]. Lower elements correspond to the (converted)
	2312	/// elements from \a __A; upper elements are zeroed.
419	2313	static __inline__ __m128i __DEFAULT_FN_ATTRS128 _mm_cvtph_bf8(__m128h __A) {
420	2314	return (__m128i)__builtin_ia32_vcvtph2bf8_128_mask(
421	2315	(__v8hf)__A, (__v16qi)(__m128i)_mm_undefined_si128(), (__mmask8)-1);
422	2316	}
423	2317
	2318	/// Convert 128-bit vector \a __A containing packed FP16 floating-point elements
	2319	/// to a 128-bit vector containing E5M2 FP8 elements. Upper elements of
	2320	/// resulting vector are zeroed. Merging mask \a __U is used to determine if
	2321	/// given element should be taken from \a __W instead.
	2322	///
	2323	/// \code{.operation}
	2324	/// FOR i := 0 to 7
	2325	/// IF __U[i]
	2326	/// dst.bf8[i] := convert_fp16_to_bf8(__A.fp16[i])
	2327	/// ELSE
	2328	/// dst.bf8[i] := __W.bf8[i]
	2329	/// FI
	2330	/// ENDFOR
	2331	///
	2332	/// dst[MAX:64] := 0
	2333	/// \endcode
	2334	///
	2335	/// \headerfile <immintrin.h>
	2336	///
	2337	/// This intrinsic corresponds to the \c VCVTPH2BF8 instruction.
	2338	///
	2339	/// \param __W
	2340	/// A 128-bit vector of [16 x bf8].
	2341	/// \param __U
	2342	/// A 8-bit merging mask.
	2343	/// \param __A
	2344	/// A 128-bit vector of [8 x fp16].
	2345	/// \returns
	2346	/// A 128-bit vector of [16 x bf8]. Lower elements correspond to the
	2347	/// (converted) elements from \a __A; upper elements are zeroed. If
	2348	/// corresponding mask bit is not set, then element from \a __W is taken instead.
424	2349	static __inline__ __m128i __DEFAULT_FN_ATTRS128
425	2350	_mm_mask_cvtph_bf8(__m128i __W, __mmask8 __U, __m128h __A) {
426	2351	return (__m128i)__builtin_ia32_vcvtph2bf8_128_mask(
427	2352	(__v8hf)__A, (__v16qi)(__m128i)__W, (__mmask8)__U);
428	2353	}
429	2354
	2355	/// Convert 128-bit vector \a __A containing packed FP16 floating-point elements
	2356	/// to a 128-bit vector containing E5M2 FP8 elements. Upper elements of
	2357	/// resulting vector are zeroed. Zeroing mask \a __U is used to determine if
	2358	/// given element should be zeroed instead.
	2359	///
	2360	/// \code{.operation}
	2361	/// FOR i := 0 to 7
	2362	/// IF __U[i]
	2363	/// dst.bf8[i] := convert_fp16_to_bf8(__A.fp16[i])
	2364	/// ELSE
	2365	/// dst.bf8[i] := 0
	2366	/// FI
	2367	/// ENDFOR
	2368	///
	2369	/// dst[MAX:64] := 0
	2370	/// \endcode
	2371	///
	2372	/// \headerfile <immintrin.h>
	2373	///
	2374	/// This intrinsic corresponds to the \c VCVTPH2BF8 instruction.
	2375	///
	2376	/// \param __U
	2377	/// A 8-bit zeroing mask.
	2378	/// \param __A
	2379	/// A 128-bit vector of [8 x fp16].
	2380	/// \returns
	2381	/// A 128-bit vector of [16 x bf8]. Lower elements correspond to the
	2382	/// (converted) elements from \a __A; upper elements are zeroed. If
	2383	/// corresponding mask bit is not set, then element is zeroed.
430	2384	static __inline__ __m128i __DEFAULT_FN_ATTRS128
431	2385	_mm_maskz_cvtph_bf8(__mmask8 __U, __m128h __A) {
432	2386	return (__m128i)__builtin_ia32_vcvtph2bf8_128_mask(
433	2387	(__v8hf)__A, (__v16qi)(__m128i)_mm_setzero_si128(), (__mmask8)__U);
434	2388	}
435	2389
436		static __inline__ __m128i __DEFAULT_FN_ATTRS256 _mm256_cvtph_bf8(__m256h __A) {
	2390	/// Convert 256-bit vector \a __A containing packed FP16 floating-point elements
	2391	/// to a 128-bit vector containing E5M2 FP8 elements.
	2392	///
	2393	/// \code{.operation}
	2394	/// FOR i := 0 to 15
	2395	/// dst.bf8[i] := convert_fp16_to_bf8(__A.fp16[i])
	2396	/// ENDFOR
	2397	///
	2398	/// dst[MAX:128] := 0
	2399	/// \endcode
	2400	///
	2401	/// \headerfile <immintrin.h>
	2402	///
	2403	/// This intrinsic corresponds to the \c VCVTPH2BF8 instruction.
	2404	///
	2405	/// \param __A
	2406	/// A 256-bit vector of [16 x fp16].
	2407	/// \returns
	2408	/// A 128-bit vector of [16 x bf8]. Resulting elements correspond to the (converted)
	2409	/// elements from \a __A.
	2410	static __inline__ __m128i __DEFAULT_FN_ATTRS256
	2411	_mm256_cvtph_bf8(__m256h __A) {
437	2412	return (__m128i)__builtin_ia32_vcvtph2bf8_256_mask(
438	2413	(__v16hf)__A, (__v16qi)(__m128i)_mm_undefined_si128(), (__mmask16)-1);
439	2414	}
440	2415
	2416	/// Convert 256-bit vector \a __A containing packed FP16 floating-point elements
	2417	/// to a 128-bit vector containing E5M2 FP8 elements. Merging mask \a __U is
	2418	/// used to determine if given element should be taken from \a __W instead.
	2419	///
	2420	/// \code{.operation}
	2421	/// FOR i := 0 to 15
	2422	/// IF __U[i]
	2423	/// dst.bf8[i] := convert_fp16_to_bf8(__A.fp16[i])
	2424	/// ELSE
	2425	/// dst.bf8[i] := __W.bf8[i]
	2426	/// FI
	2427	/// ENDFOR
	2428	///
	2429	/// dst[MAX:128] := 0
	2430	/// \endcode
	2431	///
	2432	/// \headerfile <immintrin.h>
	2433	///
	2434	/// This intrinsic corresponds to the \c VCVTPH2BF8 instruction.
	2435	///
	2436	/// \param __W
	2437	/// A 128-bit vector of [16 x bf8].
	2438	/// \param __U
	2439	/// A 16-bit merging mask.
	2440	/// \param __A
	2441	/// A 256-bit vector of [8 x fp16].
	2442	/// \returns
	2443	/// A 128-bit vector of [16 x bf8]. Resulting elements correspond to the
	2444	/// (converted) elements from \a __A. If
	2445	/// corresponding mask bit is not set, then element from \a __W is taken instead.
441	2446	static __inline__ __m128i __DEFAULT_FN_ATTRS256
442	2447	_mm256_mask_cvtph_bf8(__m128i __W, __mmask16 __U, __m256h __A) {
443	2448	return (__m128i)__builtin_ia32_vcvtph2bf8_256_mask(
444	2449	(__v16hf)__A, (__v16qi)(__m128i)__W, (__mmask16)__U);
445	2450	}
446	2451
	2452	/// Convert 256-bit vector \a __A containing packed FP16 floating-point elements
	2453	/// to a 128-bit vector containing E5M2 FP8 elements. Zeroing mask \a __U is
	2454	/// used to determine if given element should be zeroed instead.
	2455	///
	2456	/// \code{.operation}
	2457	/// FOR i := 0 to 15
	2458	/// IF __U[i]
	2459	/// dst.bf8[i] := convert_fp16_to_bf8(__A.fp16[i])
	2460	/// ELSE
	2461	/// dst.bf8[i] := 0
	2462	/// FI
	2463	/// ENDFOR
	2464	///
	2465	/// dst[MAX:128] := 0
	2466	/// \endcode
	2467	///
	2468	/// \headerfile <immintrin.h>
	2469	///
	2470	/// This intrinsic corresponds to the \c VCVTPH2BF8 instruction.
	2471	///
	2472	/// \param __U
	2473	/// A 16-bit zeroing mask.
	2474	/// \param __A
	2475	/// A 256-bit vector of [16 x fp16].
	2476	/// \returns
	2477	/// A 128-bit vector of [16 x bf8]. Resulting elements correspond to the
	2478	/// (converted) elements from \a __A. If corresponding mask bit is not set,
	2479	/// then element is zeroed instead.
447	2480	static __inline__ __m128i __DEFAULT_FN_ATTRS256
448	2481	_mm256_maskz_cvtph_bf8(__mmask16 __U, __m256h __A) {
449	2482	return (__m128i)__builtin_ia32_vcvtph2bf8_256_mask(
450	2483	(__v16hf)__A, (__v16qi)(__m128i)_mm_setzero_si128(), (__mmask16)__U);
451	2484	}
452	2485
453		static __inline__ __m128i __DEFAULT_FN_ATTRS128 _mm_cvtsph_bf8(__m128h __A) {
	2486	/// Convert 128-bit vector \a __A containing packed FP16 floating-point elements
	2487	/// to a 128-bit vector containing E5M2 FP8 elements. Upper elements of
	2488	/// resulting vector are zeroed. Results are saturated.
	2489	///
	2490	/// \code{.operation}
	2491	/// FOR i := 0 to 7
	2492	/// dst.bf8[i] := convert_fp16_to_bf8_saturate(__A.fp16[i])
	2493	/// ENDFOR
	2494	///
	2495	/// dst[MAX:64] := 0
	2496	/// \endcode
	2497	///
	2498	/// \headerfile <immintrin.h>
	2499	///
	2500	/// This intrinsic corresponds to the \c VCVTPH2BF8S instruction.
	2501	///
	2502	/// \param __A
	2503	/// A 128-bit vector of [8 x fp16].
	2504	/// \returns
	2505	/// A 128-bit vector of [16 x bf8]. Lower elements correspond to the (converted)
	2506	/// elements from \a __A; upper elements are zeroed.
	2507	static __inline__ __m128i __DEFAULT_FN_ATTRS128 _mm_cvts_ph_bf8(__m128h __A) {
454	2508	return (__m128i)__builtin_ia32_vcvtph2bf8s_128_mask(
455	2509	(__v8hf)__A, (__v16qi)(__m128i)_mm_undefined_si128(), (__mmask8)-1);
456	2510	}
457	2511
	2512	/// Convert 128-bit vector \a __A containing packed FP16 floating-point elements
	2513	/// to a 128-bit vector containing E5M2 FP8 elements. Upper elements of
	2514	/// resulting vector are zeroed. Results are saturated. Merging mask \a __U is
	2515	/// used to determine if given element should be taken from \a __W instead.
	2516	///
	2517	/// \code{.operation}
	2518	/// FOR i := 0 to 7
	2519	/// IF __U[i]
	2520	/// dst.bf8[i] := convert_fp16_to_bf8_saturate(__A.fp16[i])
	2521	/// ELSE
	2522	/// dst.bf8[i] := __W.bf8[i]
	2523	/// FI
	2524	/// ENDFOR
	2525	///
	2526	/// dst[MAX:64] := 0
	2527	/// \endcode
	2528	///
	2529	/// \headerfile <immintrin.h>
	2530	///
	2531	/// This intrinsic corresponds to the \c VCVTPH2BF8S instruction.
	2532	///
	2533	/// \param __W
	2534	/// A 128-bit vector of [16 x bf8].
	2535	/// \param __U
	2536	/// A 8-bit merging mask.
	2537	/// \param __A
	2538	/// A 128-bit vector of [8 x fp16].
	2539	/// \returns
	2540	/// A 128-bit vector of [16 x bf8]. Lower elements correspond to the
	2541	/// (converted) elements from \a __A; upper elements are zeroed. If
	2542	/// corresponding mask bit is not set, then element from \a __W is taken instead.
458	2543	static __inline__ __m128i __DEFAULT_FN_ATTRS128
459		_mm_mask_cvtsph_bf8(__m128i __W, __mmask8 __U, __m128h __A) {
	2544	_mm_mask_cvts_ph_bf8(__m128i __W, __mmask8 __U, __m128h __A) {
460	2545	return (__m128i)__builtin_ia32_vcvtph2bf8s_128_mask(
461	2546	(__v8hf)__A, (__v16qi)(__m128i)__W, (__mmask8)__U);
462	2547	}
463	2548
	2549	/// Convert 128-bit vector \a __A containing packed FP16 floating-point elements
	2550	/// to a 128-bit vector containing E5M2 FP8 elements. Upper elements of
	2551	/// resulting vector are zeroed. Results are saturated. Zeroing mask \a __U is
	2552	/// used to determine if given element should be zeroed instead.
	2553	///
	2554	/// \code{.operation}
	2555	/// FOR i := 0 to 7
	2556	/// IF __U[i]
	2557	/// dst.bf8[i] := convert_fp16_to_bf8_saturate(__A.fp16[i])
	2558	/// ELSE
	2559	/// dst.bf8[i] := 0
	2560	/// FI
	2561	/// ENDFOR
	2562	///
	2563	/// dst[MAX:64] := 0
	2564	/// \endcode
	2565	///
	2566	/// \headerfile <immintrin.h>
	2567	///
	2568	/// This intrinsic corresponds to the \c VCVTPH2BF8S instruction.
	2569	///
	2570	/// \param __U
	2571	/// A 8-bit zeroing mask.
	2572	/// \param __A
	2573	/// A 128-bit vector of [8 x fp16].
	2574	/// \returns
	2575	/// A 128-bit vector of [16 x bf8]. Lower elements correspond to the
	2576	/// (converted) elements from \a __A; upper elements are zeroed. If
	2577	/// corresponding mask bit is not set, then element is zeroed.
464	2578	static __inline__ __m128i __DEFAULT_FN_ATTRS128
465		_mm_maskz_cvtsph_bf8(__mmask8 __U, __m128h __A) {
	2579	_mm_maskz_cvts_ph_bf8(__mmask8 __U, __m128h __A) {
466	2580	return (__m128i)__builtin_ia32_vcvtph2bf8s_128_mask(
467	2581	(__v8hf)__A, (__v16qi)(__m128i)_mm_setzero_si128(), (__mmask8)__U);
468	2582	}
469	2583
470		static __inline__ __m128i __DEFAULT_FN_ATTRS256 _mm256_cvtsph_bf8(__m256h __A) {
	2584	/// Convert 256-bit vector \a __A containing packed FP16 floating-point elements
	2585	/// to a 128-bit vector containing E5M2 FP8 elements. Results are saturated.
	2586	///
	2587	/// \code{.operation}
	2588	/// FOR i := 0 to 15
	2589	/// dst.bf8[i] := convert_fp16_to_bf8_saturate(__A.fp16[i])
	2590	/// ENDFOR
	2591	///
	2592	/// dst[MAX:128] := 0
	2593	/// \endcode
	2594	///
	2595	/// \headerfile <immintrin.h>
	2596	///
	2597	/// This intrinsic corresponds to the \c VCVTPH2BF8S instruction.
	2598	///
	2599	/// \param __A
	2600	/// A 256-bit vector of [16 x fp16].
	2601	/// \returns
	2602	/// A 128-bit vector of [16 x bf8]. Resulting elements correspond to the (converted)
	2603	/// elements from \a __A.
	2604	static __inline__ __m128i __DEFAULT_FN_ATTRS256
	2605	_mm256_cvts_ph_bf8(__m256h __A) {
471	2606	return (__m128i)__builtin_ia32_vcvtph2bf8s_256_mask(
472	2607	(__v16hf)__A, (__v16qi)(__m128i)_mm_undefined_si128(), (__mmask16)-1);
473	2608	}
474	2609
	2610	/// Convert 256-bit vector \a __A containing packed FP16 floating-point elements
	2611	/// to a 128-bit vector containing E5M2 FP8 elements. Results are saturated.
	2612	/// Merging mask \a __U is used to determine if given element should be taken
	2613	/// from \a __W instead.
	2614	///
	2615	/// \code{.operation}
	2616	/// FOR i := 0 to 15
	2617	/// IF __U[i]
	2618	/// dst.bf8[i] := convert_fp16_to_bf8_saturate(__A.fp16[i])
	2619	/// ELSE
	2620	/// dst.bf8[i] := __W.bf8[i]
	2621	/// FI
	2622	/// ENDFOR
	2623	///
	2624	/// dst[MAX:128] := 0
	2625	/// \endcode
	2626	///
	2627	/// \headerfile <immintrin.h>
	2628	///
	2629	/// This intrinsic corresponds to the \c VCVTPH2BF8S instruction.
	2630	///
	2631	/// \param __W
	2632	/// A 128-bit vector of [16 x bf8].
	2633	/// \param __U
	2634	/// A 16-bit merging mask.
	2635	/// \param __A
	2636	/// A 256-bit vector of [8 x fp16].
	2637	/// \returns
	2638	/// A 128-bit vector of [16 x bf8]. Resulting elements correspond to the
	2639	/// (converted) elements from \a __A. If
	2640	/// corresponding mask bit is not set, then element from \a __W is taken instead.
475	2641	static __inline__ __m128i __DEFAULT_FN_ATTRS256
476		_mm256_mask_cvtsph_bf8(__m128i __W, __mmask16 __U, __m256h __A) {
	2642	_mm256_mask_cvts_ph_bf8(__m128i __W, __mmask16 __U, __m256h __A) {
477	2643	return (__m128i)__builtin_ia32_vcvtph2bf8s_256_mask(
478	2644	(__v16hf)__A, (__v16qi)(__m128i)__W, (__mmask16)__U);
479	2645	}
480	2646
	2647	/// Convert 256-bit vector \a __A containing packed FP16 floating-point elements
	2648	/// to a 128-bit vector containing E5M2 FP8 elements. Results are saturated.
	2649	/// Zeroing mask \a __U is used to determine if given element should be zeroed
	2650	/// instead.
	2651	///
	2652	/// \code{.operation}
	2653	/// FOR i := 0 to 15
	2654	/// IF __U[i]
	2655	/// dst.bf8[i] := convert_fp16_to_bf8_saturate(__A.fp16[i])
	2656	/// ELSE
	2657	/// dst.bf8[i] := 0
	2658	/// FI
	2659	/// ENDFOR
	2660	///
	2661	/// dst[MAX:128] := 0
	2662	/// \endcode
	2663	///
	2664	/// \headerfile <immintrin.h>
	2665	///
	2666	/// This intrinsic corresponds to the \c VCVTPH2BF8S instruction.
	2667	///
	2668	/// \param __U
	2669	/// A 16-bit zeroing mask.
	2670	/// \param __A
	2671	/// A 256-bit vector of [16 x fp16].
	2672	/// \returns
	2673	/// A 128-bit vector of [16 x bf8]. Resulting elements correspond to the
	2674	/// (converted) elements from \a __A. If corresponding mask bit is not set,
	2675	/// then element is zeroed instead.
481	2676	static __inline__ __m128i __DEFAULT_FN_ATTRS256
482		_mm256_maskz_cvtsph_bf8(__mmask16 __U, __m256h __A) {
	2677	_mm256_maskz_cvts_ph_bf8(__mmask16 __U, __m256h __A) {
483	2678	return (__m128i)__builtin_ia32_vcvtph2bf8s_256_mask(
484	2679	(__v16hf)__A, (__v16qi)(__m128i)_mm_setzero_si128(), (__mmask16)__U);
485	2680	}
486	2681
	2682	/// Convert 128-bit vector \a __A containing packed FP16 floating-point elements
	2683	/// to a 128-bit vector containing E5M2 FP8 elements. Upper elements of
	2684	/// resulting vector are zeroed.
	2685	///
	2686	/// \code{.operation}
	2687	/// FOR i := 0 to 7
	2688	/// dst.hf8[i] := convert_fp16_to_hf8(__A.fp16[i])
	2689	/// ENDFOR
	2690	///
	2691	/// dst[MAX:64] := 0
	2692	/// \endcode
	2693	///
	2694	/// \headerfile <immintrin.h>
	2695	///
	2696	/// This intrinsic corresponds to the \c VCVTPH2HF8 instruction.
	2697	///
	2698	/// \param __A
	2699	/// A 128-bit vector of [8 x fp16].
	2700	/// \returns
	2701	/// A 128-bit vector of [16 x hf8]. Lower elements correspond to the (converted)
	2702	/// elements from \a __A; upper elements are zeroed.
487	2703	static __inline__ __m128i __DEFAULT_FN_ATTRS128 _mm_cvtph_hf8(__m128h __A) {
488	2704	return (__m128i)__builtin_ia32_vcvtph2hf8_128_mask(
489	2705	(__v8hf)__A, (__v16qi)(__m128i)_mm_undefined_si128(), (__mmask8)-1);
490	2706	}
491	2707
	2708	/// Convert 128-bit vector \a __A containing packed FP16 floating-point elements
	2709	/// to a 128-bit vector containing E4M3 FP8 elements. Upper elements of
	2710	/// resulting vector are zeroed. Merging mask \a __U is used to determine if
	2711	/// given element should be taken from \a __W instead.
	2712	///
	2713	/// \code{.operation}
	2714	/// FOR i := 0 to 7
	2715	/// IF __U[i]
	2716	/// dst.hf8[i] := convert_fp16_to_hf8(__A.fp16[i])
	2717	/// ELSE
	2718	/// dst.hf8[i] := __W.hf8[i]
	2719	/// FI
	2720	/// ENDFOR
	2721	///
	2722	/// dst[MAX:64] := 0
	2723	/// \endcode
	2724	///
	2725	/// \headerfile <immintrin.h>
	2726	///
	2727	/// This intrinsic corresponds to the \c VCVTPH2HF8 instruction.
	2728	///
	2729	/// \param __W
	2730	/// A 128-bit vector of [16 x hf8].
	2731	/// \param __U
	2732	/// A 8-bit merging mask.
	2733	/// \param __A
	2734	/// A 128-bit vector of [8 x fp16].
	2735	/// \returns
	2736	/// A 128-bit vector of [16 x hf8]. Lower elements correspond to the
	2737	/// (converted) elements from \a __A; upper elements are zeroed. If
	2738	/// corresponding mask bit is not set, then element from \a __W is taken instead.
492	2739	static __inline__ __m128i __DEFAULT_FN_ATTRS128
493	2740	_mm_mask_cvtph_hf8(__m128i __W, __mmask8 __U, __m128h __A) {
494	2741	return (__m128i)__builtin_ia32_vcvtph2hf8_128_mask(
495	2742	(__v8hf)__A, (__v16qi)(__m128i)__W, (__mmask8)__U);
496	2743	}
497	2744
	2745	/// Convert 128-bit vector \a __A containing packed FP16 floating-point elements
	2746	/// to a 128-bit vector containing E4M3 FP8 elements. Upper elements of
	2747	/// resulting vector are zeroed. Zeroing mask \a __U is used to determine if
	2748	/// given element should be zeroed instead.
	2749	///
	2750	/// \code{.operation}
	2751	/// FOR i := 0 to 7
	2752	/// IF __U[i]
	2753	/// dst.hf8[i] := convert_fp16_to_hf8(__A.fp16[i])
	2754	/// ELSE
	2755	/// dst.hf8[i] := 0
	2756	/// FI
	2757	/// ENDFOR
	2758	///
	2759	/// dst[MAX:64] := 0
	2760	/// \endcode
	2761	///
	2762	/// \headerfile <immintrin.h>
	2763	///
	2764	/// This intrinsic corresponds to the \c VCVTPH2HF8 instruction.
	2765	///
	2766	/// \param __U
	2767	/// A 8-bit zeroing mask.
	2768	/// \param __A
	2769	/// A 128-bit vector of [8 x fp16].
	2770	/// \returns
	2771	/// A 128-bit vector of [16 x hf8]. Lower elements correspond to the
	2772	/// (converted) elements from \a __A; upper elements are zeroed. If
	2773	/// corresponding mask bit is not set, then element is zeroed.
498	2774	static __inline__ __m128i __DEFAULT_FN_ATTRS128
499	2775	_mm_maskz_cvtph_hf8(__mmask8 __U, __m128h __A) {
500	2776	return (__m128i)__builtin_ia32_vcvtph2hf8_128_mask(
501	2777	(__v8hf)__A, (__v16qi)(__m128i)_mm_setzero_si128(), (__mmask8)__U);
502	2778	}
503	2779
504		static __inline__ __m128i __DEFAULT_FN_ATTRS256 _mm256_cvtph_hf8(__m256h __A) {
	2780	/// Convert 256-bit vector \a __A containing packed FP16 floating-point elements
	2781	/// to a 128-bit vector containing E4M3 FP8 elements.
	2782	///
	2783	/// \code{.operation}
	2784	/// FOR i := 0 to 15
	2785	/// dst.hf8[i] := convert_fp16_to_hf8(__A.fp16[i])
	2786	/// ENDFOR
	2787	///
	2788	/// dst[MAX:128] := 0
	2789	/// \endcode
	2790	///
	2791	/// \headerfile <immintrin.h>
	2792	///
	2793	/// This intrinsic corresponds to the \c VCVTPH2HF8 instruction.
	2794	///
	2795	/// \param __A
	2796	/// A 256-bit vector of [16 x fp16].
	2797	/// \returns
	2798	/// A 128-bit vector of [16 x hf8]. Resulting elements correspond to the (converted)
	2799	/// elements from \a __A.
	2800	static __inline__ __m128i __DEFAULT_FN_ATTRS256
	2801	_mm256_cvtph_hf8(__m256h __A) {
505	2802	return (__m128i)__builtin_ia32_vcvtph2hf8_256_mask(
506	2803	(__v16hf)__A, (__v16qi)(__m128i)_mm_undefined_si128(), (__mmask16)-1);
507	2804	}
508	2805
	2806	/// Convert 256-bit vector \a __A containing packed FP16 floating-point elements
	2807	/// to a 128-bit vector containing E4M3 FP8 elements. Merging mask \a __U is
	2808	/// used to determine if given element should be taken from \a __W instead.
	2809	///
	2810	/// \code{.operation}
	2811	/// FOR i := 0 to 15
	2812	/// IF __U[i]
	2813	/// dst.hf8[i] := convert_fp16_to_hf8(__A.fp16[i])
	2814	/// ELSE
	2815	/// dst.hf8[i] := __W.hf8[i]
	2816	/// FI
	2817	/// ENDFOR
	2818	///
	2819	/// dst[MAX:128] := 0
	2820	/// \endcode
	2821	///
	2822	/// \headerfile <immintrin.h>
	2823	///
	2824	/// This intrinsic corresponds to the \c VCVTPH2HF8 instruction.
	2825	///
	2826	/// \param __W
	2827	/// A 128-bit vector of [16 x hf8].
	2828	/// \param __U
	2829	/// A 16-bit merging mask.
	2830	/// \param __A
	2831	/// A 256-bit vector of [8 x fp16].
	2832	/// \returns
	2833	/// A 128-bit vector of [16 x hf8]. Resulting elements correspond to the
	2834	/// (converted) elements from \a __A. If
	2835	/// corresponding mask bit is not set, then element from \a __W is taken instead.
509	2836	static __inline__ __m128i __DEFAULT_FN_ATTRS256
510	2837	_mm256_mask_cvtph_hf8(__m128i __W, __mmask16 __U, __m256h __A) {
511	2838	return (__m128i)__builtin_ia32_vcvtph2hf8_256_mask(
512	2839	(__v16hf)__A, (__v16qi)(__m128i)__W, (__mmask16)__U);
513	2840	}
514	2841
	2842	/// Convert 256-bit vector \a __A containing packed FP16 floating-point elements
	2843	/// to a 128-bit vector containing E4M3 FP8 elements. Zeroing mask \a __U is
	2844	/// used to determine if given element should be zeroed instead.
	2845	///
	2846	/// \code{.operation}
	2847	/// FOR i := 0 to 15
	2848	/// IF __U[i]
	2849	/// dst.hf8[i] := convert_fp16_to_hf8(__A.fp16[i])
	2850	/// ELSE
	2851	/// dst.hf8[i] := 0
	2852	/// FI
	2853	/// ENDFOR
	2854	///
	2855	/// dst[MAX:128] := 0
	2856	/// \endcode
	2857	///
	2858	/// \headerfile <immintrin.h>
	2859	///
	2860	/// This intrinsic corresponds to the \c VCVTPH2HF8 instruction.
	2861	///
	2862	/// \param __U
	2863	/// A 16-bit zeroing mask.
	2864	/// \param __A
	2865	/// A 256-bit vector of [16 x fp16].
	2866	/// \returns
	2867	/// A 128-bit vector of [16 x hf8]. Resulting elements correspond to the
	2868	/// (converted) elements from \a __A. If corresponding mask bit is not set,
	2869	/// then element is zeroed instead.
515	2870	static __inline__ __m128i __DEFAULT_FN_ATTRS256
516	2871	_mm256_maskz_cvtph_hf8(__mmask16 __U, __m256h __A) {
517	2872	return (__m128i)__builtin_ia32_vcvtph2hf8_256_mask(
518	2873	(__v16hf)__A, (__v16qi)(__m128i)_mm_setzero_si128(), (__mmask16)__U);
519	2874	}
520	2875
521		static __inline__ __m128i __DEFAULT_FN_ATTRS128 _mm_cvtsph_hf8(__m128h __A) {
	2876	/// Convert 128-bit vector \a __A containing packed FP16 floating-point elements
	2877	/// to a 128-bit vector containing E4M3 FP8 elements. Upper elements of
	2878	/// resulting vector are zeroed. Results are saturated.
	2879	///
	2880	/// \code{.operation}
	2881	/// FOR i := 0 to 7
	2882	/// dst.hf8[i] := convert_fp16_to_hf8_saturate(__A.fp16[i])
	2883	/// ENDFOR
	2884	///
	2885	/// dst[MAX:64] := 0
	2886	/// \endcode
	2887	///
	2888	/// \headerfile <immintrin.h>
	2889	///
	2890	/// This intrinsic corresponds to the \c VCVTPH2HF8S instruction.
	2891	///
	2892	/// \param __A
	2893	/// A 128-bit vector of [8 x fp16].
	2894	/// \returns
	2895	/// A 128-bit vector of [16 x hf8]. Lower elements correspond to the (converted)
	2896	/// elements from \a __A; upper elements are zeroed.
	2897	static __inline__ __m128i __DEFAULT_FN_ATTRS128 _mm_cvts_ph_hf8(__m128h __A) {
522	2898	return (__m128i)__builtin_ia32_vcvtph2hf8s_128_mask(
523	2899	(__v8hf)__A, (__v16qi)(__m128i)_mm_undefined_si128(), (__mmask8)-1);
524	2900	}
525	2901
	2902	/// Convert 128-bit vector \a __A containing packed FP16 floating-point elements
	2903	/// to a 128-bit vector containing E4M3 FP8 elements. Upper elements of
	2904	/// resulting vector are zeroed. Results are saturated. Merging mask \a __U is
	2905	/// used to determine if given element should be taken from \a __W instead.
	2906	///
	2907	/// \code{.operation}
	2908	/// FOR i := 0 to 7
	2909	/// IF __U[i]
	2910	/// dst.hf8[i] := convert_fp16_to_hf8_saturate(__A.fp16[i])
	2911	/// ELSE
	2912	/// dst.hf8[i] := __W.hf8[i]
	2913	/// FI
	2914	/// ENDFOR
	2915	///
	2916	/// dst[MAX:64] := 0
	2917	/// \endcode
	2918	///
	2919	/// \headerfile <immintrin.h>
	2920	///
	2921	/// This intrinsic corresponds to the \c VCVTPH2HF8S instruction.
	2922	///
	2923	/// \param __W
	2924	/// A 128-bit vector of [16 x hf8].
	2925	/// \param __U
	2926	/// A 8-bit merging mask.
	2927	/// \param __A
	2928	/// A 128-bit vector of [8 x fp16].
	2929	/// \returns
	2930	/// A 128-bit vector of [16 x hf8]. Lower elements correspond to the
	2931	/// (converted) elements from \a __A; upper elements are zeroed. If
	2932	/// corresponding mask bit is not set, then element from \a __W is taken instead.
526	2933	static __inline__ __m128i __DEFAULT_FN_ATTRS128
527		_mm_mask_cvtsph_hf8(__m128i __W, __mmask8 __U, __m128h __A) {
	2934	_mm_mask_cvts_ph_hf8(__m128i __W, __mmask8 __U, __m128h __A) {
528	2935	return (__m128i)__builtin_ia32_vcvtph2hf8s_128_mask(
529	2936	(__v8hf)__A, (__v16qi)(__m128i)__W, (__mmask8)__U);
530	2937	}
531	2938
	2939	/// Convert 128-bit vector \a __A containing packed FP16 floating-point elements
	2940	/// to a 128-bit vector containing E4M3 FP8 elements. Upper elements of
	2941	/// resulting vector are zeroed. Results are saturated. Zeroing mask \a __U is
	2942	/// used to determine if given element should be zeroed instead.
	2943	///
	2944	/// \code{.operation}
	2945	/// FOR i := 0 to 7
	2946	/// IF __U[i]
	2947	/// dst.hf8[i] := convert_fp16_to_hf8_saturate(__A.fp16[i])
	2948	/// ELSE
	2949	/// dst.hf8[i] := 0
	2950	/// FI
	2951	/// ENDFOR
	2952	///
	2953	/// dst[MAX:64] := 0
	2954	/// \endcode
	2955	///
	2956	/// \headerfile <immintrin.h>
	2957	///
	2958	/// This intrinsic corresponds to the \c VCVTPH2HF8S instruction.
	2959	///
	2960	/// \param __U
	2961	/// A 8-bit zeroing mask.
	2962	/// \param __A
	2963	/// A 128-bit vector of [8 x fp16].
	2964	/// \returns
	2965	/// A 128-bit vector of [16 x hf8]. Lower elements correspond to the
	2966	/// (converted) elements from \a __A; upper elements are zeroed. If
	2967	/// corresponding mask bit is not set, then element is zeroed.
532	2968	static __inline__ __m128i __DEFAULT_FN_ATTRS128
533		_mm_maskz_cvtsph_hf8(__mmask8 __U, __m128h __A) {
	2969	_mm_maskz_cvts_ph_hf8(__mmask8 __U, __m128h __A) {
534	2970	return (__m128i)__builtin_ia32_vcvtph2hf8s_128_mask(
535	2971	(__v8hf)__A, (__v16qi)(__m128i)_mm_setzero_si128(), (__mmask8)__U);
536	2972	}
537	2973
538		static __inline__ __m128i __DEFAULT_FN_ATTRS256 _mm256_cvtsph_hf8(__m256h __A) {
	2974	/// Convert 256-bit vector \a __A containing packed FP16 floating-point elements
	2975	/// to a 128-bit vector containing E4M3 FP8 elements. Results are saturated.
	2976	///
	2977	/// \code{.operation}
	2978	/// FOR i := 0 to 15
	2979	/// dst.hf8[i] := convert_fp16_to_hf8_saturate(__A.fp16[i])
	2980	/// ENDFOR
	2981	///
	2982	/// dst[MAX:128] := 0
	2983	/// \endcode
	2984	///
	2985	/// \headerfile <immintrin.h>
	2986	///
	2987	/// This intrinsic corresponds to the \c VCVTPH2HF8S instruction.
	2988	///
	2989	/// \param __A
	2990	/// A 256-bit vector of [16 x fp16].
	2991	/// \returns
	2992	/// A 128-bit vector of [16 x hf8]. Resulting elements correspond to the (converted)
	2993	/// elements from \a __A.
	2994	static __inline__ __m128i __DEFAULT_FN_ATTRS256
	2995	_mm256_cvts_ph_hf8(__m256h __A) {
539	2996	return (__m128i)__builtin_ia32_vcvtph2hf8s_256_mask(
540	2997	(__v16hf)__A, (__v16qi)(__m128i)_mm_undefined_si128(), (__mmask16)-1);
541	2998	}
542	2999
	3000	/// Convert 256-bit vector \a __A containing packed FP16 floating-point elements
	3001	/// to a 128-bit vector containing E4M3 FP8 elements. Results are saturated.
	3002	/// Merging mask \a __U is used to determine if given element should be taken
	3003	/// from \a __W instead.
	3004	///
	3005	/// \code{.operation}
	3006	/// FOR i := 0 to 15
	3007	/// IF __U[i]
	3008	/// dst.hf8[i] := convert_fp16_to_hf8_saturate(__A.fp16[i])
	3009	/// ELSE
	3010	/// dst.hf8[i] := __W.hf8[i]
	3011	/// FI
	3012	/// ENDFOR
	3013	///
	3014	/// dst[MAX:128] := 0
	3015	/// \endcode
	3016	///
	3017	/// \headerfile <immintrin.h>
	3018	///
	3019	/// This intrinsic corresponds to the \c VCVTPH2HF8S instruction.
	3020	///
	3021	/// \param __W
	3022	/// A 128-bit vector of [16 x hf8].
	3023	/// \param __U
	3024	/// A 16-bit merging mask.
	3025	/// \param __A
	3026	/// A 256-bit vector of [8 x fp16].
	3027	/// \returns
	3028	/// A 128-bit vector of [16 x hf8]. Resulting elements correspond to the
	3029	/// (converted) elements from \a __A. If
	3030	/// corresponding mask bit is not set, then element from \a __W is taken instead.
543	3031	static __inline__ __m128i __DEFAULT_FN_ATTRS256
544		_mm256_mask_cvtsph_hf8(__m128i __W, __mmask16 __U, __m256h __A) {
	3032	_mm256_mask_cvts_ph_hf8(__m128i __W, __mmask16 __U, __m256h __A) {
545	3033	return (__m128i)__builtin_ia32_vcvtph2hf8s_256_mask(
546	3034	(__v16hf)__A, (__v16qi)(__m128i)__W, (__mmask16)__U);
547	3035	}
548	3036
	3037	/// Convert 256-bit vector \a __A containing packed FP16 floating-point elements
	3038	/// to a 128-bit vector containing E4M3 FP8 elements. Results are saturated.
	3039	/// Zeroing mask \a __U is used to determine if given element should be zeroed
	3040	/// instead.
	3041	///
	3042	/// \code{.operation}
	3043	/// FOR i := 0 to 15
	3044	/// IF __U[i]
	3045	/// dst.hf8[i] := convert_fp16_to_hf8_saturate(__A.fp16[i])
	3046	/// ELSE
	3047	/// dst.hf8[i] := 0
	3048	/// FI
	3049	/// ENDFOR
	3050	///
	3051	/// dst[MAX:128] := 0
	3052	/// \endcode
	3053	///
	3054	/// \headerfile <immintrin.h>
	3055	///
	3056	/// This intrinsic corresponds to the \c VCVTPH2HF8S instruction.
	3057	///
	3058	/// \param __U
	3059	/// A 16-bit zeroing mask.
	3060	/// \param __A
	3061	/// A 256-bit vector of [16 x fp16].
	3062	/// \returns
	3063	/// A 128-bit vector of [16 x hf8]. Resulting elements correspond to the
	3064	/// (converted) elements from \a __A. If corresponding mask bit is not set,
	3065	/// then element is zeroed instead.
549	3066	static __inline__ __m128i __DEFAULT_FN_ATTRS256
550		_mm256_maskz_cvtsph_hf8(__mmask16 __U, __m256h __A) {
	3067	_mm256_maskz_cvts_ph_hf8(__mmask16 __U, __m256h __A) {
551	3068	return (__m128i)__builtin_ia32_vcvtph2hf8s_256_mask(
552	3069	(__v16hf)__A, (__v16qi)(__m128i)_mm_setzero_si128(), (__mmask16)__U);
553	3070	}
554	3071
	3072	/// Convert 128-bit vector \a __A, containing packed FP8 E5M2 floating-point
	3073	/// elements to a 128-bit vector containing FP16 elements. The conversion is exact.
	3074	///
	3075	/// \code{.operation}
	3076	/// FOR i := 0 to 7
	3077	/// dst.fp16[i] := convert_bf8_to_fp16(__A.bf8[i])
	3078	/// ENDFOR
	3079	/// \endcode
	3080	///
	3081	/// \headerfile <immintrin.h>
	3082	///
	3083	/// This intrinsic does not correspond to a single instruction.
	3084	///
	3085	/// \param __A
	3086	/// A 128-bit vector of [16 x bf8].
	3087	/// \returns
	3088	/// A 128-bit vector of [8 x fp16]. Resulting elements correspond to the
	3089	/// (converted) elements from \a __A.
555	3090	static __inline__ __m128h __DEFAULT_FN_ATTRS128 _mm_cvtbf8_ph(__m128i __A) {
556	3091	return _mm_castsi128_ph(_mm_slli_epi16(_mm_cvtepi8_epi16(__A), 8));
557	3092	}
558	3093
	3094	/// Convert 128-bit vector \a __A, containing packed FP8 E5M2 floating-point
	3095	/// elements to a 128-bit vector containing FP16 elements. The conversion is
	3096	/// exact. Merging mask \a __U is used to determine if given element should be
	3097	/// taken from \a __W instead.
	3098	///
	3099	/// \code{.operation}
	3100	/// FOR i := 0 to 7
	3101	/// IF __U[i]
	3102	/// dst.fp16[i] := convert_bf8_to_fp16(__A.bf8[i])
	3103	/// ELSE
	3104	/// dst.fp16[i] := __W.fp16[i]
	3105	/// FI
	3106	/// ENDFOR
	3107	/// \endcode
	3108	///
	3109	/// \headerfile <immintrin.h>
	3110	///
	3111	/// This intrinsic does not correspond to a single instruction.
	3112	///
	3113	/// \param __W
	3114	/// A 128-bit vector of [8 x fp16].
	3115	/// \param __U
	3116	/// A 8-bit merging mask.
	3117	/// \param __A
	3118	/// A 128-bit vector of [16 x bf8].
	3119	/// \returns
	3120	/// A 128-bit vector of [8 x fp16]. Resulting elements correspond to the
	3121	/// (converted) elements from \a __A. If corresponding mask bit is not set, then
	3122	/// element from \a __W is taken instead.
559	3123	static __inline__ __m128h __DEFAULT_FN_ATTRS128
560		_mm_mask_cvtbf8_ph(__m128h __S, __mmask8 __U, __m128i __A) {
	3124	_mm_mask_cvtbf8_ph(__m128h __W, __mmask8 __U, __m128i __A) {
561	3125	return _mm_castsi128_ph(
562		_mm_mask_slli_epi16((__m128i)__S, __U, _mm_cvtepi8_epi16(__A), 8));
563		}
564
	3126	_mm_mask_slli_epi16((__m128i)__W, __U, _mm_cvtepi8_epi16(__A), 8));
	3127	}
	3128
	3129	/// Convert 128-bit vector \a __A, containing packed FP8 E5M2 floating-point
	3130	/// elements to a 128-bit vector containing FP16 elements. The conversion is
	3131	/// exact. Zeroing mask \a __U is used to determine if given element should be
	3132	/// zeroed instead.
	3133	///
	3134	/// \code{.operation}
	3135	/// FOR i := 0 to 7
	3136	/// IF __U[i]
	3137	/// dst.fp16[i] := convert_bf8_to_fp16(__A.bf8[i])
	3138	/// ELSE
	3139	/// dst.fp16[i] := 0
	3140	/// FI
	3141	/// ENDFOR
	3142	/// \endcode
	3143	///
	3144	/// \headerfile <immintrin.h>
	3145	///
	3146	/// This intrinsic does not correspond to a single instruction.
	3147	///
	3148	/// \param __U
	3149	/// A 8-bit zeroing mask.
	3150	/// \param __A
	3151	/// A 128-bit vector of [16 x bf8].
	3152	/// \returns
	3153	/// A 128-bit vector of [8 x fp16]. Resulting elements correspond to the
	3154	/// (converted) elements from \a __A. If corresponding mask bit is not set, then
	3155	/// zero is taken instead.
565	3156	static __inline__ __m128h __DEFAULT_FN_ATTRS128
566	3157	_mm_maskz_cvtbf8_ph(__mmask8 __U, __m128i __A) {
567	3158	return _mm_castsi128_ph(_mm_slli_epi16(_mm_maskz_cvtepi8_epi16(__U, __A), 8));
568	3159	}
569	3160
	3161	/// Convert 256-bit vector \a __A, containing packed FP8 E4M3 floating-point
	3162	/// elements to a 256-bit vector containing FP16 elements. The conversion is exact.
	3163	///
	3164	/// \code{.operation}
	3165	/// FOR i := 0 to 15
	3166	/// dst.fp16[i] := convert_bf8_to_fp16(__A.bf8[i])
	3167	/// ENDFOR
	3168	/// \endcode
	3169	///
	3170	/// \headerfile <immintrin.h>
	3171	///
	3172	/// This intrinsic does not correspond to a single instruction.
	3173	///
	3174	/// \param __A
	3175	/// A 256-bit vector of [32 x bf8].
	3176	/// \returns
	3177	/// A 256-bit vector of [16 x fp16]. Resulting elements correspond to the
	3178	/// (converted) elements from \a __A.
570	3179	static __inline__ __m256h __DEFAULT_FN_ATTRS256 _mm256_cvtbf8_ph(__m128i __A) {
571	3180	return _mm256_castsi256_ph(_mm256_slli_epi16(_mm256_cvtepi8_epi16(__A), 8));
572	3181	}
573	3182
	3183	/// Convert 256-bit vector \a __A, containing packed FP8 E5M2 floating-point
	3184	/// elements to a 256-bit vector containing FP16 elements. The conversion is
	3185	/// exact. Merging mask \a __U is used to determine if given element should be
	3186	/// taken from \a __W instead.
	3187	///
	3188	/// \code{.operation}
	3189	/// FOR i := 0 to 15
	3190	/// IF __U[i]
	3191	/// dst.fp16[i] := convert_bf8_to_fp16(__A.bf8[i])
	3192	/// ELSE
	3193	/// dst.fp16[i] := __W.fp16[i]
	3194	/// FI
	3195	/// ENDFOR
	3196	/// \endcode
	3197	///
	3198	/// \headerfile <immintrin.h>
	3199	///
	3200	/// This intrinsic does not correspond to a single instruction.
	3201	///
	3202	/// \param __W
	3203	/// A 256-bit vector of [16 x fp16].
	3204	/// \param __U
	3205	/// A 16-bit merging mask.
	3206	/// \param __A
	3207	/// A 256-bit vector of [32 x bf8].
	3208	/// \returns
	3209	/// A 256-bit vector of [16 x fp16]. Resulting elements correspond to the
	3210	/// (converted) elements from \a __A. If corresponding mask bit is not set, then
	3211	/// element from \a __W is taken instead.
574	3212	static __inline__ __m256h __DEFAULT_FN_ATTRS256
575		_mm256_mask_cvtbf8_ph(__m256h __S, __mmask16 __U, __m128i __A) {
	3213	_mm256_mask_cvtbf8_ph(__m256h __W, __mmask16 __U, __m128i __A) {
576	3214	return _mm256_castsi256_ph(
577		_mm256_mask_slli_epi16((__m256i)__S, __U, _mm256_cvtepi8_epi16(__A), 8));
578		}
579
	3215	_mm256_mask_slli_epi16((__m256i)__W, __U, _mm256_cvtepi8_epi16(__A), 8));
	3216	}
	3217
	3218	/// Convert 256-bit vector \a __A, containing packed FP8 E5M2 floating-point
	3219	/// elements to a 256-bit vector containing FP16 elements. The conversion is
	3220	/// exact. Zeroing mask \a __U is used to determine if given element should be
	3221	/// zeroed instead.
	3222	///
	3223	/// \code{.operation}
	3224	/// FOR i := 0 to 15
	3225	/// IF __U[i]
	3226	/// dst.fp16[i] := convert_bf8_to_fp16(__A.bf8[i])
	3227	/// ELSE
	3228	/// dst.fp16[i] := 0
	3229	/// FI
	3230	/// ENDFOR
	3231	/// \endcode
	3232	///
	3233	/// \headerfile <immintrin.h>
	3234	///
	3235	/// This intrinsic does not correspond to a single instruction.
	3236	///
	3237	/// \param __U
	3238	/// A 16-bit zeroing mask.
	3239	/// \param __A
	3240	/// A 256-bit vector of [32 x bf8].
	3241	/// \returns
	3242	/// A 256-bit vector of [16 x fp16]. Resulting elements correspond to the
	3243	/// (converted) elements from \a __A. If corresponding mask bit is not set, then
	3244	/// zero is taken instead.
580	3245	static __inline__ __m256h __DEFAULT_FN_ATTRS256
581	3246	_mm256_maskz_cvtbf8_ph(__mmask16 __U, __m128i __A) {
582	3247	return _mm256_castsi256_ph(
583	3248	_mm256_slli_epi16(_mm256_maskz_cvtepi8_epi16(__U, __A), 8));
584	3249	}
585	3250
	3251	// clang-format on
	3252
586	3253	#undef __DEFAULT_FN_ATTRS128
587	3254	#undef __DEFAULT_FN_ATTRS256
588	3255

lib/include/avx10_2minmaxintrin.h+18-63

...	...	@@ -66,34 +66,19 @@
66	66	(__v2df)_mm_setzero_pd(), (__mmask8)(U)))
67	67
68	68	#define _mm256_minmax_pd(A, B, C) \
69		((__m256d)__builtin_ia32_vminmaxpd256_round_mask( \
	69	((__m256d)__builtin_ia32_vminmaxpd256_mask( \
70	70	(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), (int)(C), \
71		(__v4df)_mm256_setzero_pd(), (__mmask8)-1, _MM_FROUND_NO_EXC))
	71	(__v4df)_mm256_setzero_pd(), (__mmask8)-1))
72	72
73	73	#define _mm256_mask_minmax_pd(W, U, A, B, C) \
74		((__m256d)__builtin_ia32_vminmaxpd256_round_mask( \
	74	((__m256d)__builtin_ia32_vminmaxpd256_mask( \
75	75	(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), (int)(C), \
76		(__v4df)(__m256d)(W), (__mmask8)(U), _MM_FROUND_NO_EXC))
	76	(__v4df)(__m256d)(W), (__mmask8)(U)))
77	77
78	78	#define _mm256_maskz_minmax_pd(U, A, B, C) \
79		((__m256d)__builtin_ia32_vminmaxpd256_round_mask( \
	79	((__m256d)__builtin_ia32_vminmaxpd256_mask( \
80	80	(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), (int)(C), \
81		(__v4df)_mm256_setzero_pd(), (__mmask8)(U), _MM_FROUND_NO_EXC))
82
83		#define _mm256_minmax_round_pd(A, B, C, R) \
84		((__m256d)__builtin_ia32_vminmaxpd256_round_mask( \
85		(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), (int)(C), \
86		(__v4df)_mm256_undefined_pd(), (__mmask8)-1, (int)(R)))
87
88		#define _mm256_mask_minmax_round_pd(W, U, A, B, C, R) \
89		((__m256d)__builtin_ia32_vminmaxpd256_round_mask( \
90		(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), (int)(C), \
91		(__v4df)(__m256d)(W), (__mmask8)(U), (int)(R)))
92
93		#define _mm256_maskz_minmax_round_pd(U, A, B, C, R) \
94		((__m256d)__builtin_ia32_vminmaxpd256_round_mask( \
95		(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), (int)(C), \
96		(__v4df)_mm256_setzero_pd(), (__mmask8)(U), (int)(R)))
	81	(__v4df)_mm256_setzero_pd(), (__mmask8)(U)))
97	82
98	83	#define _mm_minmax_ph(A, B, C) \
99	84	((__m128h)__builtin_ia32_vminmaxph128_mask( \
...	...	@@ -111,34 +96,19 @@
111	96	(__v8hf)_mm_setzero_ph(), (__mmask8)(U)))
112	97
113	98	#define _mm256_minmax_ph(A, B, C) \
114		((__m256h)__builtin_ia32_vminmaxph256_round_mask( \
	99	((__m256h)__builtin_ia32_vminmaxph256_mask( \
115	100	(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), (int)(C), \
116		(__v16hf)_mm256_setzero_ph(), (__mmask16)-1, _MM_FROUND_NO_EXC))
	101	(__v16hf)_mm256_setzero_ph(), (__mmask16)-1))
117	102
118	103	#define _mm256_mask_minmax_ph(W, U, A, B, C) \
119		((__m256h)__builtin_ia32_vminmaxph256_round_mask( \
	104	((__m256h)__builtin_ia32_vminmaxph256_mask( \
120	105	(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), (int)(C), \
121		(__v16hf)(__m256h)(W), (__mmask16)(U), _MM_FROUND_NO_EXC))
	106	(__v16hf)(__m256h)(W), (__mmask16)(U)))
122	107
123	108	#define _mm256_maskz_minmax_ph(U, A, B, C) \
124		((__m256h)__builtin_ia32_vminmaxph256_round_mask( \
125		(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), (int)(C), \
126		(__v16hf)_mm256_setzero_ph(), (__mmask16)(U), _MM_FROUND_NO_EXC))
127
128		#define _mm256_minmax_round_ph(A, B, C, R) \
129		((__m256h)__builtin_ia32_vminmaxph256_round_mask( \
	109	((__m256h)__builtin_ia32_vminmaxph256_mask( \
130	110	(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), (int)(C), \
131		(__v16hf)_mm256_undefined_ph(), (__mmask16)-1, (int)(R)))
132
133		#define _mm256_mask_minmax_round_ph(W, U, A, B, C, R) \
134		((__m256h)__builtin_ia32_vminmaxph256_round_mask( \
135		(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), (C), \
136		(__v16hf)(__m256h)(W), (__mmask16)(U), (int)(R)))
137
138		#define _mm256_maskz_minmax_round_ph(U, A, B, C, R) \
139		((__m256h)__builtin_ia32_vminmaxph256_round_mask( \
140		(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), (int)(C), \
141		(__v16hf)_mm256_setzero_ph(), (__mmask16)(U), (int)(R)))
	111	(__v16hf)_mm256_setzero_ph(), (__mmask16)(U)))
142	112
143	113	#define _mm_minmax_ps(A, B, C) \
144	114	((__m128)__builtin_ia32_vminmaxps128_mask( \
...	...	@@ -156,34 +126,19 @@
156	126	(__v4sf)_mm_setzero_ps(), (__mmask8)(U)))
157	127
158	128	#define _mm256_minmax_ps(A, B, C) \
159		((__m256)__builtin_ia32_vminmaxps256_round_mask( \
	129	((__m256)__builtin_ia32_vminmaxps256_mask( \
160	130	(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), (int)(C), \
161		(__v8sf)_mm256_setzero_ps(), (__mmask8)-1, _MM_FROUND_NO_EXC))
	131	(__v8sf)_mm256_setzero_ps(), (__mmask8)-1))
162	132
163	133	#define _mm256_mask_minmax_ps(W, U, A, B, C) \
164		((__m256)__builtin_ia32_vminmaxps256_round_mask( \
	134	((__m256)__builtin_ia32_vminmaxps256_mask( \
165	135	(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), (int)(C), (__v8sf)(__m256)(W), \
166		(__mmask8)(U), _MM_FROUND_NO_EXC))
	136	(__mmask8)(U)))
167	137
168	138	#define _mm256_maskz_minmax_ps(U, A, B, C) \
169		((__m256)__builtin_ia32_vminmaxps256_round_mask( \
170		(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), (int)(C), \
171		(__v8sf)_mm256_setzero_ps(), (__mmask8)(U), _MM_FROUND_NO_EXC))
172
173		#define _mm256_minmax_round_ps(A, B, C, R) \
174		((__m256)__builtin_ia32_vminmaxps256_round_mask( \
175		(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), (int)(C), \
176		(__v8sf)_mm256_undefined_ps(), (__mmask8)-1, (int)(R)))
177
178		#define _mm256_mask_minmax_round_ps(W, U, A, B, C, R) \
179		((__m256)__builtin_ia32_vminmaxps256_round_mask( \
180		(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), (int)(C), (__v8sf)(__m256)(W), \
181		(__mmask8)(U), (int)(R)))
182
183		#define _mm256_maskz_minmax_round_ps(U, A, B, C, R) \
184		((__m256)__builtin_ia32_vminmaxps256_round_mask( \
	139	((__m256)__builtin_ia32_vminmaxps256_mask( \
185	140	(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), (int)(C), \
186		(__v8sf)_mm256_setzero_ps(), (__mmask8)(U), (int)(R)))
	141	(__v8sf)_mm256_setzero_ps(), (__mmask8)(U)))
187	142
188	143	#define _mm_minmax_sd(A, B, C) \
189	144	((__m128d)__builtin_ia32_vminmaxsd_round_mask( \

lib/include/avx10_2niintrin.h-1666

...	...	@@ -402,1672 +402,6 @@ static __inline__ __m256i __DEFAULT_FN_ATTRS256 _mm256_maskz_dpwuuds_epi32(
402	402	(__v8si)_mm256_setzero_si256());
403	403	}
404	404
405		/* YMM Rounding */
406		#define _mm256_add_round_pd(A, B, R) \
407		((__m256d)__builtin_ia32_vaddpd256_round((__v4df)(__m256d)(A), \
408		(__v4df)(__m256d)(B), (int)(R)))
409
410		#define _mm256_mask_add_round_pd(W, U, A, B, R) \
411		((__m256d)__builtin_ia32_selectpd_256( \
412		(__mmask8)(U), (__v4df)_mm256_add_round_pd((A), (B), (R)), \
413		(__v4df)(__m256d)(W)))
414
415		#define _mm256_maskz_add_round_pd(U, A, B, R) \
416		((__m256d)__builtin_ia32_selectpd_256( \
417		(__mmask8)(U), (__v4df)_mm256_add_round_pd((A), (B), (R)), \
418		(__v4df)_mm256_setzero_pd()))
419
420		#define _mm256_add_round_ph(A, B, R) \
421		((__m256h)__builtin_ia32_vaddph256_round((__v16hf)(__m256h)(A), \
422		(__v16hf)(__m256h)(B), (int)(R)))
423
424		#define _mm256_mask_add_round_ph(W, U, A, B, R) \
425		((__m256h)__builtin_ia32_selectph_256( \
426		(__mmask16)(U), (__v16hf)_mm256_add_round_ph((A), (B), (R)), \
427		(__v16hf)(__m256h)(W)))
428
429		#define _mm256_maskz_add_round_ph(U, A, B, R) \
430		((__m256h)__builtin_ia32_selectph_256( \
431		(__mmask16)(U), (__v16hf)_mm256_add_round_ph((A), (B), (R)), \
432		(__v16hf)_mm256_setzero_ph()))
433
434		#define _mm256_add_round_ps(A, B, R) \
435		((__m256)__builtin_ia32_vaddps256_round((__v8sf)(__m256)(A), \
436		(__v8sf)(__m256)(B), (int)(R)))
437
438		#define _mm256_mask_add_round_ps(W, U, A, B, R) \
439		((__m256)__builtin_ia32_selectps_256( \
440		(__mmask8)(U), (__v8sf)_mm256_add_round_ps((A), (B), (R)), \
441		(__v8sf)(__m256)(W)))
442
443		#define _mm256_maskz_add_round_ps(U, A, B, R) \
444		((__m256)__builtin_ia32_selectps_256( \
445		(__mmask8)(U), (__v8sf)_mm256_add_round_ps((A), (B), (R)), \
446		(__v8sf)_mm256_setzero_ps()))
447
448		#define _mm256_cmp_round_pd_mask(A, B, P, R) \
449		((__mmask8)__builtin_ia32_vcmppd256_round_mask( \
450		(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), (int)(P), (__mmask8)-1, \
451		(int)(R)))
452
453		#define _mm256_mask_cmp_round_pd_mask(U, A, B, P, R) \
454		((__mmask8)__builtin_ia32_vcmppd256_round_mask( \
455		(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), (int)(P), (__mmask8)(U), \
456		(int)(R)))
457
458		#define _mm256_cmp_round_ph_mask(A, B, P, R) \
459		((__mmask16)__builtin_ia32_vcmpph256_round_mask( \
460		(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), (int)(P), (__mmask16)-1, \
461		(int)(R)))
462
463		#define _mm256_mask_cmp_round_ph_mask(U, A, B, P, R) \
464		((__mmask16)__builtin_ia32_vcmpph256_round_mask( \
465		(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), (int)(P), (__mmask16)(U), \
466		(int)(R)))
467
468		#define _mm256_cmp_round_ps_mask(A, B, P, R) \
469		((__mmask8)__builtin_ia32_vcmpps256_round_mask( \
470		(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), (int)(P), (__mmask8)-1, \
471		(int)(R)))
472
473		#define _mm256_mask_cmp_round_ps_mask(U, A, B, P, R) \
474		((__mmask8)__builtin_ia32_vcmpps256_round_mask( \
475		(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), (int)(P), (__mmask8)(U), \
476		(int)(R)))
477
478		#define _mm256_cvt_roundepi32_ph(A, R) \
479		((__m128h)__builtin_ia32_vcvtdq2ph256_round_mask( \
480		(__v8si)(A), (__v8hf)_mm_undefined_ph(), (__mmask8)(-1), (int)(R)))
481
482		#define _mm256_mask_cvt_roundepi32_ph(W, U, A, R) \
483		((__m128h)__builtin_ia32_vcvtdq2ph256_round_mask((__v8si)(A), (__v8hf)(W), \
484		(__mmask8)(U), (int)(R)))
485
486		#define _mm256_maskz_cvt_roundepi32_ph(U, A, R) \
487		((__m128h)__builtin_ia32_vcvtdq2ph256_round_mask( \
488		(__v8si)(A), (__v8hf)_mm_setzero_ph(), (__mmask8)(U), (int)(R)))
489
490		#define _mm256_cvt_roundepi32_ps(A, R) \
491		((__m256)__builtin_ia32_vcvtdq2ps256_round_mask((__v8si)(__m256i)(A), \
492		(__v8sf)_mm256_setzero_ps(), \
493		(__mmask8)-1, (int)(R)))
494
495		#define _mm256_mask_cvt_roundepi32_ps(W, U, A, R) \
496		((__m256)__builtin_ia32_vcvtdq2ps256_round_mask( \
497		(__v8si)(__m256i)(A), (__v8sf)(__m256)(W), (__mmask8)(U), (int)(R)))
498
499		#define _mm256_maskz_cvt_roundepi32_ps(U, A, R) \
500		((__m256)__builtin_ia32_vcvtdq2ps256_round_mask((__v8si)(__m256i)(A), \
501		(__v8sf)_mm256_setzero_ps(), \
502		(__mmask8)(U), (int)(R)))
503
504		#define _mm256_cvt_roundpd_epi32(A, R) \
505		((__m128i)__builtin_ia32_vcvtpd2dq256_round_mask( \
506		(__v4df)(__m256d)(A), (__v4si)_mm_setzero_si128(), (__mmask8)-1, \
507		(int)(R)))
508
509		#define _mm256_mask_cvt_roundpd_epi32(W, U, A, R) \
510		((__m128i)__builtin_ia32_vcvtpd2dq256_round_mask( \
511		(__v4df)(__m256d)(A), (__v4si)(__m128i)(W), (__mmask8)(U), (int)(R)))
512
513		#define _mm256_maskz_cvt_roundpd_epi32(U, A, R) \
514		((__m128i)__builtin_ia32_vcvtpd2dq256_round_mask( \
515		(__v4df)(__m256d)(A), (__v4si)_mm_setzero_si128(), (__mmask8)(U), \
516		(int)(R)))
517
518		#define _mm256_cvt_roundpd_ph(A, R) \
519		((__m128h)__builtin_ia32_vcvtpd2ph256_round_mask( \
520		(__v4df)(A), (__v8hf)_mm_undefined_ph(), (__mmask8)(-1), (int)(R)))
521
522		#define _mm256_mask_cvt_roundpd_ph(W, U, A, R) \
523		((__m128h)__builtin_ia32_vcvtpd2ph256_round_mask((__v4df)(A), (__v8hf)(W), \
524		(__mmask8)(U), (int)(R)))
525
526		#define _mm256_maskz_cvt_roundpd_ph(U, A, R) \
527		((__m128h)__builtin_ia32_vcvtpd2ph256_round_mask( \
528		(__v4df)(A), (__v8hf)_mm_setzero_ph(), (__mmask8)(U), (int)(R)))
529
530		#define _mm256_cvt_roundpd_ps(A, R) \
531		((__m128)__builtin_ia32_vcvtpd2ps256_round_mask( \
532		(__v4df)(__m256d)(A), (__v4sf)_mm_setzero_ps(), (__mmask8)-1, (int)(R)))
533
534		#define _mm256_mask_cvt_roundpd_ps(W, U, A, R) \
535		((__m128)__builtin_ia32_vcvtpd2ps256_round_mask( \
536		(__v4df)(__m256d)(A), (__v4sf)(__m128)(W), (__mmask8)(U), (int)(R)))
537
538		#define _mm256_maskz_cvt_roundpd_ps(U, A, R) \
539		((__m128)__builtin_ia32_vcvtpd2ps256_round_mask((__v4df)(__m256d)(A), \
540		(__v4sf)_mm_setzero_ps(), \
541		(__mmask8)(U), (int)(R)))
542
543		#define _mm256_cvt_roundpd_epi64(A, R) \
544		((__m256i)__builtin_ia32_vcvtpd2qq256_round_mask( \
545		(__v4df)(__m256d)(A), (__v4di)_mm256_setzero_si256(), (__mmask8)-1, \
546		(int)(R)))
547
548		#define _mm256_mask_cvt_roundpd_epi64(W, U, A, R) \
549		((__m256i)__builtin_ia32_vcvtpd2qq256_round_mask( \
550		(__v4df)(__m256d)(A), (__v4di)(__m256i)(W), (__mmask8)(U), (int)(R)))
551
552		#define _mm256_maskz_cvt_roundpd_epi64(U, A, R) \
553		((__m256i)__builtin_ia32_vcvtpd2qq256_round_mask( \
554		(__v4df)(__m256d)(A), (__v4di)_mm256_setzero_si256(), (__mmask8)(U), \
555		(int)(R)))
556
557		#define _mm256_cvt_roundpd_epu32(A, R) \
558		((__m128i)__builtin_ia32_vcvtpd2udq256_round_mask( \
559		(__v4df)(__m256d)(A), (__v4su)_mm_setzero_si128(), (__mmask8)-1, \
560		(int)(R)))
561
562		#define _mm256_mask_cvt_roundpd_epu32(W, U, A, R) \
563		((__m128i)__builtin_ia32_vcvtpd2udq256_round_mask( \
564		(__v4df)(__m256d)(A), (__v4su)(__m128i)(W), (__mmask8)(U), (int)(R)))
565
566		#define _mm256_maskz_cvt_roundpd_epu32(U, A, R) \
567		((__m128i)__builtin_ia32_vcvtpd2udq256_round_mask( \
568		(__v4df)(__m256d)(A), (__v4su)_mm_setzero_si128(), (__mmask8)(U), \
569		(int)(R)))
570
571		#define _mm256_cvt_roundpd_epu64(A, R) \
572		((__m256i)__builtin_ia32_vcvtpd2uqq256_round_mask( \
573		(__v4df)(__m256d)(A), (__v4du)_mm256_setzero_si256(), (__mmask8)-1, \
574		(int)(R)))
575
576		#define _mm256_mask_cvt_roundpd_epu64(W, U, A, R) \
577		((__m256i)__builtin_ia32_vcvtpd2uqq256_round_mask( \
578		(__v4df)(__m256d)(A), (__v4du)(__m256i)(W), (__mmask8)(U), (int)(R)))
579
580		#define _mm256_maskz_cvt_roundpd_epu64(U, A, R) \
581		((__m256i)__builtin_ia32_vcvtpd2uqq256_round_mask( \
582		(__v4df)(__m256d)(A), (__v4du)_mm256_setzero_si256(), (__mmask8)(U), \
583		(int)(R)))
584
585		#define _mm256_cvt_roundph_epi32(A, R) \
586		((__m256i)__builtin_ia32_vcvtph2dq256_round_mask( \
587		(__v8hf)(A), (__v8si)_mm256_undefined_si256(), (__mmask8)(-1), \
588		(int)(R)))
589
590		#define _mm256_mask_cvt_roundph_epi32(W, U, A, R) \
591		((__m256i)__builtin_ia32_vcvtph2dq256_round_mask((__v8hf)(A), (__v8si)(W), \
592		(__mmask8)(U), (int)(R)))
593
594		#define _mm256_maskz_cvt_roundph_epi32(U, A, R) \
595		((__m256i)__builtin_ia32_vcvtph2dq256_round_mask( \
596		(__v8hf)(A), (__v8si)_mm256_setzero_si256(), (__mmask8)(U), (int)(R)))
597
598		#define _mm256_cvt_roundph_pd(A, R) \
599		((__m256d)__builtin_ia32_vcvtph2pd256_round_mask( \
600		(__v8hf)(A), (__v4df)_mm256_undefined_pd(), (__mmask8)(-1), (int)(R)))
601
602		#define _mm256_mask_cvt_roundph_pd(W, U, A, R) \
603		((__m256d)__builtin_ia32_vcvtph2pd256_round_mask((__v8hf)(A), (__v4df)(W), \
604		(__mmask8)(U), (int)(R)))
605
606		#define _mm256_maskz_cvt_roundph_pd(U, A, R) \
607		((__m256d)__builtin_ia32_vcvtph2pd256_round_mask( \
608		(__v8hf)(A), (__v4df)_mm256_setzero_pd(), (__mmask8)(U), (int)(R)))
609
610		#define _mm256_cvtx_roundph_ps(A, R) \
611		((__m256)__builtin_ia32_vcvtph2psx256_round_mask( \
612		(__v8hf)(A), (__v8sf)_mm256_undefined_ps(), (__mmask8)(-1), (int)(R)))
613
614		#define _mm256_mask_cvtx_roundph_ps(W, U, A, R) \
615		((__m256)__builtin_ia32_vcvtph2psx256_round_mask((__v8hf)(A), (__v8sf)(W), \
616		(__mmask8)(U), (int)(R)))
617
618		#define _mm256_maskz_cvtx_roundph_ps(U, A, R) \
619		((__m256)__builtin_ia32_vcvtph2psx256_round_mask( \
620		(__v8hf)(A), (__v8sf)_mm256_setzero_ps(), (__mmask8)(U), (int)(R)))
621
622		#define _mm256_cvt_roundph_epi64(A, R) \
623		((__m256i)__builtin_ia32_vcvtph2qq256_round_mask( \
624		(__v8hf)(A), (__v4di)_mm256_undefined_si256(), (__mmask8)(-1), \
625		(int)(R)))
626
627		#define _mm256_mask_cvt_roundph_epi64(W, U, A, R) \
628		((__m256i)__builtin_ia32_vcvtph2qq256_round_mask((__v8hf)(A), (__v4di)(W), \
629		(__mmask8)(U), (int)(R)))
630
631		#define _mm256_maskz_cvt_roundph_epi64(U, A, R) \
632		((__m256i)__builtin_ia32_vcvtph2qq256_round_mask( \
633		(__v8hf)(A), (__v4di)_mm256_setzero_si256(), (__mmask8)(U), (int)(R)))
634
635		#define _mm256_cvt_roundph_epu32(A, R) \
636		((__m256i)__builtin_ia32_vcvtph2udq256_round_mask( \
637		(__v8hf)(A), (__v8su)_mm256_undefined_si256(), (__mmask8)(-1), \
638		(int)(R)))
639
640		#define _mm256_mask_cvt_roundph_epu32(W, U, A, R) \
641		((__m256i)__builtin_ia32_vcvtph2udq256_round_mask((__v8hf)(A), (__v8su)(W), \
642		(__mmask8)(U), (int)(R)))
643
644		#define _mm256_maskz_cvt_roundph_epu32(U, A, R) \
645		((__m256i)__builtin_ia32_vcvtph2udq256_round_mask( \
646		(__v8hf)(A), (__v8su)_mm256_setzero_si256(), (__mmask8)(U), (int)(R)))
647
648		#define _mm256_cvt_roundph_epu64(A, R) \
649		((__m256i)__builtin_ia32_vcvtph2uqq256_round_mask( \
650		(__v8hf)(A), (__v4du)_mm256_undefined_si256(), (__mmask8)(-1), \
651		(int)(R)))
652
653		#define _mm256_mask_cvt_roundph_epu64(W, U, A, R) \
654		((__m256i)__builtin_ia32_vcvtph2uqq256_round_mask((__v8hf)(A), (__v4du)(W), \
655		(__mmask8)(U), (int)(R)))
656
657		#define _mm256_maskz_cvt_roundph_epu64(U, A, R) \
658		((__m256i)__builtin_ia32_vcvtph2uqq256_round_mask( \
659		(__v8hf)(A), (__v4du)_mm256_setzero_si256(), (__mmask8)(U), (int)(R)))
660
661		#define _mm256_cvt_roundph_epu16(A, R) \
662		((__m256i)__builtin_ia32_vcvtph2uw256_round_mask( \
663		(__v16hf)(A), (__v16hu)_mm256_undefined_si256(), (__mmask16)(-1), \
664		(int)(R)))
665
666		#define _mm256_mask_cvt_roundph_epu16(W, U, A, R) \
667		((__m256i)__builtin_ia32_vcvtph2uw256_round_mask((__v16hf)(A), (__v16hu)(W), \
668		(__mmask16)(U), (int)(R)))
669
670		#define _mm256_maskz_cvt_roundph_epu16(U, A, R) \
671		((__m256i)__builtin_ia32_vcvtph2uw256_round_mask( \
672		(__v16hf)(A), (__v16hu)_mm256_setzero_si256(), (__mmask16)(U), \
673		(int)(R)))
674
675		#define _mm256_cvt_roundph_epi16(A, R) \
676		((__m256i)__builtin_ia32_vcvtph2w256_round_mask( \
677		(__v16hf)(A), (__v16hi)_mm256_undefined_si256(), (__mmask16)(-1), \
678		(int)(R)))
679
680		#define _mm256_mask_cvt_roundph_epi16(W, U, A, R) \
681		((__m256i)__builtin_ia32_vcvtph2w256_round_mask((__v16hf)(A), (__v16hi)(W), \
682		(__mmask16)(U), (int)(R)))
683
684		#define _mm256_maskz_cvt_roundph_epi16(U, A, R) \
685		((__m256i)__builtin_ia32_vcvtph2w256_round_mask( \
686		(__v16hf)(A), (__v16hi)_mm256_setzero_si256(), (__mmask16)(U), \
687		(int)(R)))
688
689		#define _mm256_cvt_roundps_epi32(A, R) \
690		((__m256i)__builtin_ia32_vcvtps2dq256_round_mask( \
691		(__v8sf)(__m256)(A), (__v8si)_mm256_setzero_si256(), (__mmask8)-1, \
692		(int)(R)))
693
694		#define _mm256_mask_cvt_roundps_epi32(W, U, A, R) \
695		((__m256i)__builtin_ia32_vcvtps2dq256_round_mask( \
696		(__v8sf)(__m256)(A), (__v8si)(__m256i)(W), (__mmask8)(U), (int)(R)))
697
698		#define _mm256_maskz_cvt_roundps_epi32(U, A, R) \
699		((__m256i)__builtin_ia32_vcvtps2dq256_round_mask( \
700		(__v8sf)(__m256)(A), (__v8si)_mm256_setzero_si256(), (__mmask8)(U), \
701		(int)(R)))
702
703		#define _mm256_cvt_roundps_pd(A, R) \
704		((__m256d)__builtin_ia32_vcvtps2pd256_round_mask( \
705		(__v4sf)(__m128)(A), (__v4df)_mm256_undefined_pd(), (__mmask8)-1, \
706		(int)(R)))
707
708		#define _mm256_mask_cvt_roundps_pd(W, U, A, R) \
709		((__m256d)__builtin_ia32_vcvtps2pd256_round_mask( \
710		(__v4sf)(__m128)(A), (__v4df)(__m256d)(W), (__mmask8)(U), (int)(R)))
711
712		#define _mm256_maskz_cvt_roundps_pd(U, A, R) \
713		((__m256d)__builtin_ia32_vcvtps2pd256_round_mask( \
714		(__v4sf)(__m128)(A), (__v4df)_mm256_setzero_pd(), (__mmask8)(U), \
715		(int)(R)))
716
717		#define _mm256_cvt_roundps_ph(A, I) \
718		((__m128i)__builtin_ia32_vcvtps2ph256_mask((__v8sf)(__m256)(A), (int)(I), \
719		(__v8hi)_mm_undefined_si128(), \
720		(__mmask8)-1))
721
722		/* FIXME: We may use these way in future.
723		#define _mm256_cvt_roundps_ph(A, I) \
724		((__m128i)__builtin_ia32_vcvtps2ph256_round_mask( \
725		(__v8sf)(__m256)(A), (int)(I), (__v8hi)_mm_undefined_si128(), \
726		(__mmask8)-1))
727		#define _mm256_mask_cvt_roundps_ph(U, W, A, I) \
728		((__m128i)__builtin_ia32_vcvtps2ph256_round_mask( \
729		(__v8sf)(__m256)(A), (int)(I), (__v8hi)(__m128i)(U), (__mmask8)(W)))
730		#define _mm256_maskz_cvt_roundps_ph(W, A, I) \
731		((__m128i)__builtin_ia32_vcvtps2ph256_round_mask( \
732		(__v8sf)(__m256)(A), (int)(I), (__v8hi)_mm_setzero_si128(), \
733		(__mmask8)(W))) */
734
735		#define _mm256_cvtx_roundps_ph(A, R) \
736		((__m128h)__builtin_ia32_vcvtps2phx256_round_mask( \
737		(__v8sf)(A), (__v8hf)_mm_undefined_ph(), (__mmask8)(-1), (int)(R)))
738
739		#define _mm256_mask_cvtx_roundps_ph(W, U, A, R) \
740		((__m128h)__builtin_ia32_vcvtps2phx256_round_mask((__v8sf)(A), (__v8hf)(W), \
741		(__mmask8)(U), (int)(R)))
742
743		#define _mm256_maskz_cvtx_roundps_ph(U, A, R) \
744		((__m128h)__builtin_ia32_vcvtps2phx256_round_mask( \
745		(__v8sf)(A), (__v8hf)_mm_setzero_ph(), (__mmask8)(U), (int)(R)))
746
747		#define _mm256_cvt_roundps_epi64(A, R) \
748		((__m256i)__builtin_ia32_vcvtps2qq256_round_mask( \
749		(__v4sf)(__m128)(A), (__v4di)_mm256_setzero_si256(), (__mmask8)-1, \
750		(int)(R)))
751
752		#define _mm256_mask_cvt_roundps_epi64(W, U, A, R) \
753		((__m256i)__builtin_ia32_vcvtps2qq256_round_mask( \
754		(__v4sf)(__m128)(A), (__v4di)(__m256i)(W), (__mmask8)(U), (int)(R)))
755
756		#define _mm256_maskz_cvt_roundps_epi64(U, A, R) \
757		((__m256i)__builtin_ia32_vcvtps2qq256_round_mask( \
758		(__v4sf)(__m128)(A), (__v4di)_mm256_setzero_si256(), (__mmask8)(U), \
759		(int)(R)))
760
761		#define _mm256_cvt_roundps_epu32(A, R) \
762		((__m256i)__builtin_ia32_vcvtps2udq256_round_mask( \
763		(__v8sf)(__m256)(A), (__v8su)_mm256_setzero_si256(), (__mmask8)-1, \
764		(int)(R)))
765
766		#define _mm256_mask_cvt_roundps_epu32(W, U, A, R) \
767		((__m256i)__builtin_ia32_vcvtps2udq256_round_mask( \
768		(__v8sf)(__m256)(A), (__v8su)(__m256i)(W), (__mmask8)(U), (int)(R)))
769
770		#define _mm256_maskz_cvt_roundps_epu32(U, A, R) \
771		((__m256i)__builtin_ia32_vcvtps2udq256_round_mask( \
772		(__v8sf)(__m256)(A), (__v8su)_mm256_setzero_si256(), (__mmask8)(U), \
773		(int)(R)))
774
775		#define _mm256_cvt_roundps_epu64(A, R) \
776		((__m256i)__builtin_ia32_vcvtps2uqq256_round_mask( \
777		(__v4sf)(__m128)(A), (__v4du)_mm256_setzero_si256(), (__mmask8)-1, \
778		(int)(R)))
779
780		#define _mm256_mask_cvt_roundps_epu64(W, U, A, R) \
781		((__m256i)__builtin_ia32_vcvtps2uqq256_round_mask( \
782		(__v4sf)(__m128)(A), (__v4du)(__m256i)(W), (__mmask8)(U), (int)(R)))
783
784		#define _mm256_maskz_cvt_roundps_epu64(U, A, R) \
785		((__m256i)__builtin_ia32_vcvtps2uqq256_round_mask( \
786		(__v4sf)(__m128)(A), (__v4du)_mm256_setzero_si256(), (__mmask8)(U), \
787		(int)(R)))
788
789		#define _mm256_cvt_roundepi64_pd(A, R) \
790		((__m256d)__builtin_ia32_vcvtqq2pd256_round_mask( \
791		(__v4di)(__m256i)(A), (__v4df)_mm256_setzero_pd(), (__mmask8)-1, \
792		(int)(R)))
793
794		#define _mm256_mask_cvt_roundepi64_pd(W, U, A, R) \
795		((__m256d)__builtin_ia32_vcvtqq2pd256_round_mask( \
796		(__v4di)(__m256i)(A), (__v4df)(__m256d)(W), (__mmask8)(U), (int)(R)))
797
798		#define _mm256_maskz_cvt_roundepi64_pd(U, A, R) \
799		((__m256d)__builtin_ia32_vcvtqq2pd256_round_mask( \
800		(__v4di)(__m256i)(A), (__v4df)_mm256_setzero_pd(), (__mmask8)(U), \
801		(int)(R)))
802
803		#define _mm256_cvt_roundepi64_ph(A, R) \
804		((__m128h)__builtin_ia32_vcvtqq2ph256_round_mask( \
805		(__v4di)(A), (__v8hf)_mm_undefined_ph(), (__mmask8)(-1), (int)(R)))
806
807		#define _mm256_mask_cvt_roundepi64_ph(W, U, A, R) \
808		((__m128h)__builtin_ia32_vcvtqq2ph256_round_mask((__v4di)(A), (__v8hf)(W), \
809		(__mmask8)(U), (int)(R)))
810
811		#define _mm256_maskz_cvt_roundepi64_ph(U, A, R) \
812		((__m128h)__builtin_ia32_vcvtqq2ph256_round_mask( \
813		(__v4di)(A), (__v8hf)_mm_setzero_ph(), (__mmask8)(U), (int)(R)))
814
815		#define _mm256_cvt_roundepi64_ps(A, R) \
816		((__m128)__builtin_ia32_vcvtqq2ps256_round_mask( \
817		(__v4di)(__m256i)(A), (__v4sf)_mm_setzero_ps(), (__mmask8)-1, (int)(R)))
818
819		#define _mm256_mask_cvt_roundepi64_ps(W, U, A, R) \
820		((__m128)__builtin_ia32_vcvtqq2ps256_round_mask( \
821		(__v4di)(__m256i)(A), (__v4sf)(__m128)(W), (__mmask8)(U), (int)(R)))
822
823		#define _mm256_maskz_cvt_roundepi64_ps(U, A, R) \
824		((__m128)__builtin_ia32_vcvtqq2ps256_round_mask((__v4di)(__m256i)(A), \
825		(__v4sf)_mm_setzero_ps(), \
826		(__mmask8)(U), (int)(R)))
827
828		#define _mm256_cvtt_roundpd_epi32(A, R) \
829		((__m128i)__builtin_ia32_vcvttpd2dq256_round_mask( \
830		(__v4df)(__m256d)(A), (__v4si)_mm_setzero_si128(), (__mmask8)-1, \
831		(int)(R)))
832
833		#define _mm256_mask_cvtt_roundpd_epi32(W, U, A, R) \
834		((__m128i)__builtin_ia32_vcvttpd2dq256_round_mask( \
835		(__v4df)(__m256d)(A), (__v4si)(__m128i)(W), (__mmask8)(U), (int)(R)))
836
837		#define _mm256_maskz_cvtt_roundpd_epi32(U, A, R) \
838		((__m128i)__builtin_ia32_vcvttpd2dq256_round_mask( \
839		(__v4df)(__m256d)(A), (__v4si)_mm_setzero_si128(), (__mmask8)(U), \
840		(int)(R)))
841
842		#define _mm256_cvtt_roundpd_epi64(A, R) \
843		((__m256i)__builtin_ia32_vcvttpd2qq256_round_mask( \
844		(__v4df)(__m256d)(A), (__v4di)_mm256_setzero_si256(), (__mmask8)-1, \
845		(int)(R)))
846
847		#define _mm256_mask_cvtt_roundpd_epi64(W, U, A, R) \
848		((__m256i)__builtin_ia32_vcvttpd2qq256_round_mask( \
849		(__v4df)(__m256d)(A), (__v4di)(__m256i)(W), (__mmask8)(U), (int)(R)))
850
851		#define _mm256_maskz_cvtt_roundpd_epi64(U, A, R) \
852		((__m256i)__builtin_ia32_vcvttpd2qq256_round_mask( \
853		(__v4df)(__m256d)(A), (__v4di)_mm256_setzero_si256(), (__mmask8)(U), \
854		(int)(R)))
855
856		#define _mm256_cvtt_roundpd_epu32(A, R) \
857		((__m128i)__builtin_ia32_vcvttpd2udq256_round_mask( \
858		(__v4df)(__m256d)(A), (__v4su)_mm_setzero_si128(), (__mmask8)-1, \
859		(int)(R)))
860
861		#define _mm256_mask_cvtt_roundpd_epu32(W, U, A, R) \
862		((__m128i)__builtin_ia32_vcvttpd2udq256_round_mask( \
863		(__v4df)(__m256d)(A), (__v4su)(__m128i)(W), (__mmask8)(U), (int)(R)))
864
865		#define _mm256_maskz_cvtt_roundpd_epu32(U, A, R) \
866		((__m128i)__builtin_ia32_vcvttpd2udq256_round_mask( \
867		(__v4df)(__m256d)(A), (__v4su)_mm_setzero_si128(), (__mmask8)(U), \
868		(int)(R)))
869
870		#define _mm256_cvtt_roundpd_epu64(A, R) \
871		((__m256i)__builtin_ia32_vcvttpd2uqq256_round_mask( \
872		(__v4df)(__m256d)(A), (__v4du)_mm256_setzero_si256(), (__mmask8)-1, \
873		(int)(R)))
874
875		#define _mm256_mask_cvtt_roundpd_epu64(W, U, A, R) \
876		((__m256i)__builtin_ia32_vcvttpd2uqq256_round_mask( \
877		(__v4df)(__m256d)(A), (__v4du)(__m256i)(W), (__mmask8)(U), (int)(R)))
878
879		#define _mm256_maskz_cvtt_roundpd_epu64(U, A, R) \
880		((__m256i)__builtin_ia32_vcvttpd2uqq256_round_mask( \
881		(__v4df)(__m256d)(A), (__v4du)_mm256_setzero_si256(), (__mmask8)(U), \
882		(int)(R)))
883
884		#define _mm256_cvtt_roundph_epi32(A, R) \
885		((__m256i)__builtin_ia32_vcvttph2dq256_round_mask( \
886		(__v8hf)(A), (__v8si)_mm256_undefined_si256(), (__mmask8)(-1), \
887		(int)(R)))
888
889		#define _mm256_mask_cvtt_roundph_epi32(W, U, A, R) \
890		((__m256i)__builtin_ia32_vcvttph2dq256_round_mask((__v8hf)(A), (__v8si)(W), \
891		(__mmask8)(U), (int)(R)))
892
893		#define _mm256_maskz_cvtt_roundph_epi32(U, A, R) \
894		((__m256i)__builtin_ia32_vcvttph2dq256_round_mask( \
895		(__v8hf)(A), (__v8si)_mm256_setzero_si256(), (__mmask8)(U), (int)(R)))
896
897		#define _mm256_cvtt_roundph_epi64(A, R) \
898		((__m256i)__builtin_ia32_vcvttph2qq256_round_mask( \
899		(__v8hf)(A), (__v4di)_mm256_undefined_si256(), (__mmask8)(-1), \
900		(int)(R)))
901
902		#define _mm256_mask_cvtt_roundph_epi64(W, U, A, R) \
903		((__m256i)__builtin_ia32_vcvttph2qq256_round_mask((__v8hf)(A), (__v4di)(W), \
904		(__mmask8)(U), (int)(R)))
905
906		#define _mm256_maskz_cvtt_roundph_epi64(U, A, R) \
907		((__m256i)__builtin_ia32_vcvttph2qq256_round_mask( \
908		(__v8hf)(A), (__v4di)_mm256_setzero_si256(), (__mmask8)(U), (int)(R)))
909
910		#define _mm256_cvtt_roundph_epu32(A, R) \
911		((__m256i)__builtin_ia32_vcvttph2udq256_round_mask( \
912		(__v8hf)(A), (__v8su)_mm256_undefined_si256(), (__mmask8)(-1), \
913		(int)(R)))
914
915		#define _mm256_mask_cvtt_roundph_epu32(W, U, A, R) \
916		((__m256i)__builtin_ia32_vcvttph2udq256_round_mask((__v8hf)(A), (__v8su)(W), \
917		(__mmask8)(U), (int)(R)))
918
919		#define _mm256_maskz_cvtt_roundph_epu32(U, A, R) \
920		((__m256i)__builtin_ia32_vcvttph2udq256_round_mask( \
921		(__v8hf)(A), (__v8su)_mm256_setzero_si256(), (__mmask8)(U), (int)(R)))
922
923		#define _mm256_cvtt_roundph_epu64(A, R) \
924		((__m256i)__builtin_ia32_vcvttph2uqq256_round_mask( \
925		(__v8hf)(A), (__v4du)_mm256_undefined_si256(), (__mmask8)(-1), \
926		(int)(R)))
927
928		#define _mm256_mask_cvtt_roundph_epu64(W, U, A, R) \
929		((__m256i)__builtin_ia32_vcvttph2uqq256_round_mask((__v8hf)(A), (__v4du)(W), \
930		(__mmask8)(U), (int)(R)))
931
932		#define _mm256_maskz_cvtt_roundph_epu64(U, A, R) \
933		((__m256i)__builtin_ia32_vcvttph2uqq256_round_mask( \
934		(__v8hf)(A), (__v4du)_mm256_setzero_si256(), (__mmask8)(U), (int)(R)))
935
936		#define _mm256_cvtt_roundph_epu16(A, R) \
937		((__m256i)__builtin_ia32_vcvttph2uw256_round_mask( \
938		(__v16hf)(A), (__v16hu)_mm256_undefined_si256(), (__mmask16)(-1), \
939		(int)(R)))
940
941		#define _mm256_mask_cvtt_roundph_epu16(W, U, A, R) \
942		((__m256i)__builtin_ia32_vcvttph2uw256_round_mask( \
943		(__v16hf)(A), (__v16hu)(W), (__mmask16)(U), (int)(R)))
944
945		#define _mm256_maskz_cvtt_roundph_epu16(U, A, R) \
946		((__m256i)__builtin_ia32_vcvttph2uw256_round_mask( \
947		(__v16hf)(A), (__v16hu)_mm256_setzero_si256(), (__mmask16)(U), \
948		(int)(R)))
949
950		#define _mm256_cvtt_roundph_epi16(A, R) \
951		((__m256i)__builtin_ia32_vcvttph2w256_round_mask( \
952		(__v16hf)(A), (__v16hi)_mm256_undefined_si256(), (__mmask16)(-1), \
953		(int)(R)))
954
955		#define _mm256_mask_cvtt_roundph_epi16(W, U, A, R) \
956		((__m256i)__builtin_ia32_vcvttph2w256_round_mask((__v16hf)(A), (__v16hi)(W), \
957		(__mmask16)(U), (int)(R)))
958
959		#define _mm256_maskz_cvtt_roundph_epi16(U, A, R) \
960		((__m256i)__builtin_ia32_vcvttph2w256_round_mask( \
961		(__v16hf)(A), (__v16hi)_mm256_setzero_si256(), (__mmask16)(U), \
962		(int)(R)))
963
964		#define _mm256_cvtt_roundps_epi32(A, R) \
965		((__m256i)__builtin_ia32_vcvttps2dq256_round_mask( \
966		(__v8sf)(__m256)(A), (__v8si)_mm256_setzero_si256(), (__mmask8)-1, \
967		(int)(R)))
968
969		#define _mm256_mask_cvtt_roundps_epi32(W, U, A, R) \
970		((__m256i)__builtin_ia32_vcvttps2dq256_round_mask( \
971		(__v8sf)(__m256)(A), (__v8si)(__m256i)(W), (__mmask8)(U), (int)(R)))
972
973		#define _mm256_maskz_cvtt_roundps_epi32(U, A, R) \
974		((__m256i)__builtin_ia32_vcvttps2dq256_round_mask( \
975		(__v8sf)(__m256)(A), (__v8si)_mm256_setzero_si256(), (__mmask8)(U), \
976		(int)(R)))
977
978		#define _mm256_cvtt_roundps_epi64(A, R) \
979		((__m256i)__builtin_ia32_vcvttps2qq256_round_mask( \
980		(__v4sf)(__m128)(A), (__v4di)_mm256_setzero_si256(), (__mmask8)-1, \
981		(int)(R)))
982
983		#define _mm256_mask_cvtt_roundps_epi64(W, U, A, R) \
984		((__m256i)__builtin_ia32_vcvttps2qq256_round_mask( \
985		(__v4sf)(__m128)(A), (__v4di)(__m256i)(W), (__mmask8)(U), (int)(R)))
986
987		#define _mm256_maskz_cvtt_roundps_epi64(U, A, R) \
988		((__m256i)__builtin_ia32_vcvttps2qq256_round_mask( \
989		(__v4sf)(__m128)(A), (__v4di)_mm256_setzero_si256(), (__mmask8)(U), \
990		(int)(R)))
991
992		#define _mm256_cvtt_roundps_epu32(A, R) \
993		((__m256i)__builtin_ia32_vcvttps2udq256_round_mask( \
994		(__v8sf)(__m256)(A), (__v8su)_mm256_setzero_si256(), (__mmask8)-1, \
995		(int)(R)))
996
997		#define _mm256_mask_cvtt_roundps_epu32(W, U, A, R) \
998		((__m256i)__builtin_ia32_vcvttps2udq256_round_mask( \
999		(__v8sf)(__m256)(A), (__v8su)(__m256i)(W), (__mmask8)(U), (int)(R)))
1000
1001		#define _mm256_maskz_cvtt_roundps_epu32(U, A, R) \
1002		((__m256i)__builtin_ia32_vcvttps2udq256_round_mask( \
1003		(__v8sf)(__m256)(A), (__v8su)_mm256_setzero_si256(), (__mmask8)(U), \
1004		(int)(R)))
1005
1006		#define _mm256_cvtt_roundps_epu64(A, R) \
1007		((__m256i)__builtin_ia32_vcvttps2uqq256_round_mask( \
1008		(__v4sf)(__m128)(A), (__v4du)_mm256_setzero_si256(), (__mmask8)-1, \
1009		(int)(R)))
1010
1011		#define _mm256_mask_cvtt_roundps_epu64(W, U, A, R) \
1012		((__m256i)__builtin_ia32_vcvttps2uqq256_round_mask( \
1013		(__v4sf)(__m128)(A), (__v4du)(__m256i)(W), (__mmask8)(U), (int)(R)))
1014
1015		#define _mm256_maskz_cvtt_roundps_epu64(U, A, R) \
1016		((__m256i)__builtin_ia32_vcvttps2uqq256_round_mask( \
1017		(__v4sf)(__m128)(A), (__v4du)_mm256_setzero_si256(), (__mmask8)(U), \
1018		(int)(R)))
1019
1020		#define _mm256_cvt_roundepu32_ph(A, R) \
1021		((__m128h)__builtin_ia32_vcvtudq2ph256_round_mask( \
1022		(__v8su)(A), (__v8hf)_mm_undefined_ph(), (__mmask8)(-1), (int)(R)))
1023
1024		#define _mm256_mask_cvt_roundepu32_ph(W, U, A, R) \
1025		((__m128h)__builtin_ia32_vcvtudq2ph256_round_mask((__v8su)(A), (__v8hf)(W), \
1026		(__mmask8)(U), (int)(R)))
1027
1028		#define _mm256_maskz_cvt_roundepu32_ph(U, A, R) \
1029		((__m128h)__builtin_ia32_vcvtudq2ph256_round_mask( \
1030		(__v8su)(A), (__v8hf)_mm_setzero_ph(), (__mmask8)(U), (int)(R)))
1031
1032		#define _mm256_cvt_roundepu32_ps(A, R) \
1033		((__m256)__builtin_ia32_vcvtudq2ps256_round_mask( \
1034		(__v8su)(__m256i)(A), (__v8sf)_mm256_setzero_ps(), (__mmask8)-1, \
1035		(int)(R)))
1036
1037		#define _mm256_mask_cvt_roundepu32_ps(W, U, A, R) \
1038		((__m256)__builtin_ia32_vcvtudq2ps256_round_mask( \
1039		(__v8su)(__m256i)(A), (__v8sf)(__m256)(W), (__mmask8)(U), (int)(R)))
1040
1041		#define _mm256_maskz_cvt_roundepu32_ps(U, A, R) \
1042		((__m256)__builtin_ia32_vcvtudq2ps256_round_mask( \
1043		(__v8su)(__m256i)(A), (__v8sf)_mm256_setzero_ps(), (__mmask8)(U), \
1044		(int)(R)))
1045
1046		#define _mm256_cvt_roundepu64_pd(A, R) \
1047		((__m256d)__builtin_ia32_vcvtuqq2pd256_round_mask( \
1048		(__v4du)(__m256i)(A), (__v4df)_mm256_setzero_pd(), (__mmask8)-1, \
1049		(int)(R)))
1050
1051		#define _mm256_mask_cvt_roundepu64_pd(W, U, A, R) \
1052		((__m256d)__builtin_ia32_vcvtuqq2pd256_round_mask( \
1053		(__v4du)(__m256i)(A), (__v4df)(__m256d)(W), (__mmask8)(U), (int)(R)))
1054
1055		#define _mm256_maskz_cvt_roundepu64_pd(U, A, R) \
1056		((__m256d)__builtin_ia32_vcvtuqq2pd256_round_mask( \
1057		(__v4du)(__m256i)(A), (__v4df)_mm256_setzero_pd(), (__mmask8)(U), \
1058		(int)(R)))
1059
1060		#define _mm256_cvt_roundepu64_ph(A, R) \
1061		((__m128h)__builtin_ia32_vcvtuqq2ph256_round_mask( \
1062		(__v4du)(A), (__v8hf)_mm_undefined_ph(), (__mmask8)(-1), (int)(R)))
1063
1064		#define _mm256_mask_cvt_roundepu64_ph(W, U, A, R) \
1065		((__m128h)__builtin_ia32_vcvtuqq2ph256_round_mask((__v4du)(A), (__v8hf)(W), \
1066		(__mmask8)(U), (int)(R)))
1067
1068		#define _mm256_maskz_cvt_roundepu64_ph(U, A, R) \
1069		((__m128h)__builtin_ia32_vcvtuqq2ph256_round_mask( \
1070		(__v4du)(A), (__v8hf)_mm_setzero_ph(), (__mmask8)(U), (int)(R)))
1071
1072		#define _mm256_cvt_roundepu64_ps(A, R) \
1073		((__m128)__builtin_ia32_vcvtuqq2ps256_round_mask( \
1074		(__v4du)(__m256i)(A), (__v4sf)_mm_setzero_ps(), (__mmask8)-1, (int)(R)))
1075
1076		#define _mm256_mask_cvt_roundepu64_ps(W, U, A, R) \
1077		((__m128)__builtin_ia32_vcvtuqq2ps256_round_mask( \
1078		(__v4du)(__m256i)(A), (__v4sf)(__m128)(W), (__mmask8)(U), (int)(R)))
1079
1080		#define _mm256_maskz_cvt_roundepu64_ps(U, A, R) \
1081		((__m128)__builtin_ia32_vcvtuqq2ps256_round_mask((__v4du)(__m256i)(A), \
1082		(__v4sf)_mm_setzero_ps(), \
1083		(__mmask8)(U), (int)(R)))
1084
1085		#define _mm256_cvt_roundepu16_ph(A, R) \
1086		((__m256h)__builtin_ia32_vcvtuw2ph256_round_mask( \
1087		(__v16hu)(A), (__v16hf)_mm256_undefined_ph(), (__mmask16)(-1), \
1088		(int)(R)))
1089
1090		#define _mm256_mask_cvt_roundepu16_ph(W, U, A, R) \
1091		((__m256h)__builtin_ia32_vcvtuw2ph256_round_mask((__v16hu)(A), (__v16hf)(W), \
1092		(__mmask16)(U), (int)(R)))
1093
1094		#define _mm256_maskz_cvt_roundepu16_ph(U, A, R) \
1095		((__m256h)__builtin_ia32_vcvtuw2ph256_round_mask( \
1096		(__v16hu)(A), (__v16hf)_mm256_setzero_ph(), (__mmask16)(U), (int)(R)))
1097
1098		#define _mm256_cvt_roundepi16_ph(A, R) \
1099		((__m256h)__builtin_ia32_vcvtw2ph256_round_mask( \
1100		(__v16hi)(A), (__v16hf)_mm256_undefined_ph(), (__mmask16)(-1), \
1101		(int)(R)))
1102
1103		#define _mm256_mask_cvt_roundepi16_ph(W, U, A, R) \
1104		((__m256h)__builtin_ia32_vcvtw2ph256_round_mask((__v16hi)(A), (__v16hf)(W), \
1105		(__mmask16)(U), (int)(R)))
1106
1107		#define _mm256_maskz_cvt_roundepi16_ph(U, A, R) \
1108		((__m256h)__builtin_ia32_vcvtw2ph256_round_mask( \
1109		(__v16hi)(A), (__v16hf)_mm256_setzero_ph(), (__mmask16)(U), (int)(R)))
1110
1111		#define _mm256_div_round_pd(A, B, R) \
1112		((__m256d)__builtin_ia32_vdivpd256_round((__v4df)(__m256d)(A), \
1113		(__v4df)(__m256d)(B), (int)(R)))
1114
1115		#define _mm256_mask_div_round_pd(W, U, A, B, R) \
1116		((__m256d)__builtin_ia32_selectpd_256( \
1117		(__mmask8)(U), (__v4df)_mm256_div_round_pd((A), (B), (R)), \
1118		(__v4df)(__m256d)(W)))
1119
1120		#define _mm256_maskz_div_round_pd(U, A, B, R) \
1121		((__m256d)__builtin_ia32_selectpd_256( \
1122		(__mmask8)(U), (__v4df)_mm256_div_round_pd((A), (B), (R)), \
1123		(__v4df)_mm256_setzero_pd()))
1124
1125		#define _mm256_div_round_ph(A, B, R) \
1126		((__m256h)__builtin_ia32_vdivph256_round((__v16hf)(__m256h)(A), \
1127		(__v16hf)(__m256h)(B), (int)(R)))
1128
1129		#define _mm256_mask_div_round_ph(W, U, A, B, R) \
1130		((__m256h)__builtin_ia32_selectph_256( \
1131		(__mmask16)(U), (__v16hf)_mm256_div_round_ph((A), (B), (R)), \
1132		(__v16hf)(__m256h)(W)))
1133
1134		#define _mm256_maskz_div_round_ph(U, A, B, R) \
1135		((__m256h)__builtin_ia32_selectph_256( \
1136		(__mmask16)(U), (__v16hf)_mm256_div_round_ph((A), (B), (R)), \
1137		(__v16hf)_mm256_setzero_ph()))
1138
1139		#define _mm256_div_round_ps(A, B, R) \
1140		((__m256)__builtin_ia32_vdivps256_round((__v8sf)(__m256)(A), \
1141		(__v8sf)(__m256)(B), (int)(R)))
1142
1143		#define _mm256_mask_div_round_ps(W, U, A, B, R) \
1144		((__m256)__builtin_ia32_selectps_256( \
1145		(__mmask8)(U), (__v8sf)_mm256_div_round_ps((A), (B), (R)), \
1146		(__v8sf)(__m256)(W)))
1147
1148		#define _mm256_maskz_div_round_ps(U, A, B, R) \
1149		((__m256)__builtin_ia32_selectps_256( \
1150		(__mmask8)(U), (__v8sf)_mm256_div_round_ps((A), (B), (R)), \
1151		(__v8sf)_mm256_setzero_ps()))
1152
1153		#define _mm256_fcmadd_round_pch(A, B, C, R) \
1154		((__m256h)__builtin_ia32_vfcmaddcph256_round_mask3( \
1155		(__v8sf)(__m256h)(A), (__v8sf)(__m256h)(B), (__v8sf)(__m256h)(C), \
1156		(__mmask8)-1, (int)(R)))
1157
1158		#define _mm256_mask_fcmadd_round_pch(A, U, B, C, R) \
1159		((__m256h)__builtin_ia32_vfcmaddcph256_round_mask( \
1160		(__v8sf)(__m256h)(A), (__v8sf)(__m256h)(B), (__v8sf)(__m256h)(C), \
1161		(__mmask8)(U), (int)(R)))
1162
1163		#define _mm256_mask3_fcmadd_round_pch(A, B, C, U, R) \
1164		((__m256h)__builtin_ia32_vfcmaddcph256_round_mask3( \
1165		(__v8sf)(__m256h)(A), (__v8sf)(__m256h)(B), (__v8sf)(__m256h)(C), \
1166		(__mmask8)(U), (int)(R)))
1167
1168		#define _mm256_maskz_fcmadd_round_pch(U, A, B, C, R) \
1169		((__m256h)__builtin_ia32_vfcmaddcph256_round_maskz( \
1170		(__v8sf)(__m256h)(A), (__v8sf)(__m256h)(B), (__v8sf)(__m256h)(C), \
1171		(__mmask8)(U), (int)(R)))
1172
1173		#define _mm256_cmul_round_pch(A, B, R) \
1174		((__m256h)__builtin_ia32_vfcmulcph256_round_mask( \
1175		(__v8sf)(__m256h)(A), (__v8sf)(__m256h)(B), \
1176		(__v8sf)(__m256h)_mm256_undefined_ph(), (__mmask8)-1, (int)(R)))
1177
1178		#define _mm256_mask_cmul_round_pch(W, U, A, B, R) \
1179		((__m256h)__builtin_ia32_vfcmulcph256_round_mask( \
1180		(__v8sf)(__m256h)(A), (__v8sf)(__m256h)(B), (__v8sf)(__m256h)(W), \
1181		(__mmask8)(U), (int)(R)))
1182
1183		#define _mm256_maskz_cmul_round_pch(U, A, B, R) \
1184		((__m256h)__builtin_ia32_vfcmulcph256_round_mask( \
1185		(__v8sf)(__m256h)(A), (__v8sf)(__m256h)(B), \
1186		(__v8sf)(__m256h)_mm256_setzero_ph(), (__mmask8)(U), (int)(R)))
1187
1188		#define _mm256_fixupimm_round_pd(A, B, C, imm, R) \
1189		((__m256d)__builtin_ia32_vfixupimmpd256_round_mask( \
1190		(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), (__v4di)(__m256i)(C), \
1191		(int)(imm), (__mmask8)-1, (int)(R)))
1192
1193		#define _mm256_mask_fixupimm_round_pd(A, U, B, C, imm, R) \
1194		((__m256d)__builtin_ia32_vfixupimmpd256_round_mask( \
1195		(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), (__v4di)(__m256i)(C), \
1196		(int)(imm), (__mmask8)(U), (int)(R)))
1197
1198		#define _mm256_maskz_fixupimm_round_pd(U, A, B, C, imm, R) \
1199		((__m256d)__builtin_ia32_vfixupimmpd256_round_maskz( \
1200		(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), (__v4di)(__m256i)(C), \
1201		(int)(imm), (__mmask8)(U), (int)(R)))
1202
1203		#define _mm256_fixupimm_round_ps(A, B, C, imm, R) \
1204		((__m256)__builtin_ia32_vfixupimmps256_round_mask( \
1205		(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), (__v8si)(__m256i)(C), \
1206		(int)(imm), (__mmask8)-1, (int)(R)))
1207
1208		#define _mm256_mask_fixupimm_round_ps(A, U, B, C, imm, R) \
1209		((__m256)__builtin_ia32_vfixupimmps256_round_mask( \
1210		(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), (__v8si)(__m256i)(C), \
1211		(int)(imm), (__mmask8)(U), (int)(R)))
1212
1213		#define _mm256_maskz_fixupimm_round_ps(U, A, B, C, imm, R) \
1214		((__m256)__builtin_ia32_vfixupimmps256_round_maskz( \
1215		(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), (__v8si)(__m256i)(C), \
1216		(int)(imm), (__mmask8)(U), (int)(R)))
1217
1218		#define _mm256_fmadd_round_pd(A, B, C, R) \
1219		((__m256d)__builtin_ia32_vfmaddpd256_round_mask( \
1220		(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), (__v4df)(__m256d)(C), \
1221		(__mmask8)-1, (int)(R)))
1222
1223		#define _mm256_mask_fmadd_round_pd(A, U, B, C, R) \
1224		((__m256d)__builtin_ia32_vfmaddpd256_round_mask( \
1225		(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), (__v4df)(__m256d)(C), \
1226		(__mmask8)(U), (int)(R)))
1227
1228		#define _mm256_mask3_fmadd_round_pd(A, B, C, U, R) \
1229		((__m256d)__builtin_ia32_vfmaddpd256_round_mask3( \
1230		(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), (__v4df)(__m256d)(C), \
1231		(__mmask8)(U), (int)(R)))
1232
1233		#define _mm256_maskz_fmadd_round_pd(U, A, B, C, R) \
1234		((__m256d)__builtin_ia32_vfmaddpd256_round_maskz( \
1235		(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), (__v4df)(__m256d)(C), \
1236		(__mmask8)(U), (int)(R)))
1237
1238		#define _mm256_fmsub_round_pd(A, B, C, R) \
1239		((__m256d)__builtin_ia32_vfmaddpd256_round_mask( \
1240		(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), -(__v4df)(__m256d)(C), \
1241		(__mmask8)-1, (int)(R)))
1242
1243		#define _mm256_mask_fmsub_round_pd(A, U, B, C, R) \
1244		((__m256d)__builtin_ia32_vfmaddpd256_round_mask( \
1245		(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), -(__v4df)(__m256d)(C), \
1246		(__mmask8)(U), (int)(R)))
1247
1248		#define _mm256_maskz_fmsub_round_pd(U, A, B, C, R) \
1249		((__m256d)__builtin_ia32_vfmaddpd256_round_maskz( \
1250		(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), -(__v4df)(__m256d)(C), \
1251		(__mmask8)(U), (int)(R)))
1252
1253		#define _mm256_fnmadd_round_pd(A, B, C, R) \
1254		((__m256d)__builtin_ia32_vfmaddpd256_round_mask( \
1255		-(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), (__v4df)(__m256d)(C), \
1256		(__mmask8)-1, (int)(R)))
1257
1258		#define _mm256_mask3_fnmadd_round_pd(A, B, C, U, R) \
1259		((__m256d)__builtin_ia32_vfmaddpd256_round_mask3( \
1260		-(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), (__v4df)(__m256d)(C), \
1261		(__mmask8)(U), (int)(R)))
1262
1263		#define _mm256_maskz_fnmadd_round_pd(U, A, B, C, R) \
1264		((__m256d)__builtin_ia32_vfmaddpd256_round_maskz( \
1265		-(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), (__v4df)(__m256d)(C), \
1266		(__mmask8)(U), (int)(R)))
1267
1268		#define _mm256_fnmsub_round_pd(A, B, C, R) \
1269		((__m256d)__builtin_ia32_vfmaddpd256_round_mask( \
1270		-(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), -(__v4df)(__m256d)(C), \
1271		(__mmask8)-1, (int)(R)))
1272
1273		#define _mm256_maskz_fnmsub_round_pd(U, A, B, C, R) \
1274		((__m256d)__builtin_ia32_vfmaddpd256_round_maskz( \
1275		-(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), -(__v4df)(__m256d)(C), \
1276		(__mmask8)(U), (int)(R)))
1277
1278		#define _mm256_fmadd_round_ph(A, B, C, R) \
1279		((__m256h)__builtin_ia32_vfmaddph256_round_mask( \
1280		(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), (__v16hf)(__m256h)(C), \
1281		(__mmask16)-1, (int)(R)))
1282
1283		#define _mm256_mask_fmadd_round_ph(A, U, B, C, R) \
1284		((__m256h)__builtin_ia32_vfmaddph256_round_mask( \
1285		(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), (__v16hf)(__m256h)(C), \
1286		(__mmask16)(U), (int)(R)))
1287
1288		#define _mm256_mask3_fmadd_round_ph(A, B, C, U, R) \
1289		((__m256h)__builtin_ia32_vfmaddph256_round_mask3( \
1290		(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), (__v16hf)(__m256h)(C), \
1291		(__mmask16)(U), (int)(R)))
1292
1293		#define _mm256_maskz_fmadd_round_ph(U, A, B, C, R) \
1294		((__m256h)__builtin_ia32_vfmaddph256_round_maskz( \
1295		(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), (__v16hf)(__m256h)(C), \
1296		(__mmask16)(U), (int)(R)))
1297
1298		#define _mm256_fmsub_round_ph(A, B, C, R) \
1299		((__m256h)__builtin_ia32_vfmaddph256_round_mask( \
1300		(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), -(__v16hf)(__m256h)(C), \
1301		(__mmask16)-1, (int)(R)))
1302
1303		#define _mm256_mask_fmsub_round_ph(A, U, B, C, R) \
1304		((__m256h)__builtin_ia32_vfmaddph256_round_mask( \
1305		(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), -(__v16hf)(__m256h)(C), \
1306		(__mmask16)(U), (int)(R)))
1307
1308		#define _mm256_maskz_fmsub_round_ph(U, A, B, C, R) \
1309		((__m256h)__builtin_ia32_vfmaddph256_round_maskz( \
1310		(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), -(__v16hf)(__m256h)(C), \
1311		(__mmask16)(U), (int)(R)))
1312
1313		#define _mm256_fnmadd_round_ph(A, B, C, R) \
1314		((__m256h)__builtin_ia32_vfmaddph256_round_mask( \
1315		(__v16hf)(__m256h)(A), -(__v16hf)(__m256h)(B), (__v16hf)(__m256h)(C), \
1316		(__mmask16)-1, (int)(R)))
1317
1318		#define _mm256_mask3_fnmadd_round_ph(A, B, C, U, R) \
1319		((__m256h)__builtin_ia32_vfmaddph256_round_mask3( \
1320		-(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), (__v16hf)(__m256h)(C), \
1321		(__mmask16)(U), (int)(R)))
1322
1323		#define _mm256_maskz_fnmadd_round_ph(U, A, B, C, R) \
1324		((__m256h)__builtin_ia32_vfmaddph256_round_maskz( \
1325		-(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), (__v16hf)(__m256h)(C), \
1326		(__mmask16)(U), (int)(R)))
1327
1328		#define _mm256_fnmsub_round_ph(A, B, C, R) \
1329		((__m256h)__builtin_ia32_vfmaddph256_round_mask( \
1330		(__v16hf)(__m256h)(A), -(__v16hf)(__m256h)(B), -(__v16hf)(__m256h)(C), \
1331		(__mmask16)-1, (int)(R)))
1332
1333		#define _mm256_maskz_fnmsub_round_ph(U, A, B, C, R) \
1334		((__m256h)__builtin_ia32_vfmaddph256_round_maskz( \
1335		-(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), -(__v16hf)(__m256h)(C), \
1336		(__mmask16)(U), (int)(R)))
1337
1338		#define _mm256_fmadd_round_ps(A, B, C, R) \
1339		((__m256)__builtin_ia32_vfmaddps256_round_mask( \
1340		(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), (__v8sf)(__m256)(C), \
1341		(__mmask8)-1, (int)(R)))
1342
1343		#define _mm256_mask_fmadd_round_ps(A, U, B, C, R) \
1344		((__m256)__builtin_ia32_vfmaddps256_round_mask( \
1345		(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), (__v8sf)(__m256)(C), \
1346		(__mmask8)(U), (int)(R)))
1347
1348		#define _mm256_mask3_fmadd_round_ps(A, B, C, U, R) \
1349		((__m256)__builtin_ia32_vfmaddps256_round_mask3( \
1350		(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), (__v8sf)(__m256)(C), \
1351		(__mmask8)(U), (int)(R)))
1352
1353		#define _mm256_maskz_fmadd_round_ps(U, A, B, C, R) \
1354		((__m256)__builtin_ia32_vfmaddps256_round_maskz( \
1355		(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), (__v8sf)(__m256)(C), \
1356		(__mmask8)(U), (int)(R)))
1357
1358		#define _mm256_fmsub_round_ps(A, B, C, R) \
1359		((__m256)__builtin_ia32_vfmaddps256_round_mask( \
1360		(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), -(__v8sf)(__m256)(C), \
1361		(__mmask8)-1, (int)(R)))
1362
1363		#define _mm256_mask_fmsub_round_ps(A, U, B, C, R) \
1364		((__m256)__builtin_ia32_vfmaddps256_round_mask( \
1365		(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), -(__v8sf)(__m256)(C), \
1366		(__mmask8)(U), (int)(R)))
1367
1368		#define _mm256_maskz_fmsub_round_ps(U, A, B, C, R) \
1369		((__m256)__builtin_ia32_vfmaddps256_round_maskz( \
1370		(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), -(__v8sf)(__m256)(C), \
1371		(__mmask8)(U), (int)(R)))
1372
1373		#define _mm256_fnmadd_round_ps(A, B, C, R) \
1374		((__m256)__builtin_ia32_vfmaddps256_round_mask( \
1375		(__v8sf)(__m256)(A), -(__v8sf)(__m256)(B), (__v8sf)(__m256)(C), \
1376		(__mmask8)-1, (int)(R)))
1377
1378		#define _mm256_mask3_fnmadd_round_ps(A, B, C, U, R) \
1379		((__m256)__builtin_ia32_vfmaddps256_round_mask3( \
1380		-(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), (__v8sf)(__m256)(C), \
1381		(__mmask8)(U), (int)(R)))
1382
1383		#define _mm256_maskz_fnmadd_round_ps(U, A, B, C, R) \
1384		((__m256)__builtin_ia32_vfmaddps256_round_maskz( \
1385		-(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), (__v8sf)(__m256)(C), \
1386		(__mmask8)(U), (int)(R)))
1387
1388		#define _mm256_fnmsub_round_ps(A, B, C, R) \
1389		((__m256)__builtin_ia32_vfmaddps256_round_mask( \
1390		(__v8sf)(__m256)(A), -(__v8sf)(__m256)(B), -(__v8sf)(__m256)(C), \
1391		(__mmask8)-1, (int)(R)))
1392
1393		#define _mm256_maskz_fnmsub_round_ps(U, A, B, C, R) \
1394		((__m256)__builtin_ia32_vfmaddps256_round_maskz( \
1395		-(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), -(__v8sf)(__m256)(C), \
1396		(__mmask8)(U), (int)(R)))
1397
1398		#define _mm256_fmadd_round_pch(A, B, C, R) \
1399		((__m256h)__builtin_ia32_vfmaddcph256_round_mask3( \
1400		(__v8sf)(__m256h)(A), (__v8sf)(__m256h)(B), (__v8sf)(__m256h)(C), \
1401		(__mmask8)-1, (int)(R)))
1402
1403		#define _mm256_mask_fmadd_round_pch(A, U, B, C, R) \
1404		((__m256h)__builtin_ia32_vfmaddcph256_round_mask( \
1405		(__v8sf)(__m256h)(A), (__v8sf)(__m256h)(B), (__v8sf)(__m256h)(C), \
1406		(__mmask8)(U), (int)(R)))
1407
1408		#define _mm256_mask3_fmadd_round_pch(A, B, C, U, R) \
1409		((__m256h)__builtin_ia32_vfmaddcph256_round_mask3( \
1410		(__v8sf)(__m256h)(A), (__v8sf)(__m256h)(B), (__v8sf)(__m256h)(C), \
1411		(__mmask8)(U), (int)(R)))
1412
1413		#define _mm256_maskz_fmadd_round_pch(U, A, B, C, R) \
1414		((__m256h)__builtin_ia32_vfmaddcph256_round_maskz( \
1415		(__v8sf)(__m256h)(A), (__v8sf)(__m256h)(B), (__v8sf)(__m256h)(C), \
1416		(__mmask8)(U), (int)(R)))
1417
1418		#define _mm256_fmaddsub_round_pd(A, B, C, R) \
1419		((__m256d)__builtin_ia32_vfmaddsubpd256_round_mask( \
1420		(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), (__v4df)(__m256d)(C), \
1421		(__mmask8)-1, (int)(R)))
1422
1423		#define _mm256_mask_fmaddsub_round_pd(A, U, B, C, R) \
1424		((__m256d)__builtin_ia32_vfmaddsubpd256_round_mask( \
1425		(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), (__v4df)(__m256d)(C), \
1426		(__mmask8)(U), (int)(R)))
1427
1428		#define _mm256_mask3_fmaddsub_round_pd(A, B, C, U, R) \
1429		((__m256d)__builtin_ia32_vfmaddsubpd256_round_mask3( \
1430		(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), (__v4df)(__m256d)(C), \
1431		(__mmask8)(U), (int)(R)))
1432
1433		#define _mm256_maskz_fmaddsub_round_pd(U, A, B, C, R) \
1434		((__m256d)__builtin_ia32_vfmaddsubpd256_round_maskz( \
1435		(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), (__v4df)(__m256d)(C), \
1436		(__mmask8)(U), (int)(R)))
1437
1438		#define _mm256_fmsubadd_round_pd(A, B, C, R) \
1439		((__m256d)__builtin_ia32_vfmaddsubpd256_round_mask( \
1440		(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), -(__v4df)(__m256d)(C), \
1441		(__mmask8)-1, (int)(R)))
1442
1443		#define _mm256_mask_fmsubadd_round_pd(A, U, B, C, R) \
1444		((__m256d)__builtin_ia32_vfmaddsubpd256_round_mask( \
1445		(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), -(__v4df)(__m256d)(C), \
1446		(__mmask8)(U), (int)(R)))
1447
1448		#define _mm256_maskz_fmsubadd_round_pd(U, A, B, C, R) \
1449		((__m256d)__builtin_ia32_vfmaddsubpd256_round_maskz( \
1450		(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), -(__v4df)(__m256d)(C), \
1451		(__mmask8)(U), (int)(R)))
1452
1453		#define _mm256_fmaddsub_round_ph(A, B, C, R) \
1454		((__m256h)__builtin_ia32_vfmaddsubph256_round_mask( \
1455		(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), (__v16hf)(__m256h)(C), \
1456		(__mmask16)-1, (int)(R)))
1457
1458		#define _mm256_mask_fmaddsub_round_ph(A, U, B, C, R) \
1459		((__m256h)__builtin_ia32_vfmaddsubph256_round_mask( \
1460		(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), (__v16hf)(__m256h)(C), \
1461		(__mmask16)(U), (int)(R)))
1462
1463		#define _mm256_mask3_fmaddsub_round_ph(A, B, C, U, R) \
1464		((__m256h)__builtin_ia32_vfmaddsubph256_round_mask3( \
1465		(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), (__v16hf)(__m256h)(C), \
1466		(__mmask16)(U), (int)(R)))
1467
1468		#define _mm256_maskz_fmaddsub_round_ph(U, A, B, C, R) \
1469		((__m256h)__builtin_ia32_vfmaddsubph256_round_maskz( \
1470		(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), (__v16hf)(__m256h)(C), \
1471		(__mmask16)(U), (int)(R)))
1472
1473		#define _mm256_fmsubadd_round_ph(A, B, C, R) \
1474		((__m256h)__builtin_ia32_vfmaddsubph256_round_mask( \
1475		(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), -(__v16hf)(__m256h)(C), \
1476		(__mmask16)-1, (int)(R)))
1477
1478		#define _mm256_mask_fmsubadd_round_ph(A, U, B, C, R) \
1479		((__m256h)__builtin_ia32_vfmaddsubph256_round_mask( \
1480		(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), -(__v16hf)(__m256h)(C), \
1481		(__mmask16)(U), (int)(R)))
1482
1483		#define _mm256_maskz_fmsubadd_round_ph(U, A, B, C, R) \
1484		((__m256h)__builtin_ia32_vfmaddsubph256_round_maskz( \
1485		(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), -(__v16hf)(__m256h)(C), \
1486		(__mmask16)(U), (int)(R)))
1487
1488		#define _mm256_fmaddsub_round_ps(A, B, C, R) \
1489		((__m256)__builtin_ia32_vfmaddsubps256_round_mask( \
1490		(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), (__v8sf)(__m256)(C), \
1491		(__mmask8)-1, (int)(R)))
1492
1493		#define _mm256_mask_fmaddsub_round_ps(A, U, B, C, R) \
1494		((__m256)__builtin_ia32_vfmaddsubps256_round_mask( \
1495		(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), (__v8sf)(__m256)(C), \
1496		(__mmask8)(U), (int)(R)))
1497
1498		#define _mm256_mask3_fmaddsub_round_ps(A, B, C, U, R) \
1499		((__m256)__builtin_ia32_vfmaddsubps256_round_mask3( \
1500		(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), (__v8sf)(__m256)(C), \
1501		(__mmask8)(U), (int)(R)))
1502
1503		#define _mm256_maskz_fmaddsub_round_ps(U, A, B, C, R) \
1504		((__m256)__builtin_ia32_vfmaddsubps256_round_maskz( \
1505		(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), (__v8sf)(__m256)(C), \
1506		(__mmask8)(U), (int)(R)))
1507
1508		#define _mm256_fmsubadd_round_ps(A, B, C, R) \
1509		((__m256)__builtin_ia32_vfmaddsubps256_round_mask( \
1510		(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), -(__v8sf)(__m256)(C), \
1511		(__mmask8)-1, (int)(R)))
1512
1513		#define _mm256_mask_fmsubadd_round_ps(A, U, B, C, R) \
1514		((__m256)__builtin_ia32_vfmaddsubps256_round_mask( \
1515		(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), -(__v8sf)(__m256)(C), \
1516		(__mmask8)(U), (int)(R)))
1517
1518		#define _mm256_maskz_fmsubadd_round_ps(U, A, B, C, R) \
1519		((__m256)__builtin_ia32_vfmaddsubps256_round_maskz( \
1520		(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), -(__v8sf)(__m256)(C), \
1521		(__mmask8)(U), (int)(R)))
1522		#define _mm256_mask3_fmsub_round_pd(A, B, C, U, R) \
1523		((__m256d)__builtin_ia32_vfmsubpd256_round_mask3( \
1524		(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), (__v4df)(__m256d)(C), \
1525		(__mmask8)(U), (int)(R)))
1526
1527		#define _mm256_mask3_fmsubadd_round_pd(A, B, C, U, R) \
1528		((__m256d)__builtin_ia32_vfmsubaddpd256_round_mask3( \
1529		(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), (__v4df)(__m256d)(C), \
1530		(__mmask8)(U), (int)(R)))
1531
1532		#define _mm256_mask_fnmadd_round_pd(A, U, B, C, R) \
1533		((__m256d)__builtin_ia32_vfmaddpd256_round_mask( \
1534		(__v4df)(__m256d)(A), -(__v4df)(__m256d)(B), (__v4df)(__m256d)(C), \
1535		(__mmask8)(U), (int)(R)))
1536
1537		#define _mm256_mask_fnmsub_round_pd(A, U, B, C, R) \
1538		((__m256d)__builtin_ia32_vfmaddpd256_round_mask( \
1539		(__v4df)(__m256d)(A), -(__v4df)(__m256d)(B), -(__v4df)(__m256d)(C), \
1540		(__mmask8)(U), (int)(R)))
1541
1542		#define _mm256_mask3_fnmsub_round_pd(A, B, C, U, R) \
1543		((__m256d)__builtin_ia32_vfmsubpd256_round_mask3( \
1544		-(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), (__v4df)(__m256d)(C), \
1545		(__mmask8)(U), (int)(R)))
1546
1547		#define _mm256_mask3_fmsub_round_ph(A, B, C, U, R) \
1548		((__m256h)__builtin_ia32_vfmsubph256_round_mask3( \
1549		(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), (__v16hf)(__m256h)(C), \
1550		(__mmask16)(U), (int)(R)))
1551
1552		#define _mm256_mask3_fmsubadd_round_ph(A, B, C, U, R) \
1553		((__m256h)__builtin_ia32_vfmsubaddph256_round_mask3( \
1554		(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), (__v16hf)(__m256h)(C), \
1555		(__mmask16)(U), (int)(R)))
1556
1557		#define _mm256_mask_fnmadd_round_ph(A, U, B, C, R) \
1558		((__m256h)__builtin_ia32_vfmaddph256_round_mask( \
1559		(__v16hf)(__m256h)(A), -(__v16hf)(__m256h)(B), (__v16hf)(__m256h)(C), \
1560		(__mmask16)(U), (int)(R)))
1561
1562		#define _mm256_mask_fnmsub_round_ph(A, U, B, C, R) \
1563		((__m256h)__builtin_ia32_vfmaddph256_round_mask( \
1564		(__v16hf)(__m256h)(A), -(__v16hf)(__m256h)(B), -(__v16hf)(__m256h)(C), \
1565		(__mmask16)(U), (int)(R)))
1566
1567		#define _mm256_mask3_fnmsub_round_ph(A, B, C, U, R) \
1568		((__m256h)__builtin_ia32_vfmsubph256_round_mask3( \
1569		-(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), (__v16hf)(__m256h)(C), \
1570		(__mmask16)(U), (int)(R)))
1571
1572		#define _mm256_mask3_fmsub_round_ps(A, B, C, U, R) \
1573		((__m256)__builtin_ia32_vfmsubps256_round_mask3( \
1574		(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), (__v8sf)(__m256)(C), \
1575		(__mmask8)(U), (int)(R)))
1576
1577		#define _mm256_mask3_fmsubadd_round_ps(A, B, C, U, R) \
1578		((__m256)__builtin_ia32_vfmsubaddps256_round_mask3( \
1579		(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), (__v8sf)(__m256)(C), \
1580		(__mmask8)(U), (int)(R)))
1581
1582		#define _mm256_mask_fnmadd_round_ps(A, U, B, C, R) \
1583		((__m256)__builtin_ia32_vfmaddps256_round_mask( \
1584		(__v8sf)(__m256)(A), -(__v8sf)(__m256)(B), (__v8sf)(__m256)(C), \
1585		(__mmask8)(U), (int)(R)))
1586
1587		#define _mm256_mask_fnmsub_round_ps(A, U, B, C, R) \
1588		((__m256)__builtin_ia32_vfmaddps256_round_mask( \
1589		(__v8sf)(__m256)(A), -(__v8sf)(__m256)(B), -(__v8sf)(__m256)(C), \
1590		(__mmask8)(U), (int)(R)))
1591
1592		#define _mm256_mask3_fnmsub_round_ps(A, B, C, U, R) \
1593		((__m256)__builtin_ia32_vfmsubps256_round_mask3( \
1594		-(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), (__v8sf)(__m256)(C), \
1595		(__mmask8)(U), (int)(R)))
1596
1597		#define _mm256_mul_round_pch(A, B, R) \
1598		((__m256h)__builtin_ia32_vfmulcph256_round_mask( \
1599		(__v8sf)(__m256h)(A), (__v8sf)(__m256h)(B), \
1600		(__v8sf)(__m256h)_mm256_undefined_ph(), (__mmask8)-1, (int)(R)))
1601
1602		#define _mm256_mask_mul_round_pch(W, U, A, B, R) \
1603		((__m256h)__builtin_ia32_vfmulcph256_round_mask( \
1604		(__v8sf)(__m256h)(A), (__v8sf)(__m256h)(B), (__v8sf)(__m256h)(W), \
1605		(__mmask8)(U), (int)(R)))
1606
1607		#define _mm256_maskz_mul_round_pch(U, A, B, R) \
1608		((__m256h)__builtin_ia32_vfmulcph256_round_mask( \
1609		(__v8sf)(__m256h)(A), (__v8sf)(__m256h)(B), \
1610		(__v8sf)(__m256h)_mm256_setzero_ph(), (__mmask8)(U), (int)(R)))
1611
1612		#define _mm256_getexp_round_pd(A, R) \
1613		((__m256d)__builtin_ia32_vgetexppd256_round_mask( \
1614		(__v4df)(__m256d)(A), (__v4df)_mm256_undefined_pd(), (__mmask8)-1, \
1615		(int)(R)))
1616
1617		#define _mm256_mask_getexp_round_pd(W, U, A, R) \
1618		((__m256d)__builtin_ia32_vgetexppd256_round_mask( \
1619		(__v4df)(__m256d)(A), (__v4df)(__m256d)(W), (__mmask8)(U), (int)(R)))
1620
1621		#define _mm256_maskz_getexp_round_pd(U, A, R) \
1622		((__m256d)__builtin_ia32_vgetexppd256_round_mask( \
1623		(__v4df)(__m256d)(A), (__v4df)_mm256_setzero_pd(), (__mmask8)(U), \
1624		(int)(R)))
1625
1626		#define _mm256_getexp_round_ph(A, R) \
1627		((__m256h)__builtin_ia32_vgetexpph256_round_mask( \
1628		(__v16hf)(__m256h)(A), (__v16hf)_mm256_undefined_ph(), (__mmask16)-1, \
1629		(int)(R)))
1630
1631		#define _mm256_mask_getexp_round_ph(W, U, A, R) \
1632		((__m256h)__builtin_ia32_vgetexpph256_round_mask( \
1633		(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(W), (__mmask16)(U), (int)(R)))
1634
1635		#define _mm256_maskz_getexp_round_ph(U, A, R) \
1636		((__m256h)__builtin_ia32_vgetexpph256_round_mask( \
1637		(__v16hf)(__m256h)(A), (__v16hf)_mm256_setzero_ph(), (__mmask16)(U), \
1638		(int)(R)))
1639
1640		#define _mm256_getexp_round_ps(A, R) \
1641		((__m256)__builtin_ia32_vgetexpps256_round_mask( \
1642		(__v8sf)(__m256)(A), (__v8sf)_mm256_undefined_ps(), (__mmask8)-1, \
1643		(int)(R)))
1644
1645		#define _mm256_mask_getexp_round_ps(W, U, A, R) \
1646		((__m256)__builtin_ia32_vgetexpps256_round_mask( \
1647		(__v8sf)(__m256)(A), (__v8sf)(__m256)(W), (__mmask8)(U), (int)(R)))
1648
1649		#define _mm256_maskz_getexp_round_ps(U, A, R) \
1650		((__m256)__builtin_ia32_vgetexpps256_round_mask((__v8sf)(__m256)(A), \
1651		(__v8sf)_mm256_setzero_ps(), \
1652		(__mmask8)(U), (int)(R)))
1653
1654		#define _mm256_getmant_round_pd(A, B, C, R) \
1655		((__m256d)__builtin_ia32_vgetmantpd256_round_mask( \
1656		(__v4df)(__m256d)(A), (int)(((C) << 2) \| (B)), \
1657		(__v4df)_mm256_undefined_pd(), (__mmask8)-1, (int)(R)))
1658
1659		#define _mm256_mask_getmant_round_pd(W, U, A, B, C, R) \
1660		((__m256d)__builtin_ia32_vgetmantpd256_round_mask( \
1661		(__v4df)(__m256d)(A), (int)(((C) << 2) \| (B)), (__v4df)(__m256d)(W), \
1662		(__mmask8)(U), (int)(R)))
1663
1664		#define _mm256_maskz_getmant_round_pd(U, A, B, C, R) \
1665		((__m256d)__builtin_ia32_vgetmantpd256_round_mask( \
1666		(__v4df)(__m256d)(A), (int)(((C) << 2) \| (B)), \
1667		(__v4df)_mm256_setzero_pd(), (__mmask8)(U), (int)(R)))
1668
1669		#define _mm256_getmant_round_ph(A, B, C, R) \
1670		((__m256h)__builtin_ia32_vgetmantph256_round_mask( \
1671		(__v16hf)(__m256h)(A), (int)(((C) << 2) \| (B)), \
1672		(__v16hf)_mm256_undefined_ph(), (__mmask16)-1, (int)(R)))
1673
1674		#define _mm256_mask_getmant_round_ph(W, U, A, B, C, R) \
1675		((__m256h)__builtin_ia32_vgetmantph256_round_mask( \
1676		(__v16hf)(__m256h)(A), (int)(((C) << 2) \| (B)), (__v16hf)(__m256h)(W), \
1677		(__mmask16)(U), (int)(R)))
1678
1679		#define _mm256_maskz_getmant_round_ph(U, A, B, C, R) \
1680		((__m256h)__builtin_ia32_vgetmantph256_round_mask( \
1681		(__v16hf)(__m256h)(A), (int)(((C) << 2) \| (B)), \
1682		(__v16hf)_mm256_setzero_ph(), (__mmask16)(U), (int)(R)))
1683
1684		#define _mm256_getmant_round_ps(A, B, C, R) \
1685		((__m256)__builtin_ia32_vgetmantps256_round_mask( \
1686		(__v8sf)(__m256)(A), (int)(((C) << 2) \| (B)), \
1687		(__v8sf)_mm256_undefined_ps(), (__mmask8)-1, (int)(R)))
1688
1689		#define _mm256_mask_getmant_round_ps(W, U, A, B, C, R) \
1690		((__m256)__builtin_ia32_vgetmantps256_round_mask( \
1691		(__v8sf)(__m256)(A), (int)(((C) << 2) \| (B)), (__v8sf)(__m256)(W), \
1692		(__mmask8)(U), (int)(R)))
1693
1694		#define _mm256_maskz_getmant_round_ps(U, A, B, C, R) \
1695		((__m256)__builtin_ia32_vgetmantps256_round_mask( \
1696		(__v8sf)(__m256)(A), (int)(((C) << 2) \| (B)), \
1697		(__v8sf)_mm256_setzero_ps(), (__mmask8)(U), (int)(R)))
1698
1699		#define _mm256_max_round_pd(A, B, R) \
1700		((__m256d)__builtin_ia32_vmaxpd256_round((__v4df)(__m256d)(A), \
1701		(__v4df)(__m256d)(B), (int)(R)))
1702
1703		#define _mm256_mask_max_round_pd(W, U, A, B, R) \
1704		((__m256d)__builtin_ia32_selectpd_256( \
1705		(__mmask8)(U), (__v4df)_mm256_max_round_pd((A), (B), (R)), \
1706		(__v4df)(__m256d)(W)))
1707
1708		#define _mm256_maskz_max_round_pd(U, A, B, R) \
1709		((__m256d)__builtin_ia32_selectpd_256( \
1710		(__mmask8)(U), (__v4df)_mm256_max_round_pd((A), (B), (R)), \
1711		(__v4df)_mm256_setzero_pd()))
1712
1713		#define _mm256_max_round_ph(A, B, R) \
1714		((__m256h)__builtin_ia32_vmaxph256_round((__v16hf)(__m256h)(A), \
1715		(__v16hf)(__m256h)(B), (int)(R)))
1716
1717		#define _mm256_mask_max_round_ph(W, U, A, B, R) \
1718		((__m256h)__builtin_ia32_selectph_256( \
1719		(__mmask16)(U), (__v16hf)_mm256_max_round_ph((A), (B), (R)), \
1720		(__v16hf)(__m256h)(W)))
1721
1722		#define _mm256_maskz_max_round_ph(U, A, B, R) \
1723		((__m256h)__builtin_ia32_selectph_256( \
1724		(__mmask16)(U), (__v16hf)_mm256_max_round_ph((A), (B), (R)), \
1725		(__v16hf)_mm256_setzero_ph()))
1726
1727		#define _mm256_max_round_ps(A, B, R) \
1728		((__m256)__builtin_ia32_vmaxps256_round((__v8sf)(__m256)(A), \
1729		(__v8sf)(__m256)(B), (int)(R)))
1730
1731		#define _mm256_mask_max_round_ps(W, U, A, B, R) \
1732		((__m256)__builtin_ia32_selectps_256( \
1733		(__mmask8)(U), (__v8sf)_mm256_max_round_ps((A), (B), (R)), \
1734		(__v8sf)(__m256)(W)))
1735
1736		#define _mm256_maskz_max_round_ps(U, A, B, R) \
1737		((__m256)__builtin_ia32_selectps_256( \
1738		(__mmask8)(U), (__v8sf)_mm256_max_round_ps((A), (B), (R)), \
1739		(__v8sf)_mm256_setzero_ps()))
1740
1741		#define _mm256_min_round_pd(A, B, R) \
1742		((__m256d)__builtin_ia32_vminpd256_round((__v4df)(__m256d)(A), \
1743		(__v4df)(__m256d)(B), (int)(R)))
1744
1745		#define _mm256_mask_min_round_pd(W, U, A, B, R) \
1746		((__m256d)__builtin_ia32_selectpd_256( \
1747		(__mmask8)(U), (__v4df)_mm256_min_round_pd((A), (B), (R)), \
1748		(__v4df)(__m256d)(W)))
1749
1750		#define _mm256_maskz_min_round_pd(U, A, B, R) \
1751		((__m256d)__builtin_ia32_selectpd_256( \
1752		(__mmask8)(U), (__v4df)_mm256_min_round_pd((A), (B), (R)), \
1753		(__v4df)_mm256_setzero_pd()))
1754
1755		#define _mm256_min_round_ph(A, B, R) \
1756		((__m256h)__builtin_ia32_vminph256_round((__v16hf)(__m256h)(A), \
1757		(__v16hf)(__m256h)(B), (int)(R)))
1758
1759		#define _mm256_mask_min_round_ph(W, U, A, B, R) \
1760		((__m256h)__builtin_ia32_selectph_256( \
1761		(__mmask16)(U), (__v16hf)_mm256_min_round_ph((A), (B), (R)), \
1762		(__v16hf)(__m256h)(W)))
1763
1764		#define _mm256_maskz_min_round_ph(U, A, B, R) \
1765		((__m256h)__builtin_ia32_selectph_256( \
1766		(__mmask16)(U), (__v16hf)_mm256_min_round_ph((A), (B), (R)), \
1767		(__v16hf)_mm256_setzero_ph()))
1768
1769		#define _mm256_min_round_ps(A, B, R) \
1770		((__m256)__builtin_ia32_vminps256_round((__v8sf)(__m256)(A), \
1771		(__v8sf)(__m256)(B), (int)(R)))
1772
1773		#define _mm256_mask_min_round_ps(W, U, A, B, R) \
1774		((__m256)__builtin_ia32_selectps_256( \
1775		(__mmask8)(U), (__v8sf)_mm256_min_round_ps((A), (B), (R)), \
1776		(__v8sf)(__m256)(W)))
1777
1778		#define _mm256_maskz_min_round_ps(U, A, B, R) \
1779		((__m256)__builtin_ia32_selectps_256( \
1780		(__mmask8)(U), (__v8sf)_mm256_min_round_ps((A), (B), (R)), \
1781		(__v8sf)_mm256_setzero_ps()))
1782
1783		#define _mm256_mul_round_pd(A, B, R) \
1784		((__m256d)__builtin_ia32_vmulpd256_round((__v4df)(__m256d)(A), \
1785		(__v4df)(__m256d)(B), (int)(R)))
1786
1787		#define _mm256_mask_mul_round_pd(W, U, A, B, R) \
1788		((__m256d)__builtin_ia32_selectpd_256( \
1789		(__mmask8)(U), (__v4df)_mm256_mul_round_pd((A), (B), (R)), \
1790		(__v4df)(__m256d)(W)))
1791
1792		#define _mm256_maskz_mul_round_pd(U, A, B, R) \
1793		((__m256d)__builtin_ia32_selectpd_256( \
1794		(__mmask8)(U), (__v4df)_mm256_mul_round_pd((A), (B), (R)), \
1795		(__v4df)_mm256_setzero_pd()))
1796
1797		#define _mm256_mul_round_ph(A, B, R) \
1798		((__m256h)__builtin_ia32_vmulph256_round((__v16hf)(__m256h)(A), \
1799		(__v16hf)(__m256h)(B), (int)(R)))
1800
1801		#define _mm256_mask_mul_round_ph(W, U, A, B, R) \
1802		((__m256h)__builtin_ia32_selectph_256( \
1803		(__mmask16)(U), (__v16hf)_mm256_mul_round_ph((A), (B), (R)), \
1804		(__v16hf)(__m256h)(W)))
1805
1806		#define _mm256_maskz_mul_round_ph(U, A, B, R) \
1807		((__m256h)__builtin_ia32_selectph_256( \
1808		(__mmask16)(U), (__v16hf)_mm256_mul_round_ph((A), (B), (R)), \
1809		(__v16hf)_mm256_setzero_ph()))
1810
1811		#define _mm256_mul_round_ps(A, B, R) \
1812		((__m256)__builtin_ia32_vmulps256_round((__v8sf)(__m256)(A), \
1813		(__v8sf)(__m256)(B), (int)(R)))
1814
1815		#define _mm256_mask_mul_round_ps(W, U, A, B, R) \
1816		((__m256)__builtin_ia32_selectps_256( \
1817		(__mmask8)(U), (__v8sf)_mm256_mul_round_ps((A), (B), (R)), \
1818		(__v8sf)(__m256)(W)))
1819
1820		#define _mm256_maskz_mul_round_ps(U, A, B, R) \
1821		((__m256)__builtin_ia32_selectps_256( \
1822		(__mmask8)(U), (__v8sf)_mm256_mul_round_ps((A), (B), (R)), \
1823		(__v8sf)_mm256_setzero_ps()))
1824
1825		#define _mm256_range_round_pd(A, B, C, R) \
1826		((__m256d)__builtin_ia32_vrangepd256_round_mask( \
1827		(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), (int)(C), \
1828		(__v4df)_mm256_setzero_pd(), (__mmask8)-1, (int)(R)))
1829
1830		#define _mm256_mask_range_round_pd(W, U, A, B, C, R) \
1831		((__m256d)__builtin_ia32_vrangepd256_round_mask( \
1832		(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), (int)(C), \
1833		(__v4df)(__m256d)(W), (__mmask8)(U), (int)(R)))
1834
1835		#define _mm256_maskz_range_round_pd(U, A, B, C, R) \
1836		((__m256d)__builtin_ia32_vrangepd256_round_mask( \
1837		(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), (int)(C), \
1838		(__v4df)_mm256_setzero_pd(), (__mmask8)(U), (int)(R)))
1839
1840		#define _mm256_range_round_ps(A, B, C, R) \
1841		((__m256)__builtin_ia32_vrangeps256_round_mask( \
1842		(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), (int)(C), \
1843		(__v8sf)_mm256_setzero_ps(), (__mmask8)-1, (int)(R)))
1844
1845		#define _mm256_mask_range_round_ps(W, U, A, B, C, R) \
1846		((__m256)__builtin_ia32_vrangeps256_round_mask( \
1847		(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), (int)(C), (__v8sf)(__m256)(W), \
1848		(__mmask8)(U), (int)(R)))
1849
1850		#define _mm256_maskz_range_round_ps(U, A, B, C, R) \
1851		((__m256)__builtin_ia32_vrangeps256_round_mask( \
1852		(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), (int)(C), \
1853		(__v8sf)_mm256_setzero_ps(), (__mmask8)(U), (int)(R)))
1854
1855		#define _mm256_reduce_round_pd(A, B, R) \
1856		((__m256d)__builtin_ia32_vreducepd256_round_mask( \
1857		(__v4df)(__m256d)(A), (int)(B), (__v4df)_mm256_setzero_pd(), \
1858		(__mmask8)-1, (int)(R)))
1859
1860		#define _mm256_mask_reduce_round_pd(W, U, A, B, R) \
1861		((__m256d)__builtin_ia32_vreducepd256_round_mask( \
1862		(__v4df)(__m256d)(A), (int)(B), (__v4df)(__m256d)(W), (__mmask8)(U), \
1863		(int)(R)))
1864
1865		#define _mm256_maskz_reduce_round_pd(U, A, B, R) \
1866		((__m256d)__builtin_ia32_vreducepd256_round_mask( \
1867		(__v4df)(__m256d)(A), (int)(B), (__v4df)_mm256_setzero_pd(), \
1868		(__mmask8)(U), (int)(R)))
1869
1870		#define _mm256_mask_reduce_round_ph(W, U, A, imm, R) \
1871		((__m256h)__builtin_ia32_vreduceph256_round_mask( \
1872		(__v16hf)(__m256h)(A), (int)(imm), (__v16hf)(__m256h)(W), \
1873		(__mmask16)(U), (int)(R)))
1874
1875		#define _mm256_maskz_reduce_round_ph(U, A, imm, R) \
1876		((__m256h)__builtin_ia32_vreduceph256_round_mask( \
1877		(__v16hf)(__m256h)(A), (int)(imm), (__v16hf)_mm256_setzero_ph(), \
1878		(__mmask16)(U), (int)(R)))
1879
1880		#define _mm256_reduce_round_ph(A, imm, R) \
1881		((__m256h)__builtin_ia32_vreduceph256_round_mask( \
1882		(__v16hf)(__m256h)(A), (int)(imm), (__v16hf)_mm256_undefined_ph(), \
1883		(__mmask16)-1, (int)(R)))
1884
1885		#define _mm256_reduce_round_ps(A, B, R) \
1886		((__m256)__builtin_ia32_vreduceps256_round_mask( \
1887		(__v8sf)(__m256)(A), (int)(B), (__v8sf)_mm256_setzero_ps(), \
1888		(__mmask8)-1, (int)(R)))
1889
1890		#define _mm256_mask_reduce_round_ps(W, U, A, B, R) \
1891		((__m256)__builtin_ia32_vreduceps256_round_mask( \
1892		(__v8sf)(__m256)(A), (int)(B), (__v8sf)(__m256)(W), (__mmask8)(U), \
1893		(int)(R)))
1894
1895		#define _mm256_maskz_reduce_round_ps(U, A, B, R) \
1896		((__m256)__builtin_ia32_vreduceps256_round_mask( \
1897		(__v8sf)(__m256)(A), (int)(B), (__v8sf)_mm256_setzero_ps(), \
1898		(__mmask8)(U), (int)(R)))
1899
1900		#define _mm256_roundscale_round_pd(A, imm, R) \
1901		((__m256d)__builtin_ia32_vrndscalepd256_round_mask( \
1902		(__v4df)(__m256d)(A), (int)(imm), (__v4df)_mm256_undefined_pd(), \
1903		(__mmask8)-1, (int)(R)))
1904
1905		#define _mm256_mask_roundscale_round_pd(A, B, C, imm, R) \
1906		((__m256d)__builtin_ia32_vrndscalepd256_round_mask( \
1907		(__v4df)(__m256d)(C), (int)(imm), (__v4df)(__m256d)(A), (__mmask8)(B), \
1908		(int)(R)))
1909
1910		#define _mm256_maskz_roundscale_round_pd(A, B, imm, R) \
1911		((__m256d)__builtin_ia32_vrndscalepd256_round_mask( \
1912		(__v4df)(__m256d)(B), (int)(imm), (__v4df)_mm256_setzero_pd(), \
1913		(__mmask8)(A), (int)(R)))
1914
1915		#define _mm256_roundscale_round_ph(A, imm, R) \
1916		((__m256h)__builtin_ia32_vrndscaleph256_round_mask( \
1917		(__v16hf)(__m256h)(A), (int)(imm), (__v16hf)_mm256_undefined_ph(), \
1918		(__mmask16)-1, (int)(R)))
1919
1920		#define _mm256_mask_roundscale_round_ph(A, B, C, imm, R) \
1921		((__m256h)__builtin_ia32_vrndscaleph256_round_mask( \
1922		(__v16hf)(__m256h)(C), (int)(imm), (__v16hf)(__m256h)(A), \
1923		(__mmask16)(B), (int)(R)))
1924
1925		#define _mm256_maskz_roundscale_round_ph(A, B, imm, R) \
1926		((__m256h)__builtin_ia32_vrndscaleph256_round_mask( \
1927		(__v16hf)(__m256h)(B), (int)(imm), (__v16hf)_mm256_setzero_ph(), \
1928		(__mmask16)(A), (int)(R)))
1929
1930		#define _mm256_roundscale_round_ps(A, imm, R) \
1931		((__m256)__builtin_ia32_vrndscaleps256_round_mask( \
1932		(__v8sf)(__m256)(A), (int)(imm), (__v8sf)_mm256_undefined_ps(), \
1933		(__mmask8)-1, (int)(R)))
1934
1935		#define _mm256_mask_roundscale_round_ps(A, B, C, imm, R) \
1936		((__m256)__builtin_ia32_vrndscaleps256_round_mask( \
1937		(__v8sf)(__m256)(C), (int)(imm), (__v8sf)(__m256)(A), (__mmask8)(B), \
1938		(int)(R)))
1939
1940		#define _mm256_maskz_roundscale_round_ps(A, B, imm, R) \
1941		((__m256)__builtin_ia32_vrndscaleps256_round_mask( \
1942		(__v8sf)(__m256)(B), (int)(imm), (__v8sf)_mm256_setzero_ps(), \
1943		(__mmask8)(A), (int)(R)))
1944
1945		#define _mm256_scalef_round_pd(A, B, R) \
1946		((__m256d)__builtin_ia32_vscalefpd256_round_mask( \
1947		(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), \
1948		(__v4df)_mm256_undefined_pd(), (__mmask8)-1, (int)(R)))
1949
1950		#define _mm256_mask_scalef_round_pd(W, U, A, B, R) \
1951		((__m256d)__builtin_ia32_vscalefpd256_round_mask( \
1952		(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), (__v4df)(__m256d)(W), \
1953		(__mmask8)(U), (int)(R)))
1954
1955		#define _mm256_maskz_scalef_round_pd(U, A, B, R) \
1956		((__m256d)__builtin_ia32_vscalefpd256_round_mask( \
1957		(__v4df)(__m256d)(A), (__v4df)(__m256d)(B), (__v4df)_mm256_setzero_pd(), \
1958		(__mmask8)(U), (int)(R)))
1959
1960		#define _mm256_scalef_round_ph(A, B, R) \
1961		((__m256h)__builtin_ia32_vscalefph256_round_mask( \
1962		(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), \
1963		(__v16hf)_mm256_undefined_ph(), (__mmask16)-1, (int)(R)))
1964
1965		#define _mm256_mask_scalef_round_ph(W, U, A, B, R) \
1966		((__m256h)__builtin_ia32_vscalefph256_round_mask( \
1967		(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), (__v16hf)(__m256h)(W), \
1968		(__mmask16)(U), (int)(R)))
1969
1970		#define _mm256_maskz_scalef_round_ph(U, A, B, R) \
1971		((__m256h)__builtin_ia32_vscalefph256_round_mask( \
1972		(__v16hf)(__m256h)(A), (__v16hf)(__m256h)(B), \
1973		(__v16hf)_mm256_setzero_ph(), (__mmask16)(U), (int)(R)))
1974
1975		#define _mm256_scalef_round_ps(A, B, R) \
1976		((__m256)__builtin_ia32_vscalefps256_round_mask( \
1977		(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), (__v8sf)_mm256_undefined_ps(), \
1978		(__mmask8)-1, (int)(R)))
1979
1980		#define _mm256_mask_scalef_round_ps(W, U, A, B, R) \
1981		((__m256)__builtin_ia32_vscalefps256_round_mask( \
1982		(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), (__v8sf)(__m256)(W), \
1983		(__mmask8)(U), (int)(R)))
1984
1985		#define _mm256_maskz_scalef_round_ps(U, A, B, R) \
1986		((__m256)__builtin_ia32_vscalefps256_round_mask( \
1987		(__v8sf)(__m256)(A), (__v8sf)(__m256)(B), (__v8sf)_mm256_setzero_ps(), \
1988		(__mmask8)(U), (int)(R)))
1989
1990		#define _mm256_sqrt_round_pd(A, R) \
1991		((__m256d)__builtin_ia32_vsqrtpd256_round((__v4df)(__m256d)(A), (int)(R)))
1992
1993		#define _mm256_mask_sqrt_round_pd(W, U, A, R) \
1994		((__m256d)__builtin_ia32_selectpd_256( \
1995		(__mmask8)(U), (__v4df)_mm256_sqrt_round_pd((A), (R)), \
1996		(__v4df)(__m256d)(W)))
1997
1998		#define _mm256_maskz_sqrt_round_pd(U, A, R) \
1999		((__m256d)__builtin_ia32_selectpd_256( \
2000		(__mmask8)(U), (__v4df)_mm256_sqrt_round_pd((A), (R)), \
2001		(__v4df)_mm256_setzero_pd()))
2002
2003		#define _mm256_sqrt_round_ph(A, R) \
2004		((__m256h)__builtin_ia32_vsqrtph256_round((__v16hf)(__m256h)(A), (int)(R)))
2005
2006		#define _mm256_mask_sqrt_round_ph(W, U, A, R) \
2007		((__m256h)__builtin_ia32_selectph_256( \
2008		(__mmask16)(U), (__v16hf)_mm256_sqrt_round_ph((A), (R)), \
2009		(__v16hf)(__m256h)(W)))
2010
2011		#define _mm256_maskz_sqrt_round_ph(U, A, R) \
2012		((__m256h)__builtin_ia32_selectph_256( \
2013		(__mmask16)(U), (__v16hf)_mm256_sqrt_round_ph((A), (R)), \
2014		(__v16hf)_mm256_setzero_ph()))
2015
2016		#define _mm256_sqrt_round_ps(A, R) \
2017		((__m256)__builtin_ia32_vsqrtps256_round((__v8sf)(__m256)(A), (int)(R)))
2018
2019		#define _mm256_mask_sqrt_round_ps(W, U, A, R) \
2020		((__m256)__builtin_ia32_selectps_256((__mmask8)(U), \
2021		(__v8sf)_mm256_sqrt_round_ps((A), (R)), \
2022		(__v8sf)(__m256)(W)))
2023
2024		#define _mm256_maskz_sqrt_round_ps(U, A, R) \
2025		((__m256)__builtin_ia32_selectps_256((__mmask8)(U), \
2026		(__v8sf)_mm256_sqrt_round_ps((A), (R)), \
2027		(__v8sf)_mm256_setzero_ps()))
2028
2029		#define _mm256_sub_round_pd(A, B, R) \
2030		((__m256d)__builtin_ia32_vsubpd256_round((__v4df)(__m256d)(A), \
2031		(__v4df)(__m256d)(B), (int)(R)))
2032
2033		#define _mm256_mask_sub_round_pd(W, U, A, B, R) \
2034		((__m256d)__builtin_ia32_selectpd_256( \
2035		(__mmask8)(U), (__v4df)_mm256_sub_round_pd((A), (B), (R)), \
2036		(__v4df)(__m256d)(W)))
2037
2038		#define _mm256_maskz_sub_round_pd(U, A, B, R) \
2039		((__m256d)__builtin_ia32_selectpd_256( \
2040		(__mmask8)(U), (__v4df)_mm256_sub_round_pd((A), (B), (R)), \
2041		(__v4df)_mm256_setzero_pd()))
2042
2043		#define _mm256_sub_round_ph(A, B, R) \
2044		((__m256h)__builtin_ia32_vsubph256_round((__v16hf)(__m256h)(A), \
2045		(__v16hf)(__m256h)(B), (int)(R)))
2046
2047		#define _mm256_mask_sub_round_ph(W, U, A, B, R) \
2048		((__m256h)__builtin_ia32_selectph_256( \
2049		(__mmask16)(U), (__v16hf)_mm256_sub_round_ph((A), (B), (R)), \
2050		(__v16hf)(__m256h)(W)))
2051
2052		#define _mm256_maskz_sub_round_ph(U, A, B, R) \
2053		((__m256h)__builtin_ia32_selectph_256( \
2054		(__mmask16)(U), (__v16hf)_mm256_sub_round_ph((A), (B), (R)), \
2055		(__v16hf)_mm256_setzero_ph()))
2056
2057		#define _mm256_sub_round_ps(A, B, R) \
2058		((__m256)__builtin_ia32_vsubps256_round((__v8sf)(__m256)(A), \
2059		(__v8sf)(__m256)(B), (int)(R)))
2060
2061		#define _mm256_mask_sub_round_ps(W, U, A, B, R) \
2062		((__m256)__builtin_ia32_selectps_256( \
2063		(__mmask8)(U), (__v8sf)_mm256_sub_round_ps((A), (B), (R)), \
2064		(__v8sf)(__m256)(W)))
2065
2066		#define _mm256_maskz_sub_round_ps(U, A, B, R) \
2067		((__m256)__builtin_ia32_selectps_256( \
2068		(__mmask8)(U), (__v8sf)_mm256_sub_round_ps((A), (B), (R)), \
2069		(__v8sf)_mm256_setzero_ps()))
2070
2071	405	#undef __DEFAULT_FN_ATTRS256
2072	406	#undef __DEFAULT_FN_ATTRS128
2073	407

lib/include/avx10_2satcvtdsintrin.h+100-220

...	...	@@ -71,175 +71,134 @@
71	71	#endif /* __x86_64__ */
72	72
73	73	// 128 Bit : Double -> int
74		static __inline__ __m128i __DEFAULT_FN_ATTRS128 _mm_cvttspd_epi32(__m128d __A) {
	74	static __inline__ __m128i __DEFAULT_FN_ATTRS128
	75	_mm_cvtts_pd_epi32(__m128d __A) {
75	76	return ((__m128i)__builtin_ia32_vcvttpd2dqs128_mask(
76	77	(__v2df)__A, (__v4si)(__m128i)_mm_undefined_si128(), (__mmask8)(-1)));
77	78	}
78	79
79	80	static __inline__ __m128i __DEFAULT_FN_ATTRS128
80		_mm_mask_cvttspd_epi32(__m128i __W, __mmask8 __U, __m128d __A) {
	81	_mm_mask_cvtts_pd_epi32(__m128i __W, __mmask8 __U, __m128d __A) {
81	82	return ((__m128i)__builtin_ia32_vcvttpd2dqs128_mask((__v2df)__A, (__v4si)__W,
82	83	__U));
83	84	}
84	85
85	86	static __inline__ __m128i __DEFAULT_FN_ATTRS128
86		_mm_maskz_cvttspd_epi32(__mmask16 __U, __m128d __A) {
	87	_mm_maskz_cvtts_pd_epi32(__mmask16 __U, __m128d __A) {
87	88	return ((__m128i)__builtin_ia32_vcvttpd2dqs128_mask(
88	89	(__v2df)__A, (__v4si)(__m128i)_mm_setzero_si128(), __U));
89	90	}
90	91
91	92	// 256 Bit : Double -> int
92	93	static __inline__ __m128i __DEFAULT_FN_ATTRS256
93		_mm256_cvttspd_epi32(__m256d __A) {
94		return ((__m128i)__builtin_ia32_vcvttpd2dqs256_round_mask(
95		(__v4df)__A, (__v4si)_mm_undefined_si128(), (__mmask8)-1,
96		_MM_FROUND_CUR_DIRECTION));
	94	_mm256_cvtts_pd_epi32(__m256d __A) {
	95	return ((__m128i)__builtin_ia32_vcvttpd2dqs256_mask(
	96	(__v4df)__A, (__v4si)_mm_undefined_si128(), (__mmask8)-1));
97	97	}
98	98
99	99	static __inline__ __m128i __DEFAULT_FN_ATTRS256
100		_mm256_mask_cvttspd_epi32(__m128i __W, __mmask8 __U, __m256d __A) {
101		return ((__m128i)__builtin_ia32_vcvttpd2dqs256_round_mask(
102		(__v4df)__A, (__v4si)__W, __U, _MM_FROUND_CUR_DIRECTION));
	100	_mm256_mask_cvtts_pd_epi32(__m128i __W, __mmask8 __U, __m256d __A) {
	101	return ((__m128i)__builtin_ia32_vcvttpd2dqs256_mask((__v4df)__A, (__v4si)__W,
	102	__U));
103	103	}
104	104
105	105	static __inline__ __m128i __DEFAULT_FN_ATTRS256
106		_mm256_maskz_cvttspd_epi32(__mmask8 __U, __m256d __A) {
107		return ((__m128i)__builtin_ia32_vcvttpd2dqs256_round_mask(
108		(__v4df)__A, (__v4si)_mm_setzero_si128(), __U, _MM_FROUND_CUR_DIRECTION));
	106	_mm256_maskz_cvtts_pd_epi32(__mmask8 __U, __m256d __A) {
	107	return ((__m128i)__builtin_ia32_vcvttpd2dqs256_mask(
	108	(__v4df)__A, (__v4si)_mm_setzero_si128(), __U));
109	109	}
110	110
111		#define _mm256_cvtts_roundpd_epi32(__A, __R) \
112		((__m128i)__builtin_ia32_vcvttpd2dqs256_round_mask( \
113		(__v4df)(__m256d)__A, (__v4si)(__m128i)_mm_undefined_si128(), \
114		(__mmask8) - 1, (int)(__R)))
115
116		#define _mm256_mask_cvtts_roundpd_epi32(__W, __U, __A, __R) \
117		((__m128i)__builtin_ia32_vcvttpd2dqs256_round_mask( \
118		(__v4df)(__m256d)__A, (__v4si)(__m128i)__W, (__mmask8)__U, (int)(__R)))
119
120		#define _mm256_maskz_cvtts_roundpd_epi32(__U, __A, __R) \
121		((__m128i)__builtin_ia32_vcvttpd2dqs256_round_mask( \
122		(__v4df)(__m256d)__A, (__v4si)(__m128i)_mm_setzero_si128(), \
123		(__mmask8)__U, (int)(__R)))
124
125	111	// 128 Bit : Double -> uint
126		static __inline__ __m128i __DEFAULT_FN_ATTRS128 _mm_cvttspd_epu32(__m128d __A) {
	112	static __inline__ __m128i __DEFAULT_FN_ATTRS128
	113	_mm_cvtts_pd_epu32(__m128d __A) {
127	114	return ((__m128i)__builtin_ia32_vcvttpd2udqs128_mask(
128	115	(__v2df)__A, (__v4si)(__m128i)_mm_undefined_si128(), (__mmask8)(-1)));
129	116	}
130	117
131	118	static __inline__ __m128i __DEFAULT_FN_ATTRS128
132		_mm_mask_cvttspd_epu32(__m128i __W, __mmask8 __U, __m128d __A) {
	119	_mm_mask_cvtts_pd_epu32(__m128i __W, __mmask8 __U, __m128d __A) {
133	120	return ((__m128i)__builtin_ia32_vcvttpd2udqs128_mask(
134	121	(__v2df)__A, (__v4si)(__m128i)__W, (__mmask8)__U));
135	122	}
136	123
137	124	static __inline__ __m128i __DEFAULT_FN_ATTRS128
138		_mm_maskz_cvttspd_epu32(__mmask8 __U, __m128d __A) {
	125	_mm_maskz_cvtts_pd_epu32(__mmask8 __U, __m128d __A) {
139	126	return ((__m128i)__builtin_ia32_vcvttpd2udqs128_mask(
140	127	(__v2df)__A, (__v4si)(__m128i)_mm_setzero_si128(), __U));
141	128	}
142	129
143	130	// 256 Bit : Double -> uint
144	131	static __inline__ __m128i __DEFAULT_FN_ATTRS256
145		_mm256_cvttspd_epu32(__m256d __A) {
146		return ((__m128i)__builtin_ia32_vcvttpd2udqs256_round_mask(
147		(__v4df)__A, (__v4si)_mm_undefined_si128(), (__mmask8)-1,
148		_MM_FROUND_CUR_DIRECTION));
	132	_mm256_cvtts_pd_epu32(__m256d __A) {
	133	return ((__m128i)__builtin_ia32_vcvttpd2udqs256_mask(
	134	(__v4df)__A, (__v4si)_mm_undefined_si128(), (__mmask8)-1));
149	135	}
150	136
151	137	static __inline__ __m128i __DEFAULT_FN_ATTRS256
152		_mm256_mask_cvttspd_epu32(__m128i __W, __mmask8 __U, __m256d __A) {
153		return ((__m128i)__builtin_ia32_vcvttpd2udqs256_round_mask(
154		(__v4df)__A, (__v4si)__W, __U, _MM_FROUND_CUR_DIRECTION));
	138	_mm256_mask_cvtts_pd_epu32(__m128i __W, __mmask8 __U, __m256d __A) {
	139	return ((__m128i)__builtin_ia32_vcvttpd2udqs256_mask((__v4df)__A, (__v4si)__W,
	140	__U));
155	141	}
156	142
157	143	static __inline__ __m128i __DEFAULT_FN_ATTRS256
158		_mm256_maskz_cvttspd_epu32(__mmask8 __U, __m256d __A) {
159		return ((__m128i)__builtin_ia32_vcvttpd2udqs256_round_mask(
160		(__v4df)__A, (__v4si)_mm_setzero_si128(), __U, _MM_FROUND_CUR_DIRECTION));
	144	_mm256_maskz_cvtts_pd_epu32(__mmask8 __U, __m256d __A) {
	145	return ((__m128i)__builtin_ia32_vcvttpd2udqs256_mask(
	146	(__v4df)__A, (__v4si)_mm_setzero_si128(), __U));
161	147	}
162	148
163		#define _mm256_cvtts_roundpd_epu32(__A, __R) \
164		((__m128i)__builtin_ia32_vcvttpd2udqs256_round_mask( \
165		(__v4df)(__m256d)__A, (__v4si)(__m128i)_mm_undefined_si128(), \
166		(__mmask8) - 1, (int)(__R)))
167
168		#define _mm256_mask_cvtts_roundpd_epu32(__W, __U, __A, __R) \
169		((__m128i)__builtin_ia32_vcvttpd2udqs256_round_mask( \
170		(__v4df)(__m256d)__A, (__v4si)(__m128i)__W, (__mmask8)__U, (int)(__R)))
171
172		#define _mm256_maskz_cvtts_roundpd_epu32(__U, __A, __R) \
173		((__m128i)__builtin_ia32_vcvttpd2udqs256_round_mask( \
174		(__v4df)(__m256d)__A, (__v4si)(__m128i)_mm_setzero_si128(), \
175		(__mmask8)__U, (int)(__R)))
176
177	149	// 128 Bit : Double -> long
178		static __inline__ __m128i __DEFAULT_FN_ATTRS128 _mm_cvttspd_epi64(__m128d __A) {
	150	static __inline__ __m128i __DEFAULT_FN_ATTRS128
	151	_mm_cvtts_pd_epi64(__m128d __A) {
179	152	return ((__m128i)__builtin_ia32_vcvttpd2qqs128_mask(
180	153	(__v2df)__A, (__v2di)_mm_undefined_si128(), (__mmask8)-1));
181	154	}
182	155
183	156	static __inline__ __m128i __DEFAULT_FN_ATTRS128
184		_mm_mask_cvttspd_epi64(__m128i __W, __mmask8 __U, __m128d __A) {
	157	_mm_mask_cvtts_pd_epi64(__m128i __W, __mmask8 __U, __m128d __A) {
185	158	return ((__m128i)__builtin_ia32_vcvttpd2qqs128_mask((__v2df)__A, (__v2di)__W,
186	159	(__mmask8)__U));
187	160	}
188	161
189	162	static __inline__ __m128i __DEFAULT_FN_ATTRS128
190		_mm_maskz_cvttspd_epi64(__mmask8 __U, __m128d __A) {
	163	_mm_maskz_cvtts_pd_epi64(__mmask8 __U, __m128d __A) {
191	164	return ((__m128i)__builtin_ia32_vcvttpd2qqs128_mask(
192	165	(__v2df)__A, (__v2di)_mm_setzero_si128(), (__mmask8)__U));
193	166	}
194	167
195	168	// 256 Bit : Double -> long
196	169	static __inline__ __m256i __DEFAULT_FN_ATTRS256
197		_mm256_cvttspd_epi64(__m256d __A) {
198		return ((__m256i)__builtin_ia32_vcvttpd2qqs256_round_mask(
199		(__v4df)__A, (__v4di)_mm256_undefined_si256(), (__mmask8)-1,
200		_MM_FROUND_CUR_DIRECTION));
	170	_mm256_cvtts_pd_epi64(__m256d __A) {
	171	return ((__m256i)__builtin_ia32_vcvttpd2qqs256_mask(
	172	(__v4df)__A, (__v4di)_mm256_undefined_si256(), (__mmask8)-1));
201	173	}
202	174
203	175	static __inline__ __m256i __DEFAULT_FN_ATTRS256
204		_mm256_mask_cvttspd_epi64(__m256i __W, __mmask8 __U, __m256d __A) {
205		return ((__m256i)__builtin_ia32_vcvttpd2qqs256_round_mask(
206		(__v4df)__A, (__v4di)__W, __U, _MM_FROUND_CUR_DIRECTION));
	176	_mm256_mask_cvtts_pd_epi64(__m256i __W, __mmask8 __U, __m256d __A) {
	177	return ((__m256i)__builtin_ia32_vcvttpd2qqs256_mask((__v4df)__A, (__v4di)__W,
	178	__U));
207	179	}
208	180
209	181	static __inline__ __m256i __DEFAULT_FN_ATTRS256
210		_mm256_maskz_cvttspd_epi64(__mmask8 __U, __m256d __A) {
211		return ((__m256i)__builtin_ia32_vcvttpd2qqs256_round_mask(
212		(__v4df)__A, (__v4di)_mm256_setzero_si256(), __U,
213		_MM_FROUND_CUR_DIRECTION));
	182	_mm256_maskz_cvtts_pd_epi64(__mmask8 __U, __m256d __A) {
	183	return ((__m256i)__builtin_ia32_vcvttpd2qqs256_mask(
	184	(__v4df)__A, (__v4di)_mm256_setzero_si256(), __U));
214	185	}
215	186
216		#define _mm256_cvtts_roundpd_epi64(__A, __R) \
217		((__m256i)__builtin_ia32_vcvttpd2qqs256_round_mask( \
218		(__v4df)__A, (__v4di)_mm256_undefined_si256(), (__mmask8) - 1, \
219		(int)__R))
220
221		#define _mm256_mask_cvtts_roundpd_epi64(__W, __U, __A, __R) \
222		((__m256i)__builtin_ia32_vcvttpd2qqs256_round_mask((__v4df)__A, (__v4di)__W, \
223		(__mmask8)__U, (int)__R))
224
225		#define _mm256_maskz_cvtts_roundpd_epi64(__U, __A, __R) \
226		((__m256i)__builtin_ia32_vcvttpd2qqs256_round_mask( \
227		(__v4df)__A, (__v4di)_mm256_setzero_si256(), (__mmask8)__U, (int)__R))
228
229	187	// 128 Bit : Double -> ulong
230		static __inline__ __m128i __DEFAULT_FN_ATTRS128 _mm_cvttspd_epu64(__m128d __A) {
	188	static __inline__ __m128i __DEFAULT_FN_ATTRS128
	189	_mm_cvtts_pd_epu64(__m128d __A) {
231	190	return ((__m128i)__builtin_ia32_vcvttpd2uqqs128_mask(
232	191	(__v2df)__A, (__v2di)_mm_undefined_si128(), (__mmask8)-1));
233	192	}
234	193
235	194	static __inline__ __m128i __DEFAULT_FN_ATTRS128
236		_mm_mask_cvttspd_epu64(__m128i __W, __mmask8 __U, __m128d __A) {
	195	_mm_mask_cvtts_pd_epu64(__m128i __W, __mmask8 __U, __m128d __A) {
237	196	return ((__m128i)__builtin_ia32_vcvttpd2uqqs128_mask((__v2df)__A, (__v2di)__W,
238	197	(__mmask8)__U));
239	198	}
240	199
241	200	static __inline__ __m128i __DEFAULT_FN_ATTRS128
242		_mm_maskz_cvttspd_epu64(__mmask8 __U, __m128d __A) {
	201	_mm_maskz_cvtts_pd_epu64(__mmask8 __U, __m128d __A) {
243	202	return ((__m128i)__builtin_ia32_vcvttpd2uqqs128_mask(
244	203	(__v2df)__A, (__v2di)_mm_setzero_si128(), (__mmask8)__U));
245	204	}
...	...	@@ -247,105 +206,74 @@ _mm_maskz_cvttspd_epu64(__mmask8 __U, __m128d __A) {
247	206	// 256 Bit : Double -> ulong
248	207
249	208	static __inline__ __m256i __DEFAULT_FN_ATTRS256
250		_mm256_cvttspd_epu64(__m256d __A) {
251		return ((__m256i)__builtin_ia32_vcvttpd2uqqs256_round_mask(
252		(__v4df)__A, (__v4di)_mm256_undefined_si256(), (__mmask8)-1,
253		_MM_FROUND_CUR_DIRECTION));
	209	_mm256_cvtts_pd_epu64(__m256d __A) {
	210	return ((__m256i)__builtin_ia32_vcvttpd2uqqs256_mask(
	211	(__v4df)__A, (__v4di)_mm256_undefined_si256(), (__mmask8)-1));
254	212	}
255	213
256	214	static __inline__ __m256i __DEFAULT_FN_ATTRS256
257		_mm256_mask_cvttspd_epu64(__m256i __W, __mmask8 __U, __m256d __A) {
258		return ((__m256i)__builtin_ia32_vcvttpd2uqqs256_round_mask(
259		(__v4df)__A, (__v4di)__W, __U, _MM_FROUND_CUR_DIRECTION));
	215	_mm256_mask_cvtts_pd_epu64(__m256i __W, __mmask8 __U, __m256d __A) {
	216	return ((__m256i)__builtin_ia32_vcvttpd2uqqs256_mask((__v4df)__A, (__v4di)__W,
	217	__U));
260	218	}
261	219
262	220	static __inline__ __m256i __DEFAULT_FN_ATTRS256
263		_mm256_maskz_cvttspd_epu64(__mmask8 __U, __m256d __A) {
264		return ((__m256i)__builtin_ia32_vcvttpd2uqqs256_round_mask(
265		(__v4df)__A, (__v4di)_mm256_setzero_si256(), __U,
266		_MM_FROUND_CUR_DIRECTION));
	221	_mm256_maskz_cvtts_pd_epu64(__mmask8 __U, __m256d __A) {
	222	return ((__m256i)__builtin_ia32_vcvttpd2uqqs256_mask(
	223	(__v4df)__A, (__v4di)_mm256_setzero_si256(), __U));
267	224	}
268	225
269		#define _mm256_cvtts_roundpd_epu64(__A, __R) \
270		((__m256i)__builtin_ia32_vcvttpd2uqqs256_round_mask( \
271		(__v4df)__A, (__v4di)_mm256_undefined_si256(), (__mmask8) - 1, \
272		(int)__R))
273
274		#define _mm256_mask_cvtts_roundpd_epu64(__W, __U, __A, __R) \
275		((__m256i)__builtin_ia32_vcvttpd2uqqs256_round_mask( \
276		(__v4df)__A, (__v4di)__W, (__mmask8)__U, (int)__R))
277
278		#define _mm256_maskz_cvtts_roundpd_epu64(__U, __A, __R) \
279		((__m256i)__builtin_ia32_vcvttpd2uqqs256_round_mask( \
280		(__v4df)__A, (__v4di)_mm256_setzero_si256(), (__mmask8)__U, (int)__R))
281
282	226	// 128 Bit : float -> int
283		static __inline__ __m128i __DEFAULT_FN_ATTRS128 _mm_cvttsps_epi32(__m128 __A) {
	227	static __inline__ __m128i __DEFAULT_FN_ATTRS128 _mm_cvtts_ps_epi32(__m128 __A) {
284	228	return ((__m128i)__builtin_ia32_vcvttps2dqs128_mask(
285	229	(__v4sf)__A, (__v4si)(__m128i)_mm_undefined_si128(), (__mmask8)(-1)));
286	230	}
287	231
288	232	static __inline__ __m128i __DEFAULT_FN_ATTRS128
289		_mm_mask_cvttsps_epi32(__m128i __W, __mmask8 __U, __m128 __A) {
	233	_mm_mask_cvtts_ps_epi32(__m128i __W, __mmask8 __U, __m128 __A) {
290	234	return ((__m128i)__builtin_ia32_vcvttps2dqs128_mask((__v4sf)__A, (__v4si)__W,
291	235	(__mmask8)__U));
292	236	}
293	237
294	238	static __inline__ __m128i __DEFAULT_FN_ATTRS128
295		_mm_maskz_cvttsps_epi32(__mmask8 __U, __m128 __A) {
	239	_mm_maskz_cvtts_ps_epi32(__mmask8 __U, __m128 __A) {
296	240	return ((__m128i)__builtin_ia32_vcvttps2dqs128_mask(
297	241	(__v4sf)__A, (__v4si)(__m128i)_mm_setzero_si128(), (__mmask8)__U));
298	242	}
299	243
300	244	// 256 Bit : float -> int
301	245	static __inline__ __m256i __DEFAULT_FN_ATTRS256
302		_mm256_cvttsps_epi32(__m256 __A) {
303		return ((__m256i)__builtin_ia32_vcvttps2dqs256_round_mask(
304		(__v8sf)__A, (__v8si)_mm256_undefined_si256(), (__mmask8)-1,
305		_MM_FROUND_CUR_DIRECTION));
	246	_mm256_cvtts_ps_epi32(__m256 __A) {
	247	return ((__m256i)__builtin_ia32_vcvttps2dqs256_mask(
	248	(__v8sf)__A, (__v8si)_mm256_undefined_si256(), (__mmask8)-1));
306	249	}
307	250
308	251	static __inline__ __m256i __DEFAULT_FN_ATTRS256
309		_mm256_mask_cvttsps_epi32(__m256i __W, __mmask8 __U, __m256 __A) {
310		return ((__m256i)__builtin_ia32_vcvttps2dqs256_round_mask(
311		(__v8sf)__A, (__v8si)__W, __U, _MM_FROUND_CUR_DIRECTION));
	252	_mm256_mask_cvtts_ps_epi32(__m256i __W, __mmask8 __U, __m256 __A) {
	253	return ((__m256i)__builtin_ia32_vcvttps2dqs256_mask((__v8sf)__A, (__v8si)__W,
	254	__U));
312	255	}
313	256
314	257	static __inline__ __m256i __DEFAULT_FN_ATTRS256
315		_mm256_maskz_cvttsps_epi32(__mmask8 __U, __m256 __A) {
316		return ((__m256i)__builtin_ia32_vcvttps2dqs256_round_mask(
317		(__v8sf)__A, (__v8si)_mm256_setzero_si256(), __U,
318		_MM_FROUND_CUR_DIRECTION));
	258	_mm256_maskz_cvtts_ps_epi32(__mmask8 __U, __m256 __A) {
	259	return ((__m256i)__builtin_ia32_vcvttps2dqs256_mask(
	260	(__v8sf)__A, (__v8si)_mm256_setzero_si256(), __U));
319	261	}
320	262
321		#define _mm256_cvtts_roundps_epi32(__A, __R) \
322		((__m256i)__builtin_ia32_vcvttps2dqs256_round_mask( \
323		(__v8sf)(__m256)__A, (__v8si)(__m256i)_mm256_undefined_si256(), \
324		(__mmask8) - 1, (int)(__R)))
325
326		#define _mm256_mask_cvtts_roundps_epi32(__W, __U, __A, __R) \
327		((__m256i)__builtin_ia32_vcvttps2dqs256_round_mask( \
328		(__v8sf)(__m256)__A, (__v8si)(__m256i)__W, (__mmask8)__U, (int)(__R)))
329
330		#define _mm256_maskz_cvtts_roundps_epi32(__U, __A, __R) \
331		((__m256i)__builtin_ia32_vcvttps2dqs256_round_mask( \
332		(__v8sf)(__m256)__A, (__v8si)(__m256i)_mm256_setzero_si256(), \
333		(__mmask8)__U, (int)(__R)))
334
335	263	// 128 Bit : float -> uint
336		static __inline__ __m128i __DEFAULT_FN_ATTRS128 _mm_cvttsps_epu32(__m128 __A) {
	264	static __inline__ __m128i __DEFAULT_FN_ATTRS128 _mm_cvtts_ps_epu32(__m128 __A) {
337	265	return ((__m128i)__builtin_ia32_vcvttps2udqs128_mask(
338	266	(__v4sf)__A, (__v4si)(__m128i)_mm_undefined_si128(), (__mmask8)(-1)));
339	267	}
340	268
341	269	static __inline__ __m128i __DEFAULT_FN_ATTRS128
342		_mm_mask_cvttsps_epu32(__m128i __W, __mmask8 __U, __m128 __A) {
	270	_mm_mask_cvtts_ps_epu32(__m128i __W, __mmask8 __U, __m128 __A) {
343	271	return ((__m128i)__builtin_ia32_vcvttps2udqs128_mask((__v4sf)__A, (__v4si)__W,
344	272	(__mmask8)__U));
345	273	}
346	274
347	275	static __inline__ __m128i __DEFAULT_FN_ATTRS128
348		_mm_maskz_cvttsps_epu32(__mmask8 __U, __m128 __A) {
	276	_mm_maskz_cvtts_ps_epu32(__mmask8 __U, __m128 __A) {
349	277	return ((__m128i)__builtin_ia32_vcvttps2udqs128_mask(
350	278	(__v4sf)__A, (__v4si)_mm_setzero_si128(), (__mmask8)__U));
351	279	}
...	...	@@ -353,144 +281,96 @@ _mm_maskz_cvttsps_epu32(__mmask8 __U, __m128 __A) {
353	281	// 256 Bit : float -> uint
354	282
355	283	static __inline__ __m256i __DEFAULT_FN_ATTRS256
356		_mm256_cvttsps_epu32(__m256 __A) {
357		return ((__m256i)__builtin_ia32_vcvttps2udqs256_round_mask(
358		(__v8sf)__A, (__v8si)_mm256_undefined_si256(), (__mmask8)-1,
359		_MM_FROUND_CUR_DIRECTION));
	284	_mm256_cvtts_ps_epu32(__m256 __A) {
	285	return ((__m256i)__builtin_ia32_vcvttps2udqs256_mask(
	286	(__v8sf)__A, (__v8si)_mm256_undefined_si256(), (__mmask8)-1));
360	287	}
361	288
362	289	static __inline__ __m256i __DEFAULT_FN_ATTRS256
363		_mm256_mask_cvttsps_epu32(__m256i __W, __mmask8 __U, __m256 __A) {
364		return ((__m256i)__builtin_ia32_vcvttps2udqs256_round_mask(
365		(__v8sf)__A, (__v8si)__W, __U, _MM_FROUND_CUR_DIRECTION));
	290	_mm256_mask_cvtts_ps_epu32(__m256i __W, __mmask8 __U, __m256 __A) {
	291	return ((__m256i)__builtin_ia32_vcvttps2udqs256_mask((__v8sf)__A, (__v8si)__W,
	292	__U));
366	293	}
367	294
368	295	static __inline__ __m256i __DEFAULT_FN_ATTRS256
369		_mm256_maskz_cvttsps_epu32(__mmask8 __U, __m256 __A) {
370		return ((__m256i)__builtin_ia32_vcvttps2udqs256_round_mask(
371		(__v8sf)__A, (__v8si)_mm256_setzero_si256(), __U,
372		_MM_FROUND_CUR_DIRECTION));
	296	_mm256_maskz_cvtts_ps_epu32(__mmask8 __U, __m256 __A) {
	297	return ((__m256i)__builtin_ia32_vcvttps2udqs256_mask(
	298	(__v8sf)__A, (__v8si)_mm256_setzero_si256(), __U));
373	299	}
374	300
375		#define _mm256_cvtts_roundps_epu32(__A, __R) \
376		((__m256i)__builtin_ia32_vcvttps2udqs256_round_mask( \
377		(__v8sf)(__m256)__A, (__v8si)(__m256i)_mm256_undefined_si256(), \
378		(__mmask8) - 1, (int)(__R)))
379
380		#define _mm256_mask_cvtts_roundps_epu32(__W, __U, __A, __R) \
381		((__m256i)__builtin_ia32_vcvttps2udqs256_round_mask( \
382		(__v8sf)(__m256)__A, (__v8si)(__m256i)__W, (__mmask8)__U, (int)(__R)))
383
384		#define _mm256_maskz_cvtts_roundps_epu32(__U, __A, __R) \
385		((__m256i)__builtin_ia32_vcvttps2udqs256_round_mask( \
386		(__v8sf)(__m256)__A, (__v8si)(__m256i)_mm256_setzero_si256(), \
387		(__mmask8)__U, (int)(__R)))
388
389	301	// 128 bit : float -> long
390		static __inline__ __m128i __DEFAULT_FN_ATTRS128 _mm_cvttsps_epi64(__m128 __A) {
	302	static __inline__ __m128i __DEFAULT_FN_ATTRS128 _mm_cvtts_ps_epi64(__m128 __A) {
391	303	return ((__m128i)__builtin_ia32_vcvttps2qqs128_mask(
392	304	(__v4sf)__A, (__v2di)_mm_undefined_si128(), (__mmask8)-1));
393	305	}
394	306
395	307	static __inline__ __m128i __DEFAULT_FN_ATTRS128
396		_mm_mask_cvttsps_epi64(__m128i __W, __mmask8 __U, __m128 __A) {
	308	_mm_mask_cvtts_ps_epi64(__m128i __W, __mmask8 __U, __m128 __A) {
397	309	return ((__m128i)__builtin_ia32_vcvttps2qqs128_mask(
398	310	(__v4sf)__A, (__v2di)(__m128i)__W, (__mmask8)__U));
399	311	}
400	312
401	313	static __inline__ __m128i __DEFAULT_FN_ATTRS128
402		_mm_maskz_cvttsps_epi64(__mmask8 __U, __m128 __A) {
	314	_mm_maskz_cvtts_ps_epi64(__mmask8 __U, __m128 __A) {
403	315	return ((__m128i)__builtin_ia32_vcvttps2qqs128_mask(
404	316	(__v4sf)__A, (__v2di)_mm_setzero_si128(), (__mmask8)__U));
405	317	}
406	318	// 256 bit : float -> long
407	319
408	320	static __inline__ __m256i __DEFAULT_FN_ATTRS256
409		_mm256_cvttsps_epi64(__m128 __A) {
410		return ((__m256i)__builtin_ia32_vcvttps2qqs256_round_mask(
411		(__v4sf)__A, (__v4di)_mm256_undefined_si256(), (__mmask8)-1,
412		_MM_FROUND_CUR_DIRECTION));
	321	_mm256_cvtts_ps_epi64(__m128 __A) {
	322	return ((__m256i)__builtin_ia32_vcvttps2qqs256_mask(
	323	(__v4sf)__A, (__v4di)_mm256_undefined_si256(), (__mmask8)-1));
413	324	}
414	325	static __inline__ __m256i __DEFAULT_FN_ATTRS256
415		_mm256_mask_cvttsps_epi64(__m256i __W, __mmask8 __U, __m128 __A) {
416		return ((__m256i)__builtin_ia32_vcvttps2qqs256_round_mask(
417		(__v4sf)__A, (__v4di)__W, __U, _MM_FROUND_CUR_DIRECTION));
	326	_mm256_mask_cvtts_ps_epi64(__m256i __W, __mmask8 __U, __m128 __A) {
	327	return ((__m256i)__builtin_ia32_vcvttps2qqs256_mask((__v4sf)__A, (__v4di)__W,
	328	__U));
418	329	}
419	330
420	331	static __inline__ __m256i __DEFAULT_FN_ATTRS256
421		_mm256_maskz_cvttsps_epi64(__mmask8 __U, __m128 __A) {
422		return ((__m256i)__builtin_ia32_vcvttps2qqs256_round_mask(
423		(__v4sf)__A, (__v4di)_mm256_setzero_si256(), __U,
424		_MM_FROUND_CUR_DIRECTION));
	332	_mm256_maskz_cvtts_ps_epi64(__mmask8 __U, __m128 __A) {
	333	return ((__m256i)__builtin_ia32_vcvttps2qqs256_mask(
	334	(__v4sf)__A, (__v4di)_mm256_setzero_si256(), __U));
425	335	}
426	336
427		#define _mm256_cvtts_roundps_epi64(__A, __R) \
428		((__m256i)__builtin_ia32_vcvttps2qqs256_round_mask( \
429		(__v4sf)(__m128)__A, (__v4di)_mm256_undefined_si256(), (__mmask8) - 1, \
430		(int)__R))
431
432		#define _mm256_mask_cvtts_roundps_epi64(__W, __U, __A, __R) \
433		((__m256i)__builtin_ia32_vcvttps2qqs256_round_mask( \
434		(__v4sf)(__m128)__A, (__v4di)__W, (__mmask8)__U, (int)__R))
435
436		#define _mm256_maskz_cvtts_roundps_epi64(__U, __A, __R) \
437		((__m256i)__builtin_ia32_vcvttps2qqs256_round_mask( \
438		(__v4sf)(__m128)__A, (__v4di)_mm256_setzero_si256(), (__mmask8)__U, \
439		(int)__R))
440
441	337	// 128 bit : float -> ulong
442		static __inline__ __m128i __DEFAULT_FN_ATTRS128 _mm_cvttsps_epu64(__m128 __A) {
	338	static __inline__ __m128i __DEFAULT_FN_ATTRS128 _mm_cvtts_ps_epu64(__m128 __A) {
443	339	return ((__m128i)__builtin_ia32_vcvttps2uqqs128_mask(
444	340	(__v4sf)__A, (__v2di)_mm_undefined_si128(), (__mmask8)-1));
445	341	}
446	342
447	343	static __inline__ __m128i __DEFAULT_FN_ATTRS128
448		_mm_mask_cvttsps_epu64(__m128i __W, __mmask8 __U, __m128 __A) {
	344	_mm_mask_cvtts_ps_epu64(__m128i __W, __mmask8 __U, __m128 __A) {
449	345	return ((__m128i)__builtin_ia32_vcvttps2uqqs128_mask(
450	346	(__v4sf)__A, (__v2di)(__m128i)__W, (__mmask8)__U));
451	347	}
452	348
453	349	static __inline__ __m128i __DEFAULT_FN_ATTRS128
454		_mm_maskz_cvttsps_epu64(__mmask8 __U, __m128 __A) {
	350	_mm_maskz_cvtts_ps_epu64(__mmask8 __U, __m128 __A) {
455	351	return ((__m128i)__builtin_ia32_vcvttps2uqqs128_mask(
456	352	(__v4sf)__A, (__v2di)_mm_setzero_si128(), (__mmask8)__U));
457	353	}
458	354	// 256 bit : float -> ulong
459	355
460	356	static __inline__ __m256i __DEFAULT_FN_ATTRS256
461		_mm256_cvttsps_epu64(__m128 __A) {
462		return ((__m256i)__builtin_ia32_vcvttps2uqqs256_round_mask(
463		(__v4sf)__A, (__v4di)_mm256_undefined_si256(), (__mmask8)-1,
464		_MM_FROUND_CUR_DIRECTION));
	357	_mm256_cvtts_ps_epu64(__m128 __A) {
	358	return ((__m256i)__builtin_ia32_vcvttps2uqqs256_mask(
	359	(__v4sf)__A, (__v4di)_mm256_undefined_si256(), (__mmask8)-1));
465	360	}
466	361
467	362	static __inline__ __m256i __DEFAULT_FN_ATTRS256
468		_mm256_mask_cvttsps_epu64(__m256i __W, __mmask8 __U, __m128 __A) {
469		return ((__m256i)__builtin_ia32_vcvttps2uqqs256_round_mask(
470		(__v4sf)__A, (__v4di)__W, __U, _MM_FROUND_CUR_DIRECTION));
	363	_mm256_mask_cvtts_ps_epu64(__m256i __W, __mmask8 __U, __m128 __A) {
	364	return ((__m256i)__builtin_ia32_vcvttps2uqqs256_mask((__v4sf)__A, (__v4di)__W,
	365	__U));
471	366	}
472	367
473	368	static __inline__ __m256i __DEFAULT_FN_ATTRS256
474		_mm256_maskz_cvttsps_epu64(__mmask8 __U, __m128 __A) {
475		return ((__m256i)__builtin_ia32_vcvttps2uqqs256_round_mask(
476		(__v4sf)__A, (__v4di)_mm256_setzero_si256(), __U,
477		_MM_FROUND_CUR_DIRECTION));
	369	_mm256_maskz_cvtts_ps_epu64(__mmask8 __U, __m128 __A) {
	370	return ((__m256i)__builtin_ia32_vcvttps2uqqs256_mask(
	371	(__v4sf)__A, (__v4di)_mm256_setzero_si256(), __U));
478	372	}
479	373
480		#define _mm256_cvtts_roundps_epu64(__A, __R) \
481		((__m256i)__builtin_ia32_vcvttps2uqqs256_round_mask( \
482		(__v4sf)(__m128)__A, (__v4di)_mm256_undefined_si256(), (__mmask8) - 1, \
483		(int)__R))
484
485		#define _mm256_mask_cvtts_roundps_epu64(__W, __U, __A, __R) \
486		((__m256i)__builtin_ia32_vcvttps2uqqs256_round_mask( \
487		(__v4sf)(__m128)__A, (__v4di)__W, (__mmask8)__U, (int)__R))
488
489		#define _mm256_maskz_cvtts_roundps_epu64(__U, __A, __R) \
490		((__m256i)__builtin_ia32_vcvttps2uqqs256_round_mask( \
491		(__v4sf)(__m128)__A, (__v4di)_mm256_setzero_si256(), (__mmask8)__U, \
492		(int)__R))
493
494	374	#undef __DEFAULT_FN_ATTRS128
495	375	#undef __DEFAULT_FN_ATTRS256
496	376	#endif // __AVX10_2SATCVTDSINTRIN_H

lib/include/avx10_2satcvtintrin.h+112-244

...	...	@@ -14,431 +14,299 @@
14	14	#ifndef __AVX10_2SATCVTINTRIN_H
15	15	#define __AVX10_2SATCVTINTRIN_H
16	16
17		#define _mm_ipcvtbf16_epi8(A) \
	17	#define _mm_ipcvts_bf16_epi8(A) \
18	18	((__m128i)__builtin_ia32_vcvtbf162ibs128((__v8bf)(__m128bh)(A)))
19	19
20		#define _mm_mask_ipcvtbf16_epi8(W, U, A) \
	20	#define _mm_mask_ipcvts_bf16_epi8(W, U, A) \
21	21	((__m128i)__builtin_ia32_selectw_128( \
22		(__mmask8)(U), (__v8hi)_mm_ipcvtbf16_epi8(A), (__v8hi)(__m128i)(W)))
	22	(__mmask8)(U), (__v8hi)_mm_ipcvts_bf16_epi8(A), (__v8hi)(__m128i)(W)))
23	23
24		#define _mm_maskz_ipcvtbf16_epi8(U, A) \
	24	#define _mm_maskz_ipcvts_bf16_epi8(U, A) \
25	25	((__m128i)__builtin_ia32_selectw_128((__mmask8)(U), \
26		(__v8hi)_mm_ipcvtbf16_epi8(A), \
	26	(__v8hi)_mm_ipcvts_bf16_epi8(A), \
27	27	(__v8hi)_mm_setzero_si128()))
28	28
29		#define _mm256_ipcvtbf16_epi8(A) \
	29	#define _mm256_ipcvts_bf16_epi8(A) \
30	30	((__m256i)__builtin_ia32_vcvtbf162ibs256((__v16bf)(__m256bh)(A)))
31	31
32		#define _mm256_mask_ipcvtbf16_epi8(W, U, A) \
	32	#define _mm256_mask_ipcvts_bf16_epi8(W, U, A) \
33	33	((__m256i)__builtin_ia32_selectw_256((__mmask16)(U), \
34		(__v16hi)_mm256_ipcvtbf16_epi8(A), \
	34	(__v16hi)_mm256_ipcvts_bf16_epi8(A), \
35	35	(__v16hi)(__m256i)(W)))
36	36
37		#define _mm256_maskz_ipcvtbf16_epi8(U, A) \
	37	#define _mm256_maskz_ipcvts_bf16_epi8(U, A) \
38	38	((__m256i)__builtin_ia32_selectw_256((__mmask16)(U), \
39		(__v16hi)_mm256_ipcvtbf16_epi8(A), \
	39	(__v16hi)_mm256_ipcvts_bf16_epi8(A), \
40	40	(__v16hi)_mm256_setzero_si256()))
41	41
42		#define _mm_ipcvtbf16_epu8(A) \
	42	#define _mm_ipcvts_bf16_epu8(A) \
43	43	((__m128i)__builtin_ia32_vcvtbf162iubs128((__v8bf)(__m128bh)(A)))
44	44
45		#define _mm_mask_ipcvtbf16_epu8(W, U, A) \
	45	#define _mm_mask_ipcvts_bf16_epu8(W, U, A) \
46	46	((__m128i)__builtin_ia32_selectw_128( \
47		(__mmask8)(U), (__v8hi)_mm_ipcvtbf16_epu8(A), (__v8hi)(__m128i)(W)))
	47	(__mmask8)(U), (__v8hi)_mm_ipcvts_bf16_epu8(A), (__v8hi)(__m128i)(W)))
48	48
49		#define _mm_maskz_ipcvtbf16_epu8(U, A) \
	49	#define _mm_maskz_ipcvts_bf16_epu8(U, A) \
50	50	((__m128i)__builtin_ia32_selectw_128((__mmask8)(U), \
51		(__v8hi)_mm_ipcvtbf16_epu8(A), \
	51	(__v8hi)_mm_ipcvts_bf16_epu8(A), \
52	52	(__v8hi)_mm_setzero_si128()))
53	53
54		#define _mm256_ipcvtbf16_epu8(A) \
	54	#define _mm256_ipcvts_bf16_epu8(A) \
55	55	((__m256i)__builtin_ia32_vcvtbf162iubs256((__v16bf)(__m256bh)(A)))
56	56
57		#define _mm256_mask_ipcvtbf16_epu8(W, U, A) \
	57	#define _mm256_mask_ipcvts_bf16_epu8(W, U, A) \
58	58	((__m256i)__builtin_ia32_selectw_256((__mmask16)(U), \
59		(__v16hi)_mm256_ipcvtbf16_epu8(A), \
	59	(__v16hi)_mm256_ipcvts_bf16_epu8(A), \
60	60	(__v16hi)(__m256i)(W)))
61	61
62		#define _mm256_maskz_ipcvtbf16_epu8(U, A) \
	62	#define _mm256_maskz_ipcvts_bf16_epu8(U, A) \
63	63	((__m256i)__builtin_ia32_selectw_256((__mmask16)(U), \
64		(__v16hi)_mm256_ipcvtbf16_epu8(A), \
	64	(__v16hi)_mm256_ipcvts_bf16_epu8(A), \
65	65	(__v16hi)_mm256_setzero_si256()))
66	66
67		#define _mm_ipcvtph_epi8(A) \
	67	#define _mm_ipcvts_ph_epi8(A) \
68	68	((__m128i)__builtin_ia32_vcvtph2ibs128_mask( \
69	69	(__v8hf)(__m128h)(A), (__v8hu)_mm_setzero_si128(), (__mmask8)-1))
70	70
71		#define _mm_mask_ipcvtph_epi8(W, U, A) \
	71	#define _mm_mask_ipcvts_ph_epi8(W, U, A) \
72	72	((__m128i)__builtin_ia32_vcvtph2ibs128_mask((__v8hf)(__m128h)(A), \
73	73	(__v8hu)(W), (__mmask8)(U)))
74	74
75		#define _mm_maskz_ipcvtph_epi8(U, A) \
	75	#define _mm_maskz_ipcvts_ph_epi8(U, A) \
76	76	((__m128i)__builtin_ia32_vcvtph2ibs128_mask( \
77	77	(__v8hf)(__m128h)(A), (__v8hu)(_mm_setzero_si128()), (__mmask8)(U)))
78	78
79		#define _mm256_ipcvtph_epi8(A) \
	79	#define _mm256_ipcvts_ph_epi8(A) \
80	80	((__m256i)__builtin_ia32_vcvtph2ibs256_mask( \
81		(__v16hf)(__m256h)(A), (__v16hu)_mm256_setzero_si256(), (__mmask16)-1, \
82		_MM_FROUND_CUR_DIRECTION))
	81	(__v16hf)(__m256h)(A), (__v16hu)_mm256_setzero_si256(), (__mmask16)-1))
83	82
84		#define _mm256_mask_ipcvtph_epi8(W, U, A) \
	83	#define _mm256_mask_ipcvts_ph_epi8(W, U, A) \
85	84	((__m256i)__builtin_ia32_vcvtph2ibs256_mask((__v16hf)(__m256h)(A), \
86		(__v16hu)(W), (__mmask16)(U), \
87		_MM_FROUND_CUR_DIRECTION))
	85	(__v16hu)(W), (__mmask16)(U)))
88	86
89		#define _mm256_maskz_ipcvtph_epi8(U, A) \
	87	#define _mm256_maskz_ipcvts_ph_epi8(U, A) \
90	88	((__m256i)__builtin_ia32_vcvtph2ibs256_mask( \
91	89	(__v16hf)(__m256h)(A), (__v16hu)(_mm256_setzero_si256()), \
92		(__mmask16)(U), _MM_FROUND_CUR_DIRECTION))
	90	(__mmask16)(U)))
93	91
94		#define _mm256_ipcvt_roundph_epi8(A, R) \
95		((__m256i)__builtin_ia32_vcvtph2ibs256_mask((__v16hf)(__m256h)(A), \
96		(__v16hu)_mm256_setzero_si256(), \
97		(__mmask16)-1, (const int)R))
98
99		#define _mm256_mask_ipcvt_roundph_epi8(W, U, A, R) \
100		((__m256i)__builtin_ia32_vcvtph2ibs256_mask( \
101		(__v16hf)(__m256h)(A), (__v16hu)(W), (__mmask16)(U), (const int)R))
102
103		#define _mm256_maskz_ipcvt_roundph_epi8(U, A, R) \
104		((__m256i)__builtin_ia32_vcvtph2ibs256_mask((__v16hf)(__m256h)(A), \
105		(__v16hu)_mm256_setzero_si256(), \
106		(__mmask16)(U), (const int)R))
107
108		#define _mm_ipcvtph_epu8(A) \
	92	#define _mm_ipcvts_ph_epu8(A) \
109	93	((__m128i)__builtin_ia32_vcvtph2iubs128_mask( \
110	94	(__v8hf)(__m128h)(A), (__v8hu)_mm_setzero_si128(), (__mmask8)-1))
111	95
112		#define _mm_mask_ipcvtph_epu8(W, U, A) \
	96	#define _mm_mask_ipcvts_ph_epu8(W, U, A) \
113	97	((__m128i)__builtin_ia32_vcvtph2iubs128_mask((__v8hf)(__m128h)(A), \
114	98	(__v8hu)(W), (__mmask8)(U)))
115	99
116		#define _mm_maskz_ipcvtph_epu8(U, A) \
	100	#define _mm_maskz_ipcvts_ph_epu8(U, A) \
117	101	((__m128i)__builtin_ia32_vcvtph2iubs128_mask( \
118	102	(__v8hf)(__m128h)(A), (__v8hu)(_mm_setzero_si128()), (__mmask8)(U)))
119	103
120		#define _mm256_ipcvtph_epu8(A) \
	104	#define _mm256_ipcvts_ph_epu8(A) \
121	105	((__m256i)__builtin_ia32_vcvtph2iubs256_mask( \
122		(__v16hf)(__m256h)(A), (__v16hu)_mm256_setzero_si256(), (__mmask16)-1, \
123		_MM_FROUND_CUR_DIRECTION))
	106	(__v16hf)(__m256h)(A), (__v16hu)_mm256_setzero_si256(), (__mmask16)-1))
124	107
125		#define _mm256_mask_ipcvtph_epu8(W, U, A) \
	108	#define _mm256_mask_ipcvts_ph_epu8(W, U, A) \
126	109	((__m256i)__builtin_ia32_vcvtph2iubs256_mask((__v16hf)(__m256h)(A), \
127		(__v16hu)(W), (__mmask16)(U), \
128		_MM_FROUND_CUR_DIRECTION))
	110	(__v16hu)(W), (__mmask16)(U)))
129	111
130		#define _mm256_maskz_ipcvtph_epu8(U, A) \
	112	#define _mm256_maskz_ipcvts_ph_epu8(U, A) \
131	113	((__m256i)__builtin_ia32_vcvtph2iubs256_mask( \
132	114	(__v16hf)(__m256h)(A), (__v16hu)(_mm256_setzero_si256()), \
133		(__mmask16)(U), _MM_FROUND_CUR_DIRECTION))
134
135		#define _mm256_ipcvt_roundph_epu8(A, R) \
136		((__m256i)__builtin_ia32_vcvtph2iubs256_mask( \
137		(__v16hf)(__m256h)(A), (__v16hu)_mm256_setzero_si256(), (__mmask16)-1, \
138		(const int)R))
139
140		#define _mm256_mask_ipcvt_roundph_epu8(W, U, A, R) \
141		((__m256i)__builtin_ia32_vcvtph2iubs256_mask( \
142		(__v16hf)(__m256h)(A), (__v16hu)(W), (__mmask16)(U), (const int)R))
143
144		#define _mm256_maskz_ipcvt_roundph_epu8(U, A, R) \
145		((__m256i)__builtin_ia32_vcvtph2iubs256_mask( \
146		(__v16hf)(__m256h)(A), (__v16hu)_mm256_setzero_si256(), (__mmask16)(U), \
147		(const int)R))
	115	(__mmask16)(U)))
148	116
149		#define _mm_ipcvtps_epi8(A) \
	117	#define _mm_ipcvts_ps_epi8(A) \
150	118	((__m128i)__builtin_ia32_vcvtps2ibs128_mask( \
151	119	(__v4sf)(__m128)(A), (__v4su)_mm_setzero_si128(), (__mmask8)-1))
152	120
153		#define _mm_mask_ipcvtps_epi8(W, U, A) \
	121	#define _mm_mask_ipcvts_ps_epi8(W, U, A) \
154	122	((__m128i)__builtin_ia32_vcvtps2ibs128_mask((__v4sf)(__m128)(A), \
155	123	(__v4su)(W), (__mmask8)(U)))
156	124
157		#define _mm_maskz_ipcvtps_epi8(U, A) \
	125	#define _mm_maskz_ipcvts_ps_epi8(U, A) \
158	126	((__m128i)__builtin_ia32_vcvtps2ibs128_mask( \
159	127	(__v4sf)(__m128)(A), (__v4su)(_mm_setzero_si128()), (__mmask8)(U)))
160	128
161		#define _mm256_ipcvtps_epi8(A) \
	129	#define _mm256_ipcvts_ps_epi8(A) \
162	130	((__m256i)__builtin_ia32_vcvtps2ibs256_mask( \
163		(__v8sf)(__m256)(A), (__v8su)_mm256_setzero_si256(), (__mmask8)-1, \
164		_MM_FROUND_CUR_DIRECTION))
	131	(__v8sf)(__m256)(A), (__v8su)_mm256_setzero_si256(), (__mmask8)-1))
165	132
166		#define _mm256_mask_ipcvtps_epi8(W, U, A) \
	133	#define _mm256_mask_ipcvts_ps_epi8(W, U, A) \
167	134	((__m256i)__builtin_ia32_vcvtps2ibs256_mask((__v8sf)(__m256)(A), \
168		(__v8su)(W), (__mmask8)(U), \
169		_MM_FROUND_CUR_DIRECTION))
	135	(__v8su)(W), (__mmask8)(U)))
170	136
171		#define _mm256_maskz_ipcvtps_epi8(U, A) \
	137	#define _mm256_maskz_ipcvts_ps_epi8(U, A) \
172	138	((__m256i)__builtin_ia32_vcvtps2ibs256_mask( \
173		(__v8sf)(__m256)(A), (__v8su)(_mm256_setzero_si256()), (__mmask8)(U), \
174		_MM_FROUND_CUR_DIRECTION))
	139	(__v8sf)(__m256)(A), (__v8su)(_mm256_setzero_si256()), (__mmask8)(U)))
175	140
176		#define _mm256_ipcvt_roundps_epi8(A, R) \
177		((__m256i)__builtin_ia32_vcvtps2ibs256_mask((__v8sf)(__m256)(A), \
178		(__v8su)_mm256_setzero_si256(), \
179		(__mmask8)-1, (const int)R))
180
181		#define _mm256_mask_ipcvt_roundps_epi8(W, U, A, R) \
182		((__m256i)__builtin_ia32_vcvtps2ibs256_mask( \
183		(__v8sf)(__m256)(A), (__v8su)(W), (__mmask8)(U), (const int)R))
184
185		#define _mm256_maskz_ipcvt_roundps_epi8(U, A, R) \
186		((__m256i)__builtin_ia32_vcvtps2ibs256_mask((__v8sf)(__m256)(A), \
187		(__v8su)_mm256_setzero_si256(), \
188		(__mmask8)(U), (const int)R))
189
190		#define _mm_ipcvtps_epu8(A) \
	141	#define _mm_ipcvts_ps_epu8(A) \
191	142	((__m128i)__builtin_ia32_vcvtps2iubs128_mask( \
192	143	(__v4sf)(__m128)(A), (__v4su)_mm_setzero_si128(), (__mmask8)-1))
193	144
194		#define _mm_mask_ipcvtps_epu8(W, U, A) \
	145	#define _mm_mask_ipcvts_ps_epu8(W, U, A) \
195	146	((__m128i)__builtin_ia32_vcvtps2iubs128_mask((__v4sf)(__m128)(A), \
196	147	(__v4su)(W), (__mmask8)(U)))
197	148
198		#define _mm_maskz_ipcvtps_epu8(U, A) \
	149	#define _mm_maskz_ipcvts_ps_epu8(U, A) \
199	150	((__m128i)__builtin_ia32_vcvtps2iubs128_mask( \
200	151	(__v4sf)(__m128)(A), (__v4su)(_mm_setzero_si128()), (__mmask8)(U)))
201	152
202		#define _mm256_ipcvtps_epu8(A) \
	153	#define _mm256_ipcvts_ps_epu8(A) \
203	154	((__m256i)__builtin_ia32_vcvtps2iubs256_mask( \
204		(__v8sf)(__m256)(A), (__v8su)_mm256_setzero_si256(), (__mmask8)-1, \
205		_MM_FROUND_CUR_DIRECTION))
	155	(__v8sf)(__m256)(A), (__v8su)_mm256_setzero_si256(), (__mmask8)-1))
206	156
207		#define _mm256_mask_ipcvtps_epu8(W, U, A) \
	157	#define _mm256_mask_ipcvts_ps_epu8(W, U, A) \
208	158	((__m256i)__builtin_ia32_vcvtps2iubs256_mask((__v8sf)(__m256)(A), \
209		(__v8su)(W), (__mmask8)(U), \
210		_MM_FROUND_CUR_DIRECTION))
	159	(__v8su)(W), (__mmask8)(U)))
211	160
212		#define _mm256_maskz_ipcvtps_epu8(U, A) \
	161	#define _mm256_maskz_ipcvts_ps_epu8(U, A) \
213	162	((__m256i)__builtin_ia32_vcvtps2iubs256_mask( \
214		(__v8sf)(__m256)(A), (__v8su)(_mm256_setzero_si256()), (__mmask8)(U), \
215		_MM_FROUND_CUR_DIRECTION))
216
217		#define _mm256_ipcvt_roundps_epu8(A, R) \
218		((__m256i)__builtin_ia32_vcvtps2iubs256_mask((__v8sf)(__m256)(A), \
219		(__v8su)_mm256_setzero_si256(), \
220		(__mmask8)-1, (const int)R))
221
222		#define _mm256_mask_ipcvt_roundps_epu8(W, U, A, R) \
223		((__m256i)__builtin_ia32_vcvtps2iubs256_mask( \
224		(__v8sf)(__m256)(A), (__v8su)(W), (__mmask8)(U), (const int)R))
225
226		#define _mm256_maskz_ipcvt_roundps_epu8(U, A, R) \
227		((__m256i)__builtin_ia32_vcvtps2iubs256_mask((__v8sf)(__m256)(A), \
228		(__v8su)_mm256_setzero_si256(), \
229		(__mmask8)(U), (const int)R))
	163	(__v8sf)(__m256)(A), (__v8su)(_mm256_setzero_si256()), (__mmask8)(U)))
230	164
231		#define _mm_ipcvttbf16_epi8(A) \
	165	#define _mm_ipcvtts_bf16_epi8(A) \
232	166	((__m128i)__builtin_ia32_vcvttbf162ibs128((__v8bf)(__m128bh)(A)))
233	167
234		#define _mm_mask_ipcvttbf16_epi8(W, U, A) \
	168	#define _mm_mask_ipcvtts_bf16_epi8(W, U, A) \
235	169	((__m128i)__builtin_ia32_selectw_128( \
236		(__mmask8)(U), (__v8hi)_mm_ipcvttbf16_epi8(A), (__v8hi)(__m128i)(W)))
	170	(__mmask8)(U), (__v8hi)_mm_ipcvtts_bf16_epi8(A), (__v8hi)(__m128i)(W)))
237	171
238		#define _mm_maskz_ipcvttbf16_epi8(U, A) \
	172	#define _mm_maskz_ipcvtts_bf16_epi8(U, A) \
239	173	((__m128i)__builtin_ia32_selectw_128((__mmask8)(U), \
240		(__v8hi)_mm_ipcvttbf16_epi8(A), \
	174	(__v8hi)_mm_ipcvtts_bf16_epi8(A), \
241	175	(__v8hi)_mm_setzero_si128()))
242	176
243		#define _mm256_ipcvttbf16_epi8(A) \
	177	#define _mm256_ipcvtts_bf16_epi8(A) \
244	178	((__m256i)__builtin_ia32_vcvttbf162ibs256((__v16bf)(__m256bh)(A)))
245	179
246		#define _mm256_mask_ipcvttbf16_epi8(W, U, A) \
	180	#define _mm256_mask_ipcvtts_bf16_epi8(W, U, A) \
247	181	((__m256i)__builtin_ia32_selectw_256((__mmask16)(U), \
248		(__v16hi)_mm256_ipcvttbf16_epi8(A), \
	182	(__v16hi)_mm256_ipcvtts_bf16_epi8(A), \
249	183	(__v16hi)(__m256i)(W)))
250	184
251		#define _mm256_maskz_ipcvttbf16_epi8(U, A) \
	185	#define _mm256_maskz_ipcvtts_bf16_epi8(U, A) \
252	186	((__m256i)__builtin_ia32_selectw_256((__mmask16)(U), \
253		(__v16hi)_mm256_ipcvttbf16_epi8(A), \
	187	(__v16hi)_mm256_ipcvtts_bf16_epi8(A), \
254	188	(__v16hi)_mm256_setzero_si256()))
255	189
256		#define _mm_ipcvttbf16_epu8(A) \
	190	#define _mm_ipcvtts_bf16_epu8(A) \
257	191	((__m128i)__builtin_ia32_vcvttbf162iubs128((__v8bf)(__m128bh)(A)))
258	192
259		#define _mm_mask_ipcvttbf16_epu8(W, U, A) \
	193	#define _mm_mask_ipcvtts_bf16_epu8(W, U, A) \
260	194	((__m128i)__builtin_ia32_selectw_128( \
261		(__mmask8)(U), (__v8hi)_mm_ipcvttbf16_epu8(A), (__v8hi)(__m128i)(W)))
	195	(__mmask8)(U), (__v8hi)_mm_ipcvtts_bf16_epu8(A), (__v8hi)(__m128i)(W)))
262	196
263		#define _mm_maskz_ipcvttbf16_epu8(U, A) \
	197	#define _mm_maskz_ipcvtts_bf16_epu8(U, A) \
264	198	((__m128i)__builtin_ia32_selectw_128((__mmask8)(U), \
265		(__v8hi)_mm_ipcvttbf16_epu8(A), \
	199	(__v8hi)_mm_ipcvtts_bf16_epu8(A), \
266	200	(__v8hi)_mm_setzero_si128()))
267	201
268		#define _mm256_ipcvttbf16_epu8(A) \
	202	#define _mm256_ipcvtts_bf16_epu8(A) \
269	203	((__m256i)__builtin_ia32_vcvttbf162iubs256((__v16bf)(__m256bh)(A)))
270	204
271		#define _mm256_mask_ipcvttbf16_epu8(W, U, A) \
	205	#define _mm256_mask_ipcvtts_bf16_epu8(W, U, A) \
272	206	((__m256i)__builtin_ia32_selectw_256((__mmask16)(U), \
273		(__v16hi)_mm256_ipcvttbf16_epu8(A), \
	207	(__v16hi)_mm256_ipcvtts_bf16_epu8(A), \
274	208	(__v16hi)(__m256i)(W)))
275	209
276		#define _mm256_maskz_ipcvttbf16_epu8(U, A) \
	210	#define _mm256_maskz_ipcvtts_bf16_epu8(U, A) \
277	211	((__m256i)__builtin_ia32_selectw_256((__mmask16)(U), \
278		(__v16hi)_mm256_ipcvttbf16_epu8(A), \
	212	(__v16hi)_mm256_ipcvtts_bf16_epu8(A), \
279	213	(__v16hi)_mm256_setzero_si256()))
280	214
281		#define _mm_ipcvttph_epi8(A) \
	215	#define _mm_ipcvtts_ph_epi8(A) \
282	216	((__m128i)__builtin_ia32_vcvttph2ibs128_mask( \
283	217	(__v8hf)(__m128h)(A), (__v8hu)_mm_setzero_si128(), (__mmask8)-1))
284	218
285		#define _mm_mask_ipcvttph_epi8(W, U, A) \
	219	#define _mm_mask_ipcvtts_ph_epi8(W, U, A) \
286	220	((__m128i)__builtin_ia32_vcvttph2ibs128_mask((__v8hf)(__m128h)(A), \
287	221	(__v8hu)(W), (__mmask8)(U)))
288	222
289		#define _mm_maskz_ipcvttph_epi8(U, A) \
	223	#define _mm_maskz_ipcvtts_ph_epi8(U, A) \
290	224	((__m128i)__builtin_ia32_vcvttph2ibs128_mask( \
291	225	(__v8hf)(__m128h)(A), (__v8hu)(_mm_setzero_si128()), (__mmask8)(U)))
292	226
293		#define _mm256_ipcvttph_epi8(A) \
	227	#define _mm256_ipcvtts_ph_epi8(A) \
294	228	((__m256i)__builtin_ia32_vcvttph2ibs256_mask( \
295		(__v16hf)(__m256h)(A), (__v16hu)_mm256_setzero_si256(), (__mmask16)-1, \
296		_MM_FROUND_CUR_DIRECTION))
	229	(__v16hf)(__m256h)(A), (__v16hu)_mm256_setzero_si256(), (__mmask16)-1))
297	230
298		#define _mm256_mask_ipcvttph_epi8(W, U, A) \
	231	#define _mm256_mask_ipcvtts_ph_epi8(W, U, A) \
299	232	((__m256i)__builtin_ia32_vcvttph2ibs256_mask((__v16hf)(__m256h)(A), \
300		(__v16hu)(W), (__mmask16)(U), \
301		_MM_FROUND_CUR_DIRECTION))
	233	(__v16hu)(W), (__mmask16)(U)))
302	234
303		#define _mm256_maskz_ipcvttph_epi8(U, A) \
	235	#define _mm256_maskz_ipcvtts_ph_epi8(U, A) \
304	236	((__m256i)__builtin_ia32_vcvttph2ibs256_mask( \
305	237	(__v16hf)(__m256h)(A), (__v16hu)(_mm256_setzero_si256()), \
306		(__mmask16)(U), _MM_FROUND_CUR_DIRECTION))
	238	(__mmask16)(U)))
307	239
308		#define _mm256_ipcvtt_roundph_epi8(A, R) \
309		((__m256i)__builtin_ia32_vcvttph2ibs256_mask( \
310		(__v16hf)(__m256h)(A), (__v16hu)_mm256_setzero_si256(), (__mmask16)-1, \
311		(const int)R))
312
313		#define _mm256_mask_ipcvtt_roundph_epi8(W, U, A, R) \
314		((__m256i)__builtin_ia32_vcvttph2ibs256_mask( \
315		(__v16hf)(__m256h)(A), (__v16hu)(W), (__mmask16)(U), (const int)R))
316
317		#define _mm256_maskz_ipcvtt_roundph_epi8(U, A, R) \
318		((__m256i)__builtin_ia32_vcvttph2ibs256_mask( \
319		(__v16hf)(__m256h)(A), (__v16hu)_mm256_setzero_si256(), (__mmask16)(U), \
320		(const int)R))
321
322		#define _mm_ipcvttph_epu8(A) \
	240	#define _mm_ipcvtts_ph_epu8(A) \
323	241	((__m128i)__builtin_ia32_vcvttph2iubs128_mask( \
324	242	(__v8hf)(__m128h)(A), (__v8hu)_mm_setzero_si128(), (__mmask8)-1))
325	243
326		#define _mm_mask_ipcvttph_epu8(W, U, A) \
	244	#define _mm_mask_ipcvtts_ph_epu8(W, U, A) \
327	245	((__m128i)__builtin_ia32_vcvttph2iubs128_mask((__v8hf)(__m128h)(A), \
328	246	(__v8hu)(W), (__mmask8)(U)))
329	247
330		#define _mm_maskz_ipcvttph_epu8(U, A) \
	248	#define _mm_maskz_ipcvtts_ph_epu8(U, A) \
331	249	((__m128i)__builtin_ia32_vcvttph2iubs128_mask( \
332	250	(__v8hf)(__m128h)(A), (__v8hu)(_mm_setzero_si128()), (__mmask8)(U)))
333	251
334		#define _mm256_ipcvttph_epu8(A) \
	252	#define _mm256_ipcvtts_ph_epu8(A) \
335	253	((__m256i)__builtin_ia32_vcvttph2iubs256_mask( \
336		(__v16hf)(__m256h)(A), (__v16hu)_mm256_setzero_si256(), (__mmask16)-1, \
337		_MM_FROUND_CUR_DIRECTION))
	254	(__v16hf)(__m256h)(A), (__v16hu)_mm256_setzero_si256(), (__mmask16)-1))
338	255
339		#define _mm256_mask_ipcvttph_epu8(W, U, A) \
	256	#define _mm256_mask_ipcvtts_ph_epu8(W, U, A) \
340	257	((__m256i)__builtin_ia32_vcvttph2iubs256_mask((__v16hf)(__m256h)(A), \
341		(__v16hu)(W), (__mmask16)(U), \
342		_MM_FROUND_CUR_DIRECTION))
	258	(__v16hu)(W), (__mmask16)(U)))
343	259
344		#define _mm256_maskz_ipcvttph_epu8(U, A) \
	260	#define _mm256_maskz_ipcvtts_ph_epu8(U, A) \
345	261	((__m256i)__builtin_ia32_vcvttph2iubs256_mask( \
346	262	(__v16hf)(__m256h)(A), (__v16hu)(_mm256_setzero_si256()), \
347		(__mmask16)(U), _MM_FROUND_CUR_DIRECTION))
348
349		#define _mm256_ipcvtt_roundph_epu8(A, R) \
350		((__m256i)__builtin_ia32_vcvttph2iubs256_mask( \
351		(__v16hf)(__m256h)(A), (__v16hu)_mm256_setzero_si256(), (__mmask16)-1, \
352		(const int)R))
353
354		#define _mm256_mask_ipcvtt_roundph_epu8(W, U, A, R) \
355		((__m256i)__builtin_ia32_vcvttph2iubs256_mask( \
356		(__v16hf)(__m256h)(A), (__v16hu)(W), (__mmask16)(U), (const int)R))
357
358		#define _mm256_maskz_ipcvtt_roundph_epu8(U, A, R) \
359		((__m256i)__builtin_ia32_vcvttph2iubs256_mask( \
360		(__v16hf)(__m256h)(A), (__v16hu)_mm256_setzero_si256(), (__mmask16)(U), \
361		(const int)R))
	263	(__mmask16)(U)))
362	264
363		#define _mm_ipcvttps_epi8(A) \
	265	#define _mm_ipcvtts_ps_epi8(A) \
364	266	((__m128i)__builtin_ia32_vcvttps2ibs128_mask( \
365	267	(__v4sf)(__m128)(A), (__v4su)_mm_setzero_si128(), (__mmask8)-1))
366	268
367		#define _mm_mask_ipcvttps_epi8(W, U, A) \
	269	#define _mm_mask_ipcvtts_ps_epi8(W, U, A) \
368	270	((__m128i)__builtin_ia32_vcvttps2ibs128_mask((__v4sf)(__m128)(A), \
369	271	(__v4su)(W), (__mmask8)(U)))
370	272
371		#define _mm_maskz_ipcvttps_epi8(U, A) \
	273	#define _mm_maskz_ipcvtts_ps_epi8(U, A) \
372	274	((__m128i)__builtin_ia32_vcvttps2ibs128_mask( \
373	275	(__v4sf)(__m128)(A), (__v4su)(_mm_setzero_si128()), (__mmask8)(U)))
374	276
375		#define _mm256_ipcvttps_epi8(A) \
	277	#define _mm256_ipcvtts_ps_epi8(A) \
376	278	((__m256i)__builtin_ia32_vcvttps2ibs256_mask( \
377		(__v8sf)(__m256)(A), (__v8su)_mm256_setzero_si256(), (__mmask8)-1, \
378		_MM_FROUND_CUR_DIRECTION))
	279	(__v8sf)(__m256)(A), (__v8su)_mm256_setzero_si256(), (__mmask8)-1))
379	280
380		#define _mm256_mask_ipcvttps_epi8(W, U, A) \
	281	#define _mm256_mask_ipcvtts_ps_epi8(W, U, A) \
381	282	((__m256i)__builtin_ia32_vcvttps2ibs256_mask((__v8sf)(__m256)(A), \
382		(__v8su)(W), (__mmask8)(U), \
383		_MM_FROUND_CUR_DIRECTION))
	283	(__v8su)(W), (__mmask8)(U)))
384	284
385		#define _mm256_maskz_ipcvttps_epi8(U, A) \
	285	#define _mm256_maskz_ipcvtts_ps_epi8(U, A) \
386	286	((__m256i)__builtin_ia32_vcvttps2ibs256_mask( \
387		(__v8sf)(__m256)(A), (__v8su)(_mm256_setzero_si256()), (__mmask8)(U), \
388		_MM_FROUND_CUR_DIRECTION))
	287	(__v8sf)(__m256)(A), (__v8su)(_mm256_setzero_si256()), (__mmask8)(U)))
389	288
390		#define _mm256_ipcvtt_roundps_epi8(A, R) \
391		((__m256i)__builtin_ia32_vcvttps2ibs256_mask((__v8sf)(__m256)(A), \
392		(__v8su)_mm256_setzero_si256(), \
393		(__mmask8)-1, (const int)R))
394
395		#define _mm256_mask_ipcvtt_roundps_epi8(W, U, A, R) \
396		((__m256i)__builtin_ia32_vcvttps2ibs256_mask( \
397		(__v8sf)(__m256)(A), (__v8su)(W), (__mmask8)(U), (const int)R))
398
399		#define _mm256_maskz_ipcvtt_roundps_epi8(U, A, R) \
400		((__m256i)__builtin_ia32_vcvttps2ibs256_mask((__v8sf)(__m256)(A), \
401		(__v8su)_mm256_setzero_si256(), \
402		(__mmask8)(U), (const int)R))
403
404		#define _mm_ipcvttps_epu8(A) \
	289	#define _mm_ipcvtts_ps_epu8(A) \
405	290	((__m128i)__builtin_ia32_vcvttps2iubs128_mask( \
406	291	(__v4sf)(__m128)(A), (__v4su)_mm_setzero_si128(), (__mmask8)-1))
407	292
408		#define _mm_mask_ipcvttps_epu8(W, U, A) \
	293	#define _mm_mask_ipcvtts_ps_epu8(W, U, A) \
409	294	((__m128i)__builtin_ia32_vcvttps2iubs128_mask((__v4sf)(__m128)(A), \
410	295	(__v4su)(W), (__mmask8)(U)))
411	296
412		#define _mm_maskz_ipcvttps_epu8(U, A) \
	297	#define _mm_maskz_ipcvtts_ps_epu8(U, A) \
413	298	((__m128i)__builtin_ia32_vcvttps2iubs128_mask( \
414	299	(__v4sf)(__m128)(A), (__v4su)(_mm_setzero_si128()), (__mmask8)(U)))
415	300
416		#define _mm256_ipcvttps_epu8(A) \
	301	#define _mm256_ipcvtts_ps_epu8(A) \
417	302	((__m256i)__builtin_ia32_vcvttps2iubs256_mask( \
418		(__v8sf)(__m256)(A), (__v8su)_mm256_setzero_si256(), (__mmask8)-1, \
419		_MM_FROUND_CUR_DIRECTION))
	303	(__v8sf)(__m256)(A), (__v8su)_mm256_setzero_si256(), (__mmask8)-1))
420	304
421		#define _mm256_mask_ipcvttps_epu8(W, U, A) \
	305	#define _mm256_mask_ipcvtts_ps_epu8(W, U, A) \
422	306	((__m256i)__builtin_ia32_vcvttps2iubs256_mask((__v8sf)(__m256)(A), \
423		(__v8su)(W), (__mmask8)(U), \
424		_MM_FROUND_CUR_DIRECTION))
425
426		#define _mm256_maskz_ipcvttps_epu8(U, A) \
427		((__m256i)__builtin_ia32_vcvttps2iubs256_mask( \
428		(__v8sf)(__m256)(A), (__v8su)(_mm256_setzero_si256()), (__mmask8)(U), \
429		_MM_FROUND_CUR_DIRECTION))
430
431		#define _mm256_ipcvtt_roundps_epu8(A, R) \
432		((__m256i)__builtin_ia32_vcvttps2iubs256_mask( \
433		(__v8sf)(__m256)(A), (__v8su)_mm256_setzero_si256(), (__mmask8)-1, \
434		(const int)R))
435
436		#define _mm256_mask_ipcvtt_roundps_epu8(W, U, A, R) \
437		((__m256i)__builtin_ia32_vcvttps2iubs256_mask( \
438		(__v8sf)(__m256)(A), (__v8su)(W), (__mmask8)(U), (const int)R))
	307	(__v8su)(W), (__mmask8)(U)))
439	308
440		#define _mm256_maskz_ipcvtt_roundps_epu8(U, A, R) \
	309	#define _mm256_maskz_ipcvtts_ps_epu8(U, A) \
441	310	((__m256i)__builtin_ia32_vcvttps2iubs256_mask( \
442		(__v8sf)(__m256)(A), (__v8su)_mm256_setzero_si256(), (__mmask8)(U), \
443		(const int)R))
	311	(__v8sf)(__m256)(A), (__v8su)(_mm256_setzero_si256()), (__mmask8)(U)))
444	312	#endif // __AVX10_2SATCVTINTRIN_H

lib/include/avx512fp16intrin.h+2-1

...	...	@@ -553,7 +553,8 @@ static __inline__ __m512h __DEFAULT_FN_ATTRS512 _mm512_abs_ph(__m512h __A) {
553	553	}
554	554
555	555	static __inline__ __m512h __DEFAULT_FN_ATTRS512 _mm512_conj_pch(__m512h __A) {
556		return (__m512h)_mm512_xor_ps((__m512)__A, _mm512_set1_ps(-0.0f));
	556	return (__m512h)_mm512_xor_epi32((__m512i)__A,
	557	_mm512_set1_epi32(-2147483648));
557	558	}
558	559
559	560	static __inline__ __m512h __DEFAULT_FN_ATTRS512

lib/include/bmiintrin.h-4

...	...	@@ -161,8 +161,6 @@ _mm_tzcnt_64(unsigned long long __X) {
161	161
162	162	#undef __RELAXED_FN_ATTRS
163	163
164		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__BMI__)
165
166	164	/* Define the default attributes for the functions in this file. */
167	165	#if defined(__cplusplus) && (__cplusplus >= 201103L)
168	166	#define __DEFAULT_FN_ATTRS \
...	...	@@ -603,6 +601,4 @@ __blsr_u64(unsigned long long __X) {
603	601
604	602	#undef __DEFAULT_FN_ATTRS
605	603
606		#endif /* !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__BMI__) */
607
608	604	#endif /* __BMIINTRIN_H */

lib/include/cpuid.h+24-22

...	...	@@ -267,18 +267,18 @@
267	267	: "0"(__leaf), "2"(__count))
268	268	#else
269	269	/* x86-64 uses %rbx as the base register, so preserve it. */
270		#define __cpuid(__leaf, __eax, __ebx, __ecx, __edx) \
271		__asm(" xchgq %%rbx,%q1\n" \
272		" cpuid\n" \
273		" xchgq %%rbx,%q1" \
274		: "=a"(__eax), "=r" (__ebx), "=c"(__ecx), "=d"(__edx) \
	270	#define __cpuid(__leaf, __eax, __ebx, __ecx, __edx) \
	271	__asm(" xchg{q\|} {%%\|}rbx,%q1\n" \
	272	" cpuid\n" \
	273	" xchg{q\|} {%%\|}rbx,%q1" \
	274	: "=a"(__eax), "=r"(__ebx), "=c"(__ecx), "=d"(__edx) \
275	275	: "0"(__leaf))
276	276
277		#define __cpuid_count(__leaf, __count, __eax, __ebx, __ecx, __edx) \
278		__asm(" xchgq %%rbx,%q1\n" \
279		" cpuid\n" \
280		" xchgq %%rbx,%q1" \
281		: "=a"(__eax), "=r" (__ebx), "=c"(__ecx), "=d"(__edx) \
	277	#define __cpuid_count(__leaf, __count, __eax, __ebx, __ecx, __edx) \
	278	__asm(" xchg{q\|} {%%\|}rbx,%q1\n" \
	279	" cpuid\n" \
	280	" xchg{q\|} {%%\|}rbx,%q1" \
	281	: "=a"(__eax), "=r"(__ebx), "=c"(__ecx), "=d"(__edx) \
282	282	: "0"(__leaf), "2"(__count))
283	283	#endif
284	284
...	...	@@ -289,20 +289,22 @@ static __inline unsigned int __get_cpuid_max (unsigned int __leaf,
289	289	#ifdef __i386__
290	290	int __cpuid_supported;
291	291
292		__asm(" pushfl\n"
293		" popl %%eax\n"
294		" movl %%eax,%%ecx\n"
295		" xorl $0x00200000,%%eax\n"
296		" pushl %%eax\n"
297		" popfl\n"
298		" pushfl\n"
299		" popl %%eax\n"
300		" movl $0,%0\n"
301		" cmpl %%eax,%%ecx\n"
	292	__asm(" pushf{l\|d}\n"
	293	" pop{l\|} {%%\|}eax\n"
	294	" mov{l\|} {%%eax,%%ecx\|ecx,eax}\n"
	295	" xor{l\|} {$0x00200000,%%eax\|eax,0x00200000}\n"
	296	" push{l\|} {%%\|}eax\n"
	297	" popf{l\|d}\n"
	298	" pushf{l\|d}\n"
	299	" pop{l\|} {%%\|}eax\n"
	300	" mov{l\|} {$0,%0\|%0,0}\n"
	301	" cmp{l\|} {%%eax,%%ecx\|ecx,eax}\n"
302	302	" je 1f\n"
303		" movl $1,%0\n"
	303	" mov{l\|} {$1,%0\|%0,1}\n"
304	304	"1:"
305		: "=r" (__cpuid_supported) : : "eax", "ecx");
	305	: "=r"(__cpuid_supported)
	306	:
	307	: "eax", "ecx");
306	308	if (!__cpuid_supported)
307	309	return 0;
308	310	#endif

lib/include/float.h+2-11

...	...	@@ -18,21 +18,12 @@
18	18	* additional definitions provided for Windows.
19	19	* For more details see http://msdn.microsoft.com/en-us/library/y0ybw9fy.aspx
20	20	*
21		* Also fall back on Darwin and AIX to allow additional definitions and
	21	* Also fall back on AIX to allow additional definitions and
22	22	* implementation-defined values.
23	23	*/
24		#if (defined(__APPLE__) \|\| defined(__MINGW32__) \|\| defined(_MSC_VER) \|\| \
25		defined(_AIX)) && \
	24	#if (defined(__MINGW32__) \|\| defined(_MSC_VER) \|\| defined(_AIX)) && \
26	25	__STDC_HOSTED__ && __has_include_next(<float.h>)
27	26
28		/* Prior to Apple's 10.7 SDK, float.h SDK header used to apply an extra level
29		* of #include_next<float.h> to keep Metrowerks compilers happy. Avoid this
30		* extra indirection.
31		*/
32		#ifdef __APPLE__
33		#define _FLOAT_H_
34		#endif
35
36	27	# include_next <float.h>
37	28
38	29	/* Undefine anything that we'll be redefining below. */

lib/include/immintrin.h-224

...	...	@@ -16,231 +16,112 @@
16	16
17	17	#include <x86gprintrin.h>
18	18
19		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__MMX__)
20	19	#include <mmintrin.h>
21		#endif
22	20
23		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__SSE__)
24	21	#include <xmmintrin.h>
25		#endif
26	22
27		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__SSE2__)
28	23	#include <emmintrin.h>
29		#endif
30	24
31		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__SSE3__)
32	25	#include <pmmintrin.h>
33		#endif
34	26
35		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__SSSE3__)
36	27	#include <tmmintrin.h>
37		#endif
38	28
39		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| \
40		(defined(__SSE4_2__) \|\| defined(__SSE4_1__))
41	29	#include <smmintrin.h>
42		#endif
43	30
44		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| \
45		(defined(__AES__) \|\| defined(__PCLMUL__))
46	31	#include <wmmintrin.h>
47		#endif
48	32
49		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__CLFLUSHOPT__)
50	33	#include <clflushoptintrin.h>
51		#endif
52	34
53		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__CLWB__)
54	35	#include <clwbintrin.h>
55		#endif
56	36
57		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__AVX__)
58	37	#include <avxintrin.h>
59		#endif
60	38
61		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__AVX2__)
62	39	#include <avx2intrin.h>
63		#endif
64	40
65		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__F16C__)
66	41	#include <f16cintrin.h>
67		#endif
68	42
69		/* No feature check desired due to internal checks */
70	43	#include <bmiintrin.h>
71	44
72		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__BMI2__)
73	45	#include <bmi2intrin.h>
74		#endif
75	46
76		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__LZCNT__)
77	47	#include <lzcntintrin.h>
78		#endif
79	48
80		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__POPCNT__)
81	49	#include <popcntintrin.h>
82		#endif
83	50
84		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__FMA__)
85	51	#include <fmaintrin.h>
86		#endif
87	52
88		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__AVX512F__)
89	53	#include <avx512fintrin.h>
90		#endif
91	54
92		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__AVX512VL__)
93	55	#include <avx512vlintrin.h>
94		#endif
95	56
96		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__AVX512BW__)
97	57	#include <avx512bwintrin.h>
98		#endif
99	58
100		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__AVX512BITALG__)
101	59	#include <avx512bitalgintrin.h>
102		#endif
103	60
104		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__AVX512CD__)
105	61	#include <avx512cdintrin.h>
106		#endif
107	62
108		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__AVX512VPOPCNTDQ__)
109	63	#include <avx512vpopcntdqintrin.h>
110		#endif
111	64
112		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| \
113		(defined(__AVX512VL__) && defined(__AVX512VPOPCNTDQ__))
114	65	#include <avx512vpopcntdqvlintrin.h>
115		#endif
116	66
117		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__AVX512VNNI__)
118	67	#include <avx512vnniintrin.h>
119		#endif
120	68
121		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| \
122		(defined(__AVX512VL__) && defined(__AVX512VNNI__))
123	69	#include <avx512vlvnniintrin.h>
124		#endif
125	70
126		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__AVXVNNI__)
127	71	#include <avxvnniintrin.h>
128		#endif
129	72
130		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__AVX512DQ__)
131	73	#include <avx512dqintrin.h>
132		#endif
133	74
134		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| \
135		(defined(__AVX512VL__) && defined(__AVX512BITALG__))
136	75	#include <avx512vlbitalgintrin.h>
137		#endif
138	76
139		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| \
140		(defined(__AVX512VL__) && defined(__AVX512BW__))
141	77	#include <avx512vlbwintrin.h>
142		#endif
143	78
144		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| \
145		(defined(__AVX512VL__) && defined(__AVX512CD__))
146	79	#include <avx512vlcdintrin.h>
147		#endif
148	80
149		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| \
150		(defined(__AVX512VL__) && defined(__AVX512DQ__))
151	81	#include <avx512vldqintrin.h>
152		#endif
153	82
154		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__AVX512IFMA__)
155	83	#include <avx512ifmaintrin.h>
156		#endif
157	84
158		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| \
159		(defined(__AVX512IFMA__) && defined(__AVX512VL__))
160	85	#include <avx512ifmavlintrin.h>
161		#endif
162	86
163		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__AVXIFMA__)
164	87	#include <avxifmaintrin.h>
165		#endif
166	88
167		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__AVX512VBMI__)
168	89	#include <avx512vbmiintrin.h>
169		#endif
170	90
171		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| \
172		(defined(__AVX512VBMI__) && defined(__AVX512VL__))
173	91	#include <avx512vbmivlintrin.h>
174		#endif
175	92
176		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__AVX512VBMI2__)
177	93	#include <avx512vbmi2intrin.h>
178		#endif
179	94
180		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| \
181		(defined(__AVX512VBMI2__) && defined(__AVX512VL__))
182	95	#include <avx512vlvbmi2intrin.h>
183		#endif
184	96
185		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__AVX512FP16__)
186	97	#include <avx512fp16intrin.h>
187		#endif
188	98
189		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| \
190		(defined(__AVX512VL__) && defined(__AVX512FP16__))
191	99	#include <avx512vlfp16intrin.h>
192		#endif
193	100
194		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__AVX512BF16__)
195	101	#include <avx512bf16intrin.h>
196		#endif
197	102
198		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| \
199		(defined(__AVX512VL__) && defined(__AVX512BF16__))
200	103	#include <avx512vlbf16intrin.h>
201		#endif
202	104
203		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__PKU__)
204	105	#include <pkuintrin.h>
205		#endif
206	106
207		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__VPCLMULQDQ__)
208	107	#include <vpclmulqdqintrin.h>
209		#endif
210	108
211		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__VAES__)
212	109	#include <vaesintrin.h>
213		#endif
214	110
215		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__GFNI__)
216	111	#include <gfniintrin.h>
217		#endif
218	112
219		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__AVXVNNIINT8__)
220	113	#include <avxvnniint8intrin.h>
221		#endif
222	114
223		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__AVXNECONVERT__)
224	115	#include <avxneconvertintrin.h>
225		#endif
226	116
227		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__SHA512__)
228	117	#include <sha512intrin.h>
229		#endif
230	118
231		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__SM3__)
232	119	#include <sm3intrin.h>
233		#endif
234	120
235		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__SM4__)
236	121	#include <sm4intrin.h>
237		#endif
238	122
239		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__AVXVNNIINT16__)
240	123	#include <avxvnniint16intrin.h>
241		#endif
242	124
243		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__RDPID__)
244	125	/// Reads the value of the IA32_TSC_AUX MSR (0xc0000103).
245	126	///
246	127	/// \headerfile <immintrin.h>
...	...	@@ -252,9 +133,7 @@ static __inline__ unsigned int __attribute__((__always_inline__, __nodebug__, __
252	133	_rdpid_u32(void) {
253	134	return __builtin_ia32_rdpid();
254	135	}
255		#endif // __RDPID__
256	136
257		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__RDRND__)
258	137	/// Returns a 16-bit hardware-generated random value.
259	138	///
260	139	/// \headerfile <immintrin.h>
...	...	@@ -314,9 +193,7 @@ _rdrand64_step(unsigned long long *__p)
314	193	}
315	194	#endif
316	195	}
317		#endif /* __RDRND__ */
318	196
319		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__FSGSBASE__)
320	197	#ifdef __x86_64__
321	198	/// Reads the FS base register.
322	199	///
...	...	@@ -427,9 +304,6 @@ _writegsbase_u64(unsigned long long __V)
427	304	}
428	305
429	306	#endif
430		#endif /* __FSGSBASE__ */
431
432		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__MOVBE__)
433	307
434	308	/* The structs used below are to force the load/store to be unaligned. This
435	309	* is accomplished with the __packed__ attribute. The __may_alias__ prevents
...	...	@@ -543,172 +417,86 @@ _storebe_i64(void * __P, long long __D) {
543	417	((struct __storeu_i64*)__P)->__v = __builtin_bswap64((unsigned long long)__D);
544	418	}
545	419	#endif
546		#endif /* __MOVBE */
547	420
548		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__RTM__)
549	421	#include <rtmintrin.h>
550	422	#include <xtestintrin.h>
551		#endif
552	423
553		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__SHA__)
554	424	#include <shaintrin.h>
555		#endif
556	425
557		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__FXSR__)
558	426	#include <fxsrintrin.h>
559		#endif
560	427
561	428	/* No feature check desired due to internal MSC_VER checks */
562	429	#include <xsaveintrin.h>
563	430
564		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__XSAVEOPT__)
565	431	#include <xsaveoptintrin.h>
566		#endif
567	432
568		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__XSAVEC__)
569	433	#include <xsavecintrin.h>
570		#endif
571	434
572		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__XSAVES__)
573	435	#include <xsavesintrin.h>
574		#endif
575	436
576		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__SHSTK__)
577	437	#include <cetintrin.h>
578		#endif
579	438
580	439	/* Intrinsics inside adcintrin.h are available at all times. */
581	440	#include <adcintrin.h>
582	441
583		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__ADX__)
584	442	#include <adxintrin.h>
585		#endif
586	443
587		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__RDSEED__)
588	444	#include <rdseedintrin.h>
589		#endif
590	445
591		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__WBNOINVD__)
592	446	#include <wbnoinvdintrin.h>
593		#endif
594	447
595		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__CLDEMOTE__)
596	448	#include <cldemoteintrin.h>
597		#endif
598	449
599		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__WAITPKG__)
600	450	#include <waitpkgintrin.h>
601		#endif
602	451
603		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__MOVDIRI__) \|\| \
604		defined(__MOVDIR64B__)
605	452	#include <movdirintrin.h>
606		#endif
607	453
608		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__MOVRS__)
609	454	#include <movrsintrin.h>
610		#endif
611	455
612		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| \
613		(defined(__AVX10_2__) && defined(__MOVRS__))
614	456	#include <movrs_avx10_2intrin.h>
615		#endif
616	457
617		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| \
618		(defined(__AVX10_2_512__) && defined(__MOVRS__))
619	458	#include <movrs_avx10_2_512intrin.h>
620		#endif
621	459
622		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__PCONFIG__)
623	460	#include <pconfigintrin.h>
624		#endif
625	461
626		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__SGX__)
627	462	#include <sgxintrin.h>
628		#endif
629	463
630		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__PTWRITE__)
631	464	#include <ptwriteintrin.h>
632		#endif
633	465
634		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__INVPCID__)
635	466	#include <invpcidintrin.h>
636		#endif
637	467
638		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__KL__) \|\| \
639		defined(__WIDEKL__)
640	468	#include <keylockerintrin.h>
641		#endif
642	469
643		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__AMX_TILE__) \|\| \
644		defined(__AMX_INT8__) \|\| defined(__AMX_BF16__)
645	470	#include <amxintrin.h>
646		#endif
647	471
648		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__AMX_FP16__)
649	472	#include <amxfp16intrin.h>
650		#endif
651	473
652		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__AMX_COMPLEX__)
653	474	#include <amxcomplexintrin.h>
654		#endif
655	475
656		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__AMX_FP8__)
657	476	#include <amxfp8intrin.h>
658		#endif
659	477
660		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__AMX_TRANSPOSE__)
661	478	#include <amxtransposeintrin.h>
662		#endif
663	479
664		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__AMX_MOVRS__)
665	480	#include <amxmovrsintrin.h>
666		#endif
667	481
668		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| \
669		(defined(__AMX_MOVRS__) && defined(__AMX_TRANSPOSE__))
670	482	#include <amxmovrstransposeintrin.h>
671		#endif
672	483
673		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__AMX_AVX512__)
674	484	#include <amxavx512intrin.h>
675		#endif
676	485
677		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__AMX_TF32__)
678	486	#include <amxtf32intrin.h>
679		#endif
680	487
681		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| \
682		(defined(__AMX_TF32__) && defined(__AMX_TRANSPOSE__))
683	488	#include <amxtf32transposeintrin.h>
684		#endif
685	489
686		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| \
687		(defined(__AMX_BF16__) && defined(__AMX_TRANSPOSE__))
688	490	#include <amxbf16transposeintrin.h>
689		#endif
690	491
691		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| \
692		(defined(__AMX_FP16__) && defined(__AMX_TRANSPOSE__))
693	492	#include <amxfp16transposeintrin.h>
694		#endif
695	493
696		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| \
697		(defined(__AMX_COMPLEX__) && defined(__AMX_TRANSPOSE__))
698	494	#include <amxcomplextransposeintrin.h>
699		#endif
700	495
701		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| \
702		defined(__AVX512VP2INTERSECT__)
703	496	#include <avx512vp2intersectintrin.h>
704		#endif
705	497
706		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| \
707		(defined(__AVX512VL__) && defined(__AVX512VP2INTERSECT__))
708	498	#include <avx512vlvp2intersectintrin.h>
709		#endif
710	499
711		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__AVX10_2__)
712	500	#include <avx10_2bf16intrin.h>
713	501	#include <avx10_2convertintrin.h>
714	502	#include <avx10_2copyintrin.h>
...	...	@@ -716,33 +504,21 @@ _storebe_i64(void * __P, long long __D) {
716	504	#include <avx10_2niintrin.h>
717	505	#include <avx10_2satcvtdsintrin.h>
718	506	#include <avx10_2satcvtintrin.h>
719		#endif
720	507
721		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__AVX10_2_512__)
722	508	#include <avx10_2_512bf16intrin.h>
723	509	#include <avx10_2_512convertintrin.h>
724	510	#include <avx10_2_512minmaxintrin.h>
725	511	#include <avx10_2_512niintrin.h>
726	512	#include <avx10_2_512satcvtdsintrin.h>
727	513	#include <avx10_2_512satcvtintrin.h>
728		#endif
729	514
730		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| \
731		(defined(__AVX10_2_512__) && defined(__SM4__))
732	515	#include <sm4evexintrin.h>
733		#endif
734	516
735		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__ENQCMD__)
736	517	#include <enqcmdintrin.h>
737		#endif
738	518
739		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__SERIALIZE__)
740	519	#include <serializeintrin.h>
741		#endif
742	520
743		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__TSXLDTRK__)
744	521	#include <tsxldtrkintrin.h>
745		#endif
746	522
747	523	#if defined(_MSC_VER) && __has_extension(gnu_asm)
748	524	/* Define the default attributes for these intrinsics */

lib/include/intrin.h+21-4

...	...	@@ -162,8 +162,6 @@ void _Store_HLERelease(long volatile *, long);
162	162	void _Store64_HLERelease(__int64 volatile *, __int64);
163	163	void _StorePointer_HLERelease(void volatile , void *);
164	164	void _WriteBarrier(void);
165		unsigned __int32 xbegin(void);
166		void _xend(void);
167	165
168	166	/* These additional intrinsics are turned on in x64/amd64/x86_64 mode. */
169	167	#if defined(__x86_64__) && !defined(__arm64ec__)
...	...	@@ -372,10 +370,29 @@ static __inline__ void __DEFAULT_FN_ATTRS __nop(void) {
372	370	\----------------------------------------------------------------------------/
373	371	#if defined(__aarch64__) \|\| defined(__arm64ec__)
374	372	unsigned __int64 __getReg(int);
375		long _InterlockedAdd(long volatile *Addend, long Value);
376		__int64 _InterlockedAdd64(__int64 volatile *Addend, __int64 Value);
	373	unsigned char _interlockedbittestandreset_acq(long volatile *, long);
	374	unsigned char _interlockedbittestandreset_nf(long volatile *, long);
	375	unsigned char _interlockedbittestandreset_rel(long volatile *, long);
	376	unsigned char _interlockedbittestandreset64_acq(__int64 volatile *, __int64);
	377	unsigned char _interlockedbittestandreset64_nf(__int64 volatile *, __int64);
	378	unsigned char _interlockedbittestandreset64_rel(__int64 volatile *, __int64);
	379	unsigned char _interlockedbittestandset_acq(long volatile *, long);
	380	unsigned char _interlockedbittestandset_nf(long volatile *, long);
	381	unsigned char _interlockedbittestandset_rel(long volatile *, long);
	382	unsigned char _interlockedbittestandset64_acq(__int64 volatile *, __int64);
	383	unsigned char _interlockedbittestandset64_nf(__int64 volatile *, __int64);
	384	unsigned char _interlockedbittestandset64_rel(__int64 volatile *, __int64);
	385	long _InterlockedAdd(long volatile *, long);
	386	long _InterlockedAdd_acq(long volatile *, long);
	387	long _InterlockedAdd_nf(long volatile *, long);
	388	long _InterlockedAdd_rel(long volatile *, long);
	389	__int64 _InterlockedAdd64(__int64 volatile *, __int64);
	390	__int64 _InterlockedAdd64_acq(__int64 volatile *, __int64);
	391	__int64 _InterlockedAdd64_nf(__int64 volatile *, __int64);
	392	__int64 _InterlockedAdd64_rel(__int64 volatile *, __int64);
377	393	__int64 _ReadStatusReg(int);
378	394	void _WriteStatusReg(int, __int64);
	395	unsigned int __sys(int, __int64);
379	396
380	397	unsigned short __cdecl _byteswap_ushort(unsigned short val);
381	398	unsigned long __cdecl _byteswap_ulong (unsigned long val);

lib/include/keylockerintrin.h-9

...	...	@@ -28,8 +28,6 @@
28	28	#ifndef _KEYLOCKERINTRIN_H
29	29	#define _KEYLOCKERINTRIN_H
30	30
31		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__KL__)
32
33	31	/* Define the default attributes for the functions in this file. */
34	32	#define __DEFAULT_FN_ATTRS \
35	33	__attribute__((__always_inline__, __nodebug__, __target__("kl"),\
...	...	@@ -326,10 +324,6 @@ _mm_aesdec256kl_u8(__m128i* __odata, __m128i __idata, const void *__h) {
326	324
327	325	#undef __DEFAULT_FN_ATTRS
328	326
329		#endif /* !defined(__SCE__ \|\| __has_feature(modules) \|\| defined(__KL__) */
330
331		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__WIDEKL__)
332
333	327	/* Define the default attributes for the functions in this file. */
334	328	#define __DEFAULT_FN_ATTRS \
335	329	__attribute__((__always_inline__, __nodebug__, __target__("kl,widekl"),\
...	...	@@ -521,7 +515,4 @@ _mm_aesdecwide256kl_u8(__m128i __odata[8], const __m128i __idata[8], const void*
521	515
522	516	#undef __DEFAULT_FN_ATTRS
523	517
524		#endif /* !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__WIDEKL__) \
525		*/
526
527	518	#endif /* _KEYLOCKERINTRIN_H */

lib/include/llvm_libc_wrappers/stdlib.h+1-1

...	...	@@ -25,7 +25,7 @@
25	25
26	26	// The LLVM C library uses these named types so we forward declare them.
27	27	typedef void (*__atexithandler_t)(void);
28		typedef int (__bsearchcompare_t)(const void , const void *);
	28	typedef int (__search_compare_t)(const void , const void *);
29	29	typedef int (__qsortcompare_t)(const void , const void *);
30	30	typedef int (__qsortrcompare_t)(const void , const void , void );
31	31

lib/include/lzcntintrin.h+6-4

...	...	@@ -14,13 +14,15 @@
14	14	#ifndef __LZCNTINTRIN_H
15	15	#define __LZCNTINTRIN_H
16	16
17		/* Define the default attributes for the functions in this file. */
	17	/* Define the default attributes for the functions in this file.
	18	Allow using the lzcnt intrinsics even for non-LZCNT targets. Since the LZCNT
	19	intrinsics are mapped to llvm.ctlz.*, false, which can be lowered to BSR on
	20	non-LZCNT targets with zero-value input handled correctly. */
18	21	#if defined(__cplusplus) && (__cplusplus >= 201103L)
19	22	#define __DEFAULT_FN_ATTRS \
20		__attribute__((__always_inline__, __nodebug__, __target__("lzcnt"))) constexpr
	23	__attribute__((__always_inline__, __nodebug__)) constexpr
21	24	#else
22		#define __DEFAULT_FN_ATTRS \
23		__attribute__((__always_inline__, __nodebug__, __target__("lzcnt")))
	25	#define __DEFAULT_FN_ATTRS __attribute__((__always_inline__, __nodebug__))
24	26	#endif
25	27
26	28	#ifndef _MSC_VER

lib/include/module.modulemap+13

...	...	@@ -35,6 +35,14 @@ module _Builtin_intrinsics [system] [extern_c] {
35	35	}
36	36	}
37	37
	38	explicit module arm64 {
	39	requires arm64
	40	requires windows
	41
	42	header "arm64intr.h"
	43	export *
	44	}
	45
38	46	explicit module intel {
39	47	requires x86
40	48	export *
...	...	@@ -231,6 +239,11 @@ module _Builtin_stdbool [system] {
231	239	export *
232	240	}
233	241
	242	module _Builtin_stdcountof [system] {
	243	header "stdcountof.h"
	244	export *
	245	}
	246
234	247	module _Builtin_stddef [system] {
235	248	textual header "stddef.h"
236	249

lib/include/movrsintrin.h+1-1

...	...	@@ -56,4 +56,4 @@ _m_prefetchrs(volatile const void *__P) {
56	56	}
57	57
58	58	#undef __DEFAULT_FN_ATTRS
59		#endif // __MOVRSINTRIN_H
		\ No newline at end of file
	59	#endif // __MOVRSINTRIN_H

lib/include/prfchwintrin.h+10-13

...	...	@@ -14,6 +14,10 @@
14	14	#ifndef __PRFCHWINTRIN_H
15	15	#define __PRFCHWINTRIN_H
16	16
	17	#if defined(__cplusplus)
	18	extern "C" {
	19	#endif
	20
17	21	/// Loads a memory sequence containing the specified memory address into
18	22	/// all data cache levels.
19	23	///
...	...	@@ -26,11 +30,7 @@
26	30	///
27	31	/// \param __P
28	32	/// A pointer specifying the memory address to be prefetched.
29		static __inline__ void __attribute__((__always_inline__, __nodebug__))
30		_m_prefetch(void *__P)
31		{
32		__builtin_prefetch (__P, 0, 3 /* _MM_HINT_T0 */);
33		}
	33	void _m_prefetch(void *__P);
34	34
35	35	/// Loads a memory sequence containing the specified memory address into
36	36	/// the L1 data cache and sets the cache-coherency state to modified.
...	...	@@ -48,13 +48,10 @@ _m_prefetch(void *__P)
48	48	///
49	49	/// \param __P
50	50	/// A pointer specifying the memory address to be prefetched.
51		static __inline__ void __attribute__((__always_inline__, __nodebug__))
52		_m_prefetchw(volatile const void *__P)
53		{
54		#pragma clang diagnostic push
55		#pragma clang diagnostic ignored "-Wcast-qual"
56		__builtin_prefetch ((const void)__P, 1, 3 / _MM_HINT_T0 */);
57		#pragma clang diagnostic pop
58		}
	51	void _m_prefetchw(volatile const void *__P);
	52
	53	#if defined(__cplusplus)
	54	} // extern "C"
	55	#endif
59	56
60	57	#endif /* __PRFCHWINTRIN_H */

lib/include/ptrauth.h+57

...	...	@@ -42,6 +42,19 @@ typedef enum {
42	42	The extra data is always 0. */
43	43	ptrauth_key_cxx_vtable_pointer = ptrauth_key_process_independent_data,
44	44
	45	/* The key used to sign metadata pointers to Objective-C method-lists. */
	46	ptrauth_key_method_list_pointer = ptrauth_key_asda,
	47
	48	/* The key used to sign Objective-C isa and super pointers. */
	49	ptrauth_key_objc_isa_pointer = ptrauth_key_process_independent_data,
	50	ptrauth_key_objc_super_pointer = ptrauth_key_process_independent_data,
	51
	52	/* The key used to sign selector pointers */
	53	ptrauth_key_objc_sel_pointer = ptrauth_key_process_dependent_data,
	54
	55	/* The key used to sign Objective-C class_ro_t pointers. */
	56	ptrauth_key_objc_class_ro_pointer = ptrauth_key_process_independent_data,
	57
45	58	/* The key used to sign pointers in ELF .init_array/.fini_array. */
46	59	ptrauth_key_init_fini_pointer = ptrauth_key_process_independent_code,
47	60
...	...	@@ -259,6 +272,46 @@ typedef __UINTPTR_TYPE__ ptrauth_generic_signature_t;
259	272	/* The value is ptrauth_string_discriminator("init_fini") */
260	273	#define __ptrauth_init_fini_discriminator 0xd9d4
261	274
	275	/* Objective-C pointer auth ABI qualifiers */
	276	#define __ptrauth_objc_method_list_imp \
	277	__ptrauth(ptrauth_key_function_pointer, 1, 0)
	278
	279	#if __has_feature(ptrauth_objc_method_list_pointer)
	280	#define __ptrauth_objc_method_list_pointer \
	281	__ptrauth(ptrauth_key_method_list_pointer, 1, 0xC310)
	282	#else
	283	#define __ptrauth_objc_method_list_pointer
	284	#endif
	285
	286	#define __ptrauth_isa_discriminator 0x6AE1
	287	#define __ptrauth_super_discriminator 0xB5AB
	288	#define __ptrauth_objc_isa_pointer \
	289	__ptrauth(ptrauth_key_objc_isa_pointer, 1, __ptrauth_isa_discriminator)
	290	#if __has_feature(ptrauth_restricted_intptr_qualifier)
	291	#define __ptrauth_objc_isa_uintptr \
	292	__ptrauth_restricted_intptr(ptrauth_key_objc_isa_pointer, 1, \
	293	__ptrauth_isa_discriminator)
	294	#else
	295	#define __ptrauth_objc_isa_uintptr \
	296	__ptrauth(ptrauth_key_objc_isa_pointer, 1, __ptrauth_isa_discriminator)
	297	#endif
	298
	299	#define __ptrauth_objc_super_pointer \
	300	__ptrauth(ptrauth_key_objc_super_pointer, 1, __ptrauth_super_discriminator)
	301
	302	#define __ptrauth_objc_sel_discriminator 0x57c2
	303	#if __has_feature(ptrauth_objc_interface_sel)
	304	#define __ptrauth_objc_sel \
	305	__ptrauth(ptrauth_key_objc_sel_pointer, 1, __ptrauth_objc_sel_discriminator)
	306	#else
	307	#define __ptrauth_objc_sel
	308	#endif
	309
	310	#define __ptrauth_objc_class_ro_discriminator 0x61f8
	311	#define __ptrauth_objc_class_ro \
	312	__ptrauth(ptrauth_key_objc_class_ro_pointer, 1, \
	313	__ptrauth_objc_class_ro_discriminator)
	314
262	315	#else
263	316
264	317	#define ptrauth_strip(__value, __key) \
...	...	@@ -331,6 +384,10 @@ typedef __UINTPTR_TYPE__ ptrauth_generic_signature_t;
331	384	#define ptrauth_cxx_vtable_pointer(key, address_discrimination, \
332	385	extra_discrimination...)
333	386
	387	#define __ptrauth_objc_isa_pointer
	388	#define __ptrauth_objc_isa_uintptr
	389	#define __ptrauth_objc_super_pointer
	390
334	391	#endif /* __has_feature(ptrauth_intrinsics) */
335	392
336	393	#endif /* __PTRAUTH_H */

lib/include/riscv_corev_alu.h+4-4

...	...	@@ -24,13 +24,13 @@ static __inline__ long __DEFAULT_FN_ATTRS __riscv_cv_abs(long a) {
24	24	return __builtin_abs(a);
25	25	}
26	26
27		static __inline__ long __DEFAULT_FN_ATTRS __riscv_cv_alu_slet(long a, long b) {
28		return __builtin_riscv_cv_alu_slet(a, b);
	27	static __inline__ long __DEFAULT_FN_ATTRS __riscv_cv_alu_sle(long a, long b) {
	28	return __builtin_riscv_cv_alu_sle(a, b);
29	29	}
30	30
31	31	static __inline__ long __DEFAULT_FN_ATTRS
32		__riscv_cv_alu_sletu(unsigned long a, unsigned long b) {
33		return __builtin_riscv_cv_alu_sletu(a, b);
	32	__riscv_cv_alu_sleu(unsigned long a, unsigned long b) {
	33	return __builtin_riscv_cv_alu_sleu(a, b);
34	34	}
35	35
36	36	static __inline__ long __DEFAULT_FN_ATTRS __riscv_cv_alu_min(long a, long b) {

lib/include/riscv_vector.h-4

...	...	@@ -49,7 +49,6 @@ enum __RISCV_FRM {
49	49	#define __riscv_vsetvl_e32m4(avl) __builtin_rvv_vsetvli((size_t)(avl), 2, 2)
50	50	#define __riscv_vsetvl_e32m8(avl) __builtin_rvv_vsetvli((size_t)(avl), 2, 3)
51	51
52		#if __riscv_v_elen >= 64
53	52	#define __riscv_vsetvl_e8mf8(avl) __builtin_rvv_vsetvli((size_t)(avl), 0, 5)
54	53	#define __riscv_vsetvl_e16mf4(avl) __builtin_rvv_vsetvli((size_t)(avl), 1, 6)
55	54	#define __riscv_vsetvl_e32mf2(avl) __builtin_rvv_vsetvli((size_t)(avl), 2, 7)
...	...	@@ -58,7 +57,6 @@ enum __RISCV_FRM {
58	57	#define __riscv_vsetvl_e64m2(avl) __builtin_rvv_vsetvli((size_t)(avl), 3, 1)
59	58	#define __riscv_vsetvl_e64m4(avl) __builtin_rvv_vsetvli((size_t)(avl), 3, 2)
60	59	#define __riscv_vsetvl_e64m8(avl) __builtin_rvv_vsetvli((size_t)(avl), 3, 3)
61		#endif
62	60
63	61	#define __riscv_vsetvlmax_e8mf4() __builtin_rvv_vsetvlimax(0, 6)
64	62	#define __riscv_vsetvlmax_e8mf2() __builtin_rvv_vsetvlimax(0, 7)
...	...	@@ -78,7 +76,6 @@ enum __RISCV_FRM {
78	76	#define __riscv_vsetvlmax_e32m4() __builtin_rvv_vsetvlimax(2, 2)
79	77	#define __riscv_vsetvlmax_e32m8() __builtin_rvv_vsetvlimax(2, 3)
80	78
81		#if __riscv_v_elen >= 64
82	79	#define __riscv_vsetvlmax_e8mf8() __builtin_rvv_vsetvlimax(0, 5)
83	80	#define __riscv_vsetvlmax_e16mf4() __builtin_rvv_vsetvlimax(1, 6)
84	81	#define __riscv_vsetvlmax_e32mf2() __builtin_rvv_vsetvlimax(2, 7)
...	...	@@ -87,7 +84,6 @@ enum __RISCV_FRM {
87	84	#define __riscv_vsetvlmax_e64m2() __builtin_rvv_vsetvlimax(3, 1)
88	85	#define __riscv_vsetvlmax_e64m4() __builtin_rvv_vsetvlimax(3, 2)
89	86	#define __riscv_vsetvlmax_e64m8() __builtin_rvv_vsetvlimax(3, 3)
90		#endif
91	87
92	88
93	89	enum __RISCV_VXRM {

lib/include/shaintrin.h+3-2

...	...	@@ -47,8 +47,9 @@
47	47	/// An immediate value where bits [1:0] select among four possible
48	48	/// combining functions and rounding constants (not specified here).
49	49	/// \returns A 128-bit vector of [4 x i32] containing the updated SHA-1 state.
50		#define _mm_sha1rnds4_epu32(V1, V2, M) \
51		__builtin_ia32_sha1rnds4((__v4si)(__m128i)(V1), (__v4si)(__m128i)(V2), (M))
	50	#define _mm_sha1rnds4_epu32(V1, V2, M) \
	51	((__m128i)__builtin_ia32_sha1rnds4((__v4si)(__m128i)(V1), \
	52	(__v4si)(__m128i)(V2), (M)))
52	53
53	54	/// Calculates the SHA-1 state variable E from the SHA-1 state variables in
54	55	/// the 128-bit vector of [4 x i32] in \a __X, adds that to the next set of

lib/include/stdcountof.h created+15

...	...	@@ -0,0 +1,15 @@
	1	/*===---- stdcountof.h - Standard header for countof -----------------------===
	2	*
	3	* Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
	4	* See https://llvm.org/LICENSE.txt for license information.
	5	* SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
	6	*
	7	*===-----------------------------------------------------------------------===
	8	*/
	9
	10	#ifndef __STDCOUNTOF_H
	11	#define __STDCOUNTOF_H
	12
	13	#define countof _Countof
	14
	15	#endif /* __STDCOUNTOF_H */

lib/include/stdint.h+18-129

...	...	@@ -317,166 +317,55 @@ typedef __UINTMAX_TYPE__ uintmax_t;
317	317	* integer width that the target implements, so corresponding macros are
318	318	* defined below, too.
319	319	*
320		* These macros are defined using the same successive-shrinking approach as
321		* the type definitions above. It is likewise important that macros are defined
322		* in order of decending width.
323		*
324	320	* Note that C++ should not check __STDC_CONSTANT_MACROS here, contrary to the
325	321	* claims of the C standard (see C++ 18.3.1p2, [cstdint.syn]).
326	322	*/
327	323
328		#define __int_c_join(a, b) a ## b
329		#define __int_c(v, suffix) __int_c_join(v, suffix)
330		#define __uint_c(v, suffix) __int_c_join(v##U, suffix)
331
332
333		#ifdef __INT64_TYPE__
334		# undef __int64_c_suffix
335		# undef __int32_c_suffix
336		# undef __int16_c_suffix
337		# undef __int8_c_suffix
338		# ifdef __INT64_C_SUFFIX__
339		# define __int64_c_suffix __INT64_C_SUFFIX__
340		# define __int32_c_suffix __INT64_C_SUFFIX__
341		# define __int16_c_suffix __INT64_C_SUFFIX__
342		# define __int8_c_suffix __INT64_C_SUFFIX__
343		# endif /* __INT64_C_SUFFIX__ */
344		#endif /* __INT64_TYPE__ */
345
346	324	#ifdef __int_least64_t
347		# ifdef __int64_c_suffix
348		# define INT64_C(v) __int_c(v, __int64_c_suffix)
349		# define UINT64_C(v) __uint_c(v, __int64_c_suffix)
350		# else
351		# define INT64_C(v) v
352		# define UINT64_C(v) v ## U
353		# endif /* __int64_c_suffix */
	325	#define INT64_C(v) __INT64_C(v)
	326	#define UINT64_C(v) __UINT64_C(v)
354	327	#endif /* __int_least64_t */
355	328
356	329
357	330	#ifdef __INT56_TYPE__
358		# undef __int32_c_suffix
359		# undef __int16_c_suffix
360		# undef __int8_c_suffix
361		# ifdef __INT56_C_SUFFIX__
362		# define INT56_C(v) __int_c(v, __INT56_C_SUFFIX__)
363		# define UINT56_C(v) __uint_c(v, __INT56_C_SUFFIX__)
364		# define __int32_c_suffix __INT56_C_SUFFIX__
365		# define __int16_c_suffix __INT56_C_SUFFIX__
366		# define __int8_c_suffix __INT56_C_SUFFIX__
367		# else
368		# define INT56_C(v) v
369		# define UINT56_C(v) v ## U
370		# endif /* __INT56_C_SUFFIX__ */
	331	#define INT56_C(v) __INT56_C(v)
	332	#define UINT56_C(v) __UINT56_C(v)
371	333	#endif /* __INT56_TYPE__ */
372	334
373	335
374	336	#ifdef __INT48_TYPE__
375		# undef __int32_c_suffix
376		# undef __int16_c_suffix
377		# undef __int8_c_suffix
378		# ifdef __INT48_C_SUFFIX__
379		# define INT48_C(v) __int_c(v, __INT48_C_SUFFIX__)
380		# define UINT48_C(v) __uint_c(v, __INT48_C_SUFFIX__)
381		# define __int32_c_suffix __INT48_C_SUFFIX__
382		# define __int16_c_suffix __INT48_C_SUFFIX__
383		# define __int8_c_suffix __INT48_C_SUFFIX__
384		# else
385		# define INT48_C(v) v
386		# define UINT48_C(v) v ## U
387		# endif /* __INT48_C_SUFFIX__ */
	337	#define INT48_C(v) __INT48_C(v)
	338	#define UINT48_C(v) __UINT48_C(v)
388	339	#endif /* __INT48_TYPE__ */
389	340
390	341
391	342	#ifdef __INT40_TYPE__
392		# undef __int32_c_suffix
393		# undef __int16_c_suffix
394		# undef __int8_c_suffix
395		# ifdef __INT40_C_SUFFIX__
396		# define INT40_C(v) __int_c(v, __INT40_C_SUFFIX__)
397		# define UINT40_C(v) __uint_c(v, __INT40_C_SUFFIX__)
398		# define __int32_c_suffix __INT40_C_SUFFIX__
399		# define __int16_c_suffix __INT40_C_SUFFIX__
400		# define __int8_c_suffix __INT40_C_SUFFIX__
401		# else
402		# define INT40_C(v) v
403		# define UINT40_C(v) v ## U
404		# endif /* __INT40_C_SUFFIX__ */
	343	#define INT40_C(v) __INT40_C(v)
	344	#define UINT40_C(v) __UINT40_C(v)
405	345	#endif /* __INT40_TYPE__ */
406	346
407	347
408		#ifdef __INT32_TYPE__
409		# undef __int32_c_suffix
410		# undef __int16_c_suffix
411		# undef __int8_c_suffix
412		# ifdef __INT32_C_SUFFIX__
413		# define __int32_c_suffix __INT32_C_SUFFIX__
414		# define __int16_c_suffix __INT32_C_SUFFIX__
415		# define __int8_c_suffix __INT32_C_SUFFIX__
416		# endif /* __INT32_C_SUFFIX__ */
417		#endif /* __INT32_TYPE__ */
418
419	348	#ifdef __int_least32_t
420		# ifdef __int32_c_suffix
421		# define INT32_C(v) __int_c(v, __int32_c_suffix)
422		# define UINT32_C(v) __uint_c(v, __int32_c_suffix)
423		# else
424		# define INT32_C(v) v
425		# define UINT32_C(v) v ## U
426		# endif /* __int32_c_suffix */
	349	#define INT32_C(v) __INT32_C(v)
	350	#define UINT32_C(v) __UINT32_C(v)
427	351	#endif /* __int_least32_t */
428	352
429	353
430	354	#ifdef __INT24_TYPE__
431		# undef __int16_c_suffix
432		# undef __int8_c_suffix
433		# ifdef __INT24_C_SUFFIX__
434		# define INT24_C(v) __int_c(v, __INT24_C_SUFFIX__)
435		# define UINT24_C(v) __uint_c(v, __INT24_C_SUFFIX__)
436		# define __int16_c_suffix __INT24_C_SUFFIX__
437		# define __int8_c_suffix __INT24_C_SUFFIX__
438		# else
439		# define INT24_C(v) v
440		# define UINT24_C(v) v ## U
441		# endif /* __INT24_C_SUFFIX__ */
	355	#define INT24_C(v) __INT24_C(v)
	356	#define UINT24_C(v) __UINT24_C(v)
442	357	#endif /* __INT24_TYPE__ */
443	358
444	359
445		#ifdef __INT16_TYPE__
446		# undef __int16_c_suffix
447		# undef __int8_c_suffix
448		# ifdef __INT16_C_SUFFIX__
449		# define __int16_c_suffix __INT16_C_SUFFIX__
450		# define __int8_c_suffix __INT16_C_SUFFIX__
451		# endif /* __INT16_C_SUFFIX__ */
452		#endif /* __INT16_TYPE__ */
453
454	360	#ifdef __int_least16_t
455		# ifdef __int16_c_suffix
456		# define INT16_C(v) __int_c(v, __int16_c_suffix)
457		# define UINT16_C(v) __uint_c(v, __int16_c_suffix)
458		# else
459		# define INT16_C(v) v
460		# define UINT16_C(v) v ## U
461		# endif /* __int16_c_suffix */
	361	#define INT16_C(v) __INT16_C(v)
	362	#define UINT16_C(v) __UINT16_C(v)
462	363	#endif /* __int_least16_t */
463	364
464	365
465		#ifdef __INT8_TYPE__
466		# undef __int8_c_suffix
467		# ifdef __INT8_C_SUFFIX__
468		# define __int8_c_suffix __INT8_C_SUFFIX__
469		# endif /* __INT8_C_SUFFIX__ */
470		#endif /* __INT8_TYPE__ */
471
472	366	#ifdef __int_least8_t
473		# ifdef __int8_c_suffix
474		# define INT8_C(v) __int_c(v, __int8_c_suffix)
475		# define UINT8_C(v) __uint_c(v, __int8_c_suffix)
476		# else
477		# define INT8_C(v) v
478		# define UINT8_C(v) v ## U
479		# endif /* __int8_c_suffix */
	367	#define INT8_C(v) __INT8_C(v)
	368	#define UINT8_C(v) __UINT8_C(v)
480	369	#endif /* __int_least8_t */
481	370
482	371
...	...	@@ -938,8 +827,8 @@ typedef __UINTMAX_TYPE__ uintmax_t;
938	827	#endif
939	828
940	829	/* 7.18.4.2 Macros for greatest-width integer constants. */
941		#define INTMAX_C(v) __int_c(v, __INTMAX_C_SUFFIX__)
942		#define UINTMAX_C(v) __int_c(v, __UINTMAX_C_SUFFIX__)
	830	#define INTMAX_C(v) __INTMAX_C(v)
	831	#define UINTMAX_C(v) __UINTMAX_C(v)
943	832
944	833	/* C23 7.22.3.x Width of other integer types. */
945	834	#if defined(__STDC_VERSION__) && __STDC_VERSION__ >= 202311L

lib/include/vecintrin.h+5

...	...	@@ -7,6 +7,9 @@
7	7	*===-----------------------------------------------------------------------===
8	8	*/
9	9
	10	#ifndef _VECINTRIN_H
	11	#define _VECINTRIN_H
	12
10	13	#if defined(__s390x__) && defined(__VEC__)
11	14
12	15	#define __ATTRS_ai __attribute__((__always_inline__))
...	...	@@ -12861,3 +12864,5 @@ vec_search_string_until_zero_cc(__vector unsigned int __a,
12861	12864	#error "Use -fzvector to enable vector extensions"
12862	12865
12863	12866	#endif
	12867
	12868	#endif /* _VECINTRIN_H */

lib/include/x86gprintrin.h-14

...	...	@@ -10,33 +10,19 @@
10	10	#ifndef __X86GPRINTRIN_H
11	11	#define __X86GPRINTRIN_H
12	12
13		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__HRESET__)
14	13	#include <hresetintrin.h>
15		#endif
16	14
17		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__UINTR__)
18	15	#include <uintrintrin.h>
19		#endif
20	16
21		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__USERMSR__)
22	17	#include <usermsrintrin.h>
23		#endif
24	18
25		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__CRC32__)
26	19	#include <crc32intrin.h>
27		#endif
28	20
29		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__PRFCHI__)
30	21	#include <prfchiintrin.h>
31		#endif
32	22
33		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__RAOINT__)
34	23	#include <raointintrin.h>
35		#endif
36	24
37		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__CMPCCXADD__)
38	25	#include <cmpccxaddintrin.h>
39		#endif
40	26
41	27	#if defined(__i386__)
42	28	#define __SAVE_GPRBX "mov {%%ebx, %%eax \|eax, ebx};"

lib/include/x86intrin.h-18

...	...	@@ -14,40 +14,22 @@
14	14
15	15	#include <immintrin.h>
16	16
17		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__PRFCHW__)
18	17	#include <prfchwintrin.h>
19		#endif
20	18
21		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__SSE4A__)
22	19	#include <ammintrin.h>
23		#endif
24	20
25		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__FMA4__)
26	21	#include <fma4intrin.h>
27		#endif
28	22
29		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__XOP__)
30	23	#include <xopintrin.h>
31		#endif
32	24
33		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__TBM__)
34	25	#include <tbmintrin.h>
35		#endif
36	26
37		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__LWP__)
38	27	#include <lwpintrin.h>
39		#endif
40	28
41		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__MWAITX__)
42	29	#include <mwaitxintrin.h>
43		#endif
44	30
45		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__CLZERO__)
46	31	#include <clzerointrin.h>
47		#endif
48	32
49		#if !defined(__SCE__) \|\| __has_feature(modules) \|\| defined(__RDPRU__)
50	33	#include <rdpruintrin.h>
51		#endif
52	34
53	35	#endif /* __X86INTRIN_H */

lib/include/xmmintrin.h+3-2

...	...	@@ -2198,8 +2198,9 @@ _mm_storer_ps(float *__p, __m128 __a)
2198	2198	#define _MM_HINT_NTA 0
2199	2199
2200	2200	#ifndef _MSC_VER
2201		/* FIXME: We have to #define this because "sel" must be a constant integer, and
2202		Sema doesn't do any form of constant propagation yet. */
	2201	// If _MSC_VER is defined, we use the builtin variant of _mm_prefetch.
	2202	// Otherwise, we provide this macro, which includes a cast, allowing the user
	2203	// to pass a pointer of any time. The _mm_prefetch accepts char to match MSVC.
2203	2204
2204	2205	/// Loads one cache line of data from the specified address to a location
2205	2206	/// closer to the processor.